Issuu on Google+


Dibuix de la coberta: Asterina gibbosa, estrelleta, estrella de capitán, gibbous starlet de Jordi Domènech Editor Executiu / Editor Ejecutivo / Executive Editor Joan Carles Senar Secretària de Redacció / Secretaria de Redacción / Managing Editor Montserrat Ferrer Assistència Tècnica / Asistencia Técnica / Technical Assistance Eulàlia Garcia Anna Omedes Josep Piqué Francesc Uribe

Secretaria de Redacció / Secretaría de Redacción / Editorial Office Museu de Ciències Naturals de Barcelona Passeig Picasso s/n. 08003 Barcelona, Spain Tel. +34–93–3196912 Fax +34–93–3104999 E–mail abc@bcn.cat

Editors / Editores / Editors Pere Abelló Inst. de Ciències del Mar CMIMA–CSIC, Barcelona, Spain Javier Alba–Tercedor Univ. de Granada, Granada, Spain Russell Alpizar–Jara Univ. of Évora, Portugal Xavier Bellés Centre d' Investigació i Desenvolupament–CSIC, Barcelona, Spain Luís Mª Carrascal Museo Nacional de Ciencias Naturales–CSIC, Madrid, Spain Michael J. Conroy Univ. of Georgia, Athens, USA Adolfo Cordero Univ. de Vigo, Vigo, Spain Mario Díaz Univ. de Castilla–La Mancha, Toledo, Spain José Antonio Donazar Estación Biológica de Doñana–CSIC, Sevilla, Spain Jordi Figuerola Estación Biológica de Doñana–CSIC, Sevilla, Spain Gary D. Grossman Univ. of Georgia, Athens, USA Damià Jaume IMEDEA–CSIC, Univ. de les Illes Balears, Spain Jordi Lleonart Inst. de Ciències del Mar CMIMA–CSIC, Barcelona, Spain Jorge M. Lobo Museo Nacional de Ciencias Naturales–CSIC, Madrid, Spain Pablo J. López–González Univ. de Sevilla, Sevilla, Spain Juan José Negro Estación Biológica de Doñana–CSIC, Sevilla, Spain Vicente M. Ortuño Univ. de Alcalá de Henares, Alcalá de Henares, Spain Miquel Palmer IMEDEA–CSIC, Univ. de les Illes Balears, Spain Javier Perez–Barberia The Macaulay Institute, Scotland, United Kingdom Oscar Ramírez Inst. de Biologia Evolutiva UPF–CSIC, Barcelona, Spain Montserrat Ramón Inst. de Ciències del Mar CMIMA­–CSIC, Barcelona, Spain Ignacio Ribera Inst. de Biología Evolutiva CSIC–UPF, Barcelona, Spain Pedro Rincón Museo Nacional de Ciencias Naturales–CSIC, Madrid, Spain Alfredo Salvador Museo Nacional de Ciencias Naturales–CSIC, Madrid, Spain José Luís Tellería Univ. Complutense de Madrid, Madrid, Spain Francesc Uribe Museu de Ciències Naturals de Barcelona, Barcelona, Spain Carles Vilà Estación Biológica de Doñana–CSIC, Sevilla, Spain Rafael Zardoya Museo Nacional de Ciencias Naturales–CSIC, Madrid, Spain Consell Editor / Consejo Editor / Editorial Board José A. Barrientos Univ. Autònoma de Barcelona, Bellaterra, Spain Jean C. Beaucournu Univ. de Rennes, Rennes, France David M. Bird McGill Univ., Québec, Canada Mats Björklund Uppsala Univ., Uppsala, Sweden Jean Bouillon Univ. Libre de Bruxelles, Brussels, Belgium Miguel Delibes Estación Biológica de Doñana–CSIC, Sevilla, Spain Dario J. Díaz Cosín Univ. Complutense de Madrid, Madrid, Spain Alain Dubois Museum national d’Histoire naturelle–CNRS, Paris, France John Fa Durrell Wildlife Conservation Trust, Jersey, United Kingdom Marco Festa–Bianchet Univ. de Sherbrooke, Québec, Canada Rosa Flos Univ. Politècnica de Catalunya, Barcelona, Spain Josep Mª Gili Inst. de Ciències del Mar CMIMA–CSIC, Barcelona, Spain Edmund Gittenberger Rijksmuseum van Natuurlijke Historie, Leiden, The Netherlands Fernando Hiraldo Estación Biológica de Doñana–CSIC, Sevilla, Spain Patrick Lavelle Inst. Français de recherche scient. pour le develop. en cooperation, Bondy, France Santiago Mas–Coma Univ. de Valencia, Valencia, Spain Joaquín Mateu Barcelona, Spain Neil Metcalfe Univ. of Glasgow, Glasgow, United Kingdom Jacint Nadal Univ. de Barcelona, Barcelona, Spain Eduard Petitpierre Univ. de les Illes Balears, Palma de Mallorca, Spain Taylor H. Ricketts Stanford Univ., Stanford, USA Joandomènec Ros Univ. de Barcelona, Barcelona, Spain Valentín Sans–Coma Univ. de Málaga, Málaga, Spain Tore Slagsvold Univ. of Oslo, Oslo, Norway Animal Biodiversity and Conservation 34.2, 2011 © 2011 Museu de Ciències Naturals, Institut de Cultura, Ajuntament de Barcelona Autoedició: Montserrat Ferrer Fotomecànica i impressió: ISSN: 1578–665X Dipòsit legal: B–16.278–58 The journal is freely available online at: http://www.bcn.cat/ABC and http://abc.bioexplora.cat/


Animal Biodiversity and Conservation 34.2 (2011)

229

Effects of species’ traits and data characteristics on distribution models of threatened invertebrates R. M. Chefaoui, J. M. Lobo & J. Hortal

Chefaoui, R. M., Lobo, J. M. & Hortal, J., 2011. Effects of species’ traits and data characteristics on distribution models of threatened invertebrates. Animal Biodiversity and Conservation, 34.2: 229–247. Abstract Effects of species’ traits and data characteristics on distribution models of threatened invertebrates.— The lack of information about the distribution of threatened species inhibits the development of strategies for their conservation. This is a particularly important problem when considering invertebrates. Here we evaluate the effects of species’ traits and data characteristics on the accuracy of species distribution models (SDM) of 20 threatened Iberian invertebrates. We found that the accuracy of the predictions was mostly affected by the characteristics of the data. Species whose distributions were most accurately modelled were those with a greater sample size or smaller relative occurrence area (ROA). Species in habitats that were difficult to detect using GIS data, such as riparian species, tended to be more difficult to predict. Key words: Ecological traits, Geographical distribution range, Iberian peninsula, Predictive accuracy, Sample size, Species distribution modelling. Resumen Efectos de las características ecológicas y de los datos sobre los modelos de distribución de invertebrados protegidos.— La escasez de información sobre la distribución de las especies amenazadas impide el desarrollo de estrategias para su conservación, un problema particularmente importante en el caso de los invertebrados. En este trabajo se evalúan los efectos que las características ecológicas y de los datos ejercen sobre la precisión de los modelos de distribución de 20 especies ibéricas de invertebrados amenazados. Se encontró que la precisión en los modelos predictivos se ve afectada mayoritariamente por las características de los datos. Las especies que obtienen modelos de distribución más precisos son aquellas con mayor tamaño de muestra o menor área de ocurrencia relativa (ROA). Además, las especies relacionadas con hábitats difíciles de detectar mediante SIG, como las especies riparias, tienden a ser más difíciles de predecir. Palabras clave: Características ecológicas, Modelos de distribución de especies, Península ibérica, Precisión del modelo, Rango de distribución geográfica, Tamaño de muestra. (Received: 16 XII 10; Conditional acceptance: 23 II 11; Final acceptance: 29 III 11) Rosa M. Chefaoui, Jorge M. Lobo & Joaquín Hortal, Depto. de Biodiversidad y Biología Evolutiva, Museo Nacional de Ciencias Naturales–CSIC, c/ José Gutiérrez Abascal 2, 28006 Madrid, España (Spain).– J. Hortal, Depto. de Ecologia, Inst. de Ciências Biológicas, Univ. Federal de Goiás, 74001–970 Goiania, GO, Brazil. Corresponding author: Rosa M. Chefaoui. E–mail: rosa.chef@gmail.com

ISSN: 1578–665X

© 2011 Museu de Ciències Naturals de Barcelona


Chefaoui et al.

230

Introduction Including rare and threatened species in the prioritisation of protected areas is particularly challenging because of the low spatial congruence (i.e., coincidence in space) between species ranges (Grenyer et al., 2006) and the difficulties associated with mapping their distributions. Data scarcity is often ameliorated with the help of GIS–based models and analytical techniques. Species distribution modelling is nowadays a well–established set of research techniques (see Franklin, 2009 and references therein), and many studies use data from museum collections and literature to model the distributions of species (e.g., Reutter et al., 2003; Brotons et al., 2004; Elith & Leathwick, 2007). Species distribution models (SDMs) are especially important when working with hyperdiverse invertebrates, where the difficulty of conducting extensive surveys makes biodiversity databases based on data from museums and atlases a necessary alternative to obtain presence records for mapping distributions (e.g., Chefaoui et al., 2005; Lobo et al., 2006, 2010; Chefaoui & Lobo, 2007). Unfortunately, the quality of data on many species in biodiversity databases is usually compromised by sampling bias and/or deficient survey effort (Hortal et al., 2007), a problem that is particularly important for many invertebrate groups (Lobo et al., 2007; Hortal et al., 2008). Under these circumstances, systematic conservation planning for invertebrate taxa generally entails modelling species with diverse characteristics and ecological requirements using poor quality data, often with no time for detailed 'species–by–species' assessment by experts (see Cabeza et al., 2010). Using automated SDM protocols to predict the distribution of invertebrates from presence–only data is hampered by: (i) the use of heterogeneous biological data sources generally without any survey effort measure; (ii) the environmental and spatially biased character of this information; (iii) the lack of accurate absence data; (iv) the difficulty of identifying the best predictor variables for each species; and (v) the difficulty of finding a reliable accuracy measure of SDM performance that allows model success to be compared between different species (see discussion in Lobo et al., 2008, 2010; Jiménez–Valverde et al., 2008; Rocchini et al., 2011). In an attempt to understand the limitations and possibilities of SDM techniques, many studies have addressed how the characteristics of the data and different ecological or geographical species’ traits affect model accuracy. An increase in model accuracy has been related to greater sample sizes (Stockwell & Peterson, 2002; McPherson et al., 2004; Wisz et al., 2008; Mateo et al., 2010), and also to species with more specialized requirements (Brotons et al., 2004; Seoane et al., 2005), less mobility (Pöyry et al., 2008), more between–year population constancy (Carrascal et al., 2006), longer life spans in plants (Hanspach et al., 2010; Syphard & Franklin, 2010), specific types of response to fire disturbance in plants (Syphard & Franklin, 2010), and smaller geographic ranges (Stockwell & Peterson, 2002; Segurado &

Araújo, 2004; Hernández et al., 2006). Nevertheless, relationships between model performance and species traits are strongly dependent on the modelling technique, and also on the characteristics of the data itself. These characteristics refer to sample size and the proportion of the occupied area over the total area of the territory under study (the relative occurrence area or ROA; Lobo, 2008; Lobo et al., 2008; Santika, 2010). Thus, a better understanding of how species’ traits and data characteristics influence the results of different modelling methods could help refine the use of SDMs. The main aim of this study was to determine how ecological traits and data characteristics influence the predictive performance of SDMs in the case of threatened insects and other invertebrate species. More precisely, we examined the relationship between three general measures of model accuracy (AUC, sensitivity and specificity), and (i) two characteristics of the data used, namely sample size and ROA, and (ii) several ecological traits, including niche specialization (marginality), the total extent of the distribution range (herein TER), dispersal ability, trophic group, habitat type and habitat detectability. To do this, we applied three SDM procedures (Generalized Linear Models, GLMs; Generalized Additive Models, GAMs; and Neural Network Models, NNETs) to model the distribution in the Iberian Peninsula of 20 threatened invertebrate species that have different ecological traits and data characteristics. We used presence data from museum collections and atlases, and pseudo–absences (Zaniewski et al., 2002; Chefaoui & Lobo, 2008). We later evaluated the influence of the data characteristics and species traits on model performance measures using non–parametric statistical tests. Methods Study area The study area was the Ibero–Balearic region (western Mediterranean), which comprises 587,663 km2. Data on species occurrences were gathered from atlases and bibliographic sources, using 10 km–resolution (i.e., 100 km2) UTM cells due to the lack of geographical precision of most data sources. Environmental data were also referenced to the same resolution. The study area was therefore divided into 6,150 cells of 100 km2, which constitute the units of analysis. Biological data We arbitrarily selected 20 species of threatened and/ or protected (Bern Convention and Habitat Directive) invertebrates found in Spain (17 Arthropoda and 3 Mollusca; table 1). Based on different catalogues (see http://www.mma.es; Galante & Verdú, 2000; Verdú & Galante, 2006), we selected species that fulfilled two requirements: (i) their presence had been recorded in a minimum of ten 10 x 10 km grid cells; and (ii) they had different biological and ecological traits and data characteristics. Occurrence data


Animal Biodiversity and Conservation 34.2 (2011)

231

Table 1. Data and species characteristics that may influence model accuracy: N. Sample size; ROA. Relative occurrence area; M. Marginality; TER. Total extent of the distribution range, in three categories from more restricted to wider distribution (C–I. Iberian [a] and Ibero–Maghrebian [b]; C–II. European; C–III. Euroasiatic); HT. Habitat types (T–I. Woods [a] and mountainous habitats [b]; T–II. Grasslands [a], varied habitats [b] and rocky slopes [c]; T–III. Riparian [a] and humid habitats [b]); HD. Habitat detectability (H. High; L. Low); TLC. Trophic level categories (Pl. Polyphagous; Cr. Carnivorous; Om. Omnivorous; Ol. Oligophagous; Ph. Phytophagous; NPh. Non–phytophagous); FC. Flight capacity. Tabla 1. Características de los datos y de las especies que pueden influir en la precisión de los modelos: N. Tamaño de muestra; ROA. Área de presencia relativa; M. Marginalidad; TER. Extensión total del rango de distribución, en tres categorías desde la distribución más restringida a más amplia (C–I. Ibérica [a] e ibero–magrebí [b]; C–II. Europea; C–III. Euroasiática); HT. Tipos de hábitats (T–I. Bosques [a] y hábitats montañosos [b]; T–II. Praderas [a]; hábitats mixtos [b] y pendientes rocosas [c]; T–III. Hábitats húmedos [b] y riparios [a]); HD. Detectabilidad del hábitat (H. Alta; L. Baja); TLC. Categorías de nivel trófico (Pl. Polífago; Cr. Carnívoro; Om. Omnívoro; Ol. Oligófago; Ph. Fitófagas; NPh. No fitófagas); FC. Capacidad de vuelo. Species

Data characteristics

Species characteristics

N

ROA

M

TER

HT

HD

TLC

FC

Cerambyx cerdo

152

0.796

0.768

C–III

T–I(a)

H

Pl(Ph)

Yes

Coenagrion mercuriale

87

0.629

0.455

C–I(b)

T–III(a)

L

Cr(NPh)

Yes

Cupido lorquinii

87

0.267

1.201

C–I(b)

T–II(a)

H

Om(NPh)

Yes

Elona quimperiana

41

0.141

2.869

C–I(b)

T–II(b)

L

Om(NPh)

No

Eriogaster catax

12

0.067

2.538

C–III

T–I(a)

H

Pl(Ph)

Yes

Euphydryas aurinia

749

0.851

1.154

C–III

T–I(a)

H

Ol(Ph)

Yes

Geomalacus maculosus

37

0.114

2.397

C–II

T–III(b)

L

Pl(Ph)

No

Graellsia isabelae

138

0.212

2.240

C–I(a)

T–I(a)

H

Ol(Ph)

Yes

Lucanus cervus

456

0.625

1.915

C–III

T–I(a)

H

Pl(Ph)

Yes

Macromia splendens

10

0.436

1.797

C–I(b)

T–III(a)

L

Cr(NPh)

Yes

Macrothele calpeiana

92

0.076

1.624

C–I(a)

T–II(b)

L

Cr(NPh)

No

Maculinea alcon

49

0.212

2.528

C–II

T–II(a)

H

Om(NPh)

Yes

Maculinea arion

166

0.310

3.397

C–III

T–II(a)

H

Om(NPh)

Yes

Maculinea nausithous

17

0.041

4.584

C–II

T–II(a)

H

Om(NPh)

Yes

Oxygastra curtisi

21

0.612

1.971

C–II

T–III(a)

L

Om(NPh)

Yes

Parnassius apollo

314

0.459

3.600

C–III

T–I(b)

L

Pl(Ph)

Yes

Parnassius mnemosyne

42

0.017

5.897

C–III

T–I(b)

H

Ol(Ph)

Yes

Rosalia alpina

47

0.132

3.656

C–III

T–I(a)

H

Pl(Ph)

Yes

Vertigo moulinsiana

20

0.064

1.261

C–II

T–III(b)

L

Cr(NPh)

No

1,107

0.927

0.376

C–I (b)

T–II(c)

L

Ol(Ph)

Yes

Zerynthia rumina

were obtained from the abovementioned catalogues, and from a diverse array of bibliographic sources (Soria et al., 1986; Castillejo, 1990; Rosas et al., 1992; Viejo Montesinos, 1992; Grosso–Silva, 1999; Grupo de Trabajo sobre Lucanidae Ibéricos, 2000; García–Barros & Herranz, 2001; Pérez–Bote et al., 2001; Raimundo et al., 2001; López–Sebastián et al., 2002; Martínez–Orti, 2004). Because accurate absence data were not available, we used pseudo–absences to perform model

training and validation. We identified environmental pseudo–absences located outside the climatic domain, defined by the available presences (see Lobo et al., 2010). To establish such a climatic domain we used a profile technique; a multidimensional envelope containing all presence data in a multivariate environmental space (Busby, 1991; Lobo et al., 2006) was calculated for each species using the maximum and minimum scores for each topographic, climatic and lithological variable mentioned in table 2.


Chefaoui et al.

232

Table 2. Predictor variables used to generate distribution models for the species. The appropriate variables for each species were previously selected by individual logistic regression analyses (see text). Tabla 2. Variables predictivas usadas para generar los modelos de distribución de especies. Las variables apropiadas para cada especie fueron previamente seleccionadas individualmente mediante análisis de regresión logística (ver texto). Predictor variables

Minimum–Maximum values

Topographic variables Maximum elevation (m)

1–3,399

Mean elevation (m)

1–2,721

Minimum elevation (m)

0–2,521

Elevation range (m)

0–2,291

Climatic variables Winter precipitation (Jan., Feb., March) (mm)

491–9,579

Spring precipitation (April, May, June) (mm)

463–6,236

Summer precipitation (July, August, Sept.) (mm)

66–4,724

Autumn precipitation (Oct., Nov., Dec.) (mm)

607–6,140

Temperature range (ºC)

11–32

Maximum Winter Temperature (ºC)

1–18

Mean Winter Temperature (ºC)

–4–13

Minimum Winter Temperature (ºC)

–8–10

Maximum Spring Temperature (ºC)

6–23

Mean Spring Temperature (ºC)

0–17

Minimum Spring Temperature (ºC)

–5–12

Maximum Summer Temperature (ºC)

19–35

Mean Summer Temperature (ºC)

10–26

Minimum Summer Temperature (ºC)

2–20

Maximum Autumn Temperature (ºC)

9–25

Mean Autumn Temperature (ºC)

2–21

Minimum Autumn Temperature (ºC)

–3–15

Aridity

0–1.64

Lithological variables Area of acid soil (km2)

0–100

Area of calcareous soil (km )

0–100

2

Area of acid sediments (km )

0–100

2

Area of calcareous sediments (km ) 2

0–100

Spatial variables Latitude (Y)

390000–4860000

Longitude (X)

–20000–1060058

We then created environmental pseudo–absences equalling ten times the number of presences (prevalence = 0.1). This way we included as many absences as possible in the training data, while avoiding biases

caused by the inclusion of an extremely high number of absences (e.g., prevalences below 0.01) (King & Zeng, 2000; Dixon et al., 2005; Jiménez–Valverde & Lobo, 2006; Jiménez–Valverde et al., 2009).


Animal Biodiversity and Conservation 34.2 (2011)

As pseudo–absences were randomly selected from the area outside each envelope, they a priori excluded the possibility that some environmentally suitable localities where the species does not occur (either because it has not been able to colonize there or because it recently became extinct) would be counted as absences. Geographical predictions thus obtained would tend to approximate the potential distributions of the studied species rather than their realized distributions, as would occur if we were using random pseudo–absences (see Chefaoui & Lobo, 2008; Jiménez–Valverde et al., 2008; Lobo et al., 2010; see also Beaumont et al., 2009). In addition, by choosing pseudo–absences far from the environmental domain occupied by the presence data the discriminant ability of the environmental predictors would be maximized, because no pseudo–absences would be located in environmental domains similar to those occupied by the species presences. Using this kind of pseudo– absences inevitably inflates the AUC values obtained to measure model accuracy (Chefaoui & Lobo, 2008) because the localities with unsuitable environmental conditions are almost always well predicted. In this study we assumed that such inflation of AUC values was similar for all species, independently of their degree of equilibrium with the environment or how narrow their environmental tolerances were. In our case, low AUC values would highlight the inability of some predictor variables to discriminate suitable from unsuitable conditions. It should be noted here that we were interested in assessing the effects of species’ traits and data characteristics on the accuracy of models aimed at representing the potential distribution of species. Therefore, we used different techniques and/or predictors to identify any patterns that consistently emerged despite the slightly different assumptions and flexibility in functions of each modelling strategy, rather than to assess the performance of different SDM techniques. Predictor variables Due to the heterogeneity in the ecological roles, life histories and adaptations of the invertebrates studied, we selected the best set of predictor variables (table 2) for each species from a range of topographic, climatic, lithological and spatial variables by means of a selection procedure (see below). We extracted topographic variables (maximum, mean and minimum elevation) from a global digital elevation model with 1–km spatial resolution (Clark Labs, 2000); elevation range was calculated as the difference between maximum and minimum elevation in each cell. GIS–layers accounting for minimum, mean and maximum temperature and precipitation for each season at 1–km resolution based on observations from weather stations were provided by the Spanish State Agency of Meteorology (http://www.aemet.es/). We calculated aridity as Ia = 1/ (P/T + 10) x 102 where P is the mean annual precipitation and T the mean annual temperature (see Verdú & Galante, 2002). We digitized four lithology variables from a

233

lithology map (Instituto Geográfico Nacional, 1995), and subsequently calculated the area of calcareous deposits, siliceous sediments, stony acidic soils and calcareous soils on each 100 km2 UTM cell. Finally, we extracted two spatial variables per cell: latitude (Lat) and longitude (Lon) of the centroid of each cell, and generated a trend surface with the third order polynomial of longitude and latitude (i.e., Trend Surface Analysis). The inclusion of these spatial variables after environmental predictors can help to represent the effect of unaccounted–for predictors and/or other factors known to generate spatial patterns in species distributions (see Legendre & Legendre, 1998). All predictor variables were extracted and handled using IDRISI Kilimanjaro GIS software (Clark Labs, 2003) to the 10 x 10 km UTM grid cells. All these variables (including latitude and longitude) were standardized to zero mean and one standard deviation to eliminate the effect of varying measurement scales. Species distribution models Species presence data, pseudo–absences and the selected predictor variables for each species were used to generate predictive functions for species distributions, by means of three different and widely used SDM techniques: Generalized Linear Models (GLMs), Generalized Additive Models (GAMs) and Neural Network Models (NNETs). GLMs (McCullagh & Nelder, 1989) were elaborated assuming a logistic relationship between the dependent and the explanatory variables (i.e., link function), and a binomial error distribution of the dependent variable. To select the best explanatory variables for each species, presence–absence data were regressed against each one of the explanatory variables, using Statistica software (Statsoft, 2001). We evaluated the linear, quadratic and cubic functions for each variable, in order to account for possible curvilinear relationships (Austin, 1980). In addition, we chose the most appropriate spatial variables for each species after a backward–stepwise elimination of non–significant terms from the third–degree polynomial of latitude and longitude. The selected explanatory variables were used in the GAM models using penalized regression splines (Wood & Augustin, 2002) and in the NNET models fitting a single–hidden–layer neural network, with skip–layer connections (Ripley, 1996). All Species Distribution Models were fitted in R (R Development Core Team, 2008). Measures of model performance Given that the sample size for some species was small, we opted not to split it into representative training and evaluation datasets. We thus implemented a 'leave–one–out' jack–knife procedure (Olden et al., 2002) to validate models for all species. For this procedure, each observation is excluded and the model is parameterized using the remaining n – 1 observations to obtain a predicted probability score for the excluded observation; this procedure yields relatively unbiased estimates of model performance (Olden et al., 2002).


Chefaoui et al.

234

Table 3. Accuracy measures and resulting area size for each studied species and modelling technique used. All areas are measured as the number of grid cells (of 100 km2 each): GAM. Generalized additive models; GLM. Generalized linear models; NNET. Neural network models; SD. Standard deviation. Tabla 3. Medidas de precisión y tamaño de área resultante para cada especie estudiada y técnica predictiva utilizada. Todas las áreas se miden por el número de celdas (de 100 km2 cada una): GAM. Modelos aditivos generalizados; GLM. Modelos generalizados lineales; NNET. Modelos de redes neuronales; SD. Desviación estándar. Species

AUC GAM

GLM

Specificity NNET

GAM

GLM

Sensitivity NNET

GAM

GLM

Area (in grid cells) NNET

GAM

GLM

NNET

Cerambyx cerdo 0.9402

0.9557 0.8353 0.8620 0.8746 0.7565 0.8618 0.8750 0.7565 4,328

4,458

2,507

4,165

1943

859

822

551

398

4,866

5,234

4,159

4,019

747

676

962

998

3,243

3,056

848

266

369

318

938

1,024

1,141

1,074

214

172

285

677

493

557

1,622

1,884

147

125

Coenagrion mercuriale 0.9196

0.9414 0.8020 0.8275 0.8735 0.7298 0.8275 0.8735 0.7356 3,857

Cupido lorquinii 0.9730

0.8936 0.9788 0.9563 0.9827 0.9310 0.9540 0.8045 0.9310 1,021

Elona quimperiana 0.9866

0.9692 0.9885 0.9512 0.9439 0.9463 0.9512 0.9512 0.9512

594

Eriogaster catax 0.9430

0.9062 0.9840 0.9083 0.9583 0.9166 0.9166 0.8333 0.9166 4,845

Euphydryas aurinia 0.9896

0.9909 0.9835 0.9524 0.9571 0.9508 0.9519 0.9572 0.9506 4,127

Geomalacus maculosus 0.9554

0.9654 0.9385 0.8918 0.9189 0.8918 0.8918 0.9189 0.8918

784

Graellsia isabelae 0.9934

0.9927 0.9709 0.9594 0.9507 0.9275 0.9565 0.9492 0.9275 1,021

Lucanus cervus 0.9926

0.9924 0.9818 0.9700 0.9649 0.9547 0.9692 0.9649 0.9539 2,983

Macromia splendens 0.8830

0.8030 0.8570 0.7600 0.7000 0.7900 0.8000 0.7000 0.8000 1,361

Macrothele calpeiana 0.9933

0.9321 0.9626 0.9565 0.9782 0.9130 0.9565 0.8804 0.9130

454

Maculinea alcon 0.9729

0.9668 0.9536 0.9346 0.9265 0.9183 0.9387 0.9183 0.9183 1,182

Maculinea arion 0.9927

0.9912 0.9785 0.9596 0.9698 0.9337 0.9578 0.9698 0.9337 1,205

Maculinea nausithous 0.9861

0.9081 0.9892 0.9411 0.9882 0.9411 0.9411 0.8235 0.9411

Oxygastra curtisi 0.8749

0.8544 0.8920 0.8904 0.9190 0.8095 0.8095 0.7619 0.8095

Parnassius apollo 0.9932

0.9917 0.9869 0.9722 0.9746 0.9726 0.9713 0.9745 0.9713 1,850

Parnassius mnemosyne 0.9953

0.9462 0.9977 0.9761 0.9880 0.9976 0.9761 0.9047 1.0000

230


Animal Biodiversity and Conservation 34.2 (2011)

235

Table 3. (Cont.) Species

AUC GAM

GLM

Specificity NNET

GAM

GLM

NNET

Sensitivity GAM

GLM

Area (in grid cells) NNET

GAM

GLM

NNET

545

485

341

405

444

1,326

5,596

5,455

Rosalia alpina 0.9881

0.9426 0.9838 0.9446 0.9148 0.9361 0.9361 0.9148 0.9361

Vertigo moulinsiana 0.9745

0.9288 0.8575 0.9350 0.8550 0.8050 0.9500 0.8500 0.8000

Zerynthia rumina 0.9860

0.9888 0.9660 0.9428 0.9551 0.9265 0.9430 0.9548 0.9268 5,422

Mean ± SD

0.9666 0.9430 0.9444 0.9245 0.9296 0.8974 0.9230 0.8890 0.8982 1,855.2 1,813.2 1,615.4 ± 0.03

± 0.04 ± 0.05 ± 0.05 ± 0.06 ± 0.07 ± 0.05 ± 0.07 ± 0.07 ±1,716.3 ±1,823.7 ±1,638.4

After repeating this procedure n times (one per observation), we used these new jack–knife probabilities to calculate three measures of model performance: (i) the area under the ROC curve (AUC) (Zweig & Campbell, 1993; Schröder, 2004), (ii) sensitivity (proportion of correctly predicted presences) and (iii) specificity (proportion of correctly predicted absences). Sensitivity and specificity were calculated fixing the threshold probability according to the prevalence of the data (0.1; see Jiménez–Valverde & Lobo, 2006). To transform the continuous probabilities obtained in SDMs to binary results (i.e., presence–absence) we used the sensitivity–specificity sum maximizer criteria (Jiménez–Valverde & Lobo, 2006, 2007). All measures ranged from 0 (poor quality model) to 1 (excellent prediction). Data characteristics We evaluated the influence of two characteristics on model performance: sample size (N) and the Relative Occurrence Area (ROA). ROA is the ratio between the area of the distribution range of the species within the studied region, and the total area of such region (Lobo, 2008; Jiménez–Valverde et al., 2008). Here, the area of the study region is the whole area of the Ibero–Balearic region (see above), and the distribution range of the species within such region was estimated as the minimum convex polygon (i.e. the smallest polygon in which no internal angle exceeds 180 degrees) that contains all presence sites (also called convex–hull; Burgman & Fox, 2003). Thus, ROA measures whether the allocation of presence points in the study area shows a relatively wide distribution (as ROA values tend to 1) or a more restricted pattern. Species traits We examined the correlation between model accuracy and six ecological and biogeographical characteristics

of the species: niche marginality, the total extent of the distribution range (TER), habitat type, habitat detectability, trophic group, and dispersal ability. Raw data on these species’ traits were collected from published information on their life histories and biogeography, and then classified into categories. The degree of specialization of each species was estimated from its marginality scores obtained with ENFA (Hirzel et al., 2007). ENFA measures the average position of the species’ niche according to the observed localities of presence in relation to the average environmental conditions in the study area; high marginality values indicate a tendency to inhabit extreme conditions regarding the overall conditions in the considered region. TER is a qualitative variable with three categories that represent the total extent and the general distribution of the species: Iberian and Ibero–Maghrebian species (C–I), European species (C–II), and Euroasiatic species (C–III). The type of habitat generally inhabited by the species was also classified into three categories: T–I (woodlands and mountainous habitats), T–II (open habitats such as grasslands, rocky slopes, etc) and T–III (humid and riparian conditions). Habitat detectability refers to the ease of detecting suitable habitat patches for each species using GIS–based data. Each species was classified according to its belonging to habitats of either low– or high–detectability. Low–detectability habitats were considered as those that are usually smaller than the resolution used in GIS data on land cover, including microhabitats such as specific host plants, under stones or river banks. Conversely, high–detectability habitats were taken to be those that were easily identifiable using GIS data, such as extensive woodlands, grasslands or mountainous areas. Species were also classified into two trophic groups according to their trophic range, phytophagous (P) or non–phytophagous (NP) species. Finally, the dispersal ability of the species was measured as a binary variable accounting for whether they are able to fly or not (table 1).


Chefaoui et al.

236

Table 4. Relationships between the three measures of model accuracy (AUC, sensitivity and specificity) and data or species’ characteristics. Spearman rank correlation coefficients (R) used to assess the effect of continuous variables (upper rows); partial correlations (RP) used to assess the individual relevance of N and ROA (lower rows). The effects of qualitative species characteristics were assessed using Kruskal–Wallis (H) and Mann–Whitney U–test (Z), either on the direct values of the accuracy measures (upper rows), or on the regression residuals on N and ROA (lower rows): M. Marginality; HT. Habitat type; TL. Trophic level; FC. Flight capacity; HD. Habitat detectability; * Statistically significant relationships (p < 0.05). Variables significant after applying a Bonferroni correction (p < 0.0060) are shown in bold. TER is the total extent of the distribution range (see text and table 1). Tabla 4. Relaciones entre las tres medidas de precisión del modelo (AUC, sensitividad y especificidad) y las características de los datos o de las especies. Los coeficientes de correlación de Spearman (R) se usan para evaluar el efecto de las variables continuas (filas superiores); las correlaciones parciales (Rp) se usan para evaluar la relevancia individual de N y ROA (filas inferiores). Los efectos de las características cualitativas de las especies sobre los valores directos de las medidas de precisión (filas superiores), o sobre los residuos de su regresión sobre N y ROA (filas inferiores), se evaluaron mediante Kruskal–Wallis (H) y el test U de Mann–Whitney (Z): M. Marginalidad; HT. Tipo de hábitat; TL. Nivel Trófico; FC. Capacidad de vuelo; HD. Detectabilidad del hábitat; * Relaciones estadísticamente significativas (p < 0,05). Las variables significativas tras aplicar la corrección de Bonferroni (p < 0,0060) se muestran en negrita. TER es la extensión total del rango de distribución (ver texto y tabla 1).

Data characteristics N

ROA

Species characteristics M

TER

HT

TL

H = 3.39

H = 7.85

Z = 1.36

FC

HD

GAM AUC

R = 0.47 R = –0.25 R = 0.41 p = 0.28

p = 0.03*

p = 0.2

p = 0.02*

p = 0.2

p = 0.8

RP = 0.75 RP = –0.74 R = 0.16

H = 5.98

H = 8.52

Z = 1.21

Z = 0.66 Z = 2.16

p < 0.001 p < 0.001 p = 0.48

p = 0.05

p = 0.01* p = 0.23

p = 0.5 p = 0.03*

Sensitivity R = 0.52 R = –0.16 R = 0.30

H = 3.20

H = 7.13

Z = 0.14 Z = 1.21

p = 0.48

p = 0.08

Z = 0.28 Z = 1.25

Z = 0.94

p = 0.21

p = 0.02*

p = 0.19

p = 0.20

p = 0.03* p = 0.34

p = 0.88 p = 0.22

RP = 0.77 RP = –0.78 R = 0.11

H = 4.05

H = 6.17

Z = 0.45

Z = 0.66 Z = 2.01

p < 0.001 p < 0.001 p = 0.63

p = 0.13

p = 0.04* p = 0.65

p = 0.5 p = 0.04*

Specificity R = 0.51 R = –0.15 R = 0.38

H = 4.11

H = 8.21

Z = 0.18 Z = 1.40

Z = 1.21

p = 0.019* p = 0.51

p = 0.09

p = 0.13

p = 0.02* p = 0.22

p = 0.85 p = 0.16

RP = 0.66 RP = –0.66 R = 0.25

H = 3.16

H = 3.76

Z = 0.38

Z = 1.23 Z = 1.56

p = 0.002 p = 0.002 p = 0.29

p = 0.2

p = 0.15

p = 0.7

p = 0.22 p = 0.12

Z = 2.19

Z = 0.28 Z = 0.87

GLM AUC

R = 0.75

H = 2.75

H = 5.04

p < 0.001 p = 0.15

p = 0.72

p = 0.25

p = 0.08 p = 0.02* p = 0.77 p = 0.38

RP = 0.54 RP = –0.31 R = 0.19

H = 2.63

H = 2.71

Z = 0.98

Z = 0.00 Z = 1.18

p = 0.016*

p = 0.42

p = 0.27

p = 0.25

p = 0.32

p = 1.00 p = 0.24

R = 0.29 R = 0.15

Z = 2.04

Z = 0.09 Z = 0.42

Sensitivity R = 0.74

R = 0.33 R = 0.08

p = 0.2

H = 3.72

H = 4.52

p < 0.001 p = 0.21

p = 0.52

p = 0.15

p = 0.10 p = 0.04* p = 0.92 p = 0.68

RP = 0.57 RP = –0.37 R = 0.27

H = 2.93

H = 2.57

Z = 1.06 Z = –0.28 Z = 0.57

p = 0.01*

p = 0.24

p = 0.23

p = 0.27

p = 0.29

p = 0.77 p = 0.57

Specificity R = 0.26 R = –0.26 R = 0.39

H = 1.29

H = 9.15

Z = 0.38

Z = 0.75 Z = 1.71

p = 0.26

p = 0.08

p = 0.52

p = 0.01* p = 0.71

p = 0.45 p = 0.09

RP = 0.51 RP = –0.49 R = 0.26

H = 1.93

H = 2.34 Z = –0.22 Z = 1.42 Z = 1.4

p = 0.023* p = 0.03* p = 0.27

p = 0.38

p = 0.31

p = 0.12 p = 0.26

p = 0.82

p = 0.15 p = 0.16


Animal Biodiversity and Conservation 34.2 (2011)

237

Table 4. (Cont.)

Data characteristics N

ROA

Species characteristics M

TER

HT

TL

FC

HD

NNET AUC

R = 0.035 R = –0.39 R = 0.69

H = 3.45

H = 8.75

Z = 1.29

Z = 0.56 Z = 1.78

p = 0.88

p = 0.08 p < 0.001

p = 0.17

p = 0.01* p = 0.19

p = 0.57 p = 0.07

RP = 0.70 RP = –0.72 R = 0.42

H = 4.04

H = 4.53

Z = 1.32 Z = 1.63

Z = 0.45

p < 0.001 p < 0.001 p = 0.07

p = 0.13

p = 0.1

p = 0.65

p = 0.18 p = 0.10

Sensitivity R = 0.32 R = –0.15 R = 0.59

H = 4.12

H = 8.84

Z = 1.43

Z = 0.80 Z = 1.71

p = 0.17

p = 0.53 p < 0.001

p = 0.13

p = 0.01* p = 0.15

p = 0.42 p = 0.08

RP = 0.73 RP = –0.74 R = 0.49

H = 4.66

H = 5.33

Z = 1.51 Z = 1.86

Z = 0.68

p < 0.001 p < 0.001 p = 0.02*

p = 0.09

p = 0.07

p = 0.49

p = 0.13 p = 0.06

Specificity R = 0.34 R = –0.14 R = 0.57

H = 4.38

H = 8.94

Z = 1.51

Z = 0.85 Z = 1.78

p = 0.14

p = 0.56 p < 0.001

p = 0.11

p = 0.01* p = 0.13

p = 0.39 p = 0.07

RP = 0.73 RP = –0.74 R = 0.48

H = 5.03

H = 6.09

Z = 0.91

Z = 1.42 Z = 1.94

p < 0.001 p < 0.001 p = 0.03*

p = 0.08

p = 0.04* p = 0.36

p = 0.16 p = 0.05

Evaluation of the influence on model performance We individually examined whether any of the data characteristics or species’ traits correlated with the measures of model performance by using non–parametric statistical tests. The influence of continuous variables (N, ROA and marginality) was assessed using Spearman rank correlations (Rs) with each one of the accuracy measures (AUC, sensitivity and specificity). Here, partial correlation analysis was also used to estimate the single contribution of N and ROA on the variation of accuracy measures. The degree of association between model accuracy measures and the qualitative variables (TER, habitat type, habitat detectability, trophic group and dispersal ability) was established using non–parametric statistical tests such as Kruskal–Wallis or Mann–Whitney U. In addition, to eliminate the influence of data characteristics, we regressed accuracy values against N and ROA. Residuals from these regression analyses were later submitted to a new correlation –either Kruskal–Wallis or Mann–Whitney U tests– to evaluate their relationships with the studied species’ traits, applying both a standard significance level (p < 0.05) and a Bonferroni correction for multiple comparisons (p = 0.05/9 = 0.006). Results The high accuracy achieved on average (mean AUC  ±  SD = 0.951 ± 0.013; mean specificity = 0.917 ± 0.017; mean sensitivity = 0.903 ± 0.017;

table 3) was to some extent expected, as absences lay outside the envelope defined by the presences and validation data were not spatially independent (Veloz, 2009). Neither AUC nor specificity or sensitivity values differed significantly between the three SDM techniques (Kruskal–Wallis test; n = 60; AUC: H = 3.98, p = 0.14; specificity: H  =  3.26, p = 0.20; sensitivity: H  = 4.20, p = 0.10). Neither did the area calculated for the potential distribution of the studied species differ significantly between the three modeling techniques (Kruskal–Wallis test; n = 60; H = 0.31, p = 0.86). Among the considered variables N, ROA and, to a lesser extent, marginality significantly (p < 0.006) affected the accuracy of distribution models (table 4). Several traits (habitat type, trophic group and habitat detectability) were also associated with model accuracy measures (p < 0.05), although their influence was much lower than data characteristics and was not significant when a Bonferroni correction was applied. In contrast, TER and flight capacity did not seem to influence any measure of model accuracy. As expected from previous studies, species with greater N obtained higher model accuracies; AUC values and sensitivity scores were higher when models were developed from samples for which there were more than 200 records (fig. 1; see appendix 1). Partial correlation analyses of both data characteristics (N and ROA) on accuracy measures showed that while sample size was always positively and significantly correlated with model accuracy, ROA was usually negatively correlated (seven out of nine; see table 4). The species’ traits showed less influence on model performance. Marginality values showed a statisti-


Chefaoui et al.

238

1.00

0.9682

0.95

0.9166 0.8918

NNET specificity

GAM sensitivity

0,9387

0.8618 0.8275 0.8000

0.700 0

200

400 600 800 1,000 1,200 Data size (N)

Fig. 1. Correlation between sensitivity of GAM models and data size (N), an example of how the number of occurrences influences accuracy scores. Similar results were obtained with specificity and AUC metrics (see Appendix 1). Fig. 1. Correlación entre la sensitividad de los modelos GAM y el tamaño de muestra (N), un ejemplo de cómo el número de ocurrencias influye sobre los valores de precisión. Resultados similares se obtuvieron con la especificidad y los valores de AUC (ver Apéndice 1).

cally significant correlation with accuracy until the effect of N and ROA was removed. The only species’ trait that remained relevant for accuracy measures after accounting for data characteristics was habitat type; predictions for species associated with humid and riparian conditions were poorer (see figs. 2, 3 and appendix 2). However, this association was not statistically significant when a Bonferroni correction was applied. Other associations, such as the trophic range of species and GLM accuracy or habitat detectability and GAM performance, also ceased to be significant under the more restrictive Bonferroni significance levels. Discussion Several species’ traits and data characteristics have shown to influence SDM performance (e.g. Brotons et al., 2004; Segurado & Araújo, 2004; Seoane et al., 2005; Hernández et al., 2006; Marmion et al., 2008). One of these characteristics is the prevalence in the dataset, which is generally thought to affect the accuracy of models (e.g. McPherson et al., 2004; Seoane et al., 2005; Marmion et al., 2008). These effects, however, might only appear in extreme prevalence values (see Jiménez–Valverde et al., 2009). Here we

0.90 0.85 0.80 0.75 0.70

C–I

C–III C–II Habitat type

Fig. 2. Specificity of NNET results by habitat type (C–I. Woods and Mountainous habitats C–II. Grasslands and varied habitats, C–III. Riparian and humid habitats). Less accurate models are obtained for species associated to riparian and humid habitats. Similar results were obtained with sensitivity and AUC metrics (see Appendix  2). The middle point shows the median response for each habitat type and specificity score combination. The bottom and top of the box show the 25 and 75 percentiles respectively. The whiskers show minimum and maximum values. Fig. 2. Resultados de especificidad de los modelos NNET en función del tipo de hábitat (C–I. Bosques y hábitats montañosos, C–II. Praderas y hábitats mixtos, C–III. Hábitats húmedos y riparios). Las especies asociadas a hábitats húmedos y riparios obtuvieron modelos menos precisos. Se obtuvieron resultados similares con las medidas de sensitividad y AUC (ver Apéndice 2). El punto central representa el valor de la combinación de la mediana para cada tipo de hábitat con los valores de especificidad, los límites inferiores y superiores de la caja muestran los percentiles 25 y 75 respectivamente. Los bigotes señalan el valor máximo y mínimo.

deliberately equalled the prevalence of all species’ datasets to avoid its effect on model performance. We also removed the effect of data size and ROA using residual analyses, and our results showed that, when these mere methodological artefacts were controlled the supposed differences in model performance attributed to the ecological or biogeographical traits of


Animal Biodiversity and Conservation 34.2 (2011)

239

A

0

GAM

B

300 km

GLM

GAM + GLM

NNET

Fig. 3. Differences between the predictive maps produced for a riparian species, Coenagrion mercuriale (A) and a species not linked to riparian habitats, Cupido lorquinii (B). Although data for both species have the same sample size (N = 87), GAM and NNET models performed better for C. lorquinii than for C. mercuriale. Note that the difference in ROA values (C. lorquinii = 0.267; C. mercuriale = 0.629) could have also influenced this disparity. Fig. 3. Diferencias entre los mapas predictivos obtenidos para una especie riparia, Coenagrion mercuriale (A) y una especie no ligada a hábitats riparios, Cupido lorquinii (B). Aunque los datos de ambas especies tienen el mismo tamaño de muestra (N = 87), los modelos GAM y NNET obtuvieron mejores resultados para C. lorquinii que para C. mercuriale. Obsérvese que la diferencia en los valores de ROA (C. lorquinii = 0,267; C. mercuriale = 0,629) también podrían haber influido en esta disparidad.

1.0 0.8 0.6 ROA

species tended to disappear. This result coincides with the findings of Santika (2010), who examined the influence of prevalence on simulated data. Such dependence on data characteristics, and the fact that these characteristics also affect the measures of SDM performance, makes us wonder whether it is possible to find an accuracy measure able to compare the performance of SDMs among different species (see Lobo et al., 2008). Larger sample sizes have previously been shown to increase model accuracy (Stockwell & Peterson, 2002; McPherson et al., 2004; Hernández et al., 2006; Wisz et al., 2008; Mateo et al., 2010). In this work, sample size had a positive and significant effect on model performance. In spite of the significant and positive correlation between sample size and ROA (RS = 0.65, p = 0.0017; fig. 4), ROA did not show any direct relationship with model performance when analyzed individually. Ecological traits of several species also seemed to influence model performance, though to a lesser degree. The species with most restricted ecological requirements (i.e., the most marginal species) were modelled more accurately than less specialized species, but only in the case of NNET. In contrast with other studies (Brotons et al., 2004; Segurado & Araújo, 2004; Luoto et al., 2005), we did not find any strong

0.4 0.2 0.0

0

200 400 600 800 1,000 1,200 N

Fig. 4. Relationship between the values of sample size (N) and the relative occurrence area (ROA). Fig. 4. Relación entre los valores del tamaño de muestra (N) y el área de presencia relativa (ROA).


240

relationship between the performance of GAM and GLM models and niche specialization (i.e., marginality). This is in agreement with Pöyry et al. (2008) and Newbold et al. (2009), who were not able to detect any effect of the niche width of butterflies regarding model accuracy. Besides, all SDM techniques, but specially GAM and NNET, seemed to perform better with species not associated with riparian and humid conditions, a result also found by McPherson & Jetz (2007). Such poor performance may be associated with a poorer localization of wetlands in land cover maps. This hampers the inclusion of predictor variables related to the quality of aquatic habitats, thereby impeding the use of the true determinants of the distribution of riparian species. Finally, we did not detect any influence of the variables measuring flight capacity and the total extent of the distribution range of the species on model accuracy. Hence, it can be assumed that the SDM techniques used are not sensitive to either how widespread the species is outside the study area, or to its dispersal capacity. However, sample size and ROA altogether seem to interact with the influence of species’ traits on model accuracy. Once the effects of these data characteristics are removed, only a few effects of species’ traits remain. In particular, the residual analyses reveal a consistent, though weak, relationship between model performance and habitat detectability; species associated to easy–to–detect habitats are predicted more accurately by GAM models than those whose preferred habitats are smaller than the resolution of the available GIS layers. This also agrees with the results obtained by McPherson & Jetz (2007), where habitat detectability also had a secondary role on model accuracy. Besides, this result supports the idea commented above: the low detectability of riparian and humid habitats could be associated with the incapacity of the predictor variables used here (which represent the most commonly used ones) to capture the species’ response to environmental conditions. On the other hand, the weak relationship between the better performance of GLMs for phytophagous species (in comparison with non–phytophagous species) disappears after removing the effect of N and ROA, revealing that this minor relationship could be a spurious statistical artefact. Further analyses are needed to evaluate whether other species traits not considered in this work are important for the performance of SDMs, beyond the mere limitations of data characteristics such as N or ROA. The limitations of this study, such as data scarcity, low spatial resolution, and lack of reliable absence data and independent validation data sets, are common when working with rare invertebrate species. These constraints, and especially the lack of reliable absence data, are also under the common choice of using background absences, which are randomly selected from the considered extent. The use of background absences generates spatial representations of the distribution of the species that are placed in an unknown situation within the realized–potential gradient described by Jiménez–Valverde et al. (2008), depending on the Relative Occurrence Area (Lobo

Chefaoui et al.

et al., 2010). Thus, the dependency of the accuracy measures on the ROA invalidates any further assessment of the relationships between these accuracy values and the predictor variables, which are also dependent on the ROA. To minimize this drawback, instead of using background absences, here we use pseudoabsences that are a priori located under environmentally unsuitable conditions. By accounting for the limitations of AUC as a measure of model accuracy, our approach identifies some factors that are related to the performance of representing potential distributions. Our results confirm that although some species’ traits may affect SDM performance, prediction accuracy is mostly affected by the characteristics of the data. The separate effects of N and ROA are difficult to determine due to the unavoidable correlation between them (species recorded in more cells have a higher probability of being widely distributed in the region). For this reason, an unknown proportion of the effect on model performance generally attributed to low sample sizes may be due to a less relative occurrence area of presence data in the studied region; i.e., the inability to select reliable absences outside environmental domain used by the species when the number of observations is low (Austin & Meyers, 1996). Given the overall good results obtained by the three methods according to the standard measures of model evaluation, we consider more attention should be given to assessing the quality and/or adequacy of the data rather than selecting a particular SDM technique. Similar results were obtained by Syphard & Franklin (2010), who found that ecological and range characteristics of the species have a greater effect on model performance than the choice of SDM method. In this study, species were modelled more accurately when samples were larger no matter which technique was used. Moreover, ROA had an additive effect to that of sample size, showing that selecting coarse extents of analysis to model the distribution of geographically restricted species may result in trivial models. These models are able to discriminate such restricted distributions within a large geographical context and they therefore yield highly accurate measures of performance, but they are unable to provide reliable descriptions of the environmental response of the species (Lobo, 2008; Jiménez–Valverde et al., 2008; VanDerWal et al., 2009). Our results suggest researchers should avoid any between–species comparison of SDM results while selecting the most adequate technique. We alternatively suggest carrying out species by species SDMs, ensuring that the amount of data available is sufficient and that the geographical focus (i.e., extent) of the analysis is adequate to recover the environmental response of each particular species. In addition, special care should be taken while modelling species inhabiting inconspicuous habitats or strongly affected by interactions occurring at small spatial scales (see Hortal et al., 2010). The problems associated with predicting the distributions of these species should be tackled either by using more precise predictors or by resizing the scale (i.e., grain) of the analyses.


Animal Biodiversity and Conservation 34.2 (2011)

Acknowledgements This research was supported by the Spanish Ministry of Science and Innovation through the project CGL2008–03878. J. H. received funding from a CSIC JAE–Doc grant. References Austin, M. P., 1980. Searching for a model for use in vegetation analysis. Vegetatio, 42: 11–21. Austin, M. P. & Meyers, J. A., 1996. Current approaches to modelling the environmental niche of eucalypts: implications for management of forest biodiversity. Forest Ecology and Management, 85: 95–106. Beaumont, L. J., Gallagher, R. V., Downey, P. O., Thuiller, W., Leishman, M. R. & Hughes, L., 2009. Modelling the impact of Hieracium spp. on protected areas in Australia under future climates. Ecography, 32: 757–764. Brotons, L., Thuiller, W., Araújo, M. B. & Hirzel, A. H., 2004. Presence–absence versus presence–only modelling methods for predicting bird habitat suitability. Ecography, 27: 437–448. Burgman, M. A. & Fox, J. C., 2003. Bias in species range estimates from minimum convex polygons: implications for conservation and options for improved planning. Animal Conservation, 6: 19–28. Busby, J. R., 1991. BIOCLIM–a bioclimate analysis and prediction system. In: Nature Conservation: Cost effective biological surveys and data analysis: 64–68 (C. R. Margules & M. P. Austin, Eds.). CSIRO, Melbourne. Cabeza, M., Arponen, A., Jäättelä, L., Kujala, H., van Teeffelen, A. & Hanski, I., 2010. Conservation planning with insects at three different spatial scales. Ecography, 33: 54–63. Carrascal, L. M., Seoane, J., Palomino, D., Alonso, C. L., Lobo, J. M., 2006. Species–specific features affect the ability of census–derived models to map winter avian distribution. Ecological Research, 21: 681–691. Castillejo, J., 1990. Babosas de la Península Ibérica. I. Los Ariónidos. Catálogo crítico y mapas de distribución (Gastrópoda, Pulmonata, Arionidae). Iberus, 9: 331–345. Chefaoui, R. M. & Lobo, J. M., 2007. Assessing the conservation status of an Iberian moth using pseudo–absences. The Journal of Wildlife Management, 8: 2507–2516. – 2008. Assessing the effects of pseudo–absences on predictive distribution model performance. Ecological Modelling, 210: 478–486. Chefaoui, R. M., Hortal, J. & Lobo, J. M., 2005. Potential distribution modelling, niche characterization and conservation status assessment using GIS tools: a case study of Iberian Copris species. Biological Conservation, 122: 327–338. Clark Labs, 2000. Global Change Data Archive Vol. 3. 1 km Global Elevation Model. Clark University. – 2003. Idrisi Kilimanjaro. Clark Labs, Worcester, MA.

241

Dixon, P. M., Ellison, A. M. & Gotelli, J., 2005. Improving the precision of estimates of the frequency or rare events. Ecology, 86: 1114–1123. Elith, J. & Leathwick, J. R., 2007. Predicting species’ distributions from museum and herbarium records using multiresponse models fitted with multivariate adaptive regression splines. Diversity and Distributions, 13: 165–175. Franklin, J., 2009. Mapping species distributions: Spatial inference and prediction. Cambridge University Press, New York. Galante, E. & Verdú, J. R., 2000. Los Artrópodos de la 'Directiva Hábitat' en España. Ministerio de Medio Ambiente, Madrid. García–Barros, E. & Herranz, J., 2001. Nuevas localidades de Proserpinus proserpina (Pallas, 1772) y Graellsia isabelae (Graells, 1849) del centro peninsular. SHILAP Revista de Lepidopterología, 29: 183–184. Grenyer, R., Orme, C. D. L., Jackson, S. F., Thomas, G. H., Davies, R. G., Davies, T. J., Jones, K. E., Olson, V. A., Ridgely, R. S., Rasmussen, P. C., Ding, T. S., Bennett, P. M., Blackburn, T. M., Gaston, K. J., Gittleman, J. L. & Owens, I. P. F., 2006. Global distribution and conservation of rare and threatened vertebrates. Nature, 444: 93–96. Grosso–Silva, J. M., 1999. Contribuição para o conhecimento dos lucanídeos (Coleoptera, Lucanidae) de Portugal. Boletín de la S.E.A., 25: 11–15. Grupo de Trabajo sobre Lucanidae Ibéricos, 2000. Proyecto Ciervo Volante (PCV). Boletín de la S.E.A., 27: 108–109. Hanspach, J., Kühn, I., Pompe, S. & Klotz, S., 2010. Predictive performance of plant species distribution models depends on species traits. Perspectives in Plant Ecology, Evolution and Systematics, 12: 219–225. Hernández, P. A., Graham, C. H., Master, L. L. & Albert, D. L., 2006. The effect of sample size and species characteristics on performance of different species distribution modeling methods. Ecography, 29: 773–785. Hirzel, A. H., Hausser, J. & Perrin, N., 2007. Biomapper 4.0. Division of Conservation Biology, University of Bern, Laussane. Hortal, J., Jiménez–Valverde, A., Gómez, J. F., Lobo, J. M. & Baselga, A., 2008. Historical bias in biodiversity inventories affects the observed realized niche of the species. Oikos, 117: 847–858. Hortal, J., Lobo, J. M. & Jimenez–Valverde, A., 2007. Limitations of biodiversity databases: case study on seed–plant diversity in Tenerife, Canary Islands. Conservation Biology, 21: 853–863. Hortal, J., Roura–Pascual, N., Sanders, N. J. & Rahbek, C., 2010. Understanding (insect) species distributions across spatial scales. Ecography, 33: 51–53. Instituto Geográfico Nacional, 1995. Mapa de suelos del Atlas Nacional de España (Edafología). CSIC/ IRNAS, Sevilla. Jiménez–Valverde, A. & Lobo, J. M., 2006. The ghost of unbalanced species distribution data in geographical model predictions. Diversity and Distributions, 12: 521–524.


242

– 2007. Threshold criteria for conversion of probability of species presence to either–or presence– absence. Acta Oecologica, 31: 361–369. Jiménez–Valverde, A., Lobo, J. M. & Hortal, J., 2008. Not as good as they seem: the importance of concepts in species distribution modelling. Diversity and Distributions, 14: 885–890. – 2009. The effect of prevalence and its interaction with sample size on the reliability of species distribution models. Community Ecology, 10: 196–205. King, G. & Zeng, L., 2000. Logistic regression in rare events data. The Global Burden of Disease 2000 in Aging Populations, Research Paper No. 2. Havard Burden of Disease Unit, Cambridge. <http://www.hsph.harvard.edu/burdenofdisease/ publications /papers/Logistic%20Regression.pdf> 6th October 2010. Legendre, P. & Legendre, L., 1998. Numerical Ecology. Elsevier, Amsterdam. Lobo, J. M., 2008. More complex distribution models or more representative data? Biodiversity Informatics, 5: 14–19. Lobo, J. M., Baselga, A., Hortal, J., Jiménez–Valverde, A. & Gómez, J. F., 2007. How does the knowledge about the spatial distribution of Iberian dung beetle species accumulate over time? Diversity and Distributions, 13: 772–780. Lobo, J. M., Jiménez–Valverde, A. & Hortal, J., 2010. The uncertain nature of absences and their importance in species distribution modelling. Ecography, 33: 103–114. Lobo, J. M., Jiménez–Valverde, A. & Real, R., 2008. AUC: a misleading measure of the performance of predictive distribution models. Global Ecology and Biogeography, 17: 145–151. Lobo, J. M., Verdú, J. R. & Numa, C., 2006. Environmental and geographical factors affecting the Iberian distribution of flightless Jekelius species Coleoptera: (Geotrupidae). Diversity and Distributions, 12:179–188. López–Sebastián, E., López, J. C., Juan, M. J. & Selfa, J., 2002. Primeras citas de mariposa isabelina en la Comunidad Valenciana. Quercus, 193: 10–13. Luoto, M., Pöyry, J., Heikkinen, R. K. & Saarinen, K., 2005. Uncertainty of bioclimate envelope models based on the geographical distribution of species. Global Ecology and Biogeography, 14: 575–584. Marmion, M., Luoto, M., Heikkinen, R. K. & Thuiller, W., 2008. The performance of state–of–the–art modelling techniques depends on geographical distribution of species. Ecological Modelling, 220: 3512–3520. Martínez–Orti, A., 2004. Descripción de los moluscos terrestres del Valle del Najerilla. Noticiario de la Sociedad Española de Malacología, 41: 30–32. Mateo, R. G., Felicísimo, Á. M. & Muñoz, J., 2010. Effects of the number of presences on reliability and stability of MARS species distribution models: the importance of regional niche variation and ecological heterogeneity. Journal of Vegetation Science, 21: 908–922. McCullagh, P. & Nelder, J. A., 1989. Generalized

Chefaoui et al.

Linear Models. Chapman & Hall, London. McPherson, J. M. & Jetz, W., 2007. Effects of species’ ecology on the accuracy of distribution models. Ecography, 30: 135–151. McPherson, J. M., Jetz, W. & Rogers, D. J., 2004. The effects of species’ range sizes on the accuracy of distribution models: ecological phenomenon or statistical artefact? Journal of Applied Ecology, 41: 811–823. Newbold, T., Reader, T., Zalat, S., El–Gabbas, A. & Gilbert, F., 2009. Effect of characteristics of butterfly species on the accuracy of distribution models in an arid environment. Biodiversity and Conservation, 18: 3629–3641. Olden, J. D., Jackson, D. A. & Peres–Neto, P., 2002. Predictive models of fish species distributions: a note on proper validation and chance predictions. Transactions of the American Fisheries Society, 131: 329–336. Pérez–Bote, J. L., García, J. M., Ferri, F. & Moreno, J. A., 2001. Nueva cita de Lucanus cervus (Linnaeus, 1758) en Extremadura (Coleoptera: Lucanidae). Boletín de la S.E.A., 28: 130. Pöyry, J., Luoto, M., Heikkinen, R. K. & Saarinen, K., 2008. Species traits are associated with the quality of bioclimatic models. Global Ecology and Biogeography, 17: 403–414. R Development Core Team, 2008. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. <http:// www.R–project.org> 6th October 2010. Raimundo, R., Algarvio, R., Casas Novas, P. & Figuereiredo, D., 2001. Cartography of some species of xilophagous and xilomicetophagous insects in Quercus suber and Quercus rotundifolia in Alentejo using GIS. Suplemento ao Boletim da Sociedade Portuguesa de Entomología, 6: 459–468. Reutter, B., Helfer, V., Hirzel, A. H. & Vogel, P., 2003. Modelling habitat–suitability using museum collections: an example with three sympatric Apodemus species from the Alps. Journal of Biogeography, 30: 581–590. Ripley, B. D., 1996. Pattern Recognition and Neural Networks. Cambridge University Press. Rocchini, D., Hortal, J., Lengyel, S., Lobo, J. M., Jiménez–Valverde, A., Ricotta, C., Bacaro, G. & Chiarucci, A., 2011. Uncertainty in species distribution mapping and the need for maps of ignorance. Progress in Physical Geography, 35: 211–226. Rosas, G., Ramos, M. A. & García Valdecasas, A., 1992. Invertebrados españoles protegidos por convenios internacionales. ICONA, Madrid. Santika, T., 2010. Assessing the effect of prevalence on the predictive performance of species distribution models using simulated data. Global Ecology and Biogeography, 20: 181–192. Schröder, B., 2004. ROC Plotting and AUC Calculation Transferability Test. Potsdam University, Potsdam. <http://brandenburg.geoecology.uni– potsdam.de/users/schroeder/download.html> 6th October 2010. Segurado, P. & Araújo, M. B., 2004. An evaluation of methods for modelling species distributions.


Animal Biodiversity and Conservation 34.2 (2011)

Journal of Biogeography, 31: 1555–1568. Seoane, J., Carrascal, L. M., Alonso, C. L. & Palomino, D., 2005. Species–specific traits associated to prediction errors in bird habitat suitability modelling. Ecological Modelling, 185: 299–308. Soria, S., Abos, F. & Martín, E., 1986. Influencia de los tratamientos con diflubenzurón ODC 45% sobre pinares en las poblaciones de Graellsia isabelae (Graells) (Lep. Syssphingidae) y reseña de su biología. Boletín de sanidad vegetal, 12: 29–50. StatSoft, I., 2001. Statistica. Data analysis software system, Tulsa. Stockwell, D. R. B. & Peterson, A. T., 2002. Effects of sample size on accuracy of species distribution models. Ecological Modelling, 148: 1–13. Syphard, A. D. & Franklin, J., 2010. Species traits affect the performance of species distribution models for plants in southern California. Journal of Vegetation Science, 21: 177–189. VanDerWal, J., Shoo, L. P., Graham, C. & Williams, S. E., 2009. Selecting pseudo–absence data for presence–only distribution modeling: How far should you stray from what you know? Ecological Modelling, 220: 589–594. Veloz, S. D., 2009. Spatially autocorrelated sampling falsely inflates measures of accuracy for presence–only niche models. Journal of Biogeography, 36: 2290–2299.

243

Verdú, J. R. & Galante, E., 2002. Climatic stress, food availability and human activity as determinants of endemism patterns in the Mediterranean region: the case of dung beetles Coleoptera, Scarabeoidea) in the Iberian Peninsula. Diversity and Distributions, 8: 259–274. – 2006. Libro Rojo de los Invertebrados de España. Ministerio de Medio Ambiente, Madrid. Viejo Montesinos, J. L., 1992. Biografía de un naturalista y biología del lepidóptero por él descrito. Graells y la Graellsia. Quercus, 74: 22–30. Wisz, M. S., Hijmans, R. J., Li J., Peterson, A. T., Graham, C. H., Guisan, A. & N. P. S. D. W. Group, 2008. Effects of sample size on the performance of species distribution models. Diversity and Distributions, 14: 763–773. Wood, S. N. & Augustin, N. H., 2002. GAMs with integrated model selection using penalized regression splines and applications to environmental modelling. Ecological Modelling, 157: 157–177. Zaniewski, A. E., Lehmann, A. & Overton, J. M., 2002. Predicting species spatial distributions using presence–only data: a case study of native New Zealand ferns. Ecological Modelling, 157: 261–280. Zweig, M. H. & Campbell, G., 1993. Receiver– operating characteristic (ROC) plots: a fundamental evaluation tool in clinical medicine. Clinical Chemistry, 39: 561–577.


Chefaoui et al.

244

Appendix 1. Scatterplots of significant correlation analyses between accuracy measures (AUC, sensitivity and specificity) and data size (N). 1.00 0.9700 0.9524 0.9346

0.98

GAM specificity

AUC GAM

0.96 0.94 0.92 0.90 0.88

0.8904 0.8620 0.8275

0.7600

0.86 0

200 400

600 800 1,000 1,200 N

0

200 400

600

N

800 1,000 1,200

0.9976

0.9869 0.9709 0.9536

0.9547 0.9337 NNET specificity

0.9385 AUC NNET

0.9083

0.8920 0.8570 0.8353

0.9130 0.8918

0.7900 0.7565

0.8020

0.7298 0

200 400

600

N

800 1,000 1,200

0

200 400

600

N

800 1,000 1,200


Animal Biodiversity and Conservation 34.2 (2011)

245

Apéndice 1. Gráfico de dispersión de las corelaciones significativas entre las medidas de precisión (AUC, sensitividad y especificidad) y el tamaño de muestra (N). 1.00

0.9745 0.9492

0.98 0.96 GLM sensitivity

AUC GLM

0.94 0.92 0.90 0.88 0.86 0.84 0.82

0 200 400 600 800 1,000 1,200 N

1.0000

NNET sensitivity

0.8333 0.8045 0.7619 0.7000

0.80 0.78

0.9047 0.8735

0.9539 0.9337 0.9130 0.8918

0.8000 0.7565 0.7356 0 200 400 600 800 1,000 1,200 N

0 200 400 600 800 1,000 1,200 N


Chefaoui et al.

246

Appendix 2. Accuracy measures results by habitat: C–I. Woods and Mountainous habitats; C–II. Grasslands and varied habitats; C–III. Riparian and humid habitats. Less accurate models are obtained for species associated to riparian and humid habitats. The middle point shows the median response, the bottom and top of the box show the 25 and 75 percentiles respectively. The whiskers show minimum and maximum vaues). 1.00 0.98 0.96 GAM spccificity

GAM sensitivity

0.94 0.92 0.90 0.88 0.86 0.84 0.82 0.80 0.78

C–I C–III C–II Habitat type

1.00

1.00

0.98

0.95 NNET spccificity

0.96 GAM AUC

1.00 0.98 0.96 0.94 0.92 0.90 0.88 0.86 0.84 0.82 0.80 0.78 0.76 0.74

0.94 0.92 0.90

0.90 0.85 0.80 0.75

0.88 0.86

C–I C–III C–II Habitat type

C–I C–III C–II Habitat type

0.70

C–I C–III C–II Habitat type


Animal Biodiversity and Conservation 34.2 (2011)

247

Apéndice 2. Precisión obtenida por los modelos en función del tipo de hábitat: C–I. Bosque y hábitats montañosos; C–II. Praderas y hábitats mixtos; C–III. Hábitats húmedos y riparios. Las especies asociadas a hábitats húmedos y riparios obtienen modelos menos precisos. El punto central representa la mediana, los límites inferiores y superiores de la caja muestran los percentiles 25 y 75 respectivamente. Los bigotes señalan el valor máximo y mínimo. 1.00

1.00 0.98

0.95

0.96 NNET sensitivity

NNET AUC

0.94 0.92 0.90 0.88 0.86 0.84 0.82

0.90 0.85 0.80 0.75

0.80 0.78

C–I C–III C–II Habitat type

1.00

GLM spccificity

0.95 0.90 0.85 0.80 0.75 0.70 0.65

C–I C–III C–II Habitat type

0.70

C–I C–III C–II Habitat type


366


Animal Biodiversity and Conservation 34.2 (2011)

249

Modelling the feeding behavior of Grey Heron (Ardea cinerea) in a coastal wetland of NW Iberian peninsula during the wintering season A. Regos

Regos, A., 2011. Modelling the feeding behavior of Grey Heron (Ardea cinerea) in a coastal wetland of NW Iberian peninsula during the wintering season. Animal Biodiversity and Conservation, 34.2: 249–256. Abstract Modelling the Feeding Behavior of Grey Heron (Ardea cinerea) in a Coastal Wetland of NW Iberian Peninsula during the Wintering Season.— For a better understanding of the foraging behavior of Grey Heron in an intertidal area we developed predictive models of number of attempts/10' using Poisson regression. The models were obtained considering the following four variables: age of bird, tidal hours, bi–monthly period, and substrate type, obtaining a total of 15 models. The most parsimonious model obtained using the Akaike Information Criteria included tidal hour, age of bird and substrate type as predictive variables. The mean number of attempts/10' was highest in the four hours around low tide and in water and muddy substrates, while foraging activity was scarcely recorded in sandy substrates. No differences of effectiveness were found between adult and juvenile birds. Grey Heron showed preference for very small and small prey, increasing handling time with prey length. Key words: Ardea cinerea, Feeding behavior, GLMM, Tidal cycle, Substrate type, Age of birds. Resumen Modelado del comportamiento trófico de la garza real (Ardea cinerea) en un humedal costero del NO ibérico durante el periodo invernal.— Para mejorar el conocimiento de la estrategia de forrajeo de la garza real en una zona intermareal se obtuvieron modelos predictivos del número de intentos/10' empleando la regresión de Poisson. Los modelos candidatos resultaron de la consideración de cuatro variables: edad, hora mareal, periodo bimensual y tipo de sustrato. El modelo más parsimonioso obtenido de acuerdo con el Criterio de Información de Akaike incluyó la edad, la hora mareal y el tipo de sustrato como variables predictivas. El promedio de intentos/10' fue mayor en las cuatro horas más próximas a la bajamar, y en sustratos con agua o fango que en medios arenosos. No se encontraron diferencias significativas en la efectividad entre adultos y juveniles. La garza real mostró preferencia por presas pequeñas o muy pequeñas, aumentando el tiempo de manejo con el tamaño de la presa. Palabras clave: Ardea cinerea, Comportamiento trófico, GLMM, Ciclo mareal, Tipo de sustrato, Edad. (Received: 14 VI 10; Conditional acceptance: 3 X 10; Final acceptance: 14 IV 11) Adrián Regos, Dept. of Zoology and Physical Anthropology, Univ. of Santiago de Compostela, Campus Sur s/n., 15782 Galicia, España (Spain). Corresponding author: adrianregos@hotmail.com

ISSN: 1578–665X

© 2011 Museu de Ciències Naturals de Barcelona


250

Introduction In recent years, the population of Grey Heron (Ardea cinerea L., 1758) in Spain has increased considerably both in number and in geographic range and it can now be found in several areas where it was previously scarce or absent (Prieta & Campos, 2003). This population growth could be due to factors such as the protection of wetlands, the creation of new artificial wetlands (ponds and dam reservoirs) and fish farms, the introduction of exotic fishes or eutrophication of water bodies (Prieta & Campos, 2003). Grey Heron feeding behavior has been studied in several habitats, mainly during the breeding season (Owen, 1995; Campos & Lekuona, 1997; Lekuona, 1999). According to Cramps & Simons (1977) Grey Heron feed by day in some places, especially in the morning and evening, while in other areas they feed mainly around dusk and at night. Besides differences concerning feeding hours, the foraging behavior of herons is also affected by factors such as: tidal cycles (Dimalexis & Pyrovetsi, 1997; Lekuona, 1999; Matsunaga, 2000), age of birds (Lekuona, 2002a), substrate type and habitat characteristics (Hampl et al., 2005; Papakostas et al., 2005; Gwiazda & Amirowicz, 2006). In a tidal flat the availability of foraging sites varies according to tidal cycles. Previous studies have demonstrated that the daily cycle of high and low tides affected the foraging strategies of herons (Lekuona, 1999; Matsunaga, 2000). Few studies, however, have examined the effect of tidal cycles on feeding behavior in relation to substrate type and age of birds, especially in the wintering season. Regarding the age of birds, various authors have suggested that juveniles are less successful at feeding and spend longer feeding than adults (Carss, 1993; Lekuona, 2002a, 2002b; Papakostas et al., 2005). However, it is not clear whether there are differences between juvenile and adult birds regarding the effectiveness of foraging. Moreover, feeding behavior in herons changes during the breeding season because of the energy requirements of chicks during this period (Campos & Lekuona, 1997; Lekuona, 1999). Studies on the feeding ecology during the winter period, however, are scant. In this study we examined the effect of tidal cycle, age of birds and substrate type on foraging activity. We compared the effectiveness of adult and juvenile birds and analyzed this as another possible factor affecting foraging activity. We also investigated associations between foraging activity rates, prey size and success so as to improve knowledge of Grey Heron behavior in an intertidal area in the wintering period. Material and methods The study was carried out in O Bao inlet, situated in the Umia–O Grove intertidal complex of 2.561 hectares of extension, located in the northwestern Iberian peninsula (42º 28' N, 8º 51' W) (fig. 1). This wetland is characterized by its great ornithological relevance, being one of the most important wintering and mi-

Regos

grating sites of water birds in northwestern Spain. It has been declared a Site of Community Importance (SCI), Special Protection Area (SPA) and Wetland of International Importance (RAMSAR site). The area consists of a bay that is separated from the open sea by a large sandy beach–dune system. It has extensive muddy intertidal flats covered by Zostera sp., with small areas of bulrushes and sandbanks, influenced by tidal dynamics. During our study, 193 Grey Herons wintered in the area, in a mixed–species colony, together with Little Egret (Egretta garzeta), Spoonbill (Platalea leucorodia) and Great Cormorant (Phalacrocorax carbo) (Xunta de Galicia, unpublished data). Data were collected from six observation sites located at strategic locations in the inner part of the bay (fig. 1). Surveys were carried out from November 2005 to February 2006, coinciding with the wintering period of Grey Heron. Two telescopes were used (20–60x and 20–70x) for observations and 13 field visits were carried out: seven in November–December and six in January–February. The surveys were planned according to tidal cycles, so that only days allowing surveys from three hours before to three hours after diurnal low tide were selected. To avoid pseudoreplication of data, simultaneous surveys were carried out in two other sites of the wetland, during a period of observation not exceeding two hours. Birds were chosen arbitrarily and followed for 10' according to the methodology proposed by Altmann (1974) and Martin & Bateson (1986). During the monitoring we recorded the following variables: (I) age of birds, considering two categories: adults and juveniles (birds were categorized on their plumage characteristics according to Cramp & Simmons, 1997); (II) tidal hour; (III) substrate type: sand, mud, water to tibia–tarsus and water to tarsus–metatarsus; (IV) bi–monthly period, considering two periods: November–December and January–February; and (V) day: survey date. To measure the foraging effort of each bird we recorded feeding attempts and their results (successful and unsuccessful). Foraging times were recorded for each prey, using a chronometer. Prey size was calculated in relation to the size of Grey Heron´s beak: 12 cm (Cramp & Simmons, 1997). In this way, four classes of prey size were established: very small (< 6 cm), small (6–12 cm), medium (12–18 cm) and large (> 18 cm). During the survey four further behavioral variables were recorded: interspecific and intraspecific presence, number of kleptoparasitic attempts made by Grey Heron and number of attempts made by other species. In addition, four meteorological variables were measured: isolation, rainfall, temperature and wind. The number of kleptoparasitic attempts recorded was very low and was therefore not considered in the final model. Meteorological variables were also categorized. Most of the values recorded were concentrated in one or two categories of the initially defined list so none of these variables were finally taken into account in the analysis.


Animal Biodiversity and Conservation 34.2 (2011)

251

Iberian peninsula

O Grove peninsula

O Bao Inlet

A Lanzada beach

RAMSAR

N

SCI 0 750 1,500

3,000

SPA

4,500 m

Observation sites

Fig. 1. Study area, located in the north–west Iberian peninsula. Fig. 1. Área de estudio, localizada en el noroeste de la península ibérica.

Based on field data, a new variable was calculated: Effectiveness (Ef). This variable was calculated from the ratio between number of successful attempts (Pr) and total attempts (successful and unsuccessful) (At). Calculated effectiveness values ranged between 0 (minimum effectiveness) and 1 (maximum effectiveness).

The Akaike weight for each model was calculated as Wi:

Ef = Pr / At

The sum of all weights equals the unit, and the value of each Wi indicates that model i is the best overall model (Anderson et al., 2000). This model is chosen from 15 well–defined candidate models using Akaike Information Criteria (Anderson et al., 2000; Seoane & Bustamante, 2001). The importance of each variable was obtained by adding the Akaike weights to the model in which that variable was present (Burnham & Anderson, 1998). The addition of the weights of each variable was considered consequential when 3Wi > 0.5 (Taylor & Knight, 2003). To compare the foraging effectiveness of adults and juveniles, and to test the differences in foraging efforts in four types of substrate, we used non–parametric statistics (Mann–Whitney test & Kruskal–Wallis test respectively) after verifying the assumptions of normality and homogeneity of variance. We also analysed the foraging effort during 6 tidal hours (Kruskal–Wallis test). One way ANOVA was used to test differences in the handling time of four classes of prey size since data were adjusted to a normal distribution. The post hoc comparison was done using a Student–Newman– Keuls test (Quinn & Keough, 2002).

This concept explains the capacity of each bird to capture prey, whereas the term 'trophic efficiency' is used on continuation to refer to the capacity of each bird to obtain a biomass intake. Generalized Linear Mixed Models (GLMMs) were derived to model the foraging effort of birds, estimated by number of attempts/10'. As the response variable (feeding attempts/10') is count data, we used a Poisson model with a log link. The mean number of attempt/10' within different groups of variables was less than 5, so Laplace approximation was used. This analysis was conducted using a PROC GLIMMIX in SAS (Bolker et al., 2009; Dean & Nielsen, 2007). The candidate models were obtained by using four variables (age of birds, tidal hour, bi–monthly period and substrate type) as fixed effects and the variable 'day' as random effect, obtaining a total of 15 candidate models. For each model i, the values of AIC (Akaike information criteria) and ∆i were obtained, where ∆i = AICi – AICminimum model

Wi =

exp (–1/2 ∆i) R

3 exp (–1/2 ∆r)

(Anderson et al., 2000)

r=1


252

Regos

All data were analyzed using SAS System v.9.1.3. Results A total of 319 birds were studied, 124 juveniles and 195 adults. No statistically significant differences were found between juveniles and adult regarding foraging effectiveness (Z = –1.642; N = 132; P > 0.05). Foraging effort was highest in the four hours around low tide for juveniles (x2 = 43.219; d.f = 5; P < 0.05) and in the 2 hours around low tide for adults (x2 = 36.074; d.f = 5; P < 0.05) (fig. 2). The highest foraging effort was recorded in water with birds wading up to the level of the tarsus–metatarsus or tibia–tarsus (x2 = 69.091; d.f. = 3; P < 0.05). This was followed by foraging in muddy substrate, whereas foraging activity was scarcely recorded in sandy substrate (fig. 3). Of the total consumed prey, 69.2% were smaller than 6 cm; of these, 56.6% were consumed by younger birds and 43.3% by adults. Prey larger than 18 cm accounted for only 6.8% of the total, all of which were exclusively consumed by adults (fig. 4). All prey larger than 18 cm were eels (Anguilla anguilla). Handling time increased significantly with prey length (F3,132 = 89.49; P < 0.001) (fig. 5). A post hoc comparison (Student–Newman–Keuls) showed

that the average handling time differed significantly between the four classes of prey size, except for between small and medium prey. Regarding the effects of the fixed factors (age of birds, tidal hour, bi–monthly period and substrate type) on the foraging effort, once all these factors were included in the model, three of them were found to have a significant effect: age of birds (3Wage  = 0.97), tidal hour (3Wtidal hour = 1) and substrate type (3Wsubstrate type = 1). However, there was no evidence of any effect of bi–monthly period (3Wbi–monthly period =  0.26) because this variable could not be considered consequential, since its 3Wi was below 0.5. The final model chosen, that which included age of birds, tidal hour and substrate type as predictive variables (table 1), was the only one of the 15 models initially analyzed that was considered competent (∆i < 2). Discussion Previous studies have emphasized the role of tidal cycle, age of birds and substrate type as decisive factors in feeding behavior in several heron species (Matsunaga, 2000; Lekuona, 2002a; Hampl et al., 2005; Papakostas et al., 2005; Gwiazda & Amirowicz, 2006). In the present study, these factors were all

Average number of attempts/10 min

3

Juveniles Adults

n = 34

2.5

n = 15

2

n = 23 n = 25 n = 25 n = 15

1.5

1 n = 20 n = 40

0.5

n = 30 n = 36

0

n = 15

–3

n = 41

–2

–1 1 Tidal hour

2

3

Fig. 2. Average number of attempts/10 min of adult and juvenile birds for each tidal hour during the wintering season 2005–2006. Fig. 2. Promedio de intentos/10 min de adultos y juveniles por hora mareal durante el periodo invernal 2005–2006.


Animal Biodiversity and Conservation 34.2 (2011)

253

Average of attempts/10min

2.5 2

n = 23

n = 94

1.5 1

n = 125

0.5 n = 77

0

Muddy

Sandy

Water to Water to metatarsus–tarsus tibia–tarsus Substrate type

Fig. 3. Average number of attempts/10' of Grey Heron for each substrate type. Fig. 3. Promedio de intentos/10' de la garza real para cada tipo de sustrato.

found to have a significant effect on the foraging effort of Grey Heron. Our results confirm that these variables are critical factors and should be taken into account for a better understanding of the feeding behavior of Grey Heron in an intertidal area. More specifically, they help us to understand which environmental factors most directly affect their feeding strategy and how herons

respond to these factors in terms of trophic effort. To analyze the effect of these factors in the study area, we discuss each one separately. Several authors have studied the effect of water level and tidal cycle on the foraging activity of wading birds (Voslamber, 1996; Dimalexis & Pyrovetsi, 1997; Matsunaga, 2000). Lekuona (1999) demonstrated

60

Total prey

50

Juveniles Adults

40 30 20 10 0

Very small

Small

Medium Prey size

Large

Fig. 4. Total prey captured by adult and juvenile birds for each prey size in O Bao inlet during the wintering season 2005–2006. Fig. 4. Presas totales capturadas por adultos y juveniles de garza real según el tamaño de cada presa durante el periodo invernal 2005–2006 en la ensenada de O Bao.


254

Regos

Handling time (s)

300 250 200 150 100 50 0

< 6 cm 6–12 cm 12–18 cm > 18 cm Class size

Fig. 5. Average (± SE) of Grey Heron handling time in seconds (s) for each prey size. Fig. 5. Valor medio (± EE) del tiempo de manejo en segundos (s) de las presas para cada clase de tamaño.

that most waders feed in the two hours before and the two hours after low tide, and our results support this finding. Regarding the age of birds, various authors have suggested that juveniles are less successful at feeding and spend more time feeding than adults (Carss, 1993; Lekuona, 2002a, 2002b; Papakostas et al., 2005). Our findings concerning more time spent feeding coincide with these authors. We found that foraging effort was higher in the four hours around low tide for juveniles and in the two hours around low tide for adults, indicating that juveniles spend more time in foraging areas than adults. Moreover, the age of birds was a consequential variable in the model obtained to explain the foraging effort. However, and contrary to the expected results, we did not find any differences between juvenile and adult birds regarding the effectiveness of foraging. The differences in trophic effort between adults and juveniles was not due to greater effectiveness in the adults but to the fact that the adults had the ability to capture more energetically profitable prey, making them more efficient but not more effective. Herons may adopt different tactics and may achieve

variable foraging efficiencies in response to particular habitat conditions and prey characteristics (Dimalexis et al., 1997; Wong et al., 2000; Gwiazda & Amirowicz, 2006). Carss & Elston (2003) found that Grey Herons show preference for particular substrate types, and may even select areas dominated by specific algae species. Our results support this finding that substrate type is a decisive factor in the foraging strategy. In our study, the foraging effort was highest in water substrate. It was lower in muddy substrate and scarce in sandy substrate. Grey Heron usually capture fishes of 10–25 cm long, although fish up to 40 cm (Del Hoyo et al., 1992) and eels up to 60 cm may be also taken (Owen, 1995; Cramp & Simmons, 1977). According to other published works, size can play an important role in the choice of prey and herons generally show preference for larger prey (Britton & Moser, 1982; Feunteun & Marion, 1994). Gwiazda & Amirowicz (2006) concluded that larger prey were more profitable for Grey Heron to forage. However, our data show that small (6–12 cm) and very small (< 6 cm) prey were the preferred sizes selected by Grey Heron in the studied wetland. This could be because this prey size was the most abundant or the most profitable in the study area in terms of biomass per time unit, an aspect previously suggested in other studies (Lekuona, 1999; Campos & Lekuona, 2000). Campos & Lekuona (2000) suggested two possible hypotheses to explain why adults captured larger prey during the breeding season if smaller prey were really more profitable: I) larger prey would provide the energy need to compensate for the energetic cost of frequent trips to and from the breeding colonies and the feeding areas. This hypothesis can be ruled out in our case because the data were recorded in the non–breeding season; II) when small prey are scarce the herons may be forced to capture larger prey for their own food. However, small prey were the most abundant in our study area. Taking this into account we suggest that: I) older birds have greater trophic efficiency –but not effectiveness– due to their greater ability to capture large and slippery prey, like eels. This could explain why all large prey were only consumed by adults; II) adults spend less time than juvenile birds do in foraging areas because the largest prey provide the quantity of biomass they require. Handling time of prey increased with prey length. Comparing values obtained for Purple Heron in the Ebro River valley (Campos & Lekuona, 2000) and our data, we can conclude that, for similar prey size,

Table 1. Models obtained for the foraging effort showing Akaike information criteria values. Tabla 1. Modelos obtenidos para el esfuerzo de forrajeo mostrando los valores del criterio de información de Akaike. Model Age + tidal hour + substrate type

AIC

∆i

Wi

762.51

0

0.721


Animal Biodiversity and Conservation 34.2 (2011)

Grey Heron spend more handling time than Purple Heron. This can be related to the type of available prey or size differences between these two species of herons. In both species, results show that the trophic effort per time unit depends directly on the prey type and size, determining a specific feeding strategy. The feeding strategy for large and more difficult to manage prey is more time–consuming and occurs in the two hours around low tide. In contrast, for smaller prey, which require less handling time, activity continues for a further two hours. Previous studies have also shown that feeding behavior in herons changes during the breeding season (Campos & Lekuona, 1997; Lekuona, 1999). However, in the present study, the bimonthly period did not affect the trophic activity of the birds. This allowed us to conclude that their foraging effort does not vary during the wintering season and remains constant throughout this period. Acknowledgements I wish to thank Cristina González Álvarez, Ibama Pineda Josseulin and Virginia Mena Ramos for their collaboration in the field work, and especially Dr. Jesús Dominguez Conde (Department of Zoology and Physical Anthropology, USC) and Dr. Raquel Diez Arenas for their advice throughout the study. I also thank Esther Miguéns Alonso, Douglas Naismith and Giuseppe di Pinto for linguistic revision and the contribution of Oliver Valero for his advice with the statistical methods (www.uab.es/s–estadistica). The comments and constructive remarks of the journal editor and two anonymous referees are also gratefully acknowledged. References Altmann J., 1974. Observational Study of Behavior: sampling methods. Behavior, 49: 227–267. Anderson, D. R., Burnham, K. P. & Thompson, W. L., 2000. Null hypothesis testing: problems, prevalence, and an alternative. J. Wildlife Manage, 64: 912–923. Bolker, B. M., Brooks, M. E., Clark, C. J., Geange, S. W., Poulsen, J. R., Stevens M. H. H. & White, J. S., 2009. Generalized linear mixed models: a practical guide for ecology and evolution. Trends in Ecology and Evolution, 24(3): 127–135. Burnham, K. P. & Anderson, D. R., 1998. Model selection and inference: a practical information–theoretic approach. Springer–Verlag, New York. Britton, R. H. & Moser, M. E., 1982. Size specific predation by Herons and its effect on the sex ratio of natural populations of the mosquito fish Gambusia affinis Baird and Girard. Oecologia, 53: 146–151. Campos, F. & Lekuona, J. M., 1997. Temporal variations in the feeding habits of the Purple Heron Ardea purpurea during the breeding season. Ibis, 139(3): 447–451. – 2000. Fish profitability for breeding purple herons.

255

Ardeola, 47(1): 105–107. Carss, D. N., 1993. Grey heron, Ardea cinerea L., predation at cage fish farms in Argyll, western Scotland. Aquaculture and Fisheries Management, 24(1): 29–45. Carss, D. N. & Elston, D. A., 2003. Patterns of association between algae, fishes and grey herons Ardea cinerea in the rocky littoral zone of a Scottish sea loch. Estuarine Coastal and Shelf Science, 58(2): 265–277. Cramps, S. & Simmons, K. E. L., 1977. The bird of Western Palearctic, Vol. 1. Oxford University Press, Oxford. Dean, C. B. & Nielsen, J. D., 2007. Generalized linear mixed models: a review and some extensions. Lifetime Data Anal, 13: 497–512. Del Hoyo, J., Elliott, A. & Sargata, J., 1992. Handbook of de Birds of the World, Vol. 1. Lynx Edicions, Barcelona. Dimalexis, A. & Pyrovetsi, M., 1997. Effect of water level fluctuations on wading bird foraging habitat use at an irrigation reservoir, Lake Kerkini, Greece. Colonial Waterbirds, 20(2): 244–252. Dimalexis, A., Pyrovetsi, M. & Sgardelis, S., 1997. Foraging ecology of the grey heron (Ardea cinerea), great egret (Ardea alba) and little egret (Egretta garzetta) in response to habitat, at 2 Greek wetlands. Colonial Waterbirds, 20(2): 261–272. Feunteun, E. & Marion, L., 1994. Assessment of Grey Heron predation of fish communities: the case of the largest European colony. Hydrobiologia, 279/280(0): 327–344. Gwiazda, R. & Amirowicz, A., 2006. Selective foraging of Grey Heron (Ardea cinerea) in relation to density and composition of the littoral fish community in a submontane dam reservoir. Waterbirds, 29(2): 226–232. Hampl, R., Bures, S., Balaz, P., Bobek, M. & Pojer, F., 2005. Food provisioning and nestling diet of the black stork in the Czech Republic. Waterbirds, 28(1): 35–40. Lekuona, J. M., 1999. Food and foraging activity of grey herons, Ardea cinerea, in a coastal area during the breeding season. Folia Zoologica, 48(2): 123–130. – 2002a. Food intake, feeding behaviour and stock losses of cormorants, Phalacrocorax carbo, and grey herons, Ardea cinerea, at a fish farm in Arcachon Bay (Southwest France) during breeding and non– breeding season. Folia Zoologica, 51(1): 23–34. – 2002b. Kleptoparasitism in wintering grey heron Ardea cinerea. Folia Zoologica, 51(3): 215–220. Martin, P. & Bateson, P., 1986. Measuring Behaviour: an introductory guide. Cambridge, Cambridge University Press. Matsunaga, K., 2000. Effects of tidal cycle on the feeding activity and behavior of Grey Herons in a tidal flat in Notsuke Bay, northern Japan. Waterbirds, 23(2): 226–235. Owen, D. F., 1995. The food of the Heron (Ardea cinerea) in the breeding season. Ibis, 97: 276–295. Papakostas, G., Kazantzidis, S., Goutner, V. & Charalambidou, I., 2005. Factors affecting the foraging


256

behavior of the Squacco Heron. Waterbirds, 28(1): 28–34. Prieta, J. & Campos, F. 2003. Garza real, Ardea cinerea. In: Atlas de las Aves Reproductoras de España: 116–117 (R. Martí & J. C. del Moral, Eds.). Dirección General de Conservación de la Naturaleza–Sociedad Española de Ornitología, Madrid. Quinn, G. P. & Keough, M. J., 2002. Experimental design and data analysis for biologist. Cambridge University Press, Cambridge Seoane, J. & Bustamante, J., 2001. Modelos predic-

Regos

tivos de la distribución de especies: una revisión de sus limitaciones. Ecología, 15: 9–21. Taylor, A. R. & Knight, R. L., 2003. Wildlife responses to recreation and associated visitor receptions. Ecol. Appl., 13(4): 951–963. Voslamber, B., 1996. Effects of eater level management and grazing on fish eating birds. Levende Natuur, 97(1): 4–10. Wong, L. C., Corlett, R. T., Young, L. & Lee, J. S.Y., 2000. Comparative feeding ecology of Little Egrets on intertidal mudflats in Hong Kong, South China. Waterbirds, 23(2): 214–225.


Animal Biodiversity and Conservation 34.2 (2011)

257

Estructura del órgano estridulador y análisis de la emisión acústica de Agapanthia dahli (Richter, 1821) (Coleoptera, Cerambycidae, Lamiinae) J. M. Hernández

Hernández, J. M., 2011. Estructura del órgano estridulador y análisis de la emisión acústica de Agapanthia dahli (Richter, 1821) (Coleoptera, Cerambycidae, Lamiinae). Animal Biodiversity and Conservation, 34.2: 257–264. Abstract Structure of the stridulatory organ and analysis of the acoustic signal in Agapanthia dahli (Richter, 1821) (Coleoptera, Cerambycidae).— Agapanthia dahli (Richter, 1821) have a stridulatory organ consisting of a grooved plate or pars strindens at the dorsal side of mesonotum and a scraper or plectrum in the internal posterior margin of the pronotum. Sound is produced when the insect moves the pronotum, sliding the plectrum against the pars stridens. The structure of the signal is typically disyllabic, reflecting the bi–directional movement of pars strindens with respect to the plectrum. We describe the stridulatory organ and acoustic signals for the first time, and discuss the possible role of stridulation in this species. Key words: Acoustic communication, Stridulation, Coleoptera, Cerambycidae, Agapanthia dahli. Resumen Estructura del órgano estridulador y análisis de la emisión acústica de Agapanthia dahli (Richter, 1821) (Coleoptera, Cerambycidae).— La estructura del órgano estridulador de Agapanthia dahli (Richter, 1821) consiste en una placa estriada o pars stridens situada en la cara dorsal del mesonoto y un rascador o plectrum constituido por un engrosamiento del margen posterior interno del pronoto. El sonido se produce cuando el insecto mueve el pronoto con respecto al mesonoto, deslizando así el plectrum contra la pars stridens. La estructura de la señal emitida es típicamente bisilábica, reflejando el movimiento bidireccional de la pars stridens con respecto al plectrum. Se describe por primera vez el órgano estridulador y la señal acústica, y se discute el posible papel de la estridulación en esta especie. Palabras clave: Comunicación acústica, Estridulación, Coleoptera, Cerambycidae, Agapanthia dahli. (Received: 7 IV 11; Conditional acceptance: 31 V 11; Final acceptance: 14 VI 11) José M. Hernández, Depto. Zoología y Antropología Física, Fac. de Biología, Univ. Complutense de Madrid, c/Antonio Novais 2 y 4, Ciudad Universitaria, 28040–Madrid, España (Spain). E–mail: jmh@bio.ucm.es

ISSN: 1578–665X

© 2011 Museu de Ciències Naturals de Barcelona


Hernández

258

Introducción La comunicación acústica es muy frecuente en muchas familias de coleópteros (Wessel, 2006), siendo la estridulación el método más generalizado (Bailey, 1991; Dumortier, 1963a; Ewing, 1989). A pesar de que este mecanismo aparece en algunas larvas de coleópteros (Leiler, 1992; Wessel, 2006; Zunino, 1987), generalmente está restringido al estado adulto (Crowson, 1981). En el imago de los Cerambycidae, han sido descritos diversos tipos de producción de sonido, tales como vibración de los élitros (Cheng, 1993). Sin embargo, el método de emisión propio y característico es la estridulación pronotal–mesonotal. La superficie dorsal del mesonoto presenta en su zona media una estriación muy fina contra la que roza el borde posterior afilado del pronoto (Dumortier, 1963a; Hernández et al., 1997; Hernández, 2007). El sonido se produce cuando el insecto dirige, con un rápido movimiento, el pronoto y la cabeza hacia abajo y hacia arriba, produciendo el frotamiento de las estructuras mencionadas en un sentido y, a continuación, en sentido contrario. Son numerosas las especies de coleópteros y otros insectos en las que se han descrito estructuras de este tipo (Álvarez et al, 2006; Cheng, 1991, 1993; Dumortier, 1963a, 1963b; Hernández et al., 2002, 2010; Ruiz et al., 2006). La comunicación acústica juega un papel muy importante en diferentes comportamientos como alarma, defensa, reclutamiento o cópula; tanto de forma aislada como modulando otro tipo de señales (Gogala, 1985; Hernández, 2007; Kirchner, 1997; Masters, 1979; Ohya, 1996; Wessel, 2006). En el presente trabajo se describe el órgano estridulador de Agapanthia dahli, así como las características acústicas de la señal emitida. Material y métodos Los ejemplares estudiados fueron recolectados en el Lerma (Burgos) el 13 de junio de 2009 y se encuentran depositados en la Colección del Departamento de Zoología y Antropología Física de la Universidad Complutense de Madrid (UCME) (tabla 1). Las grabaciones fueron realizadas en laboratorio entre 22 y 24ºC, inmovilizando a los individuos sobre un micrófono SONY ECM F8 conectado a una tarjeta digitalizadora de audio Sound Blaster Extigy, registrándose en formato WAVE PCM monofónico a 16 bits y 44100 Hz de muestreo. Para el análisis de los sonogramas, así como del espectro de frecuencias, se utilizaron los editores de audio digital Gold Wave v. 5.0. y Audacity 1.2.6. Los espectrogramas se obtuvieron con un tamaño de FFT de 256 y función de ventana Hamming. Las grabaciones se encuentran depositadas en la Fonoteca del Departamento de Zoología y Antropología Física de la Universidad Complutense de Madrid (Facultad de Biología, c/José Antonio Novais, 2 y 4; 28040–Madrid; España) (tabla 1). Para el estudio anatómico del órgano estridulador se procedió a separar el pronoto del mesonoto en

cada individuo, dejando al descubierto la cara superior del primero y el borde posterior interno del segundo, donde se sitúan el plectrum y la pars stridens, respectivamente. Ambas piezas fueron montadas en tarjetas entomológicas de cartulina. Las medidas de la pars stridens fueron tomadas sobre moldes de laca de uñas, que se realizaron aplicando ésta sobre la estructura, retirándose tras dejarse secar 24 horas a temperatura ambiente. A continuación, el molde se montó en un portaobjetos, cubriéndose con un cubreobjetos fijado con cinta autoadhesiva. Estos moldes fueron fotografiados con una cámara Opticam Pro5 adaptada a un microscopio Jaelsa Series B1 y conectada a una tarjeta digitalizadora Nvidia GeForce 8400 GS para PC. Las variables medidas fueron longitud: distancia entre la parte anterior y la posterior, anchura: la anchura máxima, y distancia entre estrías, que se calculó midiendo desde el principio de una estría hasta el principio de la siguiente. Para la obtención de microfotografías se utilizó un microscopio electrónico de barrido (JEOL, mod. JM–6400) con microsonda electrónica de 40 Kv del Centro de Microscopía Electrónica 'Luis Bru' de la Universidad Complutense de Madrid, previa metalización de las muestras con oro. Resultados y discusión El órgano estridulador en Agapanthia dahli se encuentra constituido por una pars stridens situado en la cara superior del mesonoto, que en reposo queda total o parcialmente cubierta por el pronoto, cuyo borde posterior interno constituye el plectrum o rascador. La pars stridens consiste en una región de forma elíptica, glabra y finamente estriada (fig. 1A), situada en el centro del mesonoto y con eje mayor longitudinal; las medidas se encuentran reflejadas en la tabla 2. El macho presenta una pars stridens algo más pequeña que la hembra, tanto en longitud como en anchura, conservando la misma proporción y forma. Las estrías están formadas por estrechas costillas dispuestas regularmente de forma paralela. La anchura de estas estrías es de 3 µm en la hembra y 4 µm en el macho (tabla 2). Esta diferencia, unida al mayor tamaño de la pars stridens femenina, hace que la hembra presente netamente una mayor cantidad de estrías. La estriación pierde su regularidad en los márgenes laterales de la pars stridens, transformándose en un relieve irregular en el resto del terguito, con abundantes setas (fig. 1B). A pesar de que en otros Lamiinae hayan sido descritas diferencias entre ambos sexos (Hernández et al., 1997), los datos disponibles para Agapanthia dahli no permiten discernir si éstas son significativamente mayores a las posibles diferencias interindividuales. El plectrum está formado por el borde posterior interno del pronoto, que se encuentra engrosado formando una costilla transversal cubierta de setas en su parte posterior y glabra en la cara interna (fig. 2). El sonido se produce cuando el insecto flexiona y extiende la cabeza y el pronoto, con respecto al resto del cuerpo. De esta forma, el plectrum situado


Animal Biodiversity and Conservation 34.2 (2011)

259

Tabla 1. Datos de captura y grabaciones del material estudiado. Table 1. Capture and recording data of the studied material. Individuo UCME 11967*

Sexo ♀

Localidad

Fecha

UCME 11968*

Nº Fonoteca

Lerma (Burgos) 13 VI 2009 9720090724–189 (Grab1)

Fecha 16 VI 2009

9720090724–191 (Grab2)

17 VI 2009

Lerma (Burgos) 13 VI 2009 9720090724–190 (Grab3)

16 VI 2009

9720090724–192 (Grab4)

17 VI 2009

0,5 mm

A

B

50 µm

Fig. 1. Pars stridens de la hembra de Agapanthia dahli: A. Vista general del mesonoto (50 x); B. Margen izquierdo de la pars stridens (1.000 x). Fig. 1. Pars stridens of an Agapanthia dahlia female: A. General view of mesonote (50 x); B. Left margin of the pars stridens (1,000 x).


Hernández

260

Tabla 2. Medidas de la pars stridens en Agapanthia dahli. Table 2. Measurements of pars stridens in Agapanthia dahli.

UCME 11967–Hembra

UCME 11968–Macho

Variable

n

0

σ

n

0

σ

Longitud del pars stridens

1

886 µm

1

798 µm

Anchura del pars stridens

1

512 µm

1

498 µm

Índice longitud/anchura

1

1,7

1

1,6

Anchura estría

51

3,066 µm

0,211

46

3,967 µm

0,252

2 mm

A

300 µm

B

Fig. 2. Borde posterior interno del pronoto de la hembra de Agapanthia dahli mostrando el plectrum: A. Vista general (25 x); B. Detalle de la región central (50 x). Fig. 2. Internal posterior margin of pronotum of Agapanthia dahlia female, showing plectrum: A. General view (25 x); B. Detail of central region (50 x).


Animal Biodiversity and Conservation 34.2 (2011)

A

261

H1

H2

ei

0,1 s B

0,1 s

Fig. 3. Oscilograma de la emisión de Agapanthia dahlia: A. Duplosílaba; H1. Hemisílaba primera; H2. Hemisílaba segunda; ei. Espacio intersilábico; B. Secuencia con varias sílabas. Fig. 3. Oscillogram of acoustic emission of Agapanthia dahli: A. Duplosyllable; H1. First hemisyllable; H2. Second hemisyllable; ei. Intersyllable silence; B. Several syllable sequences.

en el pronoto se frota contra la pars stridens en el mesonoto, recorriéndolo longitudinalmente tanto durante el movimiento de flexión como en el de extensión. La estridulación se produce así en forma de secuencias correspondientes al intervalo durante el que el insecto está realizando los movimientos de cabeza y pronoto. La duración de la secuencia de emisión suele prolongarse mientras se mantiene al individuo inmovilizado, alcanzando más de un minuto de emisión ininterrumpida (fig. 3B). El sonido emitido es audible para el oído humano y puede escucharse a una distancia de varios centímetros. La secuencia de emisión se estructura en un número variable de pulsos dobles (sílabas), donde cada uno de los dos pulsos (hemisílabas) corresponde a los movimientos de flexión y extensión del pronoto con respecto al mesonoto. La estructura típica consiste en una primera hemisílaba con mayor amplitud y

una segunda menos intensa, ambas de una duración similar, entre los 41 y los 108 ms, y separadas por una pequeña pausa –espacio intrasilábico– de unos 19–53 ms. Entre una sílaba y la siguiente se efectúa una pausa de extensión variable –espacio intersilábico (tabla 3, fig. 3B). La duración de cada sílaba se encuentra directamente relacionada con la velocidad con la que el insecto realiza los movimientos, al tardar el plectrum más o menos tiempo en recorrer la extensión de la pars stridens, así como por la duración de la pausa (silencio) entre uno y otro movimiento. Esta velocidad depende del estado de excitación y fatiga del individuo (Hernández, 2007). Así, podemos observar diferencias significativas en cuanto a la duración de los pulsos, los espacios intra e intersilábicos y consecuentemente en la tasa de emisión, incluso en el mismo individuo (tabla 3).


Hernández

262

Tabla 3. Estadísiticas para las características del sonograma en Agapanthia dahli. Véase el texto para la explicación. Table 3. Statistical characteristics of the sonogram in Agapanthia dahli. See the text for explanation.

UCME 11967–Hembra

1ª Hemisilaba

2ª Hemisilaba

Espacio intrasilábico

UCME 11968–Macho

Grab1

Grab2

Total

Grab3

Grab4

Total

n

335

350

685

57

41

98

0 (ms)

70,7

41,4

55,7

108,4

87,1

99,5

σ (ms)

24,8

16,2

25,4

29,7

29,7

31,3

n

333

349

682

47

26

73

0 (ms)

70,9

41,6

55,9

110,2

70,9

96,2

σ (ms)

20,4

16,4

23,6

24,7

34,3

34

n

333

349

682

47

22

69

0 (ms)

19,5

28,1

23,9

53,1

23,9

43,7

σ (ms)

9,6

26,8

20,7

81,5

23,2

69,6

n

335

350

685

57

40

97

0 (ms)

23,3

41,9

32,8

771,4

980

857,4

σ (ms)

8,3

91,2

66,1

Tasa de emisión (total)

5,56

6,67

6,09

0,9

0,67

0,79

Tasa de emisión (secuencias)

5,56

6,67

6,09

3,83

5,16

4,65

Espacio intersilábico

1.669,3 1.067,1 1.448,7

22 20 18 16

KHz

14 12 10 8 6 4 2 0 0 1 Fig. 4. Espectrograma de Agapanthia dahlia. Fig. 4. Spectrogram of Agapanthia dahli.

2

s

3

4

5


Animal Biodiversity and Conservation 34.2 (2011)

No obstante, la variación en la extensión de ambas hemisílabas es menor que la que se produce en los silencios, especialmente durante el espacio intersilábico, que puede alcanzar varios segundos. De esta forma, la tasa de emisión varía considerablemente si consideramos todo el período de grabación (tabla 3: tasa de emisión total) o únicamente los intervalos donde el insecto está realizando movimientos de forma regular (tabla 3: tasa de emisión secuencias); en este segundo caso, la tasa de emisión es más similar y se sitúa entre los 3,8 y 6,7 sílabas por segundo de media. El macho ha presentado una mayor duración tanto en las dos hemisílabas como en los espacios intrasilábicos y especialmente intersilábicos, debido a una estridulación más lenta e irregular, estructurada en grupos de pulsos separados por intervalos de silencio de duración muy variable (tabla 3). La hembra, por el contrario, ha realizado una estridulación mucho más regular durante todas las secuencias, con unos espacios intersilábicos comparables a los intrasilábicos. Como resultado, la tasa de emisión total es casi ocho veces más elevada en la hembra, mientras que si consideramos únicamente los intervalos de estridulación, la diferencia se reduce a un 30%. No obstante, y al igual que ocurría con la estructura del órgano estridulador, las diferencias encontradas con respecto a la señal acústica emitida deben tomarse con la suficiente reserva. Al haberse estudiado un único ejemplar de cada sexo, no podemos concluir si las variaciones entre ambos individuos son debidas exclusivamente a un dimorfismo sexual o a meras variaciones interindividuales. El espectro de emisión presenta una frecuencia comprendida aproximadamente entre 2,5 y 16 kHz, con dos picos de mayor energía en torno a los 6 y 12 kHz. La frecuencia es idéntica en ambas hemisílabas, y sin variación a lo largo de la secuencia (fig. 4). La frecuencia de emisión en Agapanthia dahli resulta así algo más amplia y elevada que en otros Cerambycidae como Tetraopes tetraophthalmus (Forster, 1771) y T. femoratus LeConte, 1847, con un espectro situado entre los 1,5 y 8 kHz (Alexander et al., 1963) o en varias especies del género Iberodorcadion Breuning, 1943, situadas entre 1,5 y 14 kHz (Hernández et al., 1997; Hernández, 2007). Sin embargo en Cerambyx cerdo Linneo, 1758 ha sido descrita una estridulación que alcanza una frecuencia de 20 kHz (Dumortier, 1963b). Diferencias específicas en la frecuencia de emisión también han sido descritas en otros Cerambycidae como Tylocerina nodosus (Fabricius, 1775), Neacanthocinus obsoletus (Olivier, 1795), Monochamus titillator (Fabricius, 1775) y Plectrodera scalator (Fabricius, 1792) (Finn et al., 1972). En todos estos casos, únicamente se ha observado estridulación cuando el insecto es molestado, cesando ésta cuando finaliza el estorbo (Alexander et al., 1963; Finn et al., 1972; Hernández et al., 1997; Hernández, 2007). Estos datos coinciden con los presentados en este trabajo, dado que la emisión de sonido únicamente se produce cuando el insecto es molestado o inmovilizado, no habiéndose observado estridulación espontánea en ningún momento.

263

En el caso concreto de Agaphanthia dahli, algunos aspectos como el hecho de que ambos sexos dispongan de órgano estridulador y emitan una señal similar, el amplio espectro de frecuencia abarcada, así como la producción de sonido exclusivamente como respuesta al contacto o molestia, nos inclinan a pensar que se trata de una estridulación de defensa poco específica (Eisner et al., 1974; Gorb, 1998; Hernández et al., 1997) Referencias Alexander, R. D., Moore, T. E. & Woodruff, R. E., 1963. The evolutionary differentiation of stridulatory signals in beetles. Animal Behaviour, 11: 111–115. Álvarez, M., Martínez, M. D., Ruiz, E. & Hernández, J. M., 2006. Estudio comparado del pars stridens en las obreras de cinco nidos de Aphaenogaster senilis Mayr, 1853 (Hymenoptera, Formicidae). Boletín de la Real Sociedad Española de Historia Natural (Sección Biología), 101(1–4): 93–98. Bailey, W. J., 1991. Acoustic behaviour of insects. An evolutionary prespective. Chapman and Hall, London. Cheng, J. Q., 1991. Sound production in Longhorned Beetles: stritulation and associated behaviour of the adult (Coleoptera: Cerambycidae). Scientia Silvae Sinicae, 27(3): 234–237. – 1993. A study on the acoustical properties of thoracic stridulation and elytral vibration sounding in beetle Anoplophora horsfieldi (Hope) (Coleoptera: Cerambycidae). Acta Entomologica Sinica, 36(2): 10–15. Crowson, R. A ., 1981. The Biology of the Coleoptera. Academic Press, London. Dumortier, B., 1963a. Morpholgy of sound emission apparatus in Arthropoda. In  : Acoustic behaviour of animals. Chapter 11: 277–345 (R. G. Busnell, Ed.). Elsevier Publishing Co, New York. – 1963b. The physical characteristics of sounds emissions in Arthropoda. In  : Acoustic behaviour of animals. Chapter 12: 346–373 (R. G. Busnell, Ed.). Elsevier Publishing Co, New York. Eisner, T., Aneshansley, D., Eisner, M., Rutowsky, R., Chong, B. & Meinwald, J., 1974. Chemical defense and sound production in Australian tenebrionid beetles (Adelium spp.). Psyche, 81: 189–208. Ewing, A. W., 1989. Arthropod bioacoustics. Neurobiology and behaviour. Edinburgh University Press, Edinburgh. Finn, W. E., Mastro, V. C. & Payne, T. L., 1972. Stridulatory Apparatus and Analysis of the Acoustics of Four Species of the Subfamily Lamiinae (Coleoptera: Cerambycidae) . Annals of the Entomological Society of America, 65(3): 644–647. Gogala, M., 1985. Vibrational communication in insects (biophysical and behavioural aspects). In: Acoustic and Vibrational Communication in Insects: 117–126 (K. Kalmring & N. Elsner, Eds.). Paul Parey, Berlin. Gorb, R. N., 1998. Frictional surfaces of the elytral– to–body arresting mechanism in tenebrionid beetles (Coleoptera: Tenebrionidae) design of co–opted


264

fields of microtrichia and cuticle ultrastructure. International Journal of Insect Morphology and Embryology, 27(3): 205–225. Hernández, J. M., 2007. Estridulación provocada por la interacción entre coespecíficos en la especie Iberodorcadion (Hispanodorcadion) perezi hispanicum (Mulsant, 1851) (Coleoptera, Cerambycidae, Lamiinae). Boletín de la Asociación española de Entomología, 31(3–4): 259–269. Hernández, J. M., Gamarra, P. & Outerelo, R., 2010 Características morfológicas y bioacústicas de la estridulación en Phylan (Phylan) foveipennis foveipennis (Mulsant & Rey, 1854) (Coleoptera, Tenebrionidae). Boletín de la Real Sociedad Española de Historia Natural (Sección Biológica), 104: 85–94. Hernández, J. M., García, D. & Gamarra, P., 1997. Comunicación acústica en algunas especies de Iberodorcadion Breuning, 1943 (Coleoptera, Cerambycidae, Lamiinae). Elytron, 11: 51–62. Hernández, J. M., Martínez, M. D. & Ruiz, E., 2002. Descripción del órgano estridulador en Messor barbarus (Linneo, 1767) (Hymenoptera, Formicidae). Anales de Biología, 24: 167–174. Kirchner, W. H., 1997. Acoustical communication in social insects. In: Orientation and Communication in Arthropods: 273–300 (M. Lehrer, Ed.). Birkhäuser Verlag, Basel.

Hernández

Leiler, T. E., 1992. Ljudalstring hos Lamiinae–larver (Coleoptera, Cerambycidae). Entomologisk Tidskrift, 113: 55–56. Masters, W. M., 1979. Insect disturbance stridulation: Its Defensive role. Behavioral Ecology and Sociobiology, 5: 187–200. Ohya, E., 1996. Sound communications of two species of pleasing fungus beetles Dacne japonica and D. picta (coleoptera: Erotylidae) using two types of sound producing apparatus. Proceedings of XX International Congress of Entomology; Firenze, Italy: 374. Ruiz, E., Martínez, M. H., Martínez, M. D. & Hernández, J. M., 2006. Morphological study of the Stridulatory Organ in two species of Crematogaster genus: Crematogaster scutellaris (Olivier 1792) and Crematogaster auberti (Emery 1869) (Hymenoptera: Formicidae). Annales de la Société Entomologique de France (n.s.), 42(1): 99–105. Wessel, A., 2006. Stridulation in Coleoptera – An Overview. In Insects Sounds and Comunication. Physiology, Behaviour, Ecology and Evolution: 397–403 (S. Drosopoulos & M. F. Caridge, Eds.). Taylor & Francis, New York. Zunino, M., 1987. Larval stridulation and feeding behaviour in Trogid beetles (Coleoptera). Bolletino delle Accademia Gioenia di Scienze Naturali, 20(332): 299–300.


Animal Biodiversity and Conservation 34.2 (2011)

265

Caracteres preimaginales y aspectos bionómicos de Gyriosomus luczotii Laporte, 1840 (Coleoptera, Tenebrionidae) J. Pizarro–Araya, V. Jerez, J. Cepeda–Pizarro & F. M. Alfaro

Pizarro–Araya, J., Jerez, V., Cepeda–Pizarro, J. & Alfaro, F. M., 2011. Caracteres preimaginales y aspectos bionómicos de Gyriosomus luczotii Laporte, 1840 (Coleoptera, Tenebrionidae). Animal Biodiversity and Conservation, 34.2: 265–272. Abstract Preimaginal characters and bionomical aspects of Gyriosomus luczotii Laporte, 1840 (Coleoptera, Tenebrionidae).— We describe the morphology and microstructure of the egg chorion, and the morphology of the first instar larva of Gyriosomus luczotii Laporte, 1840 (Coleoptera, Tenebrionidae, Nycteliini). Bionomical and distributional data on this species are also provided. To obtain eggs and larvae, couples were collected in the field and kept in rearing cages until oviposture and ecclosion. The structure and adornment of the egg exochorion, and the exterior morphological features of larvae were examined with electron scanning microscopy. The eggs of G. luczotii showed a rounded micropyle and a smooth exochorion, composed of hexagonal cells without aeropyles. The larvae of G. luczotii showed morphological characteristics suited for an edaphic life similar to that of Pedobionta: digging prothoracic legs, cephalic capsule with abundant sensilla, and well–developed pygopodium. We analysed the importance of larval morphology as an element for specific diagnosis and found that interspecific differences regarding frontal sensilla, clypeus shape, and anterior part of labrum, had a taxonomic value and possibly a phylogenetic value. Key words: Tenebrionidae, Gyriosomus, Morphology, Preimaginal stages, Coastal desert, Chile. Resumen Caracteres preimaginales y aspectos bionómicos de Gyriosomus luczotii Laporte, 1840 (Coleoptera, Tenebrionidae).— Se describe la morfología y microestructura coriónica del huevo y la morfología del primer estadio larvario de Gyriosomus luczotii Laporte, 1840 (Coleoptera, Tenebrionidae, Nycteliini). También se exponen antecedentes bionómicos y distribucionales de la especie. Para la obtención de huevos y larvas, se recolectaron parejas en el terreno que fueron mantenidas en cajas de cria hasta la ovoposición y posterior eclosión. La estructura y ornamentación del exocorion del huevo y características morfológicas externas de la larva fueron analizadas mediante microscopía electrónica de barrido. Los resultados muestran que los huevos de G. luczotii presentan un micropilo redondeado y exocorion liso, con células hexagonales sin aeropilos. Las larvas de G. luczotii presentan características morfológicas adaptativas para la vida edáfica del tipo Pedobionta: cápsula cefálica con gran cantidad de sensillas, patas protorácicas de función cavadora, y pigopodio bien desarrollado. Se analiza la importancia de algunos caracteres morfológicos de la larva de primer estadio como criterio de diagnóstico específico y se establece que las diferencias interespecíficas referidas a las sensillas frontales, la forma del clípeo y el margen anterior del labro tienen valor taxonómico y probablemente filogenético. Palabras clave: Tenebrionidae, Gyriosomus, Morfología, Estadios preimaginales, Desierto costero, Chile. (Received: 17 I 11; Conditional acceptance: 17 V 11; Final acceptance: 17 VI 11) J. Pizarro–Araya, J. Cepeda–Pizarro & F. M. Alfaro, Lab. de Entomología Ecológica, Depto. de Biología, Fac. de Ciencias, Univ. de La Serena, Casilla 599, La Serena, Chile.– V. Jerez, Depto. de Zoología, Fac. de Ciencias Naturales y Oceanográficas, Univ. de Concepción, Casilla 160–C, Concepción, Chile. Corresponding author: J. Pizarro–Araya. E–mail: japizarro@userena.cl ISSN: 1578–665X

© 2011 Museu de Ciències Naturals de Barcelona


Pizarro-Araya et al.

266

Introducción Debido a que presentan diversas estrategias ecológicas y caracteres morfo–adaptativos, los coleópteros tenebriónidos constituyen un componente significativo de la biota de los ecosistemas áridos y semiáridos del mundo (Cloudsley–Thompson, 2001; Whitford, 2002; Fattorini, 2009), participando en la descomposición de materia orgánica y constituyendo un recurso trófico importante para vertebrados e invertebrados propios de estos ecosistemas (Donadio et al., 2004; Pizarro– Araya, 2010; Vidal et al., 2011). Entre los caracteres morfo–adaptativos de los tenebriónidos, se destacan estructuras como el tipo de patas (Medvedev, 1965; Krasnov et al., 1996), la presencia de una cavidad subelitral (Duncan, 2003; Byrne & Duncan, 2003), espiráculos mesotorácicos (Duncan & Byrne, 2005; Duncan & Dickman, 2009), pilosidad elitral (Gorb, 1999; Bouchard & Gorb, 2000; Norgaard & Dacke, 2010) y glándulas aedeagales (Geiselhardt et al., 2008). Sin embargo el estudio de estos caracteres se ha realizado utilizando imagos y no estadios preimaginales (Pizarro–Araya et al., 2005). El conocimiento de los estados preimaginales reviste un notable interés referido a estudios sobre edafología, debido al desarrollo hipogeo de sus estadios larvarios (Whitford, 1996). Por otra parte, en diversos géneros de Tenebrionidae los caracteres morfológicos preimaginales permiten reconocer diferencias poblacionales (Francisco & Do Prado, 2001), establecer caracteres de diagnóstico genéricos o tribales (Doyen, 1988; Iwan & Bečvář, 2000; Iwan & Banaszkiewicz, 2005), postular inferencias filogenéticas al ser utilizados como homologías (Aalbu, 1985; Schunger et al., 2003; Bouchard & Steiner, 2004; Beutel & Friedrich, 2005), o bien sugerir adaptaciones ontogénicas a hábitats particulares (Jaramillo et al., 2000; Pizarro–Araya et al., 2005, 2007). Sin embargo en relación con los estadios preimaginales de tenebriónidos chilenos, existen algunos antecedentes (Pizarro–Araya et al., 2008) y la mayoría de los trabajos han considerado aspectos morfológicos de huevos (Pizarro–Araya et al., 2005, 2007), larvas (Cekalovic & Morales, 1974; Cekalovic & Quezada, 1982; Pizarro–Araya et al., 2005, 2007) y pupas (Artigas & Brañas–Rivas, 1973; Cekalovic & Quezada, 1973). Entre los coleópteros propios de ecosistemas desérticos costeros de Chile, se destaca el género Gyriosomus Guérin–Méneville, 1834 (Tenebrionidae: Nycteliini), taxón endémico y erémico con 37 especies descritas hasta la fecha (Gebien, 1944; Kulzer, 1959; Pizarro–Araya & Flores, 2006), distribuidas desde el norte de la Reserva Nacional de Paposo (25º 05' S, 70º 29 O, en la región de Antofagasta), hasta la precordillera de Rancagua (34° 11' S, 70° 39' O, en la Región de O’Higgins) (Pizarro–Araya & Flores, 2004; Pizarro–Araya & Jerez, 2004; Alfaro et al., 2009), áreas consideradas pertenecienes a las provincias biogeográficas de Coquimbo y Santiago (Morrone, 2006). Dado que la microestructura coriónica y morfología larvaria constituyen elementos de diagnóstico

específico, adaptativos y filogenéticos, los objetivos del presente trabajo son describir la morfología y microestructura del exocorion del huevo y la morfología del primer estadio larvario de Gyriosomus luczotii Laporte, 1840 (Coleoptera, Tenebrionidae, Nycteliini), documentar antecedentes bionómicos y distribucionales de la especie y determinar caracteres que permitan postular inferencias filogenéticas. Material y métodos El material estudiado proviene de parejas de adultos recolectadas en las siguientes localidades de la Región de Coquimbo (Chile): Colina El Pino, La Serena, 29° 54' S, 71° 15' O, 90 m s.n.m., 20 XI 2003; Socos, Limarí, 30º 84' 49'' S, 71º 83'  19'' O, 200  m  s.n.m., 15 XI 2003; El Molle, Elqui, 29° 97.035' S, 70° 95.789' O, 450 m s.n.m., 14 IX 2002. Las parejas se mantuvieron en cajas de crianza con sustrato de arena, temperatura de 17°C mínima y 24°C máxima y fotoperíodo de 12 horas de luz y 12 horas de oscuridad, hasta la ovoposición y posterior eclosión de los huevos. Los insectos fueron alimentados con Erodium cicutarium (L.) (Geraniaceae), Cristaria glaucophylla Cav. (Malvaceae) más un suplemento de dieta artificial. Los huevos y larvas de primer estadio obtenidos fueron conservados en alcohol del 70% y posteriormente deshidratados en batería de alcohol, secado al punto crítico y metalizado con oro, según la técnica descrita por Pizarro–Araya et al. (2005), para su observación y obtención de fotografías con microscopio electrónico de barrido (ETEC Autoscann U1) del Laboratorio de Microscopía Electrónica de la Universidad de Concepción. Para la descripción de la estructura y microestructura coriónica, se utilizó la nomenclatura seguida por Jerez (2003) y Pizarro–Araya et al. (2005). Para la morfología larvaria se siguió a López–Sánchez et al. (1985a, 1985b, 1987), Doyen (1988) y Pizarro–Araya et al. (2007). Los materiales (huevos, larvas y adultos) están depositados en la colección del Laboratorio de Entomología Ecológica de la Universidad de La Serena, La Serena, Chile (LEULS). Resultados Descripción del huevo El huevo es alargado con polos redondeados, con una longitud de 3,65 mm (rango: 3,35–3,91 mm; n = 10) y un ancho de 1,07 mm (promedio: 0,95–1,20 mm; n = 10), exocorion de color blanco y liso (fig. 1A). El primer tercio del huevo está ornamentado con celdas subhexagonales con un diámetro de 18,2 µ (promedio: 11,8–24,6 µ; n = 10). El área micropilar de forma circular y ligeramente sobresaliente, se encuentra en la región apical del huevo, con un diámetro de 18,7 µ (promedio: 12,3–22,9 µ; n = 10) y presenta numerosos poros de diámetro 1,55 µ (promedio: 0,8–1.5 µ; n  =  10) (fig. 1a). El micropilo presenta además aeropilos con diámetro aproximado de 1,5 μ (promedio: 1–2 μ; n = 10) (fig. 1B).


Animal Biodiversity and Conservation 34.2 (2011)

A

267

B amic

amic

c ae

25 x C

2000 x D

ep

sa fr

lb cl

ea s 350 x

220 x

Fig. 1. Gyriosomus luczotii: A. Huevo en vista lateral: c. Corion; amic. Área micropilar. B. Región micropilar: amic. Área micropilar; ae. Aeropila. C. Cápsula cefálica, vista latero frontal: ep. Epicranio; fr. Frente; cl. Clípeo; lb. Labro. D. Antena: sa. Sutura antenal; ea. Esclerito antenal; s. Sensorium. Fig. 1. Gyriosomus luczotii: A. Lateral view of egg: c. Chorion; amic. Micropylar area. B. Micropylar region: ami. Micropylar area; ae. Aeropyles. C. Cephalic capsule, latero–frontal view: ep. Epicranium; fr. Frons, cl. Clypeus; lb. Labrum. D. Antenna: sa. Antennal suture; ea. Antennal sclerite; s. Sensorio.

Descripción de la larva de primer estadio Presenta el cuerpo alargado y cilíndrico con una longitud de 5,41 mm (promedio: 4,57–6,79 mm; n = 10). Región frontal deprimida dorsalmente y provista de numerosas sensillas dispuestas en hileras horizontales (fig. 1C). Patas protorácicas que duplican en tamaño a las meso y metatorácicas. Cápsula cefálica: con longitud de 0,46 mm (promedio: 0,43–0,50 mm; n = 10) y ancho de 0,71 mm (promedio: 0,59–0,77 mm; n = 10) (tabla 1), globosa y fuertemente esclerotizada con sutura epicraneal no evidente; epicráneo provisto de setas dorsales y laterales largas y densas; sutura

frontal rectilínea; sutura frontoclipeal acusada y ligeramente curvada. Región frontal deprimida dorsalmente y provista de numerosos tubérculos de ápice redondeado y dispuestos en hileras horizontales (fig. 1C). Ocelos ausentes. Antena con tres segmentos, primer segmento antenal alargado, segundo igual o más largo que la mitad del primero; tercer segmento pequeño y con una sensilla redondeada y una seta apical (fig. 1D). Clípeo subrectangular y glabro. Labro transverso, con un par de sensillas dorsales y un pecten formado por una hilera de setas gruesas; margen distal levemente sinuado y provisto de setas largas y rígidas (fig. 2A). Mandíbulas subtriangulares,


Pizarro-Araya et al.

268

Tabla 1. Medidas (en mm) de la cápsula cefálica y segmentos torácicos de la larva de primer estadio de Gyriosomus luczotii Table 1. Measurements (in mm) of the head capsule and thoracic segments of the first instar larva of Gyriosomus luczotii.

Carácter

Ancho

Promedio

Longitud

Promedio

n

Cápsula cefálica

0,71

0,59–0,77

0,46

0,43–0,50

10

Segmento torácico 1

0,77

0,70–0,84

0,51

0,31–0,66

10

Segmento torácico 2

0,74

0,67–0,82

0,30

0,20–0,40

10

Segmento torácico 3

0,74

0,66–0,83

0,34

0,21–0,44

10

quitinizadas y con dos dientes apicales de ápice obtuso (fig. 2B); mola bien desarrollada, ovalada, cóncava y con superficie lisa; porción basolateral de la superficie dorsal con un mechón de setas; largas y rígidas. Palpos maxilares trisegmentados, último segmento corto y de ápice redondeado (fig. 2B); mala redondeada y provista de numerosas setas cortas y gruesas en el margen externo. Palpos labiales con dos segmentos (fig. 2B). Tórax: primer segmento más largo que los siguientes, con una longitud de 0,51 mm (promedio: 0,31–0,66 mm; n = 10) (tabla 1); escudo protorácico subrectangular y curvado lateralmente. Meso y metatórax con tegumento poco esclerotizado, y ruptor ovi no visible. Patas protorácicas más desarrolladas que el segundo y el tercer par; coxa de forma tetragonal, un poco más ancha que larga; el margen interno de la tibia lleva tres espinas de aspecto lanceolado; margen anterior y superficie externa con numerosas espinas largas y delgadas; margen interno del fémur lleva una hilera de espinas cortas, gruesas y de ápice obtuso; tarsungulus largo y curvado (figs. 2B, 2C). Espiráculos mesotorácicos de diámetro circular, uniforos y peritrema oscuro. Abdomen con una longitud de 4,26  mm (promedio: 3,57–5,33 mm; n = 10). Segmentos con pilosidad escasa y corta; IX segmento abdominal alargado y subtriangular, con el extremo apical redondeado; dorsalmente el tergito está provisto de cerdas largas repartidas por toda la superficie; pigopodio bien desarrollado y fuertemente esclerotizado en sus extremos (fig. 2D). Diferencias interespecíficas La larva de primer estadio de G. luczotii muestra uniformidad en los caracteres morfológicos con respecto a lo descrito por Pizarro–Araya et al. (2005, 2007). Sin embargo se observan algunas diferencias en la región frontal y el aparato bucal. La frente de G. luczotii está provista de numerosos tubérculos dispuestos en hileras horizontales (figs. 1C, 2A). G. luczotii presenta un clípeo subrectangular con el margen distal cóncavo, mientras que en G. subrugatus es subrectilíneo y con algunos tubérculos dorsales (figs. 1C, 2A).

A diferencia del labro levemente sinuado en G. kingi, en G. luczotii es emarginado (figs. 1C, 2A). El último segmento de los palpos maxilares puede ser corto y de ápice redondeado como en G. kingi (fig. 2B), márgenes del segundo segmento con seta apical, primer segmento provisto de un sensorio redondeado y setas apicales, o alargado y de ápice redondeado como en G. subrugatus. Bionomía G. luczotii está asociada tróficamente a la vegetación arbustiva característica del matorral estepario costero constituido por especies como Cristaria glaucophylla Cav. (Malvaceae), Frankenia chilensis K. Presl (Frankeniaceae), Nolana brunonianus Hook. et. Arn. (Nolanaceae), Nolana sedifolia Poepp. y Haplopappus foliosus DC. (Asteraceae). El período reproductivo se inicia a mediados de septiembre con la emergencia de adultos a los que es factible encontrar desplazándose sobre el suelo o sobre las partes aéreas de plantas (e.g. C. glaucophylla y Nolana spp) (figs. 3A, 3B). Son de hábitos diurnos, epígeos y en las horas de mayor radiación solar se entierran bajo el suelo o bien se esconden bajo piedras o arbustos (fig. 3C) Las cópulas comienzan inmediatamente tras la emergencia del estado de pupa (fig. 3D), y observaciones de laboratorio muestran que las hembras inician las ovoposiciónes a finales de septiembre, ya sea en la superficie del sustrato arenoso, o bien enterrándose a una profundidad que varía entre 10 a 20 cm (fig. 3D). Cada hembra presenta un máximo de seis puestas, los huevos son colocados en grupos de siete a 10 y se encuentran unidos lateralmente. Los huevos no son visibles externamente ya que están revestidos por una película de arena que se adhiere al corion mediante un mucílago producido por la hembra en el momento de la ovoposición el cual posteriormente se seca. Hábitat Los sectores litorales en donde se distribuye mayoritariamente G. luczotii se caracterizan por presentar


Animal Biodiversity and Conservation 34.2 (2011)

A

269

B md pm pl

s

a

cl lb pp 160 x

360 x C

D

ts

t pg

f 200 x

85 x

Fig. 2. Gyriosomus luczotii: A. Detalle de la frente: s. Sensillas; cl. Clípeo; lb. Labro. B. Complejo máxilo– labial vista ventral: a. Antena; md. Mandíbula; pm. Palpo maxilar; pl. Palpo labial; pp. Pata protorácica. C. Pata protorácica: f. Fémur; t. Tibia; ts. Tarsungulus. D. Detalle del pigopodio (pg). Fig. 2. Gyriosomus luczotii: A. Detail of frons: s. Sensilla; cl. Clipeus; lb Labrum. B. Ventral view of maxillo labial complex: a. Antenna; md. Mandibles; pm. Maxillary palpi; pl. Labial palpi; pp. Prothoracic leg. C. Prothoracic leg: f. Femur; t. Tibiae; ts. Tarsungulus. D. Pygopods detail (pg).

suelos planos a nivel del mar con suelos de tipo arenoso y arenoso–pedregoso (Paskoff & Manríquez, 2004; Castro & Brignardello, 2005). El clima presente en estos sitios es de tendencia mediterránea con una oscilación térmica diaria y anual baja debido a la influencia marina (Armesto et al., 1993). La vegetación existente en estos lugares se caracteriza por responder a pulsos de humedad estacionales y a la disponibilidad de humedad edáfica. De esta manera la precipitación permiten la germinación, crecimiento vegetativo y floración de plantas anuales y geófitas

(Armesto & Vidiella, 1993; Vidiella et al., 1999), elementos que a su vez representan una fuente de alimento para los adultos de esta especie (Cepeda– Pizarro et al., 2005; Alfaro et al., 2009). Distribución geográfica G. luczotii habita desde Los Choros (29°  17'  S, 71° 18' O), hasta Alcones (30° 44' S, 71° 31'´ O) en la región de Coquimbo, Chile; asociada a planicies litorales, depresiónes intermedias y valles transversales con


Pizarro-Araya et al.

270

A

B

C

D

Fig. 3. Gyriosomus luczotii: A. Hábito del adulto macho; B. Hábito del adulto hembra; C. Conducta evasiva del adulto; D. Pareja de adultos en cópula. Fig. 3. Gyriosomus luczotii: A. Habitus of male adult; B. Habitus of female adult; C. Evasive behavior of an adult; D. Adult couple mating.

vegetación arbustiva. Pizarro–Araya & Jerez (2004) establecen la distribución del género Gyriosomus en relación con las formaciones vegetales, señalando que G. luczotii está relacionada con el matorral estepario costero y con el matorral estepario boscoso (Gajardo, 1993), formaciones que incluye el desierto costero transicional de Chile (25–32º Lat S). Discusión La monofilia mostrada por el género Gyriosomus (Flores, 2000), se explica por la presencia de cuatro sinapomorfías presentes en los adultos: paraproctos pilosos, coxitos con setas largas, espermateca con cavidad principal semianillada y longitud relativa del lóbulo medio corta (L/T ≤ 0,75); estos caracteres han sido descritos sólo para imagos de Gyriosomus. La utilización de caracteres preimaginales puede reforzar esta monofilia, considerando que el conocimiento de la microestructura coriónica (Hernández, 1991; Castillo et al., 1994; Candan et al., 2004) y de la morfología larvaria permiten establecer caracteres

de diagnóstico a nivel de familia, género y especie (Crowson, 1981; Jerez, 2003). En relación con el estado larvario, la larva de primer estadio de G. luczotii corresponde al tipo pedobionta propuesto por Keleynikova (1963) y se adscribe a la descripción morfológica señalada por Doyen (1988) para larvas de la subfamilia Pimeliinae, en el sentido de presentar estructuras adaptativas para la vida edáfica, tales como son la abundante dotación sensilar en la cápsula cefálica, patas cavadoras (especialmente las protorácicas) y pigopodio bien desarrollado. Para G. luczotii la forma de las sensillas frontales y las características del clípeo y labro de la larva de primer estadio destacan como caracteres de valor taxonómico y posiblemente también filogenético. Los resultados obtenidos en este estudio pueden contribuir a definir la posición sistemática de G. luczotii y su vinculación con los demás elementos de Nycteliini, lamentablemente la ausencia de descripciones de huevos de otros Nycteliini o tribus hermanas (Praocini y Physogasterini) (Flores, 2000), no permiten establecer inferencias filogenéticas basadas en estos caracteres.


Animal Biodiversity and Conservation 34.2 (2011)

Agradecimientos Los autores agradecen al personal del Laboratorio de Microscopía Electrónica de la Dirección de Investigación de la Universidad de Concepción, Concepción, Chile. Este trabajo fue financiado por los proyectos DIULS–PF07101 de la Universidad de La Serena, Chile (JPA) y DIUC N°203.113.061–1.0, Dirección de Investigación, Universidad de Concepción, Concepción, Chile (VJ). Referencias Aalbu, R. L., 1985. New genus and species of Triorophini, including immatures, reproductive structures, and notes on biology and phylogeny (Coleoptera: Tenebrionidae). Annals of the Entomological Society of America, 78: 541–553. Alfaro, F. M., Pizarro–Araya, J. & Flores, G. E., 2009. Epigean tenebrionids (Coleoptera: Tenebrionidae) from the Choros Archipelago (Coquimbo Region, Chile). Entomological News, 120: 125–130. Armesto, J. J. & Vidiella, P. E., 1993. Plant life forms and biogeographic relations of the flora of Lagunillas (30°S) in the fog–free pacific coastal desert. Annals of the Missouri Botanical Garden, 80: 499–511. Armesto, J. J., Vidiella, P. E. & Gutiérrez, J. R., 1993. Plant communities of the fog–free coastal desert of Chile: plant strategies in a fluctuating environment. Revista Chilena de Historia Natural, 66: 271–282. Artigas, J. & Brañas–Rivas, P., 1973. Descripción de la larva y pupa de Praocis curta Solier (Coleoptera: Tenebrionidae). Boletín de la Sociedad de Biología de Concepción, XLVI: 64–74. Beutel, R. G. & Friedrich, F., 2005. Comparative study of larvae of Tenebrionoidea (Coleoptera: Cucujiformia). European Journal of Entomology, 102: 241–264. Bouchard, P. & Gorb, S. N., 2000. The elytra–to–body binding mechanism of the flightless rainforest species Tabarus montanus Kaszab (Coleoptera: Tenebrionidea). Arthropod Structure and Development, 29: 323–331. Bouchard, P. & Steiner, W. E., 2004. First descriptions of Coelometopini pupae (Coleoptera: Tenebrionidae) from Australia, Southeast Asia and the Pacific region, with comments on phylogenetic relationships and antipredator adaptations. Systematic Entomology, 29: 101–114. Byrne, M. J. & Duncan, F. D., 2003. The role of the subelytral spiracles in respiration in the flightless dung beetle Circellium bacchus. Journal of Experimental Biology, 206: 1309–1318. Candan S., Suludere, Z., Kalender, Y. & Eryilmaz, O., 2004. Ultrastructure of the chorion of Echthistus cognatus (Loew, 1849) (Diptera, Asilidae). Ohio Journal of Science, 104: 93–96. Castillo, E. E, Jerez, V. & Artigas, J. N., 1994. Microescultura corionica en huevos de Asilidae (Diptera Asilinae, Dasypogoninae, Laphriinae y Stenopogoninae). Boletín de la Sociedad de Biología de Concepción, 65: 107–116. Castro, C. & Brignardello, L., 2005. Geomorfología

271

aplicada a la ordenación territorial de litorales arenosos. Orientaciones para la protección, usos y aprovechamiento sustentables del sector de Los Choros, Comuna de la Higuera, IV Región. Revista de Geografía Norte Grande, 33: 33–58. Cekalovic, T. & Morales, E., 1974. Descripción de la larva y pupa de Oligocara nitida Solier, 1848 (Coleoptera: Tenebrionidae). Boletín de la Sociedad de Biología de Concepción, XLVIII: 173–177. Cekalovic, T. & Quezada, A., 1973. Descripción de la larva, pupa y distribución geográfica de Emmallodera multipunctata curvidens Kulzer, 1955 (Coleoptera: Tenebrionidae). Anales del Instituto de la Patagonia, 4: 417–422. – 1982. Descripción de la larva de Nycterinus (Eunycterinus) abdominalis Eschscholtz, 1828 (Coleoptera, Tenebrionidae). Boletín de la Sociedad de Biología de Concepción, LIII: 47–51. Cepeda–Pizarro, J., Pizarro–Araya, J. & Vásquez, H., 2005. Variación en la abundancia de Arthropoda en un transecto latitudinal del desierto costero transicional de Chile, con énfasis en los tenebriónidos epígeos. Revista Chilena de Historia Natural, 78: 651–663. Cloudsley–Thompson, J. L., 2001. Thermal and water relations of desert beetles. Naturwissenschaften, 88: 447–460. Crowson, R. A., 1981. The biology of the Coleoptera. Academic Press, London. Donadio, E., Di Martino, S., Aubone, M. & Novaro, A. J., 2004. Feeding ecology of the Andean hog– nosed skunk (Conepatus chinga) in areas under different land use in north–western Patagonia. Journal of Arid Environments, 56: 709–718. Doyen, J. T., 1988. Descriptions of some phylogenetically important larvae of Tenebrionidae (Coleoptera). Coleopterists Bulletin, 42: 285–301. Duncan, F. D., 2003. The role of the subelytral cavity in respiration in a tenebrionid beetle, Onymacris multistriata (Tenebrionidae: Adesmiini). Journal of Insect Physiology, 49: 339–346. Duncan, F. D. & Byrne, M. J., 2005. The role of the mesothoracic spiracles in respiration in flighted and flightless dung beetles. Journal of Experimental Biology, 208: 907–914. Duncan, F. D. & Dickman, C. R., 2009. Respiratory strategies of tenebrionid beetles in arid Australia: does physiology beget nocturnality?. Physiological Entomology, 34: 52–60. Fattorini, S., 2009. Darkling beetle communities in two geologically contrasting biotopes: testing biodiversity patterns by microsite comparisons. Biological Journal of the Linnean Society, 98: 787–793. Flores, G. E., 2000. Cladistic analysis of the neotropical tribe Nycteliini (Coleoptera: Tenebrionidae). Journal of the New York Entomological Society, 108: 13–25. Francisco, O. & Do Prado, A. P., 2001. Characterization of the larval stages of Alphitobius diaperinus (Panzer) (Coleoptera: Tenebrionidae) using head capsule width. Revista Brasileira de Biologia, 1: 125–131. Gajardo, R., 1993. La vegetación natural de Chile. Editorial Universitaria, Santiago, Chile. Gebien, H. 1944. Die Gattung Gyriosomus Guér.


272

Coleoptera, Tenebrionidae. Mitteilungen der Münchener Entomologischen Gesellschaft, 34: 135–194. Geiselhardt, S., Jakobschy, D., Ockenfels, P. & Peschke, K., 2008. A sex pheromone in the desert tenebrionid beetle Parastizopus armaticeps. Journal of Chemical Ecology, 34: 1065–1071. Gorb, S. N., 1999. Ultrastructure of the thoracic dorso– medial field (TDM) in the elytra–to–body arresting mechanism in tenebrionid Beetles (Coleoptera: Tenebrionidae). Journal of Morphology, 240: 101–113. Hernández, J. M., 1991. Estudio de los caracteres del huevo en diversos Cerambycidae ibéricos y su interés taxonómico (Coleoptera). Graellsia, 47: 49–59. Iwan, D. & Banaszkiewicz, M., 2005. Larvae of the genus Anomalipus Latreille, 1846 (Coleoptera: Tenebrionidae). Annales Zoologici, 55: 375–381. Iwan, D. & Bečvář, S., 2000. Description of the early stage of Anomalipus plebejus plebejulus (Coleoptera: Tenebrionidae) from Zimbabwe with notes on the classification of the Opatrinae. European Journal of Entomology, 97: 403–412. Jaramillo, E., Avellanal, M. H., González, M. & Kennedy, F., 2000. Locomotor activity of Phalerisida maculata Kulzer (Coleoptera, Tenebrionidae) on Chilean sandy beaches. Revista Chilena de Historia Natural, 73: 67–77. Jerez, V., 2003. Interespecific differentiation in eggs and first instar larvae in the genus Procalus Clark 1865 (Chrysomelidae: Alticinae). In: Proccedings of the fifth international Symposium on the Chrysomelidae; Special topics in leaf beetle biology: 147–153 (D. G. Furth, Ed.). Pensoft Publishers Press, Sofia. Keleynikova, S. I., 1963. Larval types of palearctic darkling beetles (Coleoptera, Tenebrionidae). Entomological Review, 42: 292–296. Krasnov, B., Ward, D. & Shenbrot, G., 1996. Body size and leg length variation in several species of darkling beetles (Coleoptera: Tenebrionidae) along a rainfall and altitudinal gradient in the Negev Desert (Israel). Journal of Arid Environments, 34: 477–489. Kulzer, H., 1959. Neue Tenebrioniden aus Südamerika (Col.) 18 Beitrang zur Kenntnis der Tenebrioniden. I Die Gattung Gyriosomus Guérin. (Nycteliini). Entomologische Arbeiten aus dem Museum Georg Frey, 10: 523–547. López–Sánchez, S., De los Santos, A. & Montes, C., 1985a. Estudio morfológico de los estados preimaginales de Micrositus ulyssiponensis Germ. 1824 (Coleoptera: Tenebrionidae). Anales de Biología Universidad de Murcia, 3: 95–102. – 1985b. Estudio morfológico de la forma larvaria de Tentyria placyteps Stev. (1829). Eos, LXI: 173–182. – 1987. Estudio morfológico de los estados preimaginales de Pimelia costata Waltl. 1835 (Coleoptera: Tenebrionidae). Nouvelle Revue d´Entomologie, 4: 299–311. Medvedev, G. S., 1965. Adaptations of leg structure in desert darkling beetles (Coleoptera, Tenebrionidae). Entomological Review, 44: 473–485. Morrone, J. J., 2006. Biogeographic areas and tran-

Pizarro-Araya et al.

sition zones of Latin America and the Caribbean islands based on panbiogeographic and cladistic analyses of the entomofauna. Annual Review of Entomology, 51: 467–494. Norgaard, T. & Dacke, M., 2010. Fog–basking behaviour and water collection efficiency in Namib Desert Darkling beetles. Frontiers in Zoology, 7: 1–23. Paskoff, R. & Manríquez, H., 2004. Las dunas de las costas de Chile. Ediciones Instituto Geográfico Militar, Santiago, Chile. Pizarro–Araya, J., 2010. Hábitos alimenticios del género Gyriosomus Guérin–Méneville, 1834 (Coleoptera: Tenebrionidae): ¿qué comen las vaquitas del desierto costero? IDESIA, 28: 115–119. Pizarro–Araya, J., Cepeda–Pizarro, J. & Flores, G. E., 2008. Diversidad taxonómica de los artrópodos epígeos de la Región de Atacama (Chile): estado del conocimiento. En: Libro Rojo de la Flora Nativa y de los Sitios Prioritarios para su Conservación: Región de Atacama: 257–274 (F. A. Squeo, G. Arancio & J. R. Gutiérrez, Eds.). Ediciones Universidad de La Serena, La Serena, Chile. Pizarro–Araya, J. & Flores, G. E., 2004. Two new species of Gyriosomus Guérin–Méneville from Chilean coastal desert (Coleoptera: Tenebrionidae: Nycteliini). Journal of the New York Entomological Society, 112: 121–126. – 2006. La posición sistemática de Geoborus lineatus (Guérin–Méneville), comb. nov. (ex. Gyriosomus) (Coleoptera: Tenebrionidae). Revista de la Sociedad Argentina de Entomología, 65: 85–90. Pizarro–Araya, J. & Jerez, V., 2004. Distribución geográfica del género Gyriosomus Guérin–Méneville, 1834 (Coleoptera: Tenebrionidae): una aproximación biogeográfica. Revista Chilena de Historia Natural, 77: 491–500. Pizarro–Araya, J., Jerez, V. & Cepeda–Pizarro, J., 2005. Descripción de huevos y larvas de primer estadio del género Gyriosomus Guérin–Méneville, 1834 (Coleoptera, Tenebrionidae, Nycteliini). Gayana, 69(2): 277–284. – 2007. Reproducción y ultraestructura del huevo y larva de primer estadio de Gyriosomus kingi (Coleoptera: Tenebrionidae) del desierto de Atacama. Revista de Biología Tropical, 55: 637–644. Schunger, I., Beutel, R. G. & Britz, R., 2003. Morphology of immature stages of Prostomis mandibularis (Coleoptera: Tenebrionoidea: Prostomidae). European Journal of Entomology, 100: 357–370. Vidiella, P. E., Armesto, J. J. & Gutiérrez, J. R., 1999. Vegetation changes and sequential flowering after rain in the southern Atacama Desert. Journal of Arid Environments, 43: 449–458. Vidal, M. A., Pizarro–Araya, J., Jerez, V. & Ortiz, J. C., 2011. Daily activity and thermoregulation in predatoreprey interaction during the Flowering Desert in Chile. Journal of Arid Environments, 75: 802–808. Whitford, W. G., 1996. The importance of the biodiversity of soil biota in arid ecosystems. Biodiversity and Conservation, 5: 185–195. – 2002. Ecology of desert systems. Academic Press, Elsevier, London, United Kingdom.


Animal Biodiversity and Conservation 34.2 (2011)

273

Do protected areas conserve neotropical freshwater fishes? A case study of a biogeographic province in Venezuela D. Rodríguez–Olarte, D. C. Taphorn & J. Lobón–Cerviá

Rodríguez–Olarte, D. Taphorn, D. C. & Lobón–Cerviá, J., 2011. Do protected areas conserve neotropical freshwater fishes? A case study of a biogeographic province in Venezuela. Animal Biodiversity ad Conservation, 34.2: 273–285. Abstract Do protected areas conserve neotropical freshwater fishes? A case study of a biogeographic province in Venezuela.— The effectiveness of protected areas to conserve freshwater fishes is limited because these areas are not usually congruent with regional patterns of fish species richness and distribution. We compared the richness, distribution and abundance of coastal freshwater fishes in a biogeographic province of Venezuela to determine their conservation status. We also estimated the relevance of existing protected areas in conserving fishes in different physiographic units and tributaries by evaluating species richness and distribution. The ichthyofauna (72 spp., ~30% endemic, ~10% threatened) was distributed according to orography, drainage and physiographic units. Most protected areas had limited effectiveness for fish conservation, mainly because they were too small or included only fragments of tributaries or drainages, or because they were located only in highland drainages where species diversity was minimal. To adequately protect freshwater fishes in this province the existing protected areas should be modified and expanded. Key words: Aquatic biodiversity, Biogeographic province, National parks, Coastal rivers. Resumen  ¿Las áreas protegidas conservan los peces continentales neotropicales? un caso de estudio para una provincia biogeográfica en Venezuela.— La efectividad de las áreas protegidas para la conservación de peces continentales es limitada ya que generalmente estas no son congruentes con los patrones regionales de la riqueza y distribución de las especies de peces. Como caso de estudio comparamos la riqueza, distribución y abundancia de la ictiofauna en ríos costeros de una provincia biogeográfica de Venezuela para determinar su estatus de conservación. Además, también estimamos la efectividad de las áreas protegidas para la conservación de la ictiofauna según la riqueza y distribución de especies en diferentes unidades fisiográficas y afluentes. La ictiofauna (72 spp., ~30% endémicas; ~10% amenazadas) se distribuyó acorde con la orografía, cuencas y unidades fisiográficas. La mayoría de áreas protegidas evidenciaron una efectividad baja para la conservación de peces, principalmente porque eran muy pequeñas o incluían sólo fragmentos de afluentes o cuencas, o porque estaban localizadas en zonas de montaña, donde la diversidad de especies era mínima. Para proteger con eficacia adecuada a los peces continentales de la provincia, las áreas protegidas existentes deberían ser modificadas y expandidas. Palabras clave: Biodiversidad acuática, Provincia biogeográfica, Parques nacionales, Ríos costeros. (Received: 15 II 10; Conditional acceptance: 16 IX 10; Final acceptance: 22 VI 11) Douglas Rodríguez Olarte, Colección Regional de Peces, Lab. de Ecología, Depto. de Ciencias Biológicas, Decanato de Agronomía, Univ. Centroccidental Lisandro Alvarado–UCLA, Ap. postal 400, Barquisimeto (Lara), Venezuela.– Donald C. Taphorn, 1822 North Charles Street, Belleville, IL, 62221, USA.– Javier Lobón Cerviá, Museo Nacional de Ciencias Naturales–CSIC, Depto. de Ecología Evolutiva, c/. José Gutiérrez Abascal 2, 28066 Madrid, España (Spain). Corresponding author: Douglas Rodríguez Olarte. E–mail: douglasrodriguez@ucla.edu.ve ISSN: 1578–665X

© 2011 Museu de Ciències Naturals de Barcelona


274

Introduction Areas protected for conservation are often set up on the basis of available regional inventories of landscapes, biogeographic patterns of terrestrial biota, or the need to protect populations of specific, usually terrestrial, species. A lack of complete biogeographic records often leads to the creation of protected areas that are later found to exclude important habitats and species, and thus limit their relevance for conservation. This lack of foresight is most evident and worrisome when considering freshwater ecosystems and their fish fauna. Information about fishes is not usually taken into consideration when designing park and refuge systems. This is paradoxical since today we know that freshwater fishes are among the most endangered species on the planet as a result of habitat loss and degradation, water pollution, species invasion and climate change (Abell et al., 2009). Terrestrial protected areas have been shown to have inadequate design and coverage to sufficiently conserve aquatic ecosystems (Herbert et al., 2010; Barletta et al., 2010). The situation is dire in the Neotropics where the largest diversity of freshwater fishes occurs, but our scant knowledge of fish taxonomy, biology and ecology hinders the design of effective strategies for their conservation. Thus, many protected areas in South America include only fragments of watersheds or streams and so fail to include essential regions necessary to guarantee the continuity of hydrosystem function and maintenance of freshwater biodiversity. Conservation biologists have a serious interest in determining how to best evaluate the effectiveness of protected areas, but freshwater hydrosystems and fishes have only recently been taken seriously into account. Newer methods now include a simple quantification of fish distribution coverage, freshwater habitat features and their relationships with attributes of protected areas or drainages (Herbert et al., 2010; Nogueira et al., 2010) and the use of rarity, vulnerability or conservation indices for fishes (Abellán et al., 2005; Bergerot et al., 2008). Most evaluations are applied to a specific tributary or drainage, and few take the regional biogeographic context into account. In biogeographic provinces, biotas have evolved together, and they show patterns and gradients of species richness and distribution that differ at different scales (Whittaker & Fernández–Palacios, 2007). Evaluations of protected area systems that take biogeographic context into consideration will provide more useful information for effective conservation. The Western Caribbean province (wcp) –a zoogeographic unit proposed to delimit the freshwater fishes of coastal Caribbean drainages in Venezuela– includes streams that originate in the high Andes Mountains, the arid hills of the Coriano range and humid valleys in the limits of the Coastal range (Rodríguez–Olarte et al., 2009). This set of drainages comprises only around 30,000 km2 but it includes a remarkable variety of landscapes and biotas, reflected in its species–rich freshwater fish fauna and its many

Rodríguez-Olarte et al.

endemics. Several protected areas have been created in the wcp (abrae, or areas under special administration regime). These include national parks and natural monuments. In our study area these comprise mainly mountain regions, a few wildlife refuges in lowland areas near river deltas, and areas adjacent to marine parks. Also, in the wcp, two areas of special interest for the conservation of biodiversity come together: the Tropical Andes Hotspot (Myers et al., 2000), which includes several terrestrial ecoregions (Olson & Dinerstein, 2002) and the Caribbean Freshwater Ecoregion of South America and Trinidad (Abell et al., 2008). Recent work in the wcp shows that most aquatic ecosystems and their fishes are at risk and several species have been listed as threatened (Rodríguez & Rojas–Suárez, 2008; Rodríguez–Olarte et al., 2007; 2009). Since the distribution patterns of fish species vary within drainage and even more so within a biogeographic province, we hypothesised that the current protected area network of this biogeographic province did not adequately protect the variety of aquatic ecosystems and the fishes found there. To design a project that would systematically document this, we evaluated the protected areas of the wcp as a case study, with respect to the freshwater fishes, incorporating both historical and recent records for fish species distribution. Methods Historical records, fish sampling and geographical data We used museum collections (cpucla, mbucv, mcng and mhnls) and published accounts to obtain historical records of fish distribution in the province (Reis et al., 2003; Rodríguez–Olarte et al., 2009). We also consulted records from databases (Froese & Pauly, 2010). We estimate that the historical records thus obtained would be sufficient in the wcp to delimit species distributions. We also considered that data on fish abundance of historical records are from sufficiently long–term observations to infer tendencies in variation for most species and most drainages. We collected standardized samples from 32 different sites (fig. 1) among drainages of the Andes (Tocuyo), Coastal (Aroa, Yaracuy, Urama) and Coriano mountain ranges (Ricoa, Hueque, Coro, Mitare and Tucurere). At each locality, the sampling transect was about 50 m long and less than 1.5 m deep. We used electroshocking gear with hand nets and seines to capture fishes in three successive passes (Lobón–Cerviá, 1991). We collected samples (n = 120) from foothills and mountain streams, concentrating efforts towards the end of the wet and dry seasons (September–October and February–March) from 2005 to 2007. We also collected fish samples at other sites (207 localities, fig. 1), most frequently from the foothill floodplains and river mouths, using non–standardized methods, electroshockers, seines (5 x 2 m, 0.5 mm mesh),


Animal Biodiversity and Conservation 34.2 (2011)

275

Caribbean Sea 7

8

5

6 9 1

C

Golfo Triste 2 3

B 3

4

1

m a.s.l. 2

0–100 100–250 250–500

Orinoco Basin A

0

50

100 km

500–1,000

Lakes and reservoirs Only historic records

1,000–2,000 2,000–­3,500

Fig. 1. Western Caribbean province of Venezuela (WCP): ● Historical records sites; ∆ Localities where standardized sampling was applied from 2005 to 2007; j Main cities (1. Caracas; 2. Valencia; and 3. Barquisimeto). The drainages are: 1. Tocuyo; 2. Aroa; 3. Yaracuy; 4. Urama; 5. Ricoa; 6. Hueque; 7. Coro; 8. Mitare; and 9. Tucurere. Mountains shown are: A. Andes; B. Coastal range; C. Coriano range. Fig. 1. Provincia Caribe Occidental de Venezuela (WCP): ● Registros históricos; ∆ Localidades donde se aplicaron muestreos estandarizados del 2005 al 2007; j Principales ciudades (1. Caracas; 2. Valencia y 3. Barquisimeto). Las cuencas son: 1. Tocuyo; 2. Aroa; 3. Yaracuy; 4. Urama; 5. Ricoa; 6. Hueque; 7. Coro; 8. Mitare y 9. Tucurere. La orografía regional está representada por: A. Los Andes; B. Cordillera de la Costa y C. Sistema Coriano.

hand nets, pneumatic harpoons, traps, and hook and line. These samples were also considered as historical records. In a few of the smallest drainages, especially those of the Coastal range (fig. 1), standard fishing methods could not be applied, so we used only historical records. Fishes were usually identified and counted in the field and returned live, but in some cases representative samples were preserved in 10% formalin, later transferred to 70% ethanol, and deposited in the Colección Regional de Peces (cpucla–ucla) and the Museo de Ciencias Naturales Guanare (mcng– unellez) following identification. The ABRAE that we evaluated here are those designated principally for conservation: national parks, natural monuments, wildlife refuges and fauna reserves. We classified and quantified the abrae in the province according to type, location, surface area, coverage of physiographic units (lowland floodplains, foothills or mountains), tributaries and drainages. Geographic data were taken from Rodríguez et al. (2004), Lehner et al. (2006) and http://www.feow.org.

Data analysis Conservation status of the fish fauna Species were classified by distribution type: endemic to the province, or occuring in other Caribbean and/ or the Orinoco drainages. We used relative abundance from all samples standardized from 2005 to 2007, but for other samples we used five categories of abundance: abundant, common, scarce, rare and very rare. Conservation status follows IUCN criteria (IUCN, 2001, 2003, 2006), but because information needed to accurately classify many species of fish from the wcp is still lacking, non–subjective assignation to a threat category was often difficult. To assign a species to a category ('critically endangered', 'endangered', 'vulnerable', 'near threatened', 'of least concern' and 'data deficient') we used the following criteria: (a) decreasing populations, (b) size of the geographic area of distribution, (c) small population size, (d) very small populations or distribution area


276

Rodríguez-Olarte et al.

Table 1. Conservation status of freshwater fishes in WCP: EN. Endemic species to the province (E) and/ or that occur in Caribbean (C) or Orinoco (O) drainages; DI. Endemic species classified as occurring in Andean (●), Coriano (▲) or Coastal mountain ranges (■); AB. Abundance (expressed as a relative proportion for standardized samples; in non–standardized samples, abundance classes were assigned as abundant [a], common [c], scarce [e], rare [r] and very rare [vr]); TC: threat categories: critically endangered (CR), endangered (EN), vulnerable (VU), near threatened (NT), of least concern (LC) and data deficient (DD). Non–native species (nn) were not included in the categories. Tabla 1. El estatus de conservación de los peces continentals en WCP: EN. Especies endémicas a la provincia (E) y/o que habitan en cuencas del Caribe (C) u Orinoco (O); DI. Especies endémicas según su presencia en los sistemas orográficos (Andes (●), sistema Coriano (▲) y cordillera de la Costa (■); AB. Abundancia (proporción relativa en muestras estandarizadas; en muestras no estandarizadas, las tipologías de abundancia son: abundantes [a], comunes [b], escasas [e], raras [r] y muy raras [vr]); TC. Se asignan los tipos de amenaza: en peligro crítico (CR), en peligro (EN), vulnerable (VU), casi amenzado (NT), preocupación menor (LC) y datos insuficientes (DD). Las especies no nativas (nn) no fueron incluidas dentro de las categorías.

Species

EN DI AB TC

Species

EN DI AB TC CO

Astyanax viejita

C

9.8 LC

Chaetostoma milesi

Astyanax magdalenae

C

0.8 LC

Chaetostoma sp. Alto Tocuyo E

0.1 DD 8.4 LC

Astyanax metae

O

4.2 LC

Chaetostoma stanni

E

Astyanax venezuelae

O

0.2 LC

Chaetostoma yurubiense

E

0.3 LC

Bryconamericus alpha

O

1.3 LC

Chaetostoma sp. Tocuyo N E

1.0 LC

Bryconamericus charalae

E

e DD

Farlowella mariaelenae

Bryconamericus cismontanus O

12 LC

Farlowella martini

C

c DD

Farlowella vittata

Bryconamericus loisae Creagrutus crenatus

E

Creagrutus hildebrandi

C

● 0.9 LC 0.8 LC

●■ 4.6 LC

CO < 0.1 DD E

CO

Hypostomus pagei

E

Rineloricaria rupestris

C

0.2 NT r

DD

●■ 1.7 LC 0.9 LC

Creagrutus lassoi

E ●■ 6.6 LC

Pimelodus blochii

O < 0.1 nn

Creagrutus lepidus

E

Batrochoglanis mathisoni

E

■ 0.2 VU

●■ < 0.1 EN

CO

1.1 LC

Trichomycterus arleoi

E

O

– nn

Trichomycterus kneri

C

Gephyrocharax melanocheir C

3.4 LC

Creagrutus melasma Colossoma macropomum Gephyrocharax valencia

O < 0.1 LC

0.9 LC 0.6 LC

Trichomycterus sp. Tocuyo E

0.8 LC

Apteronotus sp. Yaracuy

0.1 NT

E

Gephyrocharax venezuelae C

2.0 LC

Gymnotus carapo

CO < 0.1 DD

C

8.6 LC

Brachyhypopomus diazi

CO < 0.1 NT

Hemibrycon jabonero

Hyphessobrycon fernandezi E ●■ < 0.1 NT

Oncorhynchus mykiss

Hyphessobrycon sovichthys C

Poecilia caucana

C

Poecilia dauli

C

Poecilia koperi

C < 0.1 DD

Hyphessobrycon paucilepis E Nanocheirodon insignis

e DD

● < 0.1 DD

C

e DD

c

nn

0.4 LC d

DD

Roeboides dientonito

CO

0.6 LC

Poecilia reticulata

O

Characidium chupa

CO

1.2 LC

Poecilia sphenops

C

r

DD

Steindachnerina argentea

CO < 0.1 LC

Pseudolimia heterandria

C

c

DD

Hoplias malabaricus

CO

0.2 LC

Cyprinodon dearborni

C

c

LC

Lebiasina erythrinoides

CO

3.0 LC

Austrofundulus leohoignei

E

vr

CR

Parodon apolinari

CO

1.3 LC

Rivulus hartii

C < 0.1 LC

O

– nn

Prochilodus mariae Corydoras venezuelanus

CO < 0.1 LC

Hoplosternum littorale

O

Cetopsis orinoco

O < 0.1 DD

e DD

●■

2.0 LC

Synbranchus marmoratus CO < 0.1 LC Andinoacara pulcher

C

5.5 LC

Caquetaia kraussii

C < 0.1 nn

Cichla sp.

O

nn


Animal Biodiversity and Conservation 34.2 (2011)

277

Table 1. (Cont.)

Species Cetopsorhamdia sp. Aroa

EN DI AB TC

Species

EN DI AB TC

E

■ < 0.1 DD

Crenicichla geayi

O

Cetopsorhamdia sp. Tocuyo E

● 0.2 DD

Geophagus sp.

O

Pimelodella odynea Rhamdia quelen Rhamdia guatemalensis Ancistrus gymnorhynchus Ancistrus falconensis Ancistrus triradiatus

2.3 LC nn

C

3.2 LC

Oreochromis sp.

nn

CO

2.2 LC

Eleotris pisonis

C

c

DD

C

c LC

Awaous banana

C < 0.1 LC

CO

2.8 LC

Sicydium plumieri

C

Agonostomus monticola

C < 0.1 DD

E ▲ 2.6 LC

r

LC

CO < 0.1 LC

and (e) analysis of threat of extinction. Because biological and ecological data (growth, reproduction, diet, etc.) were not available for most species we used information for similar species, field observations and historical records instead. Demographic data on fish populations are not availble to assess the imminence of extinction from natural habitat (criterion e). However, we consider that the viability of populations of fishes can be inferred from information about their very restricted distribution, and/or by evaluating the threat of habitat destruction. Biogeographical patterns For the biogeographical analyses we used only the standardized samples from rivers considered to be in pristine condition, according to our experience (17 localities; 64 collections). To compare the association of species with orography, drainages and physiographic units across the wpc we used a non–metric multidimensional scaling analysis (nms) based on the Bray–Curtis distance measure and applied to ln (x + 1) transformed data. This analytical method represents relationships between samples in a similarity matrix (Clarke & Warwick, 1994). The graphic was rotated for easier comparison. The robustness of the nms ordination was indicated by the average stress values based on a two–dimensional solution, and a Monte Carlo stress test of randomized data was done with 999 permutations. Since the conservation status of each species is linked to its distribution and associated with certain habitats or drainages, we classified distributions as occurring in physiographic units, and determined the association of these with respect to drainages with Indicator Species Analysis (ISA, Dufrêne & Legendre, 1997). Here, this analysis provides an indicator value, associated with a probability, for each species with respect to a physiographic unit or drainage. Significance was evaluated using a Monte Carlo test (10,000 permutations). The nms and ISA were evaluated using the program pc–ord 4.41 (McCune & Mefford, 1999).

Effectiveness of protected areas To evalutate the significance of species richness in protected areas network we used species–area relationships. We related the accumulated area of the abrae in chronological order according to the date of its creation and the number of species accumulated therein (recorded or estimated). We also evaluated the coverage and borders of hotspots and terrestrial and aquatic ecoregions with respect to the distribution of freshwater fishes. We classified the effectiveness of abrae by creating relative coverage indices for (a) physiographic units, (b) tributaries and (c) species richness. In the first case we estimated that abrae that included a wider variety of physiographic units would have a higher relevance for conservation. We assigned the following indices for these: lowland floodplains (60%); foothills (30%) and mountains (10%). An index value of maximum coverage, for example, suggests that an abrae includes all three physiographic units. In the same way we estimated that an abrae that included an entire tributary from its origin to its confluence with a major river or the sea would have greater impact for conservation of freshwater fishes than an abrae that only included fragments of tributaries, independent of physiographic units. Therefore, we rated abraes by comparing the total length of the major tributary associated with an abrae with the length of the tributary included within the boundaries of that abrae. In this index of tributary coverage, a maximum value indicated that the entire length of the major river associated with an abrae occurred within its boundaries. To estimate the abrae coverage of species richness we compared total species richness of the principal tributary associated with an abrae to the number of species found within its boundaries. Here, maximum values indicated that all species present in a river could be found within the abrae. The sum of these three components –physiographic units, tributaries and species richness– was the value assigned to each abrae to rate its effectiveness for the conservation of freshwater fishes. These values


278

Rodríguez-Olarte et al.

Stress: 0.19

Coro Ricoa

3

NMS axis 2

Tucurere Hueque 1

Yaracuy 2 Aroa

Tocuyo Urama Hueque NMS axis 1 Fig. 2. A non–metric multidimensional scaling (nms) applied on samples from rivers considered to have pristine conditions performed in the wpc from 2005 to 2007. The results suggest that the ichthyofauna is associated with the orography, drainages and altitude. The broken lines separate the drainages according to orography (1. Andes; 2. Coastal range; 3. Coriano range), except in the Hueque drainage. Symbols containing a black dot are localities in mountains; the rest are mostly situated in the foothills. Fig. 2. Escalado multidimensional no métrico (nms) para las muestras en ríos de condiciones originales en la wpc desde 2005 hasta 2007. Los resultados sugieren que la ictiofauna está asociada a la orografia, las cuencas y la altitud. Las líneas punteadas separan las cuencas según la orografía (1. Andes; 2. Cordillera de la Costa; 3. Sistema Coriano), excepto en la cuenca Hueque. Los símbolos que contienen un punto negro son localidades en montañas y el resto se sitúan generalmente en piedemontes.

were rated into four categories: very low (< 75%), low (75–150%), medium (151–225%) and high effectiveness (> 225%). Results About 30% of the 72 species recorded are endemic, and most of them are from Andean or Coastal ranges. The families with the most species are Characidae (24 spp.) and Loricariidae (13 spp.). About half of the loricariids are endemic (table 1). Locality records showed that several endemic species were restricted to very small drainages (< 5,000 km2) or physiographic units, usually small mountain watersheds, and others were found only in temporary rain pools in the lowlands. The drainages of the Tocuyo, Aroa and Yaracuy Rivers were recognized as having the highest species richness and endemism in the province, and as such are of highest priority for conservation. Most of the more abundant species (Bryconamericus cismontanus, Astyanax viejita, Hemibrycon jabonero) maintained high abundances and many occurred

throughout the wcp. However, some endemic species with restricted distribution had high local abundance (Chaetostoma sp. Alto Tocuyo, Creagrutus lassoi). The annual killifish Austrofundulus leohoignei, which lives in temporary rain pools in the lowlands, had the smallest known distribution of all the species studied. The pseudopimelodid catfish, Batrochoglanis mathisoni which lives among rocks in clear foothills and flooplain streams was very rarely collected (table 1). Most species fell into the conservation category 'of least concern' (63%); they had widespread distribution in one or more drainages, and they also occurred outside the province in some cases. The frequency of appearance of these species in the samples and their abundance did not show notable variation. In the category of 'near threatened' we included Brachyhypopomus diazi (table 1), which is found in the Aroa and Yaracuy drainages. A. leohoignei, known from only a few temporary rain pools in the lowlands of the Aroa and Tocuyo drainages was classified as 'critically endangered', and B. mathisoni was classified as 'endangered'. We classified about 28% of the species encountered as 'data deficient', but several


Animal Biodiversity and Conservation 34.2 (2011)

279

Table 2. The associations of species with particular drainages were statistically significant (P) according to indicator values (IV) of the indicator species analysis. The distributions were also related to orography: MR. Mountain ranges; E. Endemic species; M. Species that occur in the mountains. Tabla 2. La asociación de especies respecto a las cuencas, según los indicadores (IV) en el análisis de especies indicadoras, tuvo significancia estadística (P). Las distribuciones también estuvieron relacionadas con la orografía: MR. Cadenas montañosas; E. Especies endémicas; M. Especies que habitan en las montañas. MR

Drainages

Species

IV

P

Coriano

Tucurere

Astyanax magdalenae

62

0.004

Gephyrocharax venezuelae

62

0.002

Synbranchus marmoratus

48

0.007

Coro

Creagrutus hildebrandi

70

0.001

Mitare

Agonostomus monticola

50

0.020

Bryconamericus cismontanus M

34

0.001

Hypostomus pagei E

47

0.007

Rineloricaria rupestre

45

0.019

Hueque

Ancistrus falconensis E, M

41

0.023

Ricoa ­–

Characidium chupa M

53

0.002

Chaetostoma stanni E

31

0.046

Trichomycterus arleoi E, M

35

0.020

Coastal Urama

Apteronotus sp. Yaracuy E

35

0.027

Creagrutus lassoi E, M

35

0.031

Aroa

Chaetostoma yurubiense E, M

32

0.033

Tocuyo

Chaetostoma sp. Alto Tocuyo E, M

40

0.025

Andean

Yaracuy

of these are widespread throughout the province. We found eight introduced species; these were usually found in reservoirs and artificial ponds, but P. mariae (Prochilodontidae) and P. blochii (Pimelodidae) also occurred in the main channels of some rivers. Species association with orography, drainages and physiographic units was evident. The nms ordination accounted for 74.8% of the variance (axis 1: 44.8%; axis 2: 30%; fig. 2). The final mean stress of 0.19% indicated a potentially useful two–dimensional solution. The ordination evaluation recognized patterns of the ichthyofauna in orographic scale, and also for species restricted to mountains. About 20% of the fishes in the wcp –including about half of the endemics– showed significant indication values for almost all drainages (table 2). Of these species, about half occurred in mountains. Although our analysis showed that a little more than 17% of the entire surface area of the wcp was included in some sort of abrae, nearly 90% of these protected areas were small (less than 500 km2) and mainly included mountain regions (table 3). The drain-

ages of the Coastal range had nearly one third of their area included in protected areas, but this coverage was only about 10% of the Coriano range, and even less, about 6%, of the Andean drainages. Thus, in our study area most abrae protect highlands. Species richness did not necessarily increase with increased surface area of the abrae. The relationship between the accumulated total number of species and the accumulated area of the abrae was expressed as a curve that tended to become saturated as we neared 83% of the total species richness recorded in the wcp (fig. 3A). Excluding the abrae located in mountain regions and/or of small size, we found a direct relationship between the number of species recorded for each drainage and the number of species found in the river stretch or affluent under protection. This relationship was expressed as a linear function with high significance (R2 = 0.88; P = 0.0006, fig. 3B). Including different physiographic units revealed the effectiveness of the abrae. Lowlands were the physiographic unit with least protection and highest species diversity. Tributaries entirely included within


280

Rodríguez-Olarte et al.

Table 3. Protected areas in the Western Caribbean Province (WCP) and their effectiveness for the conservation of fishes. The coverage for drainages refers to the percentage of a local drainage covered by each ABRAE. The coverage for physiographic units (column I) are plains (A), foothills (B) and mountains (C). Coverage for tributaries is given in column II. The coverage for richness (column III) is expressed in St (total number of species) and Se (total number of endemic species) recorded throughout the course of the main tributary associated with each ABRAE. Sp is the number of species estimated to occur within the protected area. The coverage values in parentheses are percentages and the effectiveness of each ABRAE for the conservation of fishes is the sum of the coverages (Σ  =  I  +  II + III).

ABRAE

Area (km ) 2

Altitude (m a.s.l.)

Coverages

Drainages (%)

Parks

Henri Pittier 1937

1,078

0–2,436

Coastal range (95)

Ávila 1958

819

120–2,765

Coastal range (90)

Yurubí 1960

237

150–1,950

Aroa (7), Yaracuy (6)

Yacambú 1962

269

500–2,280

Tocuyo (< 5)

Quebrada del Toro 1969

49

400–1,120

Tocuyo (< 1), Hueque (< 1)

Morrocoy 1974

321

0–285

Juan Crisóstomo F. 1987

191

200–1,500

Aroa (< 5), Tocuyo (< 1)

Hueque (< 5), Ricoa (17),

Coro (15), Mitare (< 1)

San Esteban 1987

431

0–1,830

Coastal range (80)

Dinira 1988

459

1,800–3,585

Tocuyo (< 5)

Saroche 1989

346

500–1,300

Tocuyo (< 5)

María Lionza 1960

158

210–1,205

Yaracuy (< 5)

Pico Codazzi 1991

119

600–2,429

Coastal range (40)

118

0–285

Tocuyo (< 1)

178

0–40

Tucurere (43), Tocuyo (< 1)

372

0–20

Hueque (9)

Monuments

Refuges Cuare 1972 Reserve Tucurere 2001 Hueque–Sauca 2005 Total ABRAE

5,145

an abrae were quite rare, and those that did exist were usually very small drainages that emptied directly into the sea. The average coverage of total species richness protected by an abrae did not exceed 50% (table 3). The abrae with the smallest percentage of fish species protected were the smaller ones located in the mountains, but these abrae included most of the endemic species known from this physiographic unit. On the other hand, several endemic species that occurred on floodplains or in foothills tributaries had only a very small portion of their known distribution areas protected by existing abrE.

WPC (17.2)

The only abrae with a high score of effectiveness for freshwater fishes was the Henry Pittier National Park (Coastal range) which contained some 30 species in its major river, with most of those restricted to the lower stretch of the river. Several abrae in the wcp were classified as having low and very low effectiveness for conservation of freshwater fishes. The abrae associated with drainages or tributaries with higher species richness and endemism were found to have medium effectiveness for fish conservation (Aroa and Yaracuy River drainages) and protect about half of the freshwater species found in those drainages. The abrae with less restrictive protection rules, such


Animal Biodiversity and Conservation 34.2 (2011)

281

Tabla 3. Las áreas protegidas en la Provincia Caribe Occidental (WCP) y su efectividad para la conservación de la ictiofauna. La cobertura de las cuencas se refiere al porcentaje de la cuenca local cubierto por cada ABRAE. La cobertura de unidades fisiográficas (columna I) se expresa en planicies (A), estribaciones (B) y montañas (C). La cobertura de afluentes se da en la columna II. La cobertura de la riqueza (columna III) es expresada mediante el número total de especies (St) y el número total de especies endémicas (Se) registradas en todo el curso del principal afluente asociado a cada ABRAE. Sp es el número de especies que ocurren dentro del área protegida. Los valores de cobertura en paréntesis son los porcentajes y la efectividad de cada ABRAE para la conservación de la ictiofauna es la suma de las coberturas (Σ = I + II + III). Coverages

I

II

III

Efffectiveness

St

Se

Sp

Σ

Class

BC(40)

(100)

30

4

30(100)

240

High

BC(40)

(90)

30

2

22(73)

203

Medium

BC(40)

(80)

48

8

35(73)

193

Medium

C(10)

(20)

16

2

4(25)

55

Very low

C(10)

(20)

35

4

8(23)

53

Very low

B(30)

(100)

18

3

8(44)

170

Medium

C(10)

(40)

18

3

6(33)

83

Low

BC(40)

(90)

30

4

28(93)

223

Medium

C(10)

(20)

36

2

2(6)

36

Very low

BC(40)

(3)

8

1

1(13)

56

Very low

BC(40)

(100)

42

6

28(67)

207

Medium

C(10)

(20)

30

1

2(7)

37

Very low

AB(90)

(20)

28

2

12(43)

153

Medium

AB(90)

(60)

25

1

18(72)

222

Medium

AB(90)

(60)

22

1

15(68)

218

Medium

as the wildlife refuges and faunal reserves, covered part of the lowland floodplains but since they offered modest protection, they were classified as being of medium effectiveness. Discussion Of all the species of fishes included in the different threat categories for Venezuela (Rodríguez & Rojas– Suárez, 2008) a little more than 90% occur only in the coastal Caribbean drainages, and nearly 20% of those occur only in the wcp. Although many species

are not currently included in any iucn category, in the near future, many of them face real threats. For example, several species of variable abundance but endemic to one drainage, physiographic unit or tributary (Bryconamericus charalae, Creagrutus crenatus, Hyphessobrycon paucilepis, Chaetostoma yurubiense) were included in the categories of least concern' or 'data deficient'. This situation, a result of our lack of biological and biogeographic information, prohibits a strict application of the protocol for conservation classification. We believe that this is the case for many other compendia of endangered species for South American countries, such as Colombia (Mojica


282

Rodríguez-Olarte et al.

A

Species by affluent in ABRAE

Species accumulated

70 60 50 40 30 20 10 0

0

10 20 30 40 50 Area accumulation by ABRAE (x 100 km2)

70

B

60

R2 = 0.88

50 40 30 20 10 0

0

10 20 30 40 50 60 70 Species by affluent in drainage

Fig. 3. A. Species and area accumulation in the abrae. Only national parks (●) and natural monuments (▲) were plotted against the cumulative number of fish species under protection. B. Relation between total number of species by tributary in the drainage as compared to those in an abrae in the same drainage. The broken line in B corresponds to a linear regression (R2 = 0.88) excluding very small abrae and/or those located in mountains (∆,○). Fig. 3. A. Acumulación de especies y áreas en las abrae. Sólo se consideraron los parques nacionales (●) y los monumentos naturales (▲) para construir la curva acumulativa de especies de peces protegidas. B. Relación entre el número total de especies por afluente en la cuenca y las especies de la misma cuenca que se encuentran dentro de una abrae. En B la línea punteada corresponde a una regresión lineal (R2 = 0.88) excluyendo las abrae muy pequeñas y/o localizadas en montañas (∆,○).

et al., 2002), Venezuela (Rodríguez & Rojas–Suárez, 2008) and Brazil (Rosa & Lima, 2008). The pace and intensity of environmental perturbations and their effects on Neotropical fishes are still only partly evaluated and understood (Winemiller et al., 1996). The individual effects of multiple combined impacts acting on the same body of water are often difficult to assess and to replicate. The richness and distribution of freshwater fishes needs to be taken into account to evaluate the utility of areas with special interest for the conservation of biodiversity. Dramatic variation detected in patterns of fish species richness, distribution and endemism associated with gradients of orography, physiographical units and hydrographic drainages indicates effects derive from many sources, such as altitude, latitude, and historical geological processes with associated ecological differences, all of which act at different scales. For freshwater fish faunas of Caribbean slopes, regional boundaries generally coincide with the ecoregion for freshwater drainages of Caribbean South America and Trinidad proposed by Abell et al. (2008) and follow limits imposed by natural geographic characteristics of the region. However, these freshwater ecoregions did not discriminate local variability in patterns of species richness or distribution details for the species of fishes of the Caribbean provinces (fig. 4A). This also

occurred with most terrestrial ecoregions occurring in the wcp, such as xeric scrublands and montane forests of the Coastal range (Rodríguez et al., 2004) that cover fragments of the drainages where we found high species richness and endemism (fig. 4B). Similarly, the Tropical Andes hotspot (Myers et al., 2000) covers about 10% of the wcp, and was restricted to the high Andean and Coastal ranges where elevated endemism of terrestrial plants and animals has been reported, but it omits lowland floodplains where most fish species richness occurs. Most abrae in the Western Caribbean Province of Venezuela do not adequately protect the variety of aquatic ecosystems present in the region, so they cannot guarantee the continuity of the hydrobiological processes required for their conservation. Most ABRAE cover highland, arid desert landscapes, or small fragments of tributaries and drainages. Although the streams protected in the highlands offer direct protection to the species that occur there and indirect help for species occuring downstream of protected areas, they cannot provide protection for impacts outside their boundaries. Such impacts are common downstream and include, for example, channelization, dredging, and water extraction, all of which severely degrade fluvial ecosystems and their fish faunas (Rodríguez–Olarte et al., 2006). Alteration


Animal Biodiversity and Conservation 34.2 (2011)

283

A

B

Freshwater ecoregions

Terrestrial ecoregions

1 2

1

3 4

9 8

3 5

2

4

8 7

6

5

C Pertinence of ABRAE in WCP Very low

a

Low

b

1

Medium

c

2

High

3 9

6

8

4

10 d

11

5 7 Other ABRAE

Fig. 4. Areas of interest for conservation associated with the Western Caribbean Province: A. Freshwater ecoregions (1. South America Caribbean Drainages–Trinidad; 2. Maracaibo; 3. Orinoco High Andes; 4. Orinoco Llanos; and 5. Orinoco Foothills); B. Terrestrial ecoregions (1. Coastal Venezuelan mangroves; 2. Paraguaná xeric scrub; 3. Maracaibo dry forests; 4. Catatumbo moist forests; 5. Venezuelan Andes montane forest; 6. Llanos; 7. La Costa xeric scrublands; 8. Cordillera de La Costa montane forests; and 9. Lara–Falcón dry forests); the limits of the Tropical Andes hotspot are similar to terrestrial ecoregions 5 and 8; C. Protected areas abrae (1. Juan Cristóstomo Falcón; 2. Cueva de la Quebrada del Toro; 3. Morrocoy; 4. Yurubí; 5. Maria Lionza; 6. Saroche; 7. Yacambú; 8. Dinira; 9. San Esteban; 10. Henry Pittier; 11. Waraira Repano; 12. Hueque–Sauca [a], Tucurere [b] Golfete de Cuare [c] and Pico Codazzi [d]). Fig. 4. Áreas de interés para la conservación asociadas a la Provincia Caribe Occidental: A. Ecorregiones de aguas continentales (1. Cuencas de Sudamérica al Caribe y Trinidad; 2. Maracaibo; 3. Altos Andes del Orinoco; 4. Llanos del Orinoco y 5. Piedemonte del Orinoco); B. Ecorregiones terrestres (1. Manglares costeros de Venezuela; 2. Matorral xerófilo de Paraguaná; 3. Bosques secos de Maracaibo; 4. Bosques húmedos de Catatumbo; 5. Bosques montanos de los Andes de Venezuela; 6. Llanos; 7. Matorral xerófilo de la Cordillera de la Costa; 8. Bosques montanos de la Cordillera de la Costa y 9. Bosques secos de Lara–Falcón); los límites de los áreas clave de biodiveridad de los Andes Tropicales son similares a los de las ecorregiones 5 y 8; C. Áreas protegidas abrae (1. Juan Cristóstomo Falcón; 2. Cueva de la Quebrada del Toro; 3. Morrocoy; 4. Yurubí; 5. Maria Lionza; 6. Saroche; 7. Yacambú; 8. Dinira; 9. San Esteban; 10. Henry Pittier; 11. Waraira Repano; 12. Hueque–Sauca [a], Tucurere [b] Golfete de Cuare [c] y Pico Codazzi [d]).


284

of the natural flow regime affects the movements and migrations of fishes, which in turn greatly impacts local species richness gradients (Grossman et al., 2010). This could diminish the local conservation effects of a protected area, since fish diversity within its borders depends upon the conservation status of streams both upstream and downstream of their boundaries. Floodplain abrae are particularly vulnerable to disturbances that occur in headwaters. The existing protected areas in the study area offer protection to a little more than 80% of the fishes reported for the province. However, this is not an indication of high effectiveness of these abrae regarding the protection of freshwater fishes, and should be considered with caution. Existing abrae do offer partial protection to the fishes in the stream fragments they include (fig. 4C), but as we have shown, they do not effectively offer long term protection to the freshwater fishes of the region, and should be expanded or modified. Areas with high priority for conservation identified in this study are offered only partial protection by existing abrae (e.g. Yurubí National Park, table 3) but this protection does not cover the physiographic units where highest fish species richness and endemism occur. Even so, the effectiveness of these abrae would be considerably increased if they were expanded to include the entire lengths of the tributaries, currently only partially protected within them. The scenario is different for mountain abrae because there is very low species diversity but high endemism. In these cases, protection of longer stretches would not necessarily increase the number of species under protection. These changes might occur in an ideal scenario, but the current condition of terrestrial ecosystems and established land–use patterns limit the implementation of protected areas for the conservation of aquatic resources. However, to better conserve even heavily used and limited water resources, we can promote management practices at a local scale such as by implementing reforestation of buffer strips or green belts along the shores (see Saunders et al., 2002), and concentrating efforts into smaller projects that focus on local objectives. This approach, as opposed to those that might require extensive areas set aside for conservation purposes, could be more successful in economically challenged areas with limited resources. We thus suggest that wcp programs that protect specific tributaries within aquatic refuge areas should extend the protective influence of already existing abrae as this would significantly improve protection of freshwater fish diversity. Acknowledgements This study was financed by grants from cdcht–ucla (Projects 001–dag–2005) and iea–provita (Project 2006–08) and support by Colección Regional de Peces (cpucla–ucla). We greatly appreciate the help provided by Ahyran Amaro, Héctor Rivera, J. Luis Coronel, David Alfaro, Carlos E. López, Henry Agudelo and Jacinto López Velasco. We thank Francisco

Rodríguez-Olarte et al.

Provenzano (mbucv–ucv), Carlos y Oscar Lasso (mhnls–flasa) and Marcos Campo (ebrg–marn) for allowing access to their institutions´ databases for records of freshwater fish distribution. We also thank Kirk Winemiller for useful comments on improving the manuscript. References Abell, R., Blanch, S., Revenga, C. & Thieme, M., 2009. Conservation of aquatic ecosystems. In: Encyclopedia of Inland Waters (1): 249–258 (G. E. Likens, Eds.) Oxford: Elsevier. Abell, R., Thieme, M. L., Revenga, C., Bryer, M., Kottelat, M., Bogutskaya, N., Coad. B., Mandrak, N., Balderas, S. C., Bussing, W., Stiassny, M. L. J., Skelton, P., Allen, G. R., Unmack, P., Naseka, A., Ng, R., Sindorf, N., Robertson, J., Armijo, E., Higgins, J. V., Heibel, T. J., Wikramanayake, E., Olson, D., López, H. L., Reis, R. E., Lundberg, J. G., Sabaj Pérez, M. H. & Petry, P., 2008. Freshwater Ecoregions of the World: A New Map of Biogeographic Units for Freshwater Biodiversity Conservation. BioScience, 58(5): 403–414. Abellán, P., Sánchez–Fernández, D., Velasco, J. & Millán, A., 2005. Conservation of freshwater biodiversity: a comparison of different area selection methods. Biodiversity and Conservation, 14: 3457–3474. Barletta, M., Jaureguizar, A. J., Baigun, C., Fontoura, N. F., Agostinho, A. A., Almeida–Val, V., Val, A., Torres, R. A., Jimenes, L. F., Giarrizzo, T., Fabré, N. N., Batista, V., Lasso, C., Taphorn, D. C., Costa, M. F., Chaves, P. T., Vieira, J. P. & Corrêa, M. F. M., 2010. Fish and aquatic habitat conservation in South America: a continental overview with emphasis on neotropical systems. Journal of Fish Biology, 76(9): 2118–2176. Bergerot, B., Lasne, E., Vigneron, T. & Laffaille, P., 2008. Prioritization of fish assemblages with a view to conservation and restoration on a large scale European basin, the Loire (France). Biodiversity and Conservation, 17: 2247–2262. Clarke, K. R. & Warwick, W. M., 1994. Similarity–based testing for community pattern: the 2–way layout with no replication. Marine Biology, 118: 167–176. Dufrêne, M. & Legendre, P., 1997. Species assemblages and indicator species: The need for a flexible asymmetrical approach. Ecological Monographs, 67(3): 345–366. Froese, R. & Pauly, D., 2010. FishBase. http://www. fishbase.org. Cited 15 Dic 2010. Grossman, G. D., Ratajczak, R. E., Farr, M. D., Wagner, C. M. & Petty, J. D., 2010. Why There Are Fewer Fish Upstream. American Fisheries Society Symposium, 73: 63–81. Herbert, M. E., Mcintyre, P. B, Doran, P. J., Allan, J. D. & Abell, R., 2010. Terrestrial reserve networks do not adequately represent aquatic ecosystems. Conservation Biology, 24(4): 1002–1011. iucn, 2001. IUCN Red list categories and criteria: Version 3.1. IUCN. Species Survival Commission.


Animal Biodiversity and Conservation 34.2 (2011)

IUCN, Gland, Switzerland and Cambridge, UK. – 2003. Guidelines for application of IUCN red list criteria at regional Levels: Version 3.0. IUCN Species Survival Commission. IUCN, Gland, Switzerland and Cambridge, UK. – 2006. Guidelines for using the IUCN Red List categories and criteria. Version 6.2. Species Survival Commission, IUCN, Gland, Switzerland and Cambridge, UK. Lehner, B., Verdin, K. & Jarvis, A., 2006. Hydrosheds Technical Documentation. World Wildlife Fund US, Washington, DC. http://hydrosheds.cr.usgs.gov. Cited 20 Jan 2008. Lobón–Cerviá, J., 1991. Dinámica de poblaciones de peces. Pesca eléctrica y métodos de capturas sucesivas en la estima de abundancias. Monografías 3. Museo Nacional de Ciencias Naturales. CSIC, Madrid. McCune, B. & Mefford, M. J., 1999. PC–ORD. Multivariate Analysis of Ecological Data. Version 4.41. MjM Software, Gleneden Beach, Oregon, USA. Mojica, J., Castellanos, C., Usma, S. & Álvarez, R., 2002. Libro Rojo de las especies de peces dulceacuícolas de Colombia. Instituto de Ciencias Naturales Universidad Nacional de Colombia, Ministerio del Medio Ambiente, Bogotá, Colombia. Myers, N., Mittermeier, R., Mittermeier, C., da Fonseca, G. & Kent, J., 2000. Biodiversity hotspots for conservation priorities. Nature, 403: 853–858. Nogueira, C., Backup, P. A., Menezes, N. A., Oyakawa, O. T., Kasecker, T. P., Ramos Neto, M. B. & C. da Silva. J. M., 2010. Restricted–range fishes and the conservation of Brazilian freshwaters. PLoS ONE, 5(6): e11390. doi:10.1371/journal. pone.0011390. Olson, D. M. & Dinerstein, E., 2002. The Global 200: priority ecoregions for global conservation. Annals of the Missouri Botanical Garden, 89: 199–224. Reis, R. E., Kullander, S. O. & Ferraris, C. J., 2003. Check list of the freshwater fishes of South and Central America. EDIPUCRS, Porto Alegre, Brasil. Rodríguez, J. P., Lazo, R., Solórzano, L. A. & Rojas– Suárez, F., 2004. Cartografía digital básica de las Áreas Naturales Protegidas de Venezuela: Parques

285

Nacionales, Monumentos Naturales, Refugios de Fauna, Reservas de Fauna y Reservas de Biósfera. Versión 1.0, CD ROM y en–línea. Centro Internacional de Ecología Tropical (CIET), Instituto Venezolano de Investigaciones Científicas (IVIC), Conservación Internacional Venezuela, UNESCO y Oficina Nacional de Diversidad Biológica del Ministerio del Ambiente y de los Recursos Naturales (MARN). Caracas, Venezuela. http://ecosig.ivic.ve. Cited 20 Jan 2008. Rodríguez, J. P. & Rojas–Suárez, F., 2008. Libro Rojo de la Fauna Venezolana. Tercera Edición. Provita y Shell Venezuela, S. A., Caracas, Venezuela. Rodríguez–Olarte, D., Amaro, A., Coronel, J. L. & Taphorn, D. C., 2006. Integrity of fluvial fish communities is subject to environmental gradients in mountain streams, Sierra de Aroa, north Caribbean coast, Venezuela. Neotropical Ichthyology, 4(3): 319–328. Rodríguez–Olarte, D., Coronel, J. L., Taphorn, D. C. & Amaro, A., 2007. Los Peces del río Tocuyo, Vertiente del Caribe, Venezuela: un Análisis Preliminar para su conservación. Memoria de la Fundación La Salle de Ciencias Naturales, 165: 45–72. Rodríguez–Olarte, D., Taphorn, D. C. & Lobón–Cerviá, J., 2009. Patterns of Freshwater Fishes of the Caribbean Versant of Venezuela. International Review of Hydrobiology, 94(1): 67–90. Rosa, R. S. & Lima, F. C. T., 2008. Os Peixes Brasileiros Ameaçados de Extinção. In: Livro Vermelho da fauna brasileira ameaçada de extinção. Volume 2: 9–285 (A. B. M. Machado, G. M. Drummond & A. P. Paglia, Eds.). 1st edition. Brasília, DF: MMA; Belo Horizonte, MG: Fundação Biodiversitas. Saunders, D. L., Meeuwig, J. J. & Vincent, C. J., 2002. Freshwater protected areas: strategies for conservation. Conservation Biology, 16: 30–41. Whittaker, R. J. & Fernández–Palacios, J. M., 2007. Island biogeography: ecology, evolution, and conservation, 2nd edition. Oxford University Press, Oxford. Winemiller, K., Marrero, C. & D. Taphorn. 1996. Perturbaciones causadas por el hombre a las poblaciones de peces de los llanos y del piedemonte andino de Venezuela. BioLlania, 12: 13–48.


366


Animal Biodiversity and Conservation 34.2 (2011)

287

Diet of three large pelagic fishes associated with drifting fish aggregating devices (DFADs) in the western equatorial Indian Ocean M. A. Malone, K. M. Buck, G. Moreno & G. Sancho

Malone, M. A., Buck, K. M., Moreno, G. & Sancho, G., 2011. Diet of three large pelagic fishes associated with drifting fish aggregating devices (DFADs) in the western equatorial Indian Ocean. Animal Biodiversity and Conservation, 34.2: 287–294. Abstract Diet of three large pelagic fishes associated with drifting fish aggregating devices (DFADs) in the western equatorial Indian Ocean.— Several species of fish, aggregate around DFADs in marine tropical waters. We captured three predatory species: yellow fin tuna (Thunnus albacares), wahoo (Acanthocybium solandri) and dolphinfish (Coryphaena hippurus) from aggregations under DFADs in the Western Indian Ocean to characterize their diet and determine whether they fed on other DFAD associated organisms. Yellowfin tuna did not feed on DFAD–associated prey, while wahoo and dolphinfish did exploit resources aggregated by the DFADs, though they predominantly fed on other non–associated organisms. Opportunistic feeding on surface swarming stomatopod crustaceans was observed in yellowfin tuna and dolphinfish associated with FADs, but was not observed in wahoo. Key words: FAD, Yellowfin tuna, Dolphinfish, Wahoo, Stomach contents, Diet. Resumen Dieta de tres peces pelágicos de gran tamaño asociados a agregaciones de peces bajo objetos flotantes (DFADs) en el Océano Índico ecuatorial occidental.— Los objetos flotantes congregan distintas especies de peces en aguas tropicales marinas. Se capturaron tres especies depredadoras: atún de aleta amarilla (Thunnus albacares), petos (Acanthocybium solandri) y llampugas (Coryphaena hippurus) en agregaciones bajo objetos flotantes en el Océano Índico occidental, para caracterizar su dieta y determinar si se alimentan de organismos asociados a las DFADs. Los atunes de aleta amarilla no se alimentaron de presas asociadas a objetos flotantes, mientras los petos y llampugas sí explotaron los recursos agregados a estos objetos flotantes, aunque predominantemente se alimentaron de otros organismos no asociados a ellos. En atunes de aleta amarilla y en llampugas asociados a FADs se observó una predación oportunista en la superficie de agrupaciones de crustáceos estomatópodos pero no así en los petos. Palabras clave: Dispositivos agregadores de peces, Atún de aleta amarilla, Llampuga, Peto, Contenidos estomacales, Dieta. (Received: 10 X 10; Conditional acceptance: 2 II 11; Final acceptance: 6 IV 11) Margaret A. Malone, Kelly M. Buck & Gorka Sancho, College of Charleston, Grice Marine Lab., 205 Fort Johnson Road, Charleston, SC 29412, USA.– Gala Moreno, AZTI–Tecnalia, Herrera Kaia, Portualdea z/g, 20110 Pasaia, España (Spain). Corresponding Author: Gorka Sancho. E.–mail: sanchog@cofc.edu

ISSN: 1578–665X

© 2011 Museu de Ciències Naturals de Barcelona


288

Introduction Floating objects are known to aggregate fishes in tropical oceans. They can be natural or man–made objects, the latter often deployed specifically to act as Fish Aggregating Devices (FADs). Commercial tuna purse seine fisheries in tropical oceans deploy drifting FADs (DFADs) to target schools of tuna that aggregate underneath them. Other species of fish also aggregate under these floating objects, including epipelagic predatory fishes (Taquet et al., 2007). Although there has been an increase in the use of DFADs in tropical tuna fisheries in the past two decades, especially in the Western Indian Ocean (Fonteneau et al., 2004), the mechanisms underlying these floating objects are unknown (Fréon & Dagorn, 2000). Many hypotheses have been suggested to explain why these aggregations form, one of which is the ‘concentration of food supply’ hypothesis (Fréon and Dagorn, 2000). This hypothesis states that certain pelagic predators aggregate around FADs to feed upon the fauna of smaller fishes that also aggregate under these floating objects (Klima & Wickman, 1971; Fréon & Dagorn, 2000). The diet of yellowfin tuna has been studied in many regions, including the Western Indian Ocean (Roger, 1994; Ménard et al., 2000; Somvanshi, 2002; Potier et al., 2004; 2007). Yellowfin tuna can be characterized as generalist predators that feed on a wide variety of small prey, including fish larvae, epipelagic and mesopelagic fishes, squid and pelagic crustaceans. Dietary studies of yellowfin tuna associated with anchored FADs generally show no feeding on FAD associated fish communities (Brock, 1985; Buckley & Miller, 1994; Graham et al., 2007). The few studies that analyze the feeding patterns of yellowfin tuna aggregated under DFADs show high percentages of fishes with empty stomachs (Ménard et al., 2000). The feeding behavior of dolphinfish has also been extensively studied in tropical waters (i.e. Oxenford, 1999; Olson & Galvan–Magana, 2002). This epipelagic fish is characterized as a generalist predator that feeds on small epipelagic fishes, fish larvae and pelagic invertebrates. Dolphinfish have been shown to feed on prey associated with Sargassum mats (Oxenford, 1999). However, diets of dolphinfish caught under DFADs in Atlantic and Pacific waters were dominated by organisms that do not aggregate under floating objects (Oxenford, 1999; Olson & Galvan–Magana, 2002). Taquet (2004) found that a portion of the diet of dolphinfish under DFADs in the south western Indian Ocean came from prey associated with DFADs (14% when considering number of prey, 27% considering prey weight). Interestingly, juvenile flying gurnards (Dactylopteridae) and flyingfish (Exocetidae) are a common teleost prey for dolphinfish all around the world (Oxenford, 1999; Olson & Galvan–Magana, 2002; Taquet, 2004). Dietary studies of wahoo are not as common, but this species is characterized as predominantly piscivorous, consuming larger sized prey than yellowfin tuna and dolphinfish (Manooch & Hogarth, 1983). This study describes the diet of three pelagic,

Malone et al.

predatory species associated with DFADs in the Western Indian Ocean: yellowfin tuna (Thunnus albacares), dolphinfish (Coryphaena hippurus) and wahoo (Acanthocybium solandri), and addresses the ‘concentration of food supply’ hypothesis for DFADs in this environment. Material and methods Data were collected during four offshore cruises around the Seychelles archipelago (0°  01' N to 9°  06'  S) that visited multiple DFADs deployed by the French and Spanish tuna purse seine fleets. The four cruises took place in October and February of 2004 and 2005. Fish were sampled for stomach contents from a total of 17 DFADs. Upon arrival at a DFAD, abundance and species composition of fish aggregations associated with the DFADs were visually estimated by divers (Taquet et al., 2007). A total of 32 fish species were identified aggregating with these DFADs, with an average abundance of 2680 fishes per aggregation (Taquet et al., 2007). During early morning and evening hours, fishes were caught by trolling artificial lures in close proximity to the DFADs. Most of the captured fishes were tagged with acoustic tags and released so that residence times around the DFADs could be monitored (Dagorn et al., 2007). Throughout the four cruises wounded yellowfin tuna, dolphinfish and wahoo that were not suitable for tagging because of their unlikely survival after release were sacrificed for stomach content analyses. Fork length measurements were made on deck and whole stomachs were quickly removed and frozen for future analysis. In the laboratory, preserved stomachs were rinsed and opened, and their contents were removed. Mucous and gastric parasites were set aside and not taken into consideration. Extracted prey were gently blotted with paper towels, counted, weighed (wet weight) and identified down to the lowest possible taxonomic level using a dissecting scope. Dissected stomachs and their contents were fixed in 10% buffered formalin and preserved in 90% ethyl alcohol. Prey items were grouped into seven categories: FAD–associated fishes (based on diver surveys done on each FAD; Taquet et al., 2007), non–FAD associated fishes (epipelagic and mesopelagic species combined), cephalopoda (squids), stomatopoda (almost exclusively Natosquilla investigatoris), crustaceans (non–stomatopod crustaceans, mainly decapod crabs and megalopa larvae), fish larvae (pelagic larvae of benthic and pelagic species), unidentified fishes (unrecognizable, digested fish remains) and other (items not belonging to the previous categories). Mean stomach fullness was calculated as the weight of the stomach contents expressed as the percentage of the total fish weight, which was estimated using specific weight–length relationships for each species. Empty stomachs were defined as those with a stomach fullness value smaller than 0.01%. The percent abundance (%N), percent weight (%Wt), and percent frequency (%F) of each prey category were calculated as described by Hyslop (1980). This information was used to calculate an


Animal Biodiversity and Conservation 34.2 (2011)

289

Table 1. Number of stomachs analyzed, % empty stomachs and stomach fullness (stomach content weight as % of fish body weight). Natosquilla investigatoris were regularly observed by divers in October 2004. Tabla 1. Numero de estómagos analizados, % de estómagos vacíos, y contenido estomacal (peso de los contenidos estomacales expresados como % del peso total del pez). Natosquilla investogatoris fueron observados con regularidad por los buceadors en octubre del 2004.

Stomach contents (% body weight)

N

% Empty stomachs

Mean

Max.

All data

31

16.1

1.05

3.30

X 04

20

15.0

1.45

3.30

X 05, II 04, II 05

11

18.2

0.06

0.81

All data

83

25.3

0.91

6.60

X 04

20

10.0

2.36

6.60

Yellowfin

Dolphinfish

X 05, II 04, II 05

63

30.2

0.45

3.82

All data

32

34.4

0.48

4.38

X 04

16

31.2

0.49

4.38

X 05, II 04, II 05

16

37.5

0.48

3.43

Wahoo

index of relative importance (IRI) that combines all three diet estimates (Pinkas et al., 1971): IRI = (%N + %Wt) * %F A relative IRI (%IRI) was also calculated for each food category (i ) to facilitate comparisons among the existent number (n) of food categories (Cortés, 1997): %IRIi = 100IRIi /

n

S IRIi

i=1

Though these previous indices have extensively been used to interpret stomach content data (Hyslop, 1980), mean percent abundance (%MN) and weight (%MWt) of prey items were also calculated, allowing for the calculation of standard errors (Graham et al., 2007). To calculate the mean percent abundances (%MN), the percent abundance (%N) was calculated for each individual stomach by dividing the abundance of each prey taxon by the total number of prey in that stomach. The mean of these values for all samples within each species was calculated by dividing the percent abundance (%N) of all the prey taxa in an individual stomach by the total number of prey items in that stomach. Percent abundance values from individual stomachs were averaged to yield single estimates of mean percent abundance (%MN) and standard error values for each of the three predatory species studied (Graham et al., 2007). This method was also used with prey weights to obtain mean percent weights (%MWt). When analyzing stomach contents from individual cruises, data from February 2004 and February 2005 were combined due to overall lower captures in this month (n = 40 fishes in 2004 and 2005 combined).

Results We analyzed the stomach contents of 31 yellowfin tuna, 80 dolphinfish, and 32 wahoo–associated DFADs (n = 17). Captured yellowfin tuna ranged from 29 to 124 cm in fork length (average = 52.6 cm), dolphinfish from 53 to 110 cm (average = 86.1 cm) and wahoo from 80.0 and 110.0 cm (average = 95.4 cm). Yellowfin tuna had the lowest percentage of empty stomachs (16%), followed by dolphinfish (25%) and wahoo (34%) (table 1). Overall mean stomach fullness values (weight of stomach contents as percentage of total body weight) were 1.45% (yellowfin), 0.91% (dolphinfish) and 0.48% (wahoo) (table 1). The basic dietary descriptive indices (%N, %Wt and %F) for all three predatory species collected during the four cruises are presented in table 2. Yellowfin tuna exploited pelagic communities not aggregated by the FADs sampled, and no remains of FAD–associated organisms were found in yellowfin stomachs (table 2). Dolphinfish and wahoo did feed on FAD–associated fishes, though this prey category did not dominate either species’ diet. Wahoo ingested large fishes associated with FADs (%Wt = 41.15), including tuna (Scombridae) and jacks (Carangidae), but their overall importance, based on %MN and %MWt, in the diet of wahoo does not seem to be as high as cephalopods and non–FAD associated fishes (fig. 1). Dolphinfish stomachs also contained FAD associated fishes, but the FAD associated fishes were not as important a part of their diet as the non–FAD associated fishes, cephalopods and stomatopods (table 2; fig. 1). FAD– associated fishes found in dolphinfish stomachs included triggerfish (Balistidae), mackerel (Scombridae), jacks (Carangidae), seahorses (Sygnathidae) and


290

Malone et al.

Table 2. Percent abundance (%N), percent weight (%Wt), and percent frequency (%F) of prey items found in the three predatory species: yellowfin tuna (Thunnus albacares), dolphinfish (Coryphaena hippurus) and wahoo (Acanthocybium solandri). Tabla 2. Porcentaje de abundancia (%N), porcentaje de peso (%Wt) y porcentaje de frecuencia (%F) de presas encontradas en las tres especies depredadoras: atunes de aleta amarilla (Thunnus albacares), llampugas (Coryphaena hippurus) y petos (Acanthocybium solandri). Functional Prey Groups

Yellowfin tuna %N

%Wt

%F

Dolphinfish %N %Wt

%F

Wahoo %N %Wt %F

FAD–associated fishes

0.00

0.00

0.00

0.73 1.69

8.75

2.41 41.15 4.88

Non–FAD–associated fishes

0.00

0.00

0.00

2.30 26.65 25.00

10.84 40.46 19.51

Cephalopoda

0.64

0.38

9.68

14.09 10.49 33.75

54.22 7.12 41.46

Stomatopoda

93.51 92.78 58.07

73.17 42.06 15.00

0.00 0.00 0.00

Crustaceans

4.45

1.25 3.41

3.61 0.04 4.88

6.44 29.03

7.50

Larval fishes

0.13

0.02

3.23

4.07 3.50 18.75

1.20 0.03 2.44

Unidentified fishes

1.02

0.39

12.9

4.38 12.2

6.25

24.1 9.52 19.51

Other

0.00

0.00

0.00

0.00 0.00

0.00

3.61 1.68 7.32

barracuda (Sphyraenidae). It should be mentioned that the 'unidentified fishes' category consisted of very digested or partial remains that could not be identified to family level. If all unidentified fishes were to have been FAD–associated fishes, then wahoo (FAD–associated fishes %MN = 17.6; %MWt = 24.1) and dolphinfish (FAD–associated fishes %MN = 19.5; %MWt = 23.2) could be considered to prey on FAD– associated fishes as often as on fishes not associated with FADs. But if all unidentified remains were to be from non–FAD–aggregated species, then wahoo (non FAD–aggregated %MN = 28.9; %MWt = 37.5) and dolphinfish (non FAD–aggregated %MN = 32.6; %MWt = 33.6) would seem to mainly prey on fishes not associated with FADs and cephalopods, leaving FAD–associated fishes as the third most common food item. When analyzing the stomach content data by cruises, clear seasonal and interannual differences in diets were observed. During the October 2004 cruise, the yellowfin tuna and dolphinfish sampled fed almost exclusively on Natosquilla investigatoris, a pelagic swarming Stomatopod crustacean, showing %IRI values of 99.2 and 95.9 respectively (fig. 2). Wahoo did not feed on the stomatopod Natosquilla investigatoris at any time, and in October 2004 fed primarily upon cephalapods and non–FAD–associated fishes (fig. 2), with flying fish (Exocetidae) being the most abundant fish prey (62% of all identified fishes). During the three other cruises (October 2005 and February 2004/2005) yellowfin tuna associated with drifting FADs had a more diverse diet, with stomach contents dominated by pelagic crustaceans (amphipods, megalopa larvae and the pelagic portunid crab Charybdis edwardsi), with some Natosquilla investigatoris found in October

2005 (fig. 2). Dolphinfish diet also showed increased diet diversity in February and October 2005, mainly consuming cephalopods in October 2005, and fishes not associated with FADs during the February cruises. Squid remained a strong component of their diet in the winter and some FAD–associated fishes were found in February samples (%IRI = 4.5; fig. 2). During both October cruises, wahoo predominantly fed on squid, and during February cruises they mainly fed on fishes, though mostly on non FAD–associated species (fig. 2). The proportion of empty stomachs and fullness estimates did not vary seasonally for wahoo (table 1). Yellowfin tuna and dolphinfish during October 2004, when N. investigatoris dominated their diets, showed much lower proportions of empty stomachs and higher stomach fullness values than in other sampling periods (table 1). Discussion Yellowfin tuna did not feed on fishes aggregated under DFADs in the Western Indian Ocean. Wahoo did feed on DFAD aggregations, especially during the winter months when their diet was more piscivorous. Dolphinfish also fed on FAD associated fishes, but they were not a sizeable part of their diet. The diet of yellowfin tuna captured under DFADs in this study was dominated by crustaceans, but also included cephalopods and larval fishes. These results coincide with most previous studies which characterized tuna associated with DFADs as generalist predators of small organisms, not exploiting other species associated with DFADs (Brock, 1985; Buckley & Miller, 1994). Ménard et al. (2000) described how small sized


Animal Biodiversity and Conservation 34.2 (2011)

291

70

Yellowfin tuna

60

Dolphinfish

%MN

50

Wahoo

40 30 20 10 0 70 60

%MWt

50 40 30 20 10

0

1

2

3

4 5 Prey groups

6

7

8

Fig. 1. Relative importance of seven prey groups in the diets of yellowfin tuna (Thunnus albacares), dolphinfish (Coryphaena hippurus), and wahoo (Acanthocybium solandri): A. Mean percent abundance (%MN); B. Mean percent weight (%MWt) ± SE; 1. FAD–associated fishes; 2. Non–FAD–associated fishes; 3. Cephalopoda; 4. Stomatopoda; 5. Crustaceans; 6. Larval fishes; 7. Unidentified fishes; 8. Other. Fig. 1. Importancia relativa de siete grupos de presas en las dietas de atunes de aleta amarilla (Thunnus albacares), llampugas (Coryphaena hippurus) y petos (Acanthocybium solandri): A. Porcentaje de abundancia media (%MN); B. Porcentaje de peso medio (%MWt) ± EE; 1. Peces asociados a las FAD; 2. Peces no asociados a las FAD; 3. Cephalopoda; 4. Stomatopoda; 5. Crustacea; 6. Larvas de peces; 7. Peces no identificados; 8. Otros.

(< 90cm) yellowfin tuna associated with DFADs in the equatorial Atlantic fed on diverse prey not associated with DFADs. However, they observed that large (>  90  cm) yellowfin tuna associated with DFADs fed mainly on small epipelagic fishes, including small scombrids that commonly aggregate under FADs. Similar ontogenetic dietary shifts have also been observed in yellowfin tuna associated with anchored FADs in Hawaii (Graham et al., 2007), with small juvenile fishes (<  50  cm) mainly feeding on crustacean larvae and larger individuals (>  50 cm) feeding on mesopelagic shrimp, reef fish pelagic juveniles and epipelagic fishes. It is unknown whether the ingested epipelagic fishes were associated with the anchored FADs where the tuna were captured (Graham et al., 2007). All the yellowfin tuna analyzed in our study were smaller than 80 cm in fork length (average FL = 52.6 cm), except

for a single large tuna of 124 cm. Therefore we cannot address the possibility of large tuna feeding on small epipelagic fishes aggregated under DFADs in the Indian Ocean, but it is a question that should be addressed in future studies. Dolphinfish have been described as opportunistic epipelagic predators, feeding both diurnally and nocturnally (Oxenford, 1999; Olson & Galvan–Magana, 2002). Our results agree with this general assessment, showing that dolphinfish mainly fed on squids, non– FAD associated fishes and crustaceans. Dolphinfish diet showed seasonal and interannual shifts, likely due to changes in prey availability in the epipelagic zone. Only during the February cruises, when dolphinfish displayed a mostly piscivorous diet, were FAD–associated fishes found in their stomachs. However, even during these winter cruises, the majority of fishes preyed


292

on by dolphinfish were epipelagic species not found in association with the FADs where the dolphinfish were captured. Taquet (2004) found dolphinfish in the southern Indian Ocean feeding on FAD–associated fishes at higher rates than we observed in this study. This might be explained by the significant proportion of unidentifiable, digested fishes found in stomach contents in our study. These prey fish could neither be classified as species associated with FADs nor as unassociated species. Another explanation could be related to the different fish communities aggregated under DFADS in both studies (Taquet et al., 2007), Equatorial commercial DFADs harbored much larger and more diverse fish communities, including schools of tuna, than tropical experimental DFADs in the southern Indian Ocean (Taquet, 2004; Taquet et al., 2007). The diet of Indian Ocean wahoo caught around DFADs was similar to that of wahoo from the Eastern Atlantic and Gulf of Mexico, composed mainly of squid and fishes (Manooch & Hogarth, 1983). In February, wahoo had a piscivorous diet, mostly consuming fish species not aggregated by FADs (mainly flying fish), but also preying on some FAD aggregated fishes. In October, the diet of wahoo was dominated by squids, though fish remains were also recorded (from species both associated and not associated with FADs). Ménard et al. (2000) found that the proportion of empty stomachs in Atlantic Ocean yellowfin tuna captured by purse seiners under drifting FADs (65%) was higher than in yellowfin tuna captured from free swimming schools (17%). Similarly, combining data from Atlantic and Indian Oceans, Hallier & Gaertner (2008) also found a higher proportion of yellowfin empty stomachs in purse seine caught tuna under FADs (49%) than in free schools (7%). Free swimming tuna captured in the Indian Ocean captured with longlines (Potier et al., 2007) and by trolling (Roger, 1994) also show low proportions of empty stomachs (13% and 8% respectively), indicating a high feeding activity. These results suggest that yellowfin tuna do not feed intensively under drifting FADs. The present study measured a low proportion of empty stomachs (16%), which indicates a higher feeding activity of yellowfin tuna captured by trolling under drifting FADs in the Indian Ocean. Our results were similar to the proportion of empty stomachs in yellowfin tuna captured by trolling around anchored FADs in Hawaii (17%; Graham et al., 2007). We suggest that these differences in empty stomach frequency are not due to ecological differences between different tuna populations, aggregation size or FAD characteristics, but more likely to be the result of the different sampling methods used . It is imaginable that capturing tuna associated with FADs with hook and line selects those individuals that are actively feeding at the time of capture, while purse seining captures all or most of the fishes aggregated under a FAD, independently of their feeding activity. This idea is further supported by the low proportions of empty stomachs found in dolphinfish captured by trolling in this study (25%) and in the Atlantic Ocean (11%; Oxenford, 1999), and the higher proportion of empty stomachs in dolphinfish caught by purse–seining in the Pacific Ocean (58%;

Malone et al.

Olson & Galvan–Magana, 2002). Another factor could be the very likely different rates in regurgitation of stomach contents by fishes caught with hook and line and those caught in purse seines (Bowman, 1986). We hypothesize that fish in purse seines experience higher regurgitation rates because of the long time it takes to bring the fish on board. Future studies measuring feeding rates of FAD–associated predators should account for the differential selectivity of different sampling methods used. In October 2004, yellowfin tuna and dolphinfish caught around DFADs fed intensively and almost exclusively on the pelagic, swarming stomatopod Natosquilla investigatoris. High concentrations of swarming N. investigatoris were noted by divers in the surface waters of the study area at this time. Little is known about the ecology and behavior of this crustacean in pelagic environments of the western Indian Ocean (Losse & Merret, 1971). Surface blooms of pelagic N. investigatoris have been recorded as periodic occurrences and were observed in 1933, 1944, 1965–1967, 1999 and 2000 (Losse & Merret, 1971; Potier et al., 2002; Kamukuru & Mgaya, 2004). In other regions, both yellowfin tuna and dolphinfish also engage in opportunistic feeding of very abundant, small prey (Oxenford, 1999; Ménard et al., 2000). The appearance of N. investigatoris swarms commonly results in opportunistic feeding by many shallow water predators (most tuna species, dolphinfish, marlin, swordfish, snapper, lancetfish), and likely has ecological effects on the whole western Indian Ocean (Losse & Merret, 1971; Potier et al., 2004; 2007). The low proportion of empty stomachs and high stomach fullness values measured for yellowfin and dolphinfish associated with DFADs in this study further supports this concept. Opportunistic feeding events like these could play an important role in population fluctuations of predators that exploit them, and they have been linked to increased catches of yellowfin tuna (Fonteneau et al., 2004). Efforts should be made to study the swarming events of N. investigatoris in the Indian Ocean. In October 2005, no N. investigatoris were found in dolphinfish stomachs, while some were present in the diet of yellowfin tuna. Due to the epipelagic habits of dolphinfish when compared to yellowfin tuna, which are capable of feeding in deeper layers (Graham et al., 2007), the pattern observed in October 2005 could be explained if N. investigatoris were only to be found in deep waters and not swarming at the surface. Interestingly, wahoo did not take advantage of this opportunistic food resource, mainly feeding on squid in October 2004 and October 2005, and showing no seasonal changes in mean stomach fullness and empty stomach frequency. This is most likely due to the preference of wahoo for larger prey (Manooch & Hogarth, 1983). During N. investigatoris surface swarming events, interspecific competition between wahoo and other co–occurring epipelagic predatory fishes is greatly reduced. In conclusion, our data suggest that yellowfin tuna associated to DFADs in the Indian Ocean do not feed on other species aggregated by FADs, as suggested by the ‘concentration of food supply’ hypothesis. This observation complements other studies of yellowfin


Animal Biodiversity and Conservation 34.2 (2011)

80%

80%

60%

60%

40%

40%

20% 0% C

100%

20% October 2004

October 2005

February 2004/2005

October 2004

October 2005

February 2004/2005

Non–FAD–associated fishes

60%

Larval fishes Crustaceans

40%

Cephalopoda Stomatopoda

20% 0%

0%

FAD–associated fishes

80% % IRI

B 100%

% IRI

% IRI

A 100%

293

Unidentified fishes October 2004

October 2005

February 2004/2005

Other

Fig. 2 Index of relative importance (% IRI) for the diets of: A. Yellowfin tuna (Thunnus albacores); B. Dolphinfish (Coryphaena hippurus); and C. Wahoo (Acanthocybium solandri) in October 2004, October 2005, and combined February 2004/2005. Fig. 2. Indice de importancia relativa (% IRI) para las dietas de: A. Atunes de aleta amarilla (Thunnus albacares); B. Llampugas (Coryphaena hippurus) y C. Petos (Acanthocybium solandri) en octubre 2004, octubre 2005 y febrero 2004/2005 combinados.

tuna associated with FADs (Brock, 1985; Buckley & Miller, 1994; Graham et al., 2007). Yellowfin showed a relatively low proportion of empty stomachs, similar to studies that captured tuna by hook and line (Graham et al., 2006; Potier et al., 2007), but much lower than the proportion of tuna captured at FADs by purse seining (Ménard et al., 2000; Hallier & Gaertner, 2008). Yellowfin tuna displayed opportunistic intense feeding on N. investigatoris when surface swarms were observed in October 2004. Dolphinfishes have been described as opportunistic epipelagic predators, occasionally feeding on prey associated with sargassum algae and FADs (i.e. Oxenford, 1999; Olson and Galvan–Magana, 2002; Taquet, 2004). Our results support this idea, since dolphinfish did occasionally feed on FAD associated fishes, but switched diets and fed more intensively when an opportunistic resource (Natsoquilla investigatoris) became available. Wahoo seemed to seasonally exploit trophic resources available at FADs in the Indian Ocean, though their diet was always dominated by organisms not aggregated by FADs. This suggests that multiple factors are influencing the

aggregative behavior of dolphinfish and wahoo around FADs, though the concentration of potential prey items might be an important factor in the case of wahoo. Acknowledgements This study was part of the FADIO (Fish Aggregating Devices as Instrumented Observatories of pelagic ecosystems) project, funded by the DG Research (QLRI–CT–2002–02773). We would like to thank all the FADIO project members that directly assisted with the capture of specimens and making underwater fish surveys, especially David Itano, Marc Taquet, Laurent Dagorn, Erwan Josse and Charlotte Girard. Deborah Bidwell and Laurent Dagorn and two anonymous reviewers made comments on earlier drafts of this paper, and David Knott helped with the identification of many invertebrate specimens. This research was done under auspices of the College of Charleston Institutional Animal Care and Use Committee (IACUC) permit number 04–010.


294

References Bowman, R. E., 1986. Effect of regurgitation on stomach content data of marine fishes. Environmental Biology of Fishes, 16: 161–171. Brock, R. E., 1985. Preliminary study of the feeding habits of pelagic fish around Hawaiian fish aggregating devices, or can fish aggregation devices enhance local fish productivity? Bulletin of Marine Science, 37: 40–49. Buckley, T. W. & Miller B. S., 1994. Feeding habits of yellow–fin tuna associated with fish aggregation devices in American Samoa. Bulletin of Marine Science, 55: 445–459. Cortés, E., 1997. A critical review of methods of studying fish feeding based on analysis of stomach contents: application to elasmobranch fishes. Canadian Journal of Fisheries and Aquatic Sciences, 54: 726–738. Dagorn, L., Pincock, D., Girard, C., Holland, K., Taquet, M., Sancho, G., Itano, D. & Aumeeruddy, R., 2007. Satellite–linked acoustic receivers to observe behavior of fish in remote areas. Aquatic Living Resources, 20: 307–312. Fonteneau, A., Ariz, J., Hallier, P., Lucas, V., Pallares, P. & Potier, M., 2004. The Indian Ocean yellowfin stock and fisheries in 2003: overview and discussion of the present situation. Document IOTC 2004/ WTTP/02: 1–17. Fréon, P. & Dagorn, L., 2000. Review of fish associative behaviour: toward a generalisation of the meeting point hypothesis. Reviews of Fish Biology and Fisheries, 10: 183–207. Graham, B. S., Grubbs, D., Holland, K. & Popp, B. N., 2007. A rapid ontogenetic shift in the diet of juvenile yellowfin tuna from Hawaii. Marine Biology, 150: 647–658. Hallier, J. P. & Gaertner, D., 2008. Drifting fish aggregation devices could act as an ecological trap for tropical tuna species. Marine Ecology Progress Series, 353: 255–264 Hyslop, E. J., 1980. Stomach contents analysis – a review of methods and their application. Journal of Fish Biology, 17: 411–429. Kamukuru, A. T. & Mgaya, Y. D., 2004. The food and feeding habits of blackspot snapper, Lutjanus fulviflamma (Pisces: Lutjanidae) in shallow waters of Mafia Island, Tanzania. African Journal of Ecology, 42: 49–58. Klima, E. F. & Wickman, D. D., 1971. Attraction of coastal pelagic fishes with artificial structures. Transactions of the American Fisheries Society, 1: 86–99. Losse, G. F. & Merrett, N. R., 1971. The occurrence of Oratosquilla investigatoris (Crusacea: Stomapoda)

Malone et al.

in the pelagic zone of the Gulf of Aden and the equatorial western Indian Ocean. Marine Biology, 10: 244–253. Manooch, C. S. & Hogarth, W. T., 1983. Stomach contents and giant trematodes from wahoo, Acanthocybium solanderi, collected along the South Atlantic and Gulf coasts of the United States. Bulletin of Marine Science, 33: 227–238. Ménard, F., Stéquer, B., Rubin, A., Herrera, M. & Marchal, E., 2000. Food consumption of tuna in the Equatorial Atlantic Ocean: FAD–associated versus unassociated schools. Aquatic Living Resources, 13: 233–240. Olson, R. J, & Galvan–Magana, F., 2002. Food habits and consumption rates of common dolphinfish (Coryphaena hippurus) in the eastern Pacific Ocean. Fishery Bulletin, 100(2): 279–298. Oxenford, H. A., 1999. Biology of the dolphinfish (Coryphaena hippurus) in the western central Atlantic: a review. Scientia Marina, 63: 277–301. Pinkas, L., Oliphant, M. S., Iverson I. L. K., 1971. Food habits of albacore, bluefin tuna, and bonito in Californian Waters. California Fish and Game, 152: 1–105. Potier, M., Lucas, V., Marsac, F., Sabatié, R. & Ménard, F., 2002. On–going research activities on trophic ecology of tuna in equatorial ecosystems of Indian Ocean. IOTC Proceedings, 5: 368–374. Potier, M., Marsac, F., Cherel, Y., Lucas, V., Sabatié, R., Maury, O. & Ménard, F., 2007. Forage fauna in the diet of three large pelagic fishes (lancetfish, swordfish and yellowfin tuna) in the western equatorial Indian Ocean. Fisheries Research, 83: 60–72. Potier, M., Marsac, F., Lucas, V., Sabatié, R., Hallier, J. P. & Ménard, F., 2004. Feeding partitioning among tuna taken in surface and mid–water layers: the case of yellowfin (Thunnus albacares) and bigeye (T. obesus) in the Western Tropical Indian Ocean. Western Indian Ocean Journal of Marine Science, 3: 51–62. Roger, C., 1994. Relationships among yellowfin and skipjack tuna, their prey–fish and plankton in the tropical western Indian Ocean. Fisheries Oceanography, 3: 133–141. Somvanshi, V. S., 2002. Review of biological aspects of yellowfin tuna (Thunnus albacares) from the Indian Ocean. IOTC Proceedings, 5: 420–426. Taquet, M., 2004. Le comportement agrégatif de la dorade coryphène (Coryphaena hippurus) autour des objets flottants. Thèse de doctorat de l’Université de Paris 6, Océanologie biologique. Taquet, M., Aumeruddy, R., Sancho, G., Itano, D., Wendling, B., Peignon, C. & Dagorn, L., 2007. Pelagic fish communities around drifting FADs in the Indian Ocean. Aquatic Living Resources, 20: 331–341.


Animal Biodiversity and Conservation 34.2 (2011)

295

Studies on the subgenus Agraphoderus Bates of Blennidus Motschulsky from Peru: the jelskii species–group (Coleoptera, Carabidae, Pterostichini) G. Allegro & P. M. Giachino

Allegro, G. & Giachino, P. M., 2011. Studies on the subgenus Agraphoderus Bates of Blennidus Motschulsky from Peru: the jelskii species–group (Coleoptera, Carabidae, Pterostichini). Animal Biodiversity and Conservation, 34.2: 295–308. Abstract Studies on the subgenus Agraphoderus Bates of Blennidus Motschulsky from Peru: the jelskii species–group (Coleoptera, Carabidae, Pterostichini).— Four new species of Blennidus subgenus Agraphoderus are described from the Andes of Southern Peru: B. (A.) procerus n. sp., B. (A.) abramalagae n. sp., B. (A.) etontii n. sp. and B. (A.) straneoi n. sp. Together with B. (A.) jelskii (Tschitschérine, 1897), they form a very homogeneous group of probably closely related species (the jelskii group), which is distinguished from other members of the subgenus by the distinctive morphology of the aedeagus. A redescription of B. (A.) jelskii is given based on the lectotype and paralectotype designated by Straneo & Vereshagina (1991), supplementing Tschitschérine’s brief original description. The distribution pattern of the species presently included in the jelskii species–group is discussed, emphasizing distinctive traits of stenoendemic species inhabiting restricted geographical areas, and discussing their possible origin by allopatric speciation. Key words: Blennidus, Agraphoderus, Peru, Taxonomy, New species. Resumen Estudio sobre el subgénero Agraphoderus Bates de Blennidus Motschulsky de Perú: el grupo de especies jelskii (Coleoptera, Carabidae, Pterostichini).— Se describen cuatro especies nuevas de Blennidus subgen. Agraphoderus, de los Andes al sur de Perú: B. (A.) procerus sp. n., B. (A.) abramalagae sp. n., B. (A.) etontii sp. n. y B. (A.) straneoi sp. n. Estas cuatro especies, conjuntamente con B. (A.) jelskii (Tschitschérine, 1897), constituyen el grupo de especies jelskii, que es muy homogéneo y cuyas especies posiblemente estarían estrechamente emparentadas, diferenciándose de otros miembros de mismo subgénero por la peculiar morfología del edeago. Se realiza una redescripción de B. (A.) jelskii tomando como criterio el lectotipo y paralectotipo designados por Straneo & Vereshagina (1991), con la cual se completa la breve descripción original realizada por Tschitschérine. Finalmente, basándonos en el patrón de distribución de las especies del grupo jelskii, y enfatizando el rasgo característico de las especies estenoendémicas que habitan áreas geográficas restringidas, se considera que posiblemente se hayan originado por procesos de especiación alopátrica. Palabras clave: Blennidus, Agraphoderus, Perú, Taxonomía, Especies nuevas. (Received: 29 IV 11; Conditional acceptance: 8 VII 11; Final acceptance: 10 VIII 11) Gianni Allegro, CRA–PLF Unità di Ricerca per le Produzioni legnose Fuori Foresta, Strada Frassineto 35, I–15033 Casale Monferrato (AL), Italia.– Pier Mauro Giachino, Settore Fitosanitario Regionale, Environment Park, Palazzina A2, Via Livorno 60, 10144 Torino, Italia. Corresponding author: Gianni Allegro. E–mail: gianni.allegro@entecra.it

ISSN: 1578–665X

© 2011 Museu de Ciències Naturals de Barcelona


296

Introduction The genus Blennidus Motschulsky, 1866 at present includes about 120 species (Lorenz, 2005a) distributed in the Andean region of South America, extending from North Colombia to Chile. Moret (1995) considered the genera/subgenera Agraphoderus Bates, 1891, Ogmopleura Tschitschérine, 1899, Sierrobius Straneo, 1951, Pachyabaris Straneo, 1953 and Pseudocynthidia Straneo, 1953 (formerly separated based on the presence/absence of metathoracic wings as well as of a transverse sulcus on the abdominal sterna IV–VI) as synonyms of the senior name Blennidus. He regarded these characters were largely inconsistent, in particular the presence/absence of an abdominal transverse sulcus. This view–point was accepted by Lorenz (2005a). In a later paper, Moret (2005) used three ‘convenience subgenera’ (Blennidus s. str., Sierrobius Straneo, 1951 and Agraphoderus Bates, 1891) and described a new subgenus (Jasinskiellus Moret, 2005). His aim was to keep the lineages of some well characterized species separate, even though these taxa lacked any phyletic value (Moret, in litteris 2011). Among these four subgenera, the subgenus Agraphoderus is the richest in species. It includes the micropterous species which usually display a transverse sulcus and/or puncture rows only laterally on abdominal sterna IV–VI, with a gap in the middle. They were formerly attributed to Ogmopleura Tschitschérine, 1899. We agree with Moret (1995) on the inconsistency of such characters, as many species in Agraphoderus display only very superficial and hardly visible lateral impressions or very tiny punctures. Moreover, high variability can be observed among species which are probably related based on the similarities of their external morphology and, above all, their genitalia. For these reasons, assessment of the phyletic relationships among species within the genus is urgently required. In the meantime, we prefer to retain these ‘convenience taxa’, as the species concerned in this article can all be included in the subgenus Agraphoderus sensu Moret (2005). A total of 34 Agraphoderus species are currently recorded from Peru (Straneo, 1993; Allegro, 2010). They display a rather uniform ‘harpaloid’ habitus (Straneo, 1993) and share common and peculiar features such as the enlarged basal half of the aedeagus, just above the insertion of the parameres. Most species have a stumpy, convex body with short appendages; only a few species have a slender and/or depressed body and long appendages (Allegro, 2010). Their habitat and altitudinal distribution are also very similar; they mainly inhabit high altitude Andean grasslands at 3,300–4,800 m a.s.l. During two stays on the mountains of the Cordillera Blanca, Peru, one in November–December 2005 and the other in June–July 2008, one of the authors (G. Allegro) had the opportunity to collect abundant material of Carabidae. Data concerning some of this material have already been published (Allegro et al., 2008; Allegro, 2010). He explored an area included in the Dept. of Ancash (Provinces of Huaráz, Asunción, Yungay

Allegro & Giachino

and Fitzcarrald) where he collected abundant material of the Blennidus subgenus Agraphoderus. The other author (P. M. Giachino) is in possession of Pterostichine material collected in Peru by M. Etonti. This material includes several specimens that belong to the same genus and subgenus. A few specimens collected by M. Etonti were also deposited in the Mateu Collection at Museo Regionale di Scienze Naturali, Torino, Italia. The study of this material and its comparison with the type material of the Straneo Collection (at Museo Civico di Storia Naturale, Milano, Italia) and the type material described by Tschitschérine (Zoological Institute of the Russian Academy of Sciences, St. Petersburg, Russia) allowed us to recognise groups of species that can be distinguished by their very homogeneous male genitalia morphology. The study also allowed us to recognise many undescribed species. These groups, which we will consider one at a time in separate articles, seem to be congruent from a geographical point of view, as they include species ranging in clearly delimited areas of the country. The same approach was adopted by Moret (1995), who grouped the Ecuadorian Blennidus into eight different species–groups. However, some Peruvian species cannot yet be placed in any defined species–group and only further surveys in the Andean countries will provide material and information for reliable phylogenetic analysis. This work deals with the species belonging, in our opinion, to the jelskii species–group of the genus Blennidus, all of them localized in the Andes of Southern Peru. In addition to Blennidus (Agraphoderus) jelskii (Tschitschérine, 1897), this group includes four new species which are described, illustrated and discussed here. B. (A.) jelskii is redescribed, as the original description (Tschitschérine, 1897) is incomplete concerning some characters. Material and methods We examined material of the genus Blennidus from the following Museums and private Collections: MRST. Museo Regionale di Scienze Naturali, Torino, Italia; MSNM. Museo Civico di Storia Naturale, Milano, Italia; BMNH. The Natural History Museum, London, United Kingdom; ZIRA. Zoological Institute Russian Academy of Sciences, St. Petersburg, Russia; CAl. Allegro Collection, Moncalvo, Italia; CCa. Casale Collection, Torino, Italia; CGi. Giachino Collection at Settore Fitosanitario Regionale, Regione Piemonte, Torino, Italia; CMa. Mateu Collection at MRST; CMo. Moret Collection, Toulouse, France; CSc. Sciaky Collection, Milano, Italia; CSo. Solsky Collection at ZIRA; CSt. Straneo Collection at MSNM; CTs. Tschitschérine Collection at ZIRA. The following acronyms were used for the type material: HT. Holotype; PT, PTT. Paratype(s); LT. Lectotype; PLT. Paralectotype. Locality labels of the material examined are quoted in their original form. Drawings of the genitalia were made using a camera lucida connected to a Leica MZ 12.5 stereo microscope. The habitus drawings are by G. Allegro.


Animal Biodiversity and Conservation 34.2 (2011)

Taxonomy B. (Agraphoderus) jelskii species–group Straneo (1993) considered B. (A.) jelskii easy to distinguish on account of the basally sinuate sides of the pronotum as he knew of no other Peruvian species with such a character. Probably for this reason, he labelled 'Ogmopleura jelskii' two specimens from Callanga (ex Staudinger) showing this feature, putting them in his collection together with a paralectotype of B. jelskii from the Solsky Collection (ZIRA). It should be noted that in the same paper Straneo wrongly attributed the specimens of the type series to the locality Lima, when they were undoubtedly from Puno (green label handwritten by Jelski specifying 'Puno Peru–Jelski 1870'), as correctly reported in Straneo & Vereshagina (1991). More detailed examination revealed that the specimens from Callanga belonged to an undescribed species with similar external morphology but a different aedeagus (although of similar structure). Moreover, in the material collected by Etonti, we were able to find three further species that had a uniform structure of male genitalia, all from a restricted area of the Cuzco Region. Callanga, the type locality of Blennidus (Agraphoderus) straneoi n. sp., probably refers not to the town of that name near Lima but to the ancient archaeological site in the Cuzco Dept., from where four of the five species of the jelskii group are recorded. This is confirmed by the presence of two further specimens from Callanga (ex Staudinger) in the type series of Blennidus (Agraphoderus) mesotibialis (Straneo, 1993) (in CSt), a species which is only recorded from the Cuzco area. This evidence led us to consider the five species we attribute to the jelskii species–group as probably strictly related, probably sharing a common ancestor and distributed only, according to present knowledge, in the Southern Andean districts of Peru. As far as diagnostic characters are concerned, four species show sinuate (or nearly straight) sides at the base of pronotum (B. (A.) jelskii, B. (A.) straneoi n. sp., B. (A.) procerus n. sp. and B. (A.) abramalagae n. sp.). A fifth species, B. (A.) etontii n. sp., in spite of its similarity with the other species of the group on account of structure of male genitalia, displays rounded sides. The elytral striae are equally impressed in all these species. In some cases males possess a faint metallic lustre. Females are distinguished for their markedly dull elytra with strong polygonal microsculpture. The postangular seta of pronotum is always placed at the posterior angle. The mesotibiae of males are not distally swollen or provided with spines; only one species (B. (A.) etontii n. sp.) displays preapically swollen male metatibiae. The abdominal sterna IV–VI are usually smooth, showing at most very tiny punctures at the sides; only B. (A.) jelskii has a short transverse impression interrupted in the middle, together with a few punctures. Sternum VII has a pair of apical setae in males and two pairs in females. The character which defines this group is the distinct and very homogeneous morphology of the aedeagus: it is slender in

297

lateral view, sometimes ventrally depressed, distally very thin and flat (not laterally sinuate), and bent downwards but turning up at apex; it also shows a triangular apical blade, that is more or less acutely pointed in the dorsal view. Moreover, the ostium is always large, covering almost the whole dorsal part, and very long, extending to the basal bulb. Such a long ostium is only known in a few other Peruvian Agraphoderus species, including Blennidus (Agraphoderus) aulacostigma (Tschitschérine, 1897), a taxon with uncertain affinities recorded from Puno. The left paramere is always in discoid in shape, while the right paramere is narrow, curved (almost straight only in B. (A.) jelskii) and apically spatulate. A few specimens of B. (A.) procerus sp. n. display flattened eyes. This character had previously been noted by Straneo (1993) relating to B. (A.) jelskii (the two examined specimens of B. (A.) jelskii from the type series show normally convex eyes). As far as we know, no other Blennidus species in other groups have such a character. The following species are included, in the present status of knowledge, in the jelskii group of the genus Blennidus: B. (Agraphoderus) jelskii (Tschitschérine, 1897); B. Agraphoderus) straneoi n. sp.; B. (Agraphoderus) etontii n. sp.; B. (Agraphoderus) procerus n. sp; B. (Agraphoderus) abramalagae n. sp. A dichotomous key is provided to make their identification easier. Referring to the key by Straneo (1993), statement #25 should also include B. (A.) straneoi n. sp., B. (A.) procerus n. sp. and B. (A.) abramalagae n. sp., in addition to B. (A.) jelskii, whereas B. (A.) etontii n. sp. should be connected with statement #31. Blennidus (Agraphoderus) jelskii (Tschitschérine, 1897) (figs. 1, 6, 15, 20)

Feronia jelskii Tschitschérine, 1897: 290. Feronia jelskii Tschitschérine, 1897: Tschitschérine, 1898: 146 (note). Ogmopleura jelskii (Tschitschérine, 1897): Straneo & Vereshagina, 1991: 203. Ogmopleura jelskii (Tschitschérine, 1897): Straneo, 1993: 379. Blennidus jelskii (Tschitschérine, 1897): Lorenz, 2005a: 262. Blennidus jelskii (Tschitschérine, 1897): Lorenz, 2005b: 539.

Material examined LT ♂, Puno, Peru, Jelski 1870 (CSo–ZIRA); PLT ♀, same data as LT (CSt–MSNM). Type material In the original description Tschitschérine (1897) described this species based on three specimens (2 ♂♂ and 1 ♀). Curiously, Straneo & Vereshagina (1991) mention four specimens of the type series (3 ♂♂ and 1 ♀), designating a Lectotype (♂) and three Paralectotypes. Finally, Straneo (1993) affirms to have examined three specimens (LT ♂ and 2 PLT with sex not specified) of the type series, quoting a fourth specimen from the original description (?). Type locality Puno and Lima, Peru (Tschitschérine 1897). The two examined specimens (LT ♂ and PLT ♀) are labelled 'Puno' by Jelski (handwritten green label), and we


298

Allegro & Giachino

Identification key for the species of the jelskii group of the genus Blennidus (Agraphoderus): Clave de identificación de las especies del grupo jelskii del género Blennidus (Agraphoderus):

1. Sides of pronotum sinuate at the base or nearly straight; male metatibiae not preapically swollen >2 Sides of pronotum rounded over entire length; male metatibiae preapically swollen (sometimes scarcely evident) B. (Agraphoderus) etontii n. sp. 2. Impressed elytral striae with feebly convex intervals in ♂♂. Distinctly pear–shaped elytra. Aedeagus median lobe regularly curved and not depressed ventrally. Species from the Puno area B. (Agraphoderus) jelskii (Tschitschérine, 1897) Superficial elytral striae with flat intervals in both sexes. Elytra not pear–shaped. Aedeagus median lobe angulately inserted on the basal bulb and ventrally somewhat depressed. Species from the Cuzco area >3 3. Relatively larger (8.0–9.0 mm). A faint metallic lustre is usually present in ♂♂. Sides of pronotum markedly sinuate at base. In lateral view, median lobe of aedeagus larger and distally nearly straight B. (Agraphoderus) abramalagae n. sp. Relatively smaller (6.6–8.8 mm). Metallic lustre usually absent. Sides of pronotum weakly sinuate or nearly straight at base. Median lobe of aedeagus smaller and angulately bent downwards in the distal part >4 4. Habitus more slender (length/width of elytra = 1.50–1.57). Elytra less convex, with apical declivity less marked. Aedeagus smaller even in the largest specimens B. (Agraphoderus) procerus n. sp. Habitus oval (length/width of elytra = 1.45). Elytra more convex, with marked apical declivity. Aedeagus more slender and larger B. (Agraphoderus) straneoi n. sp.

could not locate any material from Lima. According to Straneo & Vereshagina (1991), only the LT ♂ is labelled 'Puno', whilst three other type specimens are generically labelled 'Peru'; these authors mentioned the presence in CSt of two further specimens from Callanga, not belonging to the type series, which are described in this paper as a new species (B. (A.) straneoi n. sp.). Straneo (1993) subsequently, and rather surprisingly, attributed the LT and two specimens of the type series to the locality 'Lima', quoting a specimen labelled 'Puno' from the original description. In our opinion, assuming the Southern distribution of the species of the jelskii group in the Peruvian Andes and based on the labels by Jelski in the examined material, B. (A.) jelskii probably occurs only in the Puno area. Differential diagnosis Among the Peruvian Agraphoderus, only four species display sinuate (or nearly straight) sides at the base of pronotum: B. (A.) jelskii, B. (A.) abramalagae n.

sp., B. (A.) straneoi n. sp. and B. (A.) procerus n. sp. In comparison with the other species, B. (A.) jelskii is distinguished by more deeply impressed elytral striae, more convex intervals, especially in ♂♂, and distinctly pear–shaped elytra. Moreover, the median lobe of the aedeagus of B. (A.) jelskii is nearly cylindrical, regularly curved and not ventrally depressed. Re–description Habitus as in fig. 1. Overall length of the LT ♂ 7.7 mm (PLT ♀ 7.8 mm). Dorsal surface dark brown, dull (♀) or moderately shiny (♂) with marked polygonal microsculpture, much more evident on elytra. Antennae, legs and mouth parts reddish–brown. Brachypterous. Head moderately large, eyes convex, temples as long as 1/3 of eyes. Clypeus bisetose; labrum transverse, 6–setose. Frontal impressions short and superficial. Frons between eyes smooth and shiny, with sparse tiny punctures. Terminal labial


palpomere with very thin and sparse hairs; penultimate palpomere bisetose, with a short apical seta. Median tooth of mentum prominent and moderately excavate at apex. Antennae short, hardly reaching the base of pronotum, with antennomeres 4–10 only a little longer than wide. Pronotum decidedly wider than long (width/ length = 1.33). Microsculpture evident only at sides, disk smooth and shiny, with regular tiny punctures on the whole surface. One basal impression on each side, superficial, linear and impunctate. Mid longitudinal line well impressed between anterior and posterior submarginal sulci, which are scarcely evident. Lateral margins narrowly bordered on overall length, rounded and markedly sinuate in basal 3rd. Anterior and posterior margins not beaded. Front angles scarcely prominent; hind angles right (fig. 7). Two lateral setae on each side, one at hind angles and one at about ¾ from base. Prosternal process glabrous, cuneate and not margined at apex. Elytra oval elongate (length from basal margin to apex/width = ♂ 1.49, ♀ 1.51), pear–shaped and moderately convex on disk. Shoulders obtuse, without denticles. Scutellar stria usually evident between striae 1 and 2. No setigerous punctures near base. Sides rounded and sinuate near apex; lateral border narrow. Usually three setae on each elytra, the 1st at basal 5th and in the 3rd interval or on 3rd stria, the following adjoining the 2nd stria. Striae smooth or weakly punctate, superficial but evident to apex, all equally impressed but more impressed in males than in females. Intervals flat (♀) or hardly convex (♂); 2nd interval wider than 1st and as wide as 3rd. Metepisterna short, slightly longer than wide. Abdominal sterna IV–VI glabrous except for the pair of central setae; a short transverse impression together with sparse, hardly visible punctures are present at each side. Legs stout. Mesotibiae crenulate at the external edge; male mesotibiae and metatibiae distally not swollen and without inner spines or denticles. Metatrochanters less than half length of femora. 5th tarsomeres with one pair of setae above and three pairs beneath. Male protarsomeres 1–3 triangular and strongly dilated. Metatarsomeres 1–4 externally furrowed. Aedeagus (fig. 15) slender (length 1.95 mm), with enlarged basal bulb and median lobe long, almost cylindrical and in lateral view regularly rounded, distally very thin and bent downwards, but turning up at apex; the apical blade is, in dorsal view, triangular and bluntly pointed (fig. 20). Ostium in dorsal position, large and very long, extended to the basal bulb. Right paramere almost straight. Distribution and habitat The only known specimens of B. (A.) jelskii are the type series described by Tschitschérine (1997). According to the above considerations this material probably originates from the Puno area. Nothing is known of the environmental conditions of the sites inhabited by this species or its ecology.

299

1 mm

Animal Biodiversity and Conservation 34.2 (2011)

Fig. 1. B. (Agraphoderus) jelskii (Tschitschérine, 1897). Fig. 1. B. (Agraphoderus) jelskii (Tschitschérine, 1897).

Blennidus (Agraphoderus) procerus n. sp. (figs. 2, 7, 11–14, 17, 22, 32) Type locality Peru, Cuzco, Abra Lares, m 4400. Type material HT ♂, Peru, Cuzco, Abra Lares, m 4400, 2 II 1994, M. Etonti leg. (CGi). PTT: 23 ♂♂ 13 ♀♀, same data as HT (CAl, CCa, CGi, CMo, MSNM, BMNH); 3 ♂♂ 2 ♀♀, Peru, Accha Alta, Abra Lares, Calca/Cuzco, m 4450, 29 XII 1999, M. Etonti legit (CMa). Etymology The specific epithet refers to the slender, oval elongate habitus.


1 mm

300

Fig. 2. B. (Agraphoderus) procerus n. sp. Fig. 2. B. (Agraphoderus) procerus sp. n.

Differential diagnosis Blennidus (Agraphoderus) procerus n. sp. is one of the four Peruvian Agraphoderus species with sinuate (or nearly straight) sides at the base of pronotum; this species, however, displays some variability, from sinuate to nearly straight sides (figs. 11–14). It is distinguished from B. (A.) abramalagae n. sp. by a more slender habitus and more weakly sinuate sides of pronotum; from B. (A.) straneoi n. sp. by a more slender and less convex habitus as well as a smaller aedeagus (the aedeagus is always small even in the largest specimens); from B. (A.) jelskii by the more superficial elytral striae, the elytra not pear–shaped and the median lobe of the aedeagus, which is more angulately inserted on the basal bulb, more slender at middle and ventrally depressed. Description Habitus as in fig. 2. Overall length of the HT ♂ 7.5 mm (PTT ♂♂ 6.5–8.3, ♀♀ 7.3–8.8 mm). Dorsal surface dark brown, a little paler at base and apex

Allegro & Giachino

of pronotum, dull (♀) or moderately shiny (♂) with marked polygonal microsculpture, much more evident on elytra. Antennae, legs and mouth parts reddish– brown. Brachypterous. Head large, eyes moderately convex, sometimes more or less flattened in both ♂♂ and ♀♀; temples as long as 1/3 of eyes. Clypeus bisetose; labrum transverse, 6–setose. Frontal impressions short and superficial. Frons between eyes smooth and shiny, with sparse tiny punctures. Terminal labial palpomere with thin and sparse hairs; penultimate palpomere bisetose and with a short apical seta. Median tooth of mentum prominent and not excavate at apex. Antennae short, hardly reaching the base of pronotum, with antennomeres 4–10 only a little longer than wide. Pronotum moderately transverse and very variable in shape (width/length = 1.18–1.33). Microsculpture evident only at sides, nearly absent on disk, with regular and very tiny punctures on the whole surface. In some specimens some lateral transverse wrinkles are evident. One basal impression on each side, superficial, linear and impunctate. Mid longitudinal line superficial, sometimes barely visible; submarginal sulci scarcely evident. Lateral margins narrowly bordered on overall length, weakly sinuate to nearly straight at basal 3rd (figs 7 and 11–14). Anterior margin unbordered, the posterior bordered only at sides. Base sinuate to nearly straight at sides. Front angles very slightly prominent; hind angles from nearly right to obtuse (figs. 7, 11–14). Two lateral setae on each side, one at hind angles and one at about ¾ from base. Prosternal process glabrous, cuneate and not margined at apex. Elytra slender, oval elongate (length from basal margin to apex/width  =  1.50–1.58), narrow at base and subdepressed on disk (fig. 32). Microsculpture markedly impressed in both sexes. Shoulders obtuse, without denticles. Scutellar stria usually evident between striae 1 and 2. No setigerous punctures near base. Sides rounded and scarcely sinuate near apex; lateral border narrow. Usually 3 setae on each elytra, the 1st often at basal 5th and in the 3rd interval or on 3rd stria (sometimes very close to 2nd stria), the following adjoining the 2nd stria. Striae smooth, superficial but evident to apex, all equally impressed, but more so in males than in females. Intervals flat (♀) or slightly convex (♂); 2nd interval wider than 1st and as wide as or narrower than 3rd. Metepisterna short, a little longer than wide. Abdominal sterna IV–VI glabrous except for the pair of central setae; no transverse impressions nor punctures are evident at sides. Legs stout. Mesotibiae crenulate at the external edge; male mesotibiae and metatibiae distally not swollen and without inner spines or denticles. Metatrochanters shorter than half femora. 5th tarsomeres with one pair of setae above and 2 pairs beneath. Male protarsomeres 1–3 triangular and strongly dilated. Metatarsomeres 1–4 externally not furrowed. Aedeagus (fig. 17) slender (length 1.85 mm), with median lobe sharply inserted on the large basal bulb, in lateral view slightly curved, distally very


Animal Biodiversity and Conservation 34.2 (2011)

301

thin and angulately bent downwards, but turning up at apex; in dorsal view, the median lobe is wide at base and the apical blade is triangular and bluntly pointed (fig. 22); a certain degree of variability, especially in the angulation of the apical blade, can be observed within populations even from a single site (figs. 25–30). Ostium in dorsal position, large and very long, extended nearly to the basal bulb. Right paramere curved. Distribution and habitat At present B. (A.) procerus sp. n. is only known from the type locality, the Abra Lares pass near Cuzco, Southern Peru. At this site (4,400 m a.s.l.), which is characterized by Andean grassland, B. (A.) procerus n. sp. was collected together with B. (A.) mesotibialis (Etonti legit, in CMa). Blennidus (Agraphoderus) abramalagae n. sp. (figs. 3, 10, 18, 23)

Type material HT ♂, Peru, Cuzco, Abra Malaga, m 4400, 20 IV 1990, M. Etonti leg. (CGi). PTT: 2 ♂♂ 3 ♀♀, same data as HT (CGi, CAl); 1 ♀ Cordillera Vilcabamba, Salcantay, m 4200, IV 1992, Divàk leg. (CSc).

1 mm

Type locality Peru, Cuzco, Abra Malaga, m 4,400.

Etymology The specific epithet derives from the type locality, the Abra Malaga pass, near Cuzco, as a noun, in the genitive case. Differential diagnosis Blennidus (Agraphoderus) abramalagae n. sp. is one of the four Peruvian Agraphoderus species with sinuate (or nearly straight) sides at the base of pronotum. It is distinguished from the other species by an on average larger size and relatively wider elytra (L/W  =  1.42–1.43); it differs from B. (A.) procerus n. sp. and B. (A.) straneoi n. sp. by more markedly sinuate sides of pronotum and from B. (A.) jelskii by slightly impressed elytral striae and nearly flat intervals in both sexes (more impressed striae and more convex intervals in B. (A.) jelskii ♂♂). Moreover, the median lobe of the aedeagus is slender and distally almost straight (angulately bent downwards in B. (A.) procerus, B. (A.) straneoi and B. (A.) jelskii). Description Habitus as in fig. 3. Overall length of the HT ♂ 8.7 mm (PTT ♂♂ 8.0–8.1, ♀♀ 8.2–9.0 mm). Dorsal surface dark brown, sometimes with bluish lustre, moderately shiny (♂) or dull (♀) with marked polygonal microsculpture, much more evident on elytra. Antennae, legs and mouth parts reddish–brown. Brachypterous. Head moderately large, eyes convex in both sexes; temples as long as 1/2.5 of eyes. Clypeus bisetose, a little excavate at middle; labrum transverse, 6–setose. Frontal impressions superficial and barely visible. Frons between eyes smooth and shiny, with sparse

Fig. 3. B. (Agraphoderus) abramalagae n. sp. Fig. 3. B. (Agraphoderus) abramalagae sp. n.

tiny punctures. Terminal labial palpomere with thin, sparse hairs; penultimate palpomere bisetose and with a short apical seta. Median tooth of mentum prominent and excavate at apex. Antennae short, hardly reaching the base of pronotum, with antennomeres 4–10 only a little longer than wide. Pronotum transverse (width from basal margin to apex/length = 1.30–1.35). Microsculpture evident only at sides, disk smooth and shiny. One basal impression on each side, superficial, linear and impunctate. Mid longitudinal line superficial, sometimes hardly visible, only impressed between the submarginal sulci, which are barely visible. Lateral margins narrowly bordered on overall length, sinuate at basal 3rd. Anterior margin unbordered, the posterior bordered at sides; base markedly sinuate at sides. Front angles very scarcely prominent; hind angles nearly right or obtuse (fig. 10). Two lateral setae on each side, one at hind angles and one at about ¾ from base. Prosternal process glabrous, cuneate and widely margined at apex.


302

Allegro & Giachino

one pair of setae above and two or three pairs beneath. Male protarsomeres 1–3 triangular and strongly dilated. Metatarsomeres 1–4 not furrowed externally. Aedeagus (fig. 18) slender (length 2.12 mm), with median lobe sharply inserted on the large basal bulb, in lateral view slightly curved, ventrally depressed, distally thin and almost straight, slightly turning up at apex; in dorsal view, the median lobe is wide at base and the apical blade is triangular and bluntly pointed (fig. 23). Ostium in dorsal position, large and very long, extended nearly to the basal bulb. Right paramere curved. Distribution and habitat At present B. (A.) abramalagae n. sp. is known from the type locality, the Abra Malaga pass near Cuzco, and from Nevado Salcantay, not far from Abra Malaga, in Southern Peru. These sites (4,200–4,400 m a.s.l.) are characterized by Andean grassland.

1 mm

Blennidus (Agraphoderus) etontii n. sp. (figs. 4, 9, 19, 24)

Fig. 4. B. (Agraphoderus) etontii n. sp. Fig. 4. B. (Agraphoderus) etontii sp. n.

Elytra oval (length/width  =  1.42–1.43), subdepressed on disk. Microsculpture more impressed in ♀♀. Shoulders obtuse, without denticles. Scutellar stria usually evident between striae 1 and 2. No setigerous punctures near base. Sides rounded and scarcely sinuate near apex; lateral keel moderately large. Usually 3 setigerous punctures on each elytron, the 1st foveate at basal 5th and in the 3rd interval, the second just behind middle and adjoining the 2nd stria; the third before apex and on the 2nd stria. Striae smooth, superficial but evident to apex, all equally and weakly impressed in both sexes. Intervals flat in both sexes; 2nd interval wider than 1st and as wide as 3rd. Metepisterna longer than wide. Abdominal sterna IV–VI glabrous except for the pair of central setae; no transverse impressions or punctures evident at sides. Legs stout. Mesotibiae crenulate at the external edge; male mesotibiae distally not swollen and without inner spines or denticles. Metatrochanters shorter than half of femora. 5 th tarsomeres with

Type locality Peru, Lares, Abra Lares, m 4,000. Type material HT ♂, Peru, Lares, Abra Lares, m 4000, 28 XII 1998, M. Etonti leg. (CGi). PTT: 1 ♀, same data as HT (CAl); 7 ♂♂ 2 ♀♀, Peru, Pampa Corral, P. Lares– Cuzco, m 200, 26  XII  1999, legit M. Etonti (CAl, CGi, CMa); 2 ♀♀, Peru, Pampa Corral, P. Lares, m 4250, 29 XII 1998, legit M. Etonti; 1 ♀, Peru, Pampa Corral, P. Lares, m 4180, 29 XII 1998, legit M. Etonti (CAl, CGi). Etymology We are pleased to dedicate this species to Mirto Etonti, collector of the specimens forming the type series of this species. Differential diagnosis Blennidus (Agraphoderus) etontii n. sp. is the only species in the jelskii group, as far as we know, with completely rounded sides of pronotum. Moreover, it is the only species with male metatibiae preapically swollen. It is distinguished from the other Peruvian Agraphoderus that have rounded sides of the pronotum by the morphology of the median lobe of aedeagus, which suggests its relationship with the species of the jelskii group. Description Habitus as in fig. 4. Overall length of the HT ♂ 7.7 mm (PTT ♂♂ 7.6–8.3, ♀♀ 7.7–9.3 mm). Dorsal surface dark brown, with elytra a little paler than head and pronotum, moderately shiny (♂) or dull (♀); marked polygonal microsculpture, much more evident on elytra. Antennae, legs and mouth parts reddish–brown. Brachypterous. Head moderately large, eyes convex in both sexes; temples as long as 1/2.75 of eyes. Clypeus


bisetose, slightly excavate at middle; labrum transverse, 6–setose. Frontal impressions superficial and barely visible. Frons between eyes smooth and shiny, with sparse tiny punctures. Terminal labial palpomere with sparse, thin hairs; penultimate palpomere bisetose and with a short apical seta. Median tooth of mentum prominent and excavate at apex. Antennae short, hardly reaching the base of pronotum, with antennomeres 4–10 only a little longer than wide. Pronotum transverse (width/length = ♂ 1.30, ♀  1.34). Microsculpture evident only at sides, disk smooth and shiny. One basal impression on each side, superficial, linear and impunctate. Mid longitudinal line superficial, only impressed between the submarginal sulci, which are barely visible. Lateral margins rounded and narrowly bordered on overall length. Anterior margin unbordered, the posterior bordered at sides; base sinuate at sides. Front angles very scarcely prominent; hind angles obtuse (fig. 9). Two lateral setae on each side, one at hind angles and one about ¾ from base. Prosternal process glabrous, cuneate and widely margined at apex. Elytra oval elongate (length from basal margin to apex/width =1.47–1.52), moderately convex. Microsculpture more impressed in ♀♀. Shoulders obtuse, without denticles. Scutellar stria superficial and scarcely evident between striae 1 and 2. No setigerous punctures near base. Sides rounded and scarcely sinuate near apex; lateral keel narrow. Usually three setae on each elytron (the HT with four setae on left elytron), the 1st at basal 4th and in the 3rd interval or on 3rd stria, the following adjoining the 2nd stria. Striae smooth, superficial but evident to apex, all equally and weakly impressed in both sexes. Intervals flat in both sexes; 2nd interval wider than 1st and as wide as 3rd. Metepisterna longer than wide. Abdominal sterna IV–VI glabrous except for the pair of central setae; only a few tiny punctures are hardly visible at sides. Legs stout. Mesotibiae crenulate at the external edge; male mesotibiae distally not swollen and without inner spines or denticles; male metatibiae preapically moderately swollen. Metatrochanters shorter than half femora. 5th tarsomeres with one pair of setae above and 2 pairs beneath. Male protarsomeres 1–3 triangular and strongly dilated. Metatarsomeres 1–4 externally not furrowed. Aedeagus (fig. 19) slender (length 2.10 mm), with median lobe sharply inserted on the large basal bulb, in lateral view hardly curved, ventrally depressed, distally thin and slightly bent downwards, turning up at apex; in dorsal view, the median lobe is wide at base and the apical blade is triangular and acutely pointed (fig. 24). Ostium in dorsal position, large and very long, extended nearly to the basal bulb. Right paramere curved. Distribution and habitat At present B. (A.) etontii n. sp. is only known from the type locality, the Abra Lares pass near Lares, Cuzco Dept., Southern Peru. The collecting site (4,000  m  a.s.l.) is characterized by Andean grassland. This species is sympatric but not syntopic

303

1 mm

Animal Biodiversity and Conservation 34.2 (2011)

Fig. 5. B. (Agraphoderus) straneoi n. sp. Fig. 5. B. (Agraphoderus) straneoi sp. n.

with B. (A.) procerus sp. n., which seems to live at a higher altitude (4,400 m a.s.l.); on the contrary, it was collected in syntopy with B. (A.) mesotibialis (Etonti legit, in CMa). Blennidus (Agraphoderus) straneoi n. sp. (figs. 5, 8, 16, 21, 31) Type locality Peru, Callanga. Type material HT ♂, Peru, Callanga (CSt at MSNM). PTT: 1 ♀, same data as HT (CAl). Note: the two specimens of the type series (ex Staudinger according to Straneo & Vereshagina, 1991) were found in the Straneo Collection, wrongly attributed to B. (A.) jelskii. Etymology We dedicate this species to the author of the most important study concerning the Peruvian Agraphoderus, Stefano L. Straneo, renowned specialist of world Pterostichinae.


304

Allegro & Giachino

6

7

8

9 11 10

12

13

14

1 mm

Figs. 6–14. Basal angles of the pronotum of Blennidus spp.: 6. B. (A.) jelskii (Tschitsch.); 7. B. (A.) procerus n. sp.; 8. B. (A.) straneoi n. sp.; 9. B. (A.) etontii n. sp.; 10. B. (A.) abramalagae n. sp.; 11–14. Variability in B. (A.) procerus n. sp. from the same locality. Figs. 6–14. Ángulos basales del pronoto de Blennidus spp.: 6. B. (A.) jelskii (Tschitsch.); 7. B. (A.) procerus sp. n.; 8. B. (A.) straneoi sp. n.; 9. B. (A.) etontii sp. n.; 10. B. (A.) abramalagae sp. n.; 11–14. Variabilidad en B. (A.) procerus sp. n. de la misma localidad.

Differential diagnosis Blennidus (Agraphoderus) straneoi n. sp. is one of the four Peruvian Agraphoderus species with sinuate (or nearly straight) sides at the base of pronotum. Respect to B. (A.) abramalagae n. sp. and B. (A.) jelskii, the sides are more weakly sinuate; moreover, B. (A.) straneoi n. sp. is distinguished from B. (A.) jelskii also by slightly impressed elytral striae and nearly flat intervals in both sexes (more impressed striae and more convex intervals in B. (A.) jelskii ♂♂) as well as not pear–shaped elytra. It differs from B. (A.) procerus n. sp. by the oval, less slender habitus, by the more convex elytra, with marked apical declivity, and by a larger aedeagus as well (B. (A.) procerus n. sp. displays a smaller aedeagus even in the largest specimens). Description Habitus as in fig. 5. Overall length of the HT ♂ 7.7 mm (PT ♀ 8.7 mm). Dorsal surface brown, moderately shiny (♂) or dull (♀) with marked polygonal microsculpture, much more evident on elytra. Antennae, legs and mouth parts reddish–brown. Brachypterous. Head moderately large, eyes convex in both sexes; temples as long as 1/3 of eyes. Clypeus bisetose, moderately concave at middle; labrum transverse, 6–setose. Frontal impressions superficial and directed inward. Frons between eyes smooth and shiny, with sparse very tiny punctures. Terminal labial palpomere with thin and sparse hairs; penultimate palpomere bisetose and with a short apical seta. Median tooth of mentum prominent and

excavate at apex. Antennae short, hardly reaching the base of pronotum, with antennomeres 4–10 only a little longer than wide. Pronotum transverse (width/length  =  ♂ 1.25, ♀  1.35). Microsculpture more evident at sides, where some transverse wrinkles are evident; disk smooth. One basal impression on each side, superficial, linear and impunctate. Mid longitudinal line superficial, only impressed between the submarginal sulci, which are scarcely evident. Lateral margins narrowly bordered on overall length, nearly linear (♂) or weakly sinuate (♀) at basal 3rd. Anterior margin unbordered, the posterior bordered at sides; base sinuate at sides. Front angles scarcely prominent; hind angles obtuse (fig. 8). Two lateral setae on each side, one at hind angles and one at about 4/5 from base. Prosternal process glabrous, cuneate and not margined at apex. Elytra oval (length from basal margin to apex/width = 1.45 in both sexes), moderately convex on disk with marked apical declivity (fig. 31). Microsculpture more impressed in ♀. Shoulders obtuse, without denticles. Scutellar stria usually evident between striae 1 and 2. No setigerous punctures near base. Sides rounded and scarcely sinuate near apex; lateral keel moderately large. Usually three setae on each elytra, the 1st at basal 5th and in the 3rd interval or on 3rd stria, the following adjoining the 2nd stria. Striae smooth or weakly punctuate (♀), superficial but evident to apex, all equally and slightly impressed in both sexes. Intervals flat in both sexes; 2nd interval wider than 1st and wider than 3rd.


Animal Biodiversity and Conservation 34.2 (2011)

305

15

16

17

18

19

1 mm

Figs. 15–19. Aedeagus in lateral view of Blennidus spp. (holotypes): 15. B. (A.) jelskii (Tschitsch.); 16. B. (A.) straneoi n. sp.; 17. B. (A.) procerus n. sp.; 18. B. (A.) abramalagae n. sp.; 19. B. (A.) etontii n. sp. Figs. 15–19. Edeago en vista lateral de Blennidus spp. (holotipos): 15. B. (A.) jelskii (Tschitsch.); 16. B. (A.) straneoi sp. n.; 17. B. (A.) procerus sp. n.; 18. B. (A.) abramalagae sp. n.; 19. B. (A.) etontii sp. n.

Metepisterna longer than wide. Abdominal sterna IV–VI glabrous except for the pair of central setae; no transverse impressions or punctures are evident at sides. Legs stout. Mesotibiae crenulate at the external edge; male mesotibiae distally not swollen and without inner spines or denticles. Metatrochanters shorter than half femora. 5th tarsomeres with one pair of setae above and two pairs beneath. Male protarsomeres 1–3 triangular and strongly dilated. Aedeagus (fig. 16) slender (length 2.0 mm), with median lobe roundly inserted on the large basal bulb, in lateral view hardly curved at middle,

thin and ventrally depressed, distally tapering and slightly bent downwards, turning up at apex; in dorsal view, the median lobe is wide at base and the apical blade is triangular and moderately sharply pointed (fig. 21). Ostium in dorsal position, large and very long, extended nearly to the basal bulb. Right paramere curved. Distribution and habitat At present B. (A.) straneoi sp. n. is only known from the type locality, Callanga. This locality probably refers not to the town near Lima but to the ancient archaeological site in the Cuzco Dept. (Southern


306

Allegro & Giachino

20

21

22

23

24

1 mm Figs. 20–24. Aedeagus in dorsal view of Blennidus spp. (holotypes): 20. B. (A.) jelskii (Tschitsch.); 21. B. (A.) straneoi n. sp.; 22. B. (A.) procerus n. sp.; 23. B. (A.) abramalagae n. sp.; 24. B. (A.) etontii n. sp. Figs. 20–24. Edeago en vista dorsal de Blennidus spp. (holotipos): 20. B. (A.) jelskii (Tschitsch.); 21. B. (A.) straneoi sp. n.; 22. B. (A.) procerus sp. n.; 23. B. (A.) abramalagae sp. n.; 24. B. (A.) etontii sp. n.

Peru), where four of the five species of the jelskii group are found. Also two specimens of B. (A.) mesotibialis (ex Staudinger, in CSt) are labelled 'Callanga' and this species is only known from the Cuzco area. Nothing is known of the environment inhabited by this species, nor of its ecology. Discussion Straneo (1993) lists only three Blennidus (Agraphoderus) species from the Southern Peruvian Andes: B. (A.) mesotibialis (Straneo, 1993) from an area between Puno and Cuzco, B. (A.) aulacostigma (Tschitschérine, 1897) and B. (A.) jelskii (Tschitschérine, 1897) from Puno. In this paper we raise this number to seven, as four new species from the Cuzco Dept. are described, all of them with probable affinities with B. (A.) jelskii. The affinities of B. (A.) mesotibialis and B. (A.) aulacostigma remain unclear. Unfortunately, almost nothing is known about the Bolivian Blennidus species, which could be southern relatives

of the species belonging to the jelskii group. The Blennidus species belonging to the jelskii group are probably related to one another based on the uniform morphology of male genitalia, which are characterized in lateral view by a thin apex, distally tapering and not sinuate, and in dorsal view by a triangular apical blade as well as a large dorsal ostium. Inside this group, B. (A.) procerus n. sp., B. (A.) abramalagae n. sp., B. (A.) etontii n. sp. and B. (A.) straneoi n. sp., all distributed in the Cuzco Dept., are morphologically more similar to each other than to B. (A.) jelskii, which is also more isolated from a geographical point of view (Puno). The apparently separate distribution of the Blennidus (Agraphoderus) species of the jelskii group, each one inhabiting a specific Andean valley or district, suggests that these flightless high–altitude carabids constitute a complex of stenoendemic species living in restricted geographic areas and in narrow altitudinal ranges. They probably originated by allopatric speciation due to the isolating effects of Pleistocene climatic cycles combined with the


Animal Biodiversity and Conservation 34.2 (2011)

307

25

26

27

28

29

30

HT

1 mm Figs. 25–30. Variability of the aedeagus in lateral view of B. (A.) procerus n. sp. from the same locality (HT, fig. 25). Figs. 25–30. Variabilidad del edeago en vista lateral de B. (A.) procerus sp. n. de la misma localidad (HT, fig. 25).

effects of orographic barriers, in particular when a warmer climate shifted the montane biota upwards during the interglacials, causing its fragmentation and isolation (Simpson–Vuilleumier, 1971). The same differentiation processes occurred to other high–altitude Andean carabid genera such as Trechisibus and Oxytrechus (Allegro et al., 2008; Moret, 2005). According to our preliminary observations carried out on the Blennidus (Agraphoderus) populations inhabiting the Cordillera Blanca, many other species

of this genus must be considered as stenoendemics. The data of Straneo (1993) referring to species covering wide geographical ranges need careful reexamination, as they could refer to specimens wrongly attributed to a single species (Allegro & Giachino, unpublished). Based on these considerations, it is very likely that new surveys in unexplored areas of Andes together with research into other entomological collections will bring to light many more unknown taxa.


308

Allegro & Giachino

31

32

Figs. 31–32. Schematic habitus in lateral view of: 31. B. (A.) straneoi n. sp.; 32. B. (A.) procerus n. sp. Figs. 31–32. Esquemas del hábitus en vista lateral de: 31. B. (A.) straneoi sp. n.; 32. B. (A.) procerus sp. n.

Acknowledgements The authors wish to thank Achille Casale and Pierre Moret for their valuable advice and for revising the manuscript; Luca Picciau (MRST), Maurizio Pavesi (MSNM) and Boris Kataev (ZIRA) for the loan of the types of the Mateu, Straneo and Tschitschérine collections; Kipling Will and Sergio Roig–Juñent for their critical suggestions as referees of this article; Max Barclay (BMNH) for linguistic revision. Moreover, G. Allegro is grateful to the missionaries and volunteers of Operazione Mato Grosso, for their fundamental support and for making this research possible. References Allegro, G., 2010. A peculiar new species of Blennidus Motschulsky subgen. Agraphoderus Bates from the Andes of the Cordillera Blanca (Peru) (Coleoptera: Carabidae: Pterostichini). Studies and Reports of District Museum Prague–East, Taxonomical Series, 6(1–2): 1–7. Allegro, G., Giachino, P.M. & Sciaky R., 2008. Notes on some Trechini (Coleoptera Carabidae) of South America with description of new species from Chile, Ecuador and Peru. In: Biodiversity of South America, I. Memoirs on Biodiversity, 1: 131–171 (P. M. Giachino, Ed.). World Biodiversity Association onlus, Verona, Italy. Lorenz, W., 2005a. Systematic list of extant ground beetles of the world (Insecta Coleoptera ‘Geadephaga’: Trachypachidae and Carabidae incl. Paussi-

nae, Cicindelinae, Rhysodinae). Second Edition. Tutzing. – 2005b. Nomina Carabidarum. A directory of the scientific names of ground beetles (Insecta Coleoptera ‘Geadephaga’: Trachypachidae and Carabidae incl. Paussinae, Cicindelinae, Rhysodinae). Second Edition, Tutzing. Moret, P., 1995. Contribution à la connaissance du genre néotropical Blennidus Motschulsky, 1865. 1ère partie (Coleoptera, Harpalidae, Pterostichinae). Bulletin de la Societé entomologique de France, 100(5): 489–500. – 2005. Los coleópteros Carabidae del páramo en los Andes del Ecuador. Sistemática, ecología y biogeografía. Quito, Pontificia Universidad Católica del Ecuador, Centro de Biodiversidad y Ambiente, Monografía 2, Gruppo Editoriale il Capitello, Torino. Simpson–Vuilleumier, B., 1971. Pleistocene changes in the fauna and flora of South–America. Science, 173: 771–780. Straneo, S. L., 1993. Nuove specie del genere Ogmopleura Tschitschérine (Coleoptera, Carabidae, Pterostichinae) del Perù e dell’Ecuador e chiave per la loro determinazione. Annali del Museo civico di Storia naturale 'G. Doria', 89: 351–399. Straneo, S. L. & Vereshagina, T., 1991. Sui tipi del genere Ogmopleura Tschitschérine descritti da Tschitschérine nel 1896 e 1898. Bollettino della Società entomologica italiana, Genova, 122(3): 195–204. Tschitschérine, T., 1897. Matériaux pour servir à l’étude des Féroniens. III. Horae Societatis Entomologicae Rossicae, 30: 260–351.


Animal Biodiversity and Conservation 34.2 (2011)

Review 309

Ecological genetics of freshwater fish: a short review of the genotype–phenotype connection O. Vidal & J. L. García–Marín

Vidal, O. & García–Marín, J. L., 2011. Ecological genetics of freshwater fish: a short review of the genotype– phenotype connection. Animal Biodiversity and Conservation, 34.2: 309–317. Abstract Ecological genetics of freshwater fish: a short review of the genotype–phenotype connection.— Molecular ecology or ecological genetics is an expanding application of population genetics which has flourished in the last two decades but it is dominated by systematic and phylogeographic studies, with relatively little emphasis on the study of the genetic basis of the process of adaptation to different ecological conditions. The relationship between genotype and adaptive phenotypes is weak because populations are often difficult to quantify and experiments are logistically challenging or unfeasible. Interestingly, in freshwater fish, studies to characterize the genetic architecture of adaptive traits are not as rare as in other vertebrate groups. In this review, we summarize the few cases where the relationship between the ecology and genetics of freshwater fish is more developed, namely the relationship between genetic markers and ecological phenotypes. Key words: Ecological genetics, Molecular ecology, Genotype–phenotype relationship, Adaptation, Landscape genetics, Species introduction. Resumen Genética ecológica de los peces de agua dulce: una breve revisión de la conexión genotipo–fenotipo.— La ecología molecular o la genética ecológica es una aplicación de la genética de poblaciones que durante las dos últimas décadas ha sufrido un proceso de expansión. Sin embargo, en la ecología molecular predominan los estudios sistemáticos y filogeográficos, con relativamente poco énfasis en el análisis de la base genética del proceso de adaptación a diferentes condiciones ecológicas. Esta relación entre genotipo y fenotipo adaptativo es poco evidente, porque las poblaciones son difíciles de cuantificar y los experimentos son logísticamente complicados. Es interesante destacar que en peces de agua dulce estos estudios no son tan poco frecuentes como en otros grupos de vertebrados. En esta revisión, nuestra intención es resumir los pocos casos en los cuales la relación entre ecología y genética de peces continentales está más desarrollada, principalmente entre marcadores genéticos y fenotipos ecológicos. Palabras clave: Genética ecológica, Ecología molecular, Interacción genotipo–fenotipo, Adaptación, Genética del paisaje, Introducción de especies. (Received: 17 I 11; Conditional acceptance: 30 V 11; Final acceptance: 28 IX 11) O. Vidal & J. L. García–Marín, Lab. d’Ictiologia Genètica, Dept of Biology, University of Girona, E–17071 Girona, Catalonia, Spain. Corresponding author: O. Vidal. E–mail: oriol.vidal@udg.cat

ISSN: 1578–665X

© 2011 Museu de Ciències Naturals de Barcelona


Vidal & García–Marín

310

Introduction

Genotype–phenotype relationships

Ecological genetics or molecular ecology is the application of molecular genetic tools to ecological problems and the field has developed enormously in the last two decades. Molecular ecology is, however, dominated by systematic and phylogeographic studies, and there is relatively little emphasis on the study of the genetic basis of the process of adaptation to different ecological conditions (Bowen, 1999). In the words of Avise (2006), 'mechanistic connections between observable genotypes and ecologically relevant phenotypes often remain black boxes' and 'molecular ecology will become a more mature discipline when it also incorporates ecologically germane information on genotype–phenotype connections.' In this review we intend to summarize studies where the relationship between genotypes and phenotypes may be regarded as adaptative. We focus on genetic studies of ecologically important traits in freshwater fish. The link between observable genotypes and ecologically relevant phenotypes is difficult to assess in most species. First, population sizes and demographic statistics are often difficult to quantify, and experiments are logistically challenging and often unfeasible. Second, although new molecular methods are becoming available (i.e., RAD–tags), molecular markers or gene sequences may be difficult to obtain. And third, identification of ecologically relevant phenotypes is not easy. In most cases a trait is considered adaptive when it is suspected to improve survival of the individual or its fitness, but very few studies to date have focused on the effects that alternative phenotypes may have from an evolutionary point of view. Identifying the genetic architecture of such traits (and the causal mutations of the alternative phenotypes) allows detection of evidence for selection in the DNA, and it may therefore highlight the evolutionary importance of that genotype. Interestingly, because in some freshwater fish species it is possible to study population dynamics as well as to breed several generations in a few years, studies to characterize the genetic architecture of adaptive traits are not as rare as those in other vertebrate groups. A review focused on these well described cases may thus provide useful information for other non–model organisms for which little previous data is available. We also briefly discuss two important topics that, from our point of view, may provide valuable information on the genotype–phenotype relationship: landscape genetics and invasive species. Landscape genetics is a new discipline focused on how physical landscape influences genetic traits of the populations (Guillot et al., 2005). Although not directly related to adaptedness or phenotype modelling, such interactions could alter the viability of the population and thus have drastic evolutionary consequences. Furthermore, biological invasions exemplify the process of species adaptation to new environments, and its analysis can add novel information about the relationship between ecology and genetics.

Several freshwater fish species are used as models to study the genetic architecture of putative adaptive traits. Such traits are related to survival and thus fitness of individuals. Although some traits (e.g. coloration) can safely be assumed to reflect adaptation, in other cases it is not obvious whether a particular phenotype is adaptive. Genetic analyses focused on identifying the causal mutations of a phenotype can help to determine if the trait is really adaptive through a further detection of evidence for selection in the DNA sequence (Nielsen, 2005). In general, the most frequent strategy to undertake this approach requires fully interfertile diverse populations showing different phenotypic traits, development of molecular markers in the targeted species (or close relatives) and setting up experimental crosses, which allow the identification of quantitative trait loci (or QTL) and the actual causative mutations. All these requirements restrict the range of species being studied, although some models are now fully established. Gasterosteus aculeatus The three–spine stickleback (Gasterosteus aculeatus) is a complex species that shows repeated episodes of colonization of freshwater habitats, including both lakes and rivers, from marine stocks (reviewed in McKinnon & Rundle, 2002). These colonization events have caused divergence in several traits (which could be adaptive) such as morphology, behaviour and physiology. The variety of environments inhabited by three–spine stickleback (ranging from fresh water to ocean) may explain the importance of behavioural and physiological traits that may be critical for adaption to new physical and chemical conditions. In contrast, morphological traits have been linked to the different predators they might encounter (Marchinko, 2009). Interestingly, all marine and fresh water forms are interfertile. An extensive set of microsatellite markers and over 45.000 single nucleotide polymorphisms (SNPs) are available (Hohenlohe et al., 2010) to carry out genome–wide typing assays. Several experimental crosses have been developed, yielding different QTLs with effects on gill structure, spine length, number of lateral plates (Peichel et al., 2001), armour (Colosimo et al., 2004, 2005) and pigmentation (Miller et al., 2007). In some of these studies, particular genes have been statistically associated with phenotypic changes, linking Ectodysplasin to changes in armour and Kit ligand to changes in pigmentation. The only phenotype with known causal mutations is pelvic size, which is modified by regulatory mutations in Pitx1 (Shapiro et al., 2004). Characterization of the genetic architecture of these traits has allowed novel approaches to confirm the traits as adaptations and to analyze the dynamics of such genotypes in populations (Kitano et al., 2008; Makinen et al., 2008; Marchinko, 2009; Chan et al., 2010). QTL analyses in these species have also confirmed a role for sex chromosomes in speciation, related to the location of loci involved in behavioural


Animal Biodiversity and Conservation 34.2 (2011)

isolation along the X chromosome (Kitano et al. 2009). Other studies have focused on specific genotypes, using a candidate gene approach. Generally, following this strategy, the function of the targeted gene is known and it is considered to be adaptive. The analysis of sweet taste receptors (Hashiguchi et al., 2007) or immunity related genes (Reusch et al., 2001; Wegner et al., 2008) would fall within this category. More recently, the analysis of plasma level mRNA expression and the genomic signature for selection on one gene in the thyroid hormone signalling pathway, TSHβ2, have shown significant differences between ancestral marine and stream– resident ecotypes (Kitano et al., 2010). These results suggest that evolutionary changes in hormonal signalling may have played an important role in the adaptive divergence of sticklebacks. Astyanax mexicanus The Mexican tetra (Astyanax mexicanus) has two forms, the surface and the cave–dwelling forms. Interestingly, the troglodyte form (or cavefish) presents some phenotypic traits shared with other animals adapted to the dark environment of caves: reduced eyes (or even complete regression) and loss of pigmentation, including albinism in some cases. Moreover, cavefish probably have more taste buds, larger jaw size and more fat reserves (reviewed in Jeffery, 2009). All these traits seem relevant for survival under some obvious cave conditions, especially absolute darkness and the low quantity of nutrients. At least thirty different caves in Mexico have been noted to harbour cave forms of A. mexicanus, and there is evidence of parallel evolution and several independent origins for the cave forms (Borowsky, 2008). Blind Astyanax mexicanus from different caves are already used as a model organism to study molecular pathways of eye development (Jeffery, 2008). In addition, several experimental crosses have been designed to identify the genetic architecture of other interesting traits. Protas et al. (2006) identified cave– specific mutations of OCA2 causing albinism, thus demonstrating evolution by convergence. Interestingly, although pigmentation loss is usual in cavefish, albinos are not present in all caves. Some depigmentated (or brown) phenotypes have been found to be caused by two different mutations of MC1R (Gross et al., 2009), although other brown mutations are likely to exist. As well as confirming convergence this pattern indicates that certain genes are frequent targets of mutation, at least in regressive phenotypes. Other traits that have been analyzed in experimental crosses include eye size and development of the jaw, teeth and taste buds (Protas et al., 2007) as well as total length and metabolism (Protas et al., 2008). Although the causative mutations for these phenotypes have not been identified and only QTLs have been located, these results represent a first approximation to describe pathways of molecular evolution. Moreover, they suggest some hypotheses about the evolutionary mechanisms behind the phenotypes, such as genetic drift and indirect selection through pleiotropy.

311

Xiphophorus spp. The genus Xiphophorus includes two of the most popular aquarium fishes: swordtails and platyfishes. Both these fishes, and their intercrosses were used in early genetic mapping studies of pigmentation (Gordon, 1931), which have since allowed the identification of at least 33 loci linked to coloration (reviewed in Basolo, 2006). However, no causal mutations have been identified so far, and some specific (adaptive?) functions of color patterns are still not fully understood (Price et al., 2008). Interestingly, the analysis of pigmentation led to the genetic characterization of melanoma formation. The mechanism involves the interaction of two loci, Tu and R, (Baudler et al., 1997), and while the Tu locus has been identified as the Xmrk gene (Wittbrodt et al., 1989), the R locus remains unknown. As hybridization is the trigger for melanoma appearance in Xiphophorus, these genes may have had a role in early speciation as a postzygotic isolation mechanism (Schartl, 2008). Sexual selection, based on female preference for the color phenotype linked to melanoma (Fernandez & Morris, 2008), larger males carrying the melanoma–related allele (Fernandez & Bowser, 2010), and intrasexual selection (Fernandez, 2010) have also been suggested to be the main forces maintaining oncogenes segregating in natural populations. Swordtails and platyfish are also model organisms in the study of sexual behavior, but information about the genetic basis of such traits is limited (Rosenthal & Garcia de Leon, 2006). Interestingly, Xmrk males display increased aggression in mirror image trials, and thus may experience a competitive advantage over wild–type males (Fernandez, 2010). Moreover, there is strong evidence for some reproductive tactics being inheritable (Zimmerer, 1989) and the P locus, affecting sexual maturity, fecundity and size (Kallman and Borkoski, 1978) has been linked to MC4R (Lampert et al., 2010). The sword of swordtails is a classical example of a sexually selected trait because females (as well as predators) prefer larger swords (Basolo, 1990; Rosenthal et al., 2001). Through the analysis of gene expression in developing and regenerating caudal fins of swordtails and platyfish, the main signalling pathway (involving the fgfr1 gene) has been identified (Offen et al., 2008). However, there is no information available on the effects of this sexual selection at a molecular level. Other species Even though zebrafish (Danio rerio) is a model species in genetics and development, natural variation has not been analyzed thoroughly. To our knowledge, only one experimental cross between wild and lab strains has been carried out (Wright et al., 2006b) to detect QTLs affecting anti–predator behavioural and morphological differentiation. Due to the design of the cross, these traits might be more related to domestication than to adaptation to natural environments. The same experimental cross has also been used


Vidal & García–Marín

312

to detect epistatic regulation acting on these traits (Wright et al., 2006a). Interestingly, the comparison of colour and stripe pattern development between different species of the genus Danio has shown the importance of genes such as kit or fms in the evolutionary change of adult phenotypes (Quigley et al., 2004; Mills et al., 2007). African cichlid fishes (family Cichlidae) are a classical example of adaptive radiation, and there are probably more than one thousand species in just three lakes in East Africa (Lakes Malawi, Tanganyika and Victoria) (Turner et al., 2001). They have been used to investigate the molecular mechanisms of adaptation and speciation, focusing on traits that may have relevant roles in diversification. The huge variation in their colour patterns is a central feature in the behaviour and evolution of these species and may have contributed to their explosive speciation. Although the full genetic architecture of coloration is not known, the orange blotch (OB) phenotype, associated to one single QTL near the c–ski1 gene (Streelman et al., 2003), is due to a cis–regulatory mutation of the Pax7 gene (Roberts et al., 2009). It has recently been suggested that sensory adaptation could be a key feature in the radiation of these species, and the genetic causes of visual pigment diversity have been identified (Hofmann et al., 2009; Carleton et al., 2010). Interestingly, regulatory changes (and not structural mutations) have been linked to the parallel evolution of fish visual systems in Lake Tanganika and Lake Malawi (O’Quin et al., 2010). Other interesting traits are related to morphological variations. Cichlid fishes forage by different modes, which are related to the functional design of the feeding apparatus. Thus, QTLs under directional selection related to jaw and teeth development have been described, and bmp4 has been suggested as a causative gene (Albertson et al., 2003, 2005). Several other freshwater fishes are currently cultured to supply both ornamental and food market demands. Although these domesticated populations do not provide information about the genetic architecture of natural variation or adaptation, they have been used in several experimental crosses. Tilapia (Oreochromis spp.) crosses have allowed identification of a genetic mechanism for sex determination (Shirak et al., 2006) and characterization of QTLs related to immune response and growth (Cnaani et al., 2004); medaka fish (Oryzias latipes) havebeen used to characterize the role of tyr gene in albinism (Inagaki et al., 1998; Koga et al., 1995; Tsutsumi et al., 2006); and multiple experimental crosses of rainbow trout (Oncorhynchus mykiss) have been carried out to focus on the genetics of a variety of traits, including the immune response to pathogens (Nichols et al., 2003; Johnson et al., 2008), maturation (Haidle et al., 2008) and smoltification (Nichols et al., 2008). Other farmed fishes have also been used to characterize immune responses because pathogens can be an issue in intensive production systems. Although most works have focused on seawater species, some freshwater studies are worth mentioning, such as the genetic characterization of brook charr (Salvelinus

fontinalis) resistance to Aeromonas salmonicida (Croisetière et al., 2008), and the herpesvirus resistance of common carp (Cyprinius carpio) (Rakus et al., 2009). These domestic populations are also providing valuable information about the effects of captive breeding. Strikingly, wild–born offspring of captive Oncorhynchus mykiss show a rapid fitness decline (Araki et al., 2007). If these results are confirmed in other organisms, they could seriously compromise some conservation strategies of endangered species. New molecular tools The recent development of high throughput genetic techniques has made genomic information accessible and affordable for a great number of organisms through next generation sequencing and DNA microarrays. In general, these new DNA sequencing technologies require a reference genome to align the massive number of short sequences that they produce. This alignment will yield high quality results only when the sample genome does not differ from the reference genome (Frith et al., 2010). However, as only a few fish species have annotated genomes (five in the ENSEMBL database, http://www.ensembl. org/index.html) these techniques have limited applications, but they have a very promising future in the field of comparative genomics. Genomes have been analysed in attempts to identify regions of synteny in some species, such as African cichlids, where low–coverage genomes of five phenotypically and ecologically diverse Lake Malawi species have been compared (Loh et al., 2008), and also in Danio rerio and Astyanax mexicanus, for which no genetic map was available until recently (Gross et al., 2008). Genomic analyses have also been carried out in G. aculeatus using next–generation sequencing of RAD markers (Baird et al., 2008; Hohenlohe et al., 2010), a very promising approach. This methodology requires the construction and massive sequencing of a RAD tag library, with DNA fragments having a restriction site at one end and randomly sheared at the other. Without using any previous genomic information, this arrangement reliably identifies new SNP markers. To date, these markers have been used in genetic mapping, phylogeography, population genomics and even whole–genome sequencing (reviewed in Rowe et al., 2011). Although they have been used in only one freshwater fish species, G. aculeatus, they may yield very interesting results in other non–model species. The lake whitefish species complex, Coregonus clupeaformis, is another case study for adaptive radiation (Bernatchez et al., 1999) and for which several genomic tools have been developed, including cDNA microarrays (Rise et al., 2004). Experimental crosses have already been used to identify QTLs affecting reproductive isolation (Rogers & Bernatchez, 2006), and to characterize loci under parallel selection within the complex (Rogers & Bernatchez, 2005).Genome– wide expression patterns have also been described using microarray technology, resulting in the first transcriptome analyses focused on speciation. This


Animal Biodiversity and Conservation 34.2 (2011)

technology provided the first evidence that parallel phenotypic evolution in C. clupeaformis also involves parallel transcriptional changes (Derome et al., 2006). This has been confirmed in another species of the same genus, C. artedi (Derome & Bernatchez, 2006), and in African cichlids (O’Quin et al., 2010), stressing the importance of gene regulation in rapid phenotypic divergence. However, one study focusing specifically on transcriptional divergence of a set of six candidate genes in three species of the genus (C. clupeaformis, C. artedi and C. albula) points out that parallelism on gene expression is not preserved among species (Jeukens et al., 2009). The availability of this microarray has also allowed the characterization of expression QTLs (eQTLs), showing a sex bias in the transcriptional genetic architecture of lake whitefish (Whiteley et al., 2008). Future research in this field may also help to disentangle whether adaptation is mainly driven by cis–regulatory mutations or by polymorphisms in coding regions. Landscape genetics Neutral molecular markers are of little relevance for the study of adaptive processes. However, the integration of population genetic data obtained from those markers with landscape ecology in the new emerging field of landscape genetics (Manel et al., 2003) represents a new way to integrate ecology and genetics (Holderegger & Wagner, 2006; Sork & Waits, 2010). Understanding how landscape variables and environmental features explain gene flow and genetic discontinuities between populations contributes to our understanding of biological processes such as metapopulation dynamics or speciation (Meeuwig et al., 2010). Such studies are also of applied conservation value in identifying current anthropogenic barriers that reduce gene flow and genetic diversity, in predicting the effects of proposed management alternatives on genetic variation and population connectivity, and in identifying potential biological corridors to assist with reserve design (Storfer et al., 2006, 2010). In fish species, for example, such studies have provided evidence for the role of contemporary landscape features in shaping the observed patterns of genetic diversity at smaller geographic scales in Salvelinus alpinus (Castric et al., 2001) and Oncorhynchus clarki (Neville et al., 2006). Other studies have also shown constrained gene flow due to local adaptation in Salmo salar (Dionne et al., 2008) and selection gradients as responsible for cryptic population divergence in Gasterosteus aculeatus (McCairns & Bernatchez, 2008). Species introductions Invasive species are a leading cause of biodiversity loss and global change, particularly in freshwater ecosystems (Mack et al., 2000; Clavero & García–Berthou, 2005), and they are also an underappreciated tool to study ecology and evolution at large spatial and

313

temporal scales (Rice & Sax, 2005; García–Berthou, 2007). Invasive species are a unique model to characterize adaptation and ecological dynamics in newly occupied territories. Nevertheless, genetic studies of invasive species have generally focused on identifying source populations and the routes of spread and the role of genetic diversity, again without much integration with ecological studies. An exception is the study of rapid evolution and adaptation following introductions of fruit flies and salmonids (Huey et al., 2000, 2005). The paradigm of conservation biology is that populations have demographic and genetic thresholds below which non–adaptative, random forces (e.g. genetic drift) prevail over adaptive processes and extinction risk increases (Soulé, 1985). Invasive species seem to defeat this paradigm, given the often low number of individuals initially introduced and low genetic diversity (Lindholm et al., 2005; Poulet et al., 2009; Vidal et al., 2009). This apparent paradox is explained by an admixture of often unadvertised, multiple source populations and mechanisms that mitigate the impact of low genetic diversity (Roman & Darling, 2007). The guppy Poecilia reticulata is a model system in the study of evolutionary ecology, sexual selection, and behaviour (Endler, 1995; Bronikowski et al., 2002; Reznick et al., 2004; Ghalambor et al., 2004; Magurran, 2009). Experimental introductions of guppy in Trinidad are a textbook example of rapid, predation–mediated life history evolution. The literature on this species is beyond the scope of our review but illustrates the wealth of information that can be gained with genetic tools, including, for instance, evidence of: i) genetic basis in mate–choice (Brooks & Endler, 2001) and shoaling (Huizinga et al., 2009) behaviours; ii) high levels of multiple mating in wild populations (Hain & Neff, 2007); and iii) high rates of natural selection and evolution in the wild (Reznick & Ghalambor, 2005). Conclusions Genetics and ecology have a long history of little mutual appreciation and exchange (Berry & Bradshaw, 1992; Cain & Provine, 1992). Despite the prosperity of conservation genetics and phylogeography, particularly in large vertebrates, studies that combine both disciplines are still rare. We have tried to give an overview of the information available on the adaptive value of genetic variation in freshwater fish. Although this topic has been investigated for a few species, for most species we do not know how genetic variation affects survival and fitness in the wild. This knowledge is vital to understand the potential response of freshwater fish to global environmental change and to mitigate the impacts of the latter. For a number of reasons, however, we are optimistic about the future of ecological genetics of freshwater fish. The studies summarized reflect the importance and benefits of such an approach. Advances in the biology of fish are often delayed in comparison with those in birds or plants (e.g., García–Berthou, 2007), but with the development of new genetic resources and genomics tools this field is likely to expand greatly in years to come.


314

Acknowledgements We wish to thank E. García–Berthou, the associate editor C. Vilà and an anonymous reviewer for their valuable comments and suggestions to improve the quality of the paper. Financial support for our research was provided by the Spanish Ministry of Science (CGL2006–11652– C02/BOS and CGL2009–12877–C02). References Albertson, R. C., Streelman, J. T. & Kocher, T. D., 2003. Directional selection has shaped the oral jaws of Lake Malawi cichlid fishes. Proceedings of the National Academy of Sciences USA, 100: 5252–5257. Albertson, R. C., Streelman, J. T., Kocher, T. D. & Yelick, P. C., 2005. Integration and evolution of the cichlid mandible: the molecular basis of alternate feeding strategies. Proceedings of the National Academy of Sciences USA, 102: 16287–16292. Araki, H., Cooper, B. & Blouin, M. S., 2007. Genetic effects of captive breeding cause a rapid, cumulative fitness decline in the wild. Science, 318: 100–103. Avise, J. C., 2006. The ontogeny of molecular ecology: A review of Molecular Ecology by Joanna R. Freeland. Molecular Ecology, 15: 2687–2689. Baird, N. A., Etter, P. D., Atwood, T. S., Currey, M. C., Shiver, A. L., Lewis, Z. A., Selker, E. U., Cresko, W. A. & Johnson, E. A., 2008. Rapid SNP discovery and genetic mapping using sequenced RAD markers. PLoS One, 3:e3376. Basolo, A. L., 1990. Female preference predates the evolution of the sword in swordtail fish. Science, 250: 808–810. – 2006. Genetic linkage and color polymorphism in the southern platyfish (Xiphophorus maculatus): a model system for studies of color pattern evolution. Zebrafish, 3: 65–83. Baudler, M., Duschl, J., Winkler, C., Schartl, M. & Altschmied, J., 1997. Activation of transcription of the melanoma inducing Xmrk oncogene by a GC box element. Journal of Biological Chemistry, 272: 131–137. Bernatchez, L., Chouinard, A. & Lu, G., 1999. Integrating molecular genetics and ecology in studies of adaptive radiation: whitefish, Coregonus sp., as a case study. Biological Journal of the Linnean Society, 68: 173–194. Berry, R. J. & Bradshaw, A. D., 1992. Genes in the real world. In: Genes in Ecology: 431–449 (R. J. Berry, T. J. Crawford & G. M. Hewitt, Eds.). Blackwell Science, Oxford. Borowsky, R., 2008. Restoring sight in blind cavefish. Current Biology, 18: R23–R24. Bowen, B. W., 1999. Preserving genes, species, or ecosystems? Healing the fractured foundations of conservation policy. Molecular Ecology, 8: S5–10. Bronikowski, A. M., Clark, M. E., Rodd, F. H. & Reznick, D. N., 2002. Population–dynamic consequences of predator–induced life history variation in the guppy (Poecilia reticulata). Ecology, 83:

Vidal & García–Marín

2194–2204. Brooks, R. & Endler, J. A., 2001. Direct and indirect sexual selection and quantitative genetics of male traits in guppies (Poecilia reticulata). Evolution, 55: 1002–1015. Cain, A. J. & Provine, W. B., 1992. Genes and ecology in history. In Genes in Ecology: 3–28 (R. J. Berry, T. J. Crawford & G. M. Hewitt, Eds.). Blackwell Science, Oxford. Carleton, K. L., Hofmann, C. M., Klisz, C., Patel, Z., Chircus, L. M., Simenauer, L. H., Soodoo, N., Albertson, R. C. & Ser, J. R., 2010. Genetic basis of differential opsin gene expression in cichlid fishes. Journal of Evolutionary Biology, 23: 840–853. Castric, V., Bonney, F. & Bernatchez, L., 2001. Landscape structure and hierarchical genetic diversity in the brook charr, Salvelinus fontinalis. Evolution, 55: 1016–1028. Chan, Y. F., Marks, M. E., Jones, F. C., Villarreal, G. Jr., Shapiro, M. D., Brady, S. D., Southwick, A. M., Absher, D. M., Grimwood, J., Schmutz, J., Myers, R. M., Petrov, D., Jónsson, B., Schluter, D., Bell, M. A. & Kingsley, D. M., 2010. Adaptive evolution of pelvic reduction in sticklebacks by recurrent deletion of a Pitx1 enhancer. Science, 327: 302–305. Clavero, M. & García–Berthou, E., 2005. Invasive species are a leading cause of animal extinctions. Trends in Ecology and Evolution, 20: 110. Cnaani, A., Zilberman, N., Tinman, S., Hulata, G. & Ron, M., 2004. Genome–scan analysis for quantitative trait loci in an F2 tilapia hybrid. Molecular Genetics and Genomics, 272: 162–172. Colosimo, P. F., Hosemann, K. E., Balabhadra, S., Villarreal, G., Jr., Dickson, M., Grimwood, J., Schmutz, J., Myers, R. M., Schluter, D. & Kingsley, D. M., 2005. Widespread parallel evolution in sticklebacks by repeated fixation of Ectodysplasin alleles. Science, 307: 1928–1933. Colosimo, P. F., Peichel, C. L., Nereng, K., Blackman, B. K., Shapiro, M. D., Schluter, D. & Kingsley, D. M., 2004. The genetic architecture of parallel armor plate reduction in threespine sticklebacks. PLoS Biology, 2: E109. Croisetière, S., Tarte, P. D., Bernatchez, L. & Belhumeur, P., 2008. Identification of MHC class IIbeta resistance/susceptibility alleles to Aeromonas salmonicida in brook charr (Salvelinus fontinalis). Molecular Immunology, 45: 3107–3116. Derome, N. & Bernatchez, L., 2006. The transcriptomics of ecological convergence between 2 limnetic coregonine fishes (Salmonidae). Molecular Biology and Evolution, 23: 2370–2378. Derome, N., Duchesne, P. & Bernatchez, L., 2006. Parallelism in gene transcription among sympatric lake whitefish (Coregonus clupeaformis Mitchill) ecotypes. Molecular Ecology, 15: 1239–1249. Dionne, M., Caron, F., Dodson, J. J. & Bernatchez, L., 2008. Landscape genetics and hierarchical genetic structure in Atlantic salmon: the interaction of gene flow and local adaptation. Molecular Ecology, 17: 2382–2396. Endler, J. A., 1995. Multiple–trait coevolution and environmental gradients in guppies. Trends in


Animal Biodiversity and Conservation 34.2 (2011)

Ecology & Evolution, 10: 22–29. Fernandez, A. A., 2010 A cancer–causing gene is positively correlated with male aggression in Xiphophorus cortezi. Journal of Evolutionary Biology, 23: 386–396. Fernandez, A. A. & Bowser, P. R., 2010 Selection for a dominant oncogene and large male size as a risk factor for melanoma in the Xiphophorus animal model. Molecular Ecology, 19: 3114–23. Fernandez, A. A. & Morris, M. R., 2008.Mate choice for more melanin as a mechanism to maintain a functional oncogene. PNAS, 105: 13503–13507. Frith, M. C., Wan, R. & Horton, P., 2010 Incorporating sequence quality data into alignment improves DNA read mapping. Nucleic Acids Research, 38: e100. García–Berthou, E., 2007 The characteristics of invasive fishes: what has been learned so far? Journal of Fish Biology, 71 (Suppl. D): 33–55. Ghalambor, C. K., Reznick, D. N. & Walker, J. A., 2004. Constraints on adaptive evolution: the functional trade–off between reproduction and fast–start swimming performance in the Trinidadian guppy (Poecilia reticulata). American Naturalist, 164: 38–50. Gordon, M., 1931. The hereditary basis for melanosis in hybrids of Mexican killifishes. Proceedings of the National Academy of Sciences USA, 17: 276–280. Gross, J. B., Borowsky, R. & Tabin, C. J., 2009. A novel role for Mc1r in the parallel evolution of depigmentation in independent populations of the cavefish Astyanax mexicanus. PLoS Genetics, 5: e1000326. Gross, J. B., Protas, M., Conrad, M., Scheid, P. E., Vidal, O., Jeffery, W. R., Borowsky, R. & Tabin, C. J., 2008. Synteny and candidate gene prediction using an anchored linkage map of Astyanax mexicanus. Proceedings of the National Academy of Sciences USA, 105: 20106–20111. Guillot, G., Estoup, A., Mortier, F. & Cosson J. F., 2005 A spatial statistical model for landscape genetics. Genetics, 170: 1261–1280. Haidle, L., Janssen, J. E., Gharbi, K., Moghadam, H. K., Ferguson, M. M. & Danzmann, R. G., 2008. Determination of quantitative trait loci (QTL) for early maturation in rainbow trout (Oncorhynchus mykiss). Marine Biotechnology (NY), 10: 579–592. Hain, T. J. A. & Neff, B. D., 2007. Multiple paternity and kin recognition mechanisms in a guppy population. Molecular Ecology, 16: 3938–3946. Hashiguchi, Y., Furuta, Y., Kawahara, R. & Nishida, M., 2007. Diversification and adaptive evolution of putative sweet taste receptors in threespine stickleback. Gene, 396: 170–179. Hofmann, C. M., O’Quin, K. E., Marshall, N. J., Cronin, T. W., Seehausen, O. & Carleton, K. L., 2009. The eyes have it: regulatory and structural changes both underlie cichlid visual pigment diversity. PLoS Biology, 7: e1000266. Hohenlohe, P. A., Bassham, S., Etter, P. D., Stiffler, N., Johnson, E. A. & Cresko, W. A., 2010. Population genomics of parallel adaptation in threespine stickleback using sequenced RAD tags. PLoS Genetics, 6: e1000862.

315

Holderegger, R. & Wagner, H. H., 2006. A brief guide to landscape genetics. Landscape Ecology, 21: 793–796. Huey, R. B., Gilchrist, G. W., Carlson, M. L., Berrigan, D. & Serra, L., 2000. Rapid evolution of a geographic cline in size in an introduced fly. Science, 287: 308–309. Huey, R. B., Gilchrist, G. W. & Hendry, A. P., 2005. Using invasive species to study evolution: case studies with Drosophila and salmon. In Species Invasions: Insights into Ecology, Evolution, and Biogeography: 139–164 (D. F. Sax, J. J. Stachowicz & S. D. Gaines, Eds.). Sinauer, Sunderland. Huizinga, M., Ghalambor, C. K. & Reznick, D. N., 2009. The genetic and environmental basis of adaptive differences in shoaling behaviour among populations of Trinidadian guppies, Poecilia reticulata. Journal of Evolutionary Biology, 22: 1860–1866. Inagaki, H., Koga, A., Bessho, Y. & Hori, H., 1998. The tyrosinase gene from medakafish: transgenic expression rescues albino mutation. Pigment Cell Research, 11: 283–290. Jeffery, W. R., 2008. Emerging model systems in evodevo: cavefish and microevolution of development. Evolution & Development, 10: 265–272. – 2009. Chapter 8. Evolution and development in the cavefish Astyanax. Current Topics in Developmental Biology, 86: 191–221. Jeukens, J., Bittner, D., Knudsen, R. & Bernatchez, L., 2009. Candidate genes and adaptive radiation: insights from transcriptional adaptation to the limnetic niche among coregonine fishes (Coregonus spp., Salmonidae). Molecular Biology and Evolution, 26: 155–166. Johnson, N. A., Vallejo, R. L., Silverstein, J. T., Welch, T. J., Wiens, G. D., Hallerman, E. M. & Palti, Y., 2008. Suggestive association of major histocompatibility IB genetic markers with resistance to bacterial cold water disease in rainbow trout (Oncorhynchus mykiss). Marine Biotechnology, 10: 429–437. Kallman, K. D. & Borkoski, V., 1978. A sex–linked gene controlling the onset of sexual maturity in female and male platyfish (Xiphophorus maculatus), fecundity in females and adult size in males. Genetics, 89: 79–119. Kitano, J., Bolnick, D. I., Beauchamp, D. A., Mazur, M. M., Mori, S., Nakano, T. & Peichel, C. L., 2008. Reverse evolution of armor plates in the threespine stickleback. Current Biology, 18: 769–774. Kitano, J., Lema, S. C., Luckenbach, J. A., Mori, S., Kawagishi, Y., Kusakabe, M., Swanson, P. & Peichel, C. L., 2010. Adaptive divergence in the thyroid hormone signaling pathway in the stickleback radiation. Current Biology, 20: 2124–2130. Kitano, J., Ross, J. A., Mori, S., Kume, M., Jones, F. C., Chan, Y. F., Absher, D. M., Grimwood, J., Schmutz, J., Myers, R. M., Kingsley, D. M. & Peichel, C. L., 2009. A role for a neo–sex chromosome in stickleback speciation. Nature, 461: 1079–1083. Koga, A., Inagaki, H., Bessho, Y. & Hori, H., 1995. Insertion of a novel transposable element in the tyrosinase gene is responsible for an albino mutation in the medaka fish, Oryzias latipes. Molecular


316

& General Genetics, 249: 400–405. Lampert, K. P., Schmidt, C., Fischer, P., Volff, J. N., Hoffmann, C., Muck, J., Lohse, M. J., Ryan, M. J. & Schartl, M., 2010. Determination of onset of sexual maturation and mating behavior by melanocortin receptor 4 polymorphisms. Current Biology, 20: 1729–1734. Lindholm, A. K., Breden, F., Alexander, H. J., Chan, W. K., Thakurta, S. G. & Brooks, R., 2005. Invasion success and genetic diversity of introduced populations of guppies Poecilia reticulata in Australia. Molecular Ecology, 14: 3671–3682. Loh, Y. H., Katz, L. S., Mims, M. C., Kocher, T. D., Yi, S. V. & Streelman, J. T., 2008. Comparative analysis reveals signatures of differentiation amid genomic polymorphism in Lake Malawi cichlids. Genome Biology, 9: R113. Mack, R. N., Simberloff, D., Lonsdale, W. M., Evans, H., Clout, M. & Bazzaz, F. A., 2000. Biotic invasions: causes, epidemiology, global consequences, and control. Ecological Applications, 10: 689–710. Magurran, A. E., 2009. Evolutionary ecology: The Trinidadian guppy. Oxford University Press, Oxford. Makinen, H. S., Cano, J. M. & Merila, J., 2008. Identifying footprints of directional and balancing selection in marine and freshwater three–spined stickleback (Gasterosteus aculeatus) populations. Molecular Ecology, 17: 3565–3582. Manel, S., Schwartz, M. K., Luikart, G. & Taberlet, P., 2003. Landscape genetics: combining landscape ecology and population genetics. Trends in Ecology & Evolution, 18: 189–197. Marchinko, K. B., 2009. Predation’s role in repeated phenotypic and genetic divergence of armor in threespine stickleback. Evolution, 63: 127–138. Meeuwig, M. H., Guy, C.S., Kalinowski, S. T. & Fredenberg W. A., 2010. Landscape influences on genetic differentiation among bull trout populations in a stream–lake network. Molecular Ecology, 19: 3620–3633. McCairns, R. & Bernatchez, L., 2008. Landscape genetic analyses reveal cryptic population structure and putative selection gradients in a large–scale estuarine environment. Molecular Ecology, 17: 3901–3916. McKinnon, J. S. & Rundle, H. D., 2002. Speciation in nature: the threespine stickleback model systems. Trends in Ecology & Evolution, 17: 480–488. Miller, C. T., Beleza, S., Pollen, A. A., Schluter, D., Kittles, R. A., Shriver, M. D. & Kingsley, D. M., 2007. Cis–Regulatory changes in Kit ligand expression and parallel evolution of pigmentation in sticklebacks and humans. Cell, 131: 1179–1189. Mills, M. G., Nuckels, R. J. & Parichy, D. M., 2007. Deconstructing evolution of adult phenotypes: genetic analyses of kit reveal homology and evolutionary novelty during adult pigment pattern development of Danio fishes. Development, 134: 1081–1090. Neville, H. M., Dunham, J. B. & Peacock, M. M., 2006. Landscape attributes and life history variability shape genetic structure of trout populations in a stream network. Landscape Ecology, 21: 901–916. Nichols, K. M., Bartholomew, J. & Thorgaard, G. H.,

Vidal & García–Marín

2003. Mapping multiple genetic loci associated with Ceratomyxa shasta resistance in Oncorhynchus mykiss. Diseases of Aquatic Organisms, 56: 145–154. Nichols, K. M., Edo, A. F., Wheeler, P. A. & Thorgaard, G. H., 2008. The genetic basis of smoltification– related traits in Oncorhynchus mykiss. Genetics, 179: 1559–1575. Nielsen, R., 2005 Molecular signatures of natural selection. Annual Review of Genetics, 39: 197–218. Offen, N., Blum, N., Meyer, A. & Begemann, G., 2008. Fgfr1 signalling in the development of a sexually selected trait in vertebrates, the sword of swordtail fish. BMC Developmental Biology, 8: 98. O’Quin, K. E., Hofmann, C. M., Hofmann, H. A. & Carleton, K. L., 2010. Parallel evolution of opsin gene expression in African cichlid fishes. Molecular Biology and Evolution, 27: 2839–2854. Peichel, C. L., Nereng, K. S., Ohgi, K. A., Cole, B. L., Colosimo, P. F., Buerkle, C. A., Schluter, D. & Kingsley, D. M., 2001. The genetic architecture of divergence between threespine stickleback species. Nature, 414: 901–905. Poulet, N., Balaresque, P., Aho, T. & Björklund, M., 2009. Genetic structure and dynamics of a small introduced population: the pikeperch, Sander lucioperca, in the Rhône delta. Genetica, 135: 77–86. Price, A. C., Weadick, C. J., Shim, J. & Rodd, F. H., 2008. Pigments, patterns, and fish behavior. Zebrafish, 5: 297–307. Protas, M., Conrad, M., Gross, J. B., Tabin, C. & Borowsky, R., 2007. Regressive evolution in the Mexican cave tetra, Astyanax mexicanus. Current Biology, 17: 452–454. Protas, M., Tabansky, I., Conrad, M., Gross, J. B., Vidal, O., Tabin, C. J. & Borowsky, R., 2008. Multi– trait evolution in a cave fish, Astyanax mexicanus. Evolution & Development, 10: 196–209. Protas, M. E., Hersey, C., Kochanek, D., Zhou, Y., Wilkens, H., Jeffery, W. R., Zon, L. I., Borowsky, R. & Tabin, C. J., 2006. Genetic analysis of cavefish reveals molecular convergence in the evolution of albinism. Nature Genetics, 38: 107–111. Quigley, I. K., Manuel, J. L., Roberts, R. A., Nuckels, R. J., Herrington, E. R., MacDonald, E. L. & Parichy, D. M., 2004. Evolutionary diversification of pigment pattern in Danio fishes: differential fms dependence and stripe loss in D. albolineatus. Development, 132: 89–104. Rakus, K. Ł., Wiegertjes, G. F., Adamek, M., Siwicki, A. K., Lepa, A. & Irnazarow, I., 2009. Resistance of common carp (Cyprinus carpio L.) to Cyprinid herpesvirus–3 is influenced by major histocompatibility (MH) class II B gene polymorphism. Fish & Shellfish Immunology, 26: 737–743. Reusch, T. B., Haberli, M. A., Aeschlimann, P. B. & Milinski, M., 2001. Female sticklebacks count alleles in a strategy of sexual selection explaining MHC polymorphism. Nature, 414: 300–302. Reznick, D. N., Bryant, M. J., Roff, D., Ghalambor, C. K. & Ghalambor, D. E., 2004. Effect of extrinsic mortality on the evolution of senescence in guppies. Nature, 431: 1095–1099.


Animal Biodiversity and Conservation 34.2 (2011)

Reznick, D. N. & Ghalambor, C. K., 2005. Selection in nature: experimental manipulations of natural populations. Integrative and Comparative Biology, 45: 456–462. Rice, J. A. & Sax, D. F., 2005 Testing fundamental evolutionary questions at large spatial and demographic scales: species invasions as an underappreciated tool. In Species Invasions: Insights into Ecology, Evolution, and Biogeography: 291–308 (D. F. Sax, J. J. Stachowicz & S. D. Gaines, Eds.). Sinauer, Sunderland. Rise, M. L., von Schalburg, K. R., Brown, G. D., Mawer, M. A., Devlin, R. H., Kuipers, N., Busby, M., Beetz–Sargent, M., Alberto, R., Gibbs, A. R., Hunt, P., Shukin, R., Zeznik, J. A., Nelson, C., Jones, S. R., Smailus, D. E., Jones, S. J., Schein, J. E., Marra, M. A., Butterfield, Y. S., Stott, J. M., Ng, S. H., Davidson, W. S. & Koop, B. F., 2004. Development and application of a salmonid EST database and cDNA microarray: data mining and interspecific hybridization characteristics. Genome Research, 14: 478–490. Roberts, R. B., Ser, J. R. & Kocher, T. D., 2009. Sexual conflict resolved by invasion of a novel sex determiner in Lake Malawi cichlid fishes. Science, 326: 998–1001. Rogers, S. M. & Bernatchez, L., 2005. Integrating QTL mapping and genome scans towards the characterization of candidate loci under parallel selection in the lake whitefish (Coregonus clupeaformis). Molecular Ecology, 14: 351–361. – 2006. The genetic basis of intrinsic and extrinsic post–zygotic reproductive isolation jointly promoting speciation in the lake whitefish species complex (Coregonus clupeaformis). Journal of Evolutionary Biology, 19: 1979–1994. Roman, J. & Darling, J. A., 2007. Paradox lost: genetic diversity and the success of aquatic invasions. Trends in Ecology & Evolution, 22: 454–464. Rosenthal, G. G., Flores Martinez, T. Y., Garcia de Leon, F. J. & Ryan, M. J., 2001. Shared preferences by predators and females for male ornaments in swordtails. American Naturalist, 158: 146–154. Rosenthal, G. G. & Garcia de Leon, F. J., 2006. Sexual behavior, genes, and evolution in Xiphophorus. Zebrafish, 3: 85–90. Rowe, H. C., Renaut, S. & Guggisbertc, A., 2011. RAD in the realm of next–generation sequencing technologies. Molecular Ecology, 20: 3499–3502. Schartl, M., 2008. Evolution of Xmrk: an oncogene, but also a speciation gene? Bioessays, 30: 822–832. Shapiro, M. D., Marks, M. E., Peichel, C. L., Blackman, B. K., Nereng, K. S., Jonsson, B., Schluter, D. & Kingsley, D. M., 2004. Genetic and developmental basis of evolutionary pelvic reduction in threespine sticklebacks. Nature, 428: 717–723. Shirak, A., Seroussi, E., Cnaani, A., Howe, A. E., Domokhovsky, R., Zilberman, N., Kocher, T. D., Hulata, G. & Ron, M., 2006. Amh and Dmrta2 genes map to tilapia (Oreochromis spp.) linkage group 23 within quantitative trait locus regions for sex deter-

317

mination. Genetics, 174: 1573–1581. Soulé, M. E., 1985. What is conservation biology? Bioscience, 35: 727–734. Sork, V. L. & Waits, L., 2010. Contributions of landscape genetics – approaches, insights, and future potential. Molecular Ecology, 19: 3489–3495. Storfer, A., Murphy, M. A., Evans, J. S., Goldberg, C. S., Robinson, S., Spear, S. F., Dezzani, R., Delmelle, E., Vierling, L. & Waits, L. P., 2006. Putting the ‘landscape’ in landscape genetics. Heredity, 98: 128–142. Storfer, A., Murphy, M. A., Spear, S. F., Holderegger, R. & Waits, L. P., 2010. Landscape genetics: where are we now? Molecular Ecology, 19: 3496–3514. Streelman, J. T., Albertson, R. C. & Kocher, T. D., 2003. Genome mapping of the orange blotch colour pattern in cichlid fishes. Molecular Ecology, 12: 2465–2471. Tsutsumi, M., Imai, S., Kyono-Hamaguchi, Y., Hamaguchi, S., Koga, A. & Hori, H., 2006. Color reversion of the albino medaka fish associated with spontaneous somatic excision of the Tol–1 transposable element from the tyrosinase gene. Pigment Cell Research, 19: 243–247. Turner, G. F., Seehausen, O., Knight, M. E., Allender, C. J. & Robinson, R. L., 2001. How many species of cichlid fishes are there in African lakes? Molecular Ecology, 10: 793–806. Vidal, O., García–Berthou, E., Tedesco, P. A. & García–Marín, J. L., 2009. Origin and genetic diversity of mosquitofish (Gambusia holbrooki) introduced to Europe. Biological Invasions, 12: 841–851. Wegner, K. M., Kalbe, M., Milinski, M. & Reusch, T. B., 2008. Mortality selection during the 2003 European heat wave in three–spined sticklebacks: effects of parasites and MHC genotype. BMC Evolutionary Biology, 8: 124. Whiteley, A. R., Derome, N., Rogers, S. M., St–Cyr, J., Laroche, J., Labbe, A., Nolte, A., Renaut, S., Jeukens, J. & Bernatchez, L., 2008. The phenomics and expression quantitative trait locus mapping of brain transcriptomes regulating adaptive divergence in lake whitefish species pairs (Coregonus sp.). Genetics, 180: 147–164. Wittbrodt, J., Adam, D., Malitschek, B., Maueler, W., Raulf, F., Telling, A., Robertson, S. M. & Schartl, M., 1989. ovel putative receptor tyrosine kinase encoded by the melanoma–inducing Tu locus in Xiphophorus. Nature, 341: 415–421. Wright, D., Butlin, R. K. & Carlborg, O., 2006a. Epistatic regulation of behavioural and morphological traits in the zebrafish (Danio rerio). Behaviour Genetics, 36: 914–922. Wright, D., Nakamichi, R., Krause, J. & Butlin, R. K., 2006b. QTL analysis of behavioral and morphological differentiation between wild and laboratory zebrafish (Danio rerio). Behaviour Genetics, 36: 271–284. Zimmerer, E. J., 1989. Genetic basis for alternative reproductive tactics in the pygmy swordtail, Xiphophorus nigrensis. Evolution, 43: 1298–1307.


366


Animal Biodiversity and Conservation 34.2 (2011)

319

Present status of the endangered limpet Cymbula nigra (Gastropoda, Patellidae) in Ceuta: how do substrate heterogeneity and area accessibility affect population structure?

G. A. Rivera–Ingraham, F. Espinosa & J. C. García–Gómez Rivera–Ingraham, G. A., Espinosa, F. & García–Gómez, J. C., 2011. Present status of the endangered limpet Cymbula nigra (Gastropoda, Patellidae) in Ceuta: how do substrate heterogeneity and area accessibility affect population structure? Animal Biodiversity and Conservation, 34.2: 319–330. Abstract Present status of the endangered limpet Cymbula nigra (Gastropoda, Patellidae) in Ceuta: how do substrate heterogeneity and area accessibility affect population structure?— Cymbula nigra (Gastropoda, Patellidae) is a threatened giant patellid limpet found on the North African coast from Namibia to Algeria. The objective of this study was to estimate the total number of individuals present in Ceuta (Strait of Gibraltar) and to determine the effect of certain physical parameters on population structure and abundance. Between 2006 and 2010 we conducted an exhaustive census in the area. Results indicate that Ceuta could be home to 48,473 individuals. The most important populations were recorded on the North Bay, characterized by its Atlantic influence. While for other similar species, such as Patella ferruginea, human accessibility to the area plays an important role in determining the structure of populations, we found that substrate roughness (small scale topographic heterogeneity) is the main determining factor in this species. Populations located on medium to low topographic heterogeneity substrates showed higher percentages of medium and large size individuals. However, recruitment rates did not differ between substrata of different roughness. Finally, and through the analysis of the C. nigra populations located on some recently constructed jetties, we obtained interesting new data regarding individual growth rates, thus contributing to our knowledge of the population structure of the species. Key words: Limpet, Endangered species, Cymbula nigra, Substrate heterogeneity, Strait of Gibraltar, Ceuta. Resumen Situación actual en Ceuta de la lapa Cymbula nigra (Gastropoda, Patellidae), una especie en peligro: ¿cómo afecta la heterogeneidad del substrato y la accesibilidad del área a la estructura de las poblaciones?— Cymbula nigra (Gastropoda, Patellidae) es una lapa gigante amenazada que se encuentra distribuida por las costas norteafricanas desde Namibia a Argelia. El objetivo del presente estudio era estimar el número total de individuos presentes en Ceuta (Estrecho de Gibraltar) y determinar el efecto de ciertos parámetros físicos sobre la abundancia y estructura de las poblaciones. Para conseguirlo, entre 2006 y 2010, se llevó a cabo un censo exhaustivo en la zona. Los resultados indicaron que Ceuta podría albergar unos 48.473 individuos. Las poblaciones más importantes fueron registradas en la Bahía del Norte de la ciudad, caracterizada por su influencia atlántica. Mientras que en otras especies similares, como Patella ferruginea, la accesibilidad de la zona por parte del hombre juega un papel importante en la determinación de la abundancia y estructura de las poblaciones, nuestros resultados indicaron que en esta especie el principal factor determinante es la rugosidad del sustrato (heterogeneidad topográfica a pequeña escala). En este sentido, aquellas poblaciones localizadas sobre sustratos de media a baja complejidad mostraron mayores porcentajes de individuos de tamaño mediano y grande. Sin embargo, las tasas de reclutamiento no variaron entre sustratos de diferente rugosidad. Finalmente, a través del análisis de poblaciones de C. nigra localizadas sobre algunos diques de construcción reciente, se obtuvieron nuevos e interesantes datos relacionados con tasas individuales de crecimiento, contribuyendo así al conocimiento de la estructura de poblaciones de esta especie. Palabras clave: Lapas, Especies amenazadas, Cymbula nigra, Heterogeneidad del sustrato, Estrecho de Gibraltar, Ceuta. (Received: 28 VI 11; Conditional acceptance: 10 XI 11; Final acceptance: 18 XI 11) Georgina Alexandra Rivera–Ingraham, Free Espinosa, José Carlos García–Gómez, Lab. de Biología Marina, Depto. de Fisiología y Zoología, Univ. de Sevilla, Av. Reina Mercedes 6, 41012, Sevilla, España (Spain). Corresponding author: G. A. Rivera–Ingraham. E–mail: grivera@us.es ISSN: 1578–665X

© 2011 Museu de Ciències Naturals de Barcelona


Rivera–Ingraham et al.

320

Introduction The commonly known Black limpet (Cymbula nigra) (da Costa, 1771) (former Patella nigra) is the largest limpet in Europe, reaching up to 13.3 cm (Rivera–Ingraham et al., 2011a). It is present in the southernmost area of the Iberian peninsula, having been located in Algeciras (Guerra–García et al., 2006; Rivera–Ingraham, 2010), Fuengirola (Spada & Maldonado Quiles, 1974; Grandfils & Vega, 1982; Christiaens, 1983), Mijas (Martínez & Peñas, 1996) and, more recently, in Granada (Moreno & Arroyo, 2008). It has also been found at Alboran Island (Moreno, 2006; Peñas et al., 2006; Templado et al., 2006; Moreno & Arroyo, 2008), although its presence here is considered rare. Along the north African coast, the species occurs from Namibia to Algeria (Koufopanou et al., 1999) and has large populations in Ceuta (Templado et al., 2004; Espinosa et al., 2007) and Melilla (Pasteur–Humbert, 1962). The coast of Senegal was thought to be the centre of the species’ dispersion (Christiaens, 1974) but more recent morphological and molecular studies suggest that the species might have originated on the south–western African coasts (Ridgway et al., 1998; Koufopanou et al., 1999; Templado et al., 2004). Cymbula nigra is currently catalogued as an 'endangered or threatened species' by the Barcelona Convention (Annex II, 1993), as 'strictly protected' by the Berne Convention (Annex II, 1995), and as 'vulnerable' by the Andalusian Red List of Threatened Invertebrate Species (Moreno & Arroyo, 2008). It is surprising therefore that little is known about the biology and ecology of the species. Most of the recent literature that mentions C. nigra is related to phylogenetic studies of the family Patellidae (Ridgway et al., 1998; Koufopanou et al., 1999; Sá–Pinto et al., 2005) or genetic aspects of the species (Espinosa et al., 2010; Nakano & Espinosa, 2010). Recent studies indicate that C. nigra recruits in the upper intertidal levels while the largest individuals occupy the lower intertidal areas (Rivera–Ingraham et al., 2011a) and can be found up to 5 m deep (Rivera–Ingraham, 2010). It is a protandric hermaphrodite species that reproduces mainly in late autumn but also, to a lesser extent, in spring (Frenkiel, 1975; Rivera–Ingraham, 2010). Cymbula nigra is a territorial species, and individuals defend a well defined area from other macroorganisms (Rivera–Ingraham, 2010). Despite the interest in quantifying populations of endangered species (e.g. Paracuellos et al., 2003) this has not been done to date for C. nigra. The present study describes the population of C. nigra from the coast of Ceuta, and estimates the total number of individuals. Moreover, and taking into account that C. nigra has been previously observed to be especially abundant on smooth surfaces (Rivera–Ingraham et al., 2011a), special attention was paid to the influence of substrate heterogeneity on population density. We also analysed the effect of the area’s accessibility on the distribution and population structures, and we compared the results with those previously obtained for other similar species, such as Patella ferruginea. Finally, we studied C. nigra populations located on recently created jetties.Knowing

the date when these structures were finished allowed us to obtain average growth rates and compare these with previous data available for the species. Material and methods Study site The present study was conducted in Ceuta, located on the African coast of the Strait of Gibraltar (fig. 1A) . This area is known to have large C. nigra populations (Rivera–Ingraham, 2010). The coasts of this city are composed of natural rocky shores, beaches and islets. There are also many breakwaters and jetties, and a commercial port. Sampling methods To estimate the total number of C. nigra individuals present in Ceuta we used the methodology described in Rivera–Ingraham et al. (2011b).The complete coastline was divided into 14 sectors (see fig. 1B) and for each one, and only for rocky shores, a total of 10 transects of 10 m were randomly established on the coastline. These transects occupied the whole vertical distribution range of the species (intertidal areas and the upper sublitoral regions, up to 5 m depth). Surveys were always carried out during low tide (tide amplitudes usually ranging from 0.41 to 1.01) and subtidal regions were sampled by diving. Each C. nigra individual located in these transects was measured to the nearest millimetre using a calliper as in Guerra–García et al. (2004), Espinosa (2009) or Rivera–Ingraham (2010). The number of individuals present in these transects was extrapolated to the total length of coastline in the sector, which was measured using 1:9,000 maps. For the specific case of islets, jetties and breakwaters, a complete census was carried out when possible. The complete coastline was inspected from 2006 to 2010. Additionally, and for each of the prospected locations, some physical parameters were recorded and sectors were classified based on: i) type of substrate (natural/artificial) ii) accessibility by humans (high/medium/low) and iii) substrate roughness or small scale 'topographical heterogeneity index' (THI) (high/medium/low) (refer to Rivera–Ingraham et al. (2011b) for further details). It is also important to note that in the present study, individuals present in Ceuta are considered part of a metapopulation composed of genetically connected sub–populations (see review by Badii & Abreu, 2006). Analyses Univariate analyses were carried out using the statistical package SPSS 15.0. Multivariate analyses were also performed to compare size distributions among populations, taking into account that the reduction of data to summarise statistics (such as means, medians, etc.) can significantly reduce the amount of available information (Sagarin et al., 2007). We used the total number of individuals for each size class (1  cm


Animal Biodiversity and Conservation 34.2 (2011)

321

A 14º 9º 45º Atlantic Ocean

40º

40º Mediterranean Sea

35º

35º

14º B

5.23º N

5.22º

5.21º 5.20º

5.19º 5.18º

G

L

J

C

35.53º A

B

5.20º

F

H I

Ceuta

2.5 km

5.17º

North Bay

K

Morocco

1,000 km

35.55º

M

35.54º

D

E 35.53º

South Bay

5.19º 5.18º

35.52º 5.17º

Fig. 1. A. Location of Ceuta; B. Sectors into which the study site was divided to estimate the total number of individuals of C. nigra. Fig. 1. A. Localización de Ceuta; B. Sectores en los que se dividió la zona de estudio para la estimación del número total de individuos de C. nigra.

intervals) and sector. This methodology has been satisfactorily used by other authors (e.g. Sagarin et al., 2007; Espinosa, 2009; Rivera–Ingraham et al., 2011b). As the total length of shoreline inspected varied considerably between locations, these frequency values were standardized by transforming them to percentages (over the total number of recorded individuals in the sector). Additionally, these data were later transformed to log (x + 1) to homogenize variances. An MDS (multi–dimensional scaling) analysis was carried out using PRIMER–E v.6.0 and based on the UPGMA (Unweigh Pair–Group Method using arithmetic means) method and the Bray–Curtis similarity index (Bray & Curtis, 1957). The Kruskal stress coefficient was used to determine ordination (Kruskal & Wish, 1978).

Results Species’ distribution and estimates of total number of individuals A total of 3,076 individuals were counted during our survey (including both intertidal and subtidal levels) which covered 1,961 m of the coast of Ceuta (12.13% of the total rocky shore that could potentially house C. nigra individuals). We estimated that 48,473 individuals can be found in the intertidal and subtidal areas of this coastline (table 1). The average densities recorded for the main locations surveyed are shown in figure 2. The species had larger populations in the North Bay and populations


Rivera–Ingraham et al.

322

Table 1. Data corresponding to each of the considered sectors: Tr. Total rocky shore length (m); N. Number of individuals recorded; Sl. Shore length inspected (m); Te. Total number of individuals estimated Tabla 1. Datos correspondientes a cada uno de los sectores considerados: Tr. Longitud total de la orilla rocosa (m); N. Número de individuos censados; Sl. Longitud de costa inspeccionada (m); Te. Número total estimado de individuos.

Sector Sector name

Tr

N

Sl

Te

A

Frontier shoreline and 'Pineo' islet

300

2

100

6

B

'Piedras Gordas' islets and 'Brazo' islet

311

36

311

36

C

Chorrillo and Foso jetties

712

14

200

50

D

Fuentecaballos jetty

350

34

350

34

E

Mellizos–Desnarigado

3,202

28

100

897

F

Desnarigado–Point Sirena

3,472

112

100

3,889

G

San Amaro

959

342

100

3,280

H

Dique de Levante

555

84

100

466

I

Parque del Mediterráneo

463

296

100

1,371

J

Dique de Poniente (concrete block section)

1,156

1,038

50

23,999

J

Dique de Poniente (limestone section)

1,115

367

50

8,184

K

Benítez

316

160

100

506

L

Desaladora–Point Bermeja

1,217

239

100

2909

M

Point–Bermeja–Point Blanca

1,138

84

100

956

N

Point Blanca–Benzú

900

240

100

2,160

1,961

48,743

Total

became scarcer towards the South Bay. A Kruskal Wallis test indicated that the average density of individuals in populations in North Bay (6.26 ± 10.80 ind./m) was significantly higher than that recorded for populations in South Bay (0.42 ± 0.69 ind./m) (K = 9.86; p = 0.002). C. nigra showed an aggregated distribution that was especially patent in areas with the highest number of individuals. Population structure All individuals recorded were classified in size classes (established at 1 cm intervals). Figure 3 and table 2 show size structures and general descriptive parameters for each sector (see fig. 3, table 2). Effect of physical parameters on population structure A Pearson correlation test was carried out between several physical parameters (nature of substrate, natural/artificial, and substrate heterogeneity) and the most common population parameters (density of individuals per size class, average shell size, maximum and minimum shell size and total density). Results showed the following patterns:

16,167

3,076

Nature of substrate: Substrate did not seem to influence any of the population parameters, such as recruit density (F = 2.38; p = 0.14), average shell size (F = 2.77; p = 0.11), maximum shell size (F = 1.03; p = 0.33), or overall density (F = 1.40; p = 0.25). Effect of substrate heterogeneity on population structure Figure 4 was produced after conducting the multivariate analyses (based on population size structures). Figure 4A shows how a total of 4 subgroups of populations can be differentiated for a similarity of 70%: sectors I and J were both characterized for presenting the largest percentage of large individuals (75.67% and 67.40% respectively); sector C had the highest percentage of recruits recorded (50%); sectors E, H and M all presented between 25–50% of recruits and more than 65% of medium size individuals; the last group was composed of the remaining sectors, with 45–65% of medium size individuals and up to 41% of large individuals. Taking the results of the MDS analyses into account, we found no clear influence of substrate heterogeneity and accessibility on popula-


Animal Biodiversity and Conservation 34.2 (2011)

323

North Bay

Ceuta

Morocco

2.5 km

South Bay

0 ind. < 2 ind./m 2–4 ind./m 4–6 ind./m > 6 ind./m

Fig. 2. General distribution of C. nigra in Ceuta. Circle diameter corresponds to the density of individuals. Light grey circles indicate the density recorded for each transect established (except for sectors C, D, H, I and K where only some transects are represented because of lack of space). Darker circles represent the density of individuals in an area where a complete census of the coastline was carried out. Coastline sections plotted with thicker lines indicate the location of non–suitable areas (e.g. beaches) for the species. Fig. 2. Distribución general de C. nigra en Ceuta. El diámetro de los círculos se corresponde con la densidad de individuos. Los círculos claros indican la densidad registrada en cada uno de los transectos realizados (excepto para los sectores C, D, H, I y K para los que sólo se ha representado algún transecto por falta de espacio). Los círculos oscuros representan la densidad de individuos en un área costera tras la realización de un censo completo. Aquellas secciones de costa marcadas con trama más gruesa indican la localización de zonas no útiles para la especie (p.ej. playas).

tion structure. However, the data in figure 5 show how substrate heterogeneity correlated with the density of individuals. Recruitment rates did not vary between substrates, but the density of medium and large size individuals was significantly higher in substrates of low heterogeneity (fig. 5). To corroborate these results we considered the specific case of ‘Dique de Poniente’. This sector presents the same physical parameters throughout the area, except for the fact that half of the breakwater is composed of rocks with high surface roughness, while the other half is made of smooth 3 x 3 m cement blocks. The Kruskal–Wallis test evidenced that the area with the lowest THI had significantly higher densities of C. nigra individuals (34.10 ± 8.62 ind./m) than the pier area composed of rugged rocks (7.34 ± 1.23 ind./m) (K = 5.00; p = 0.025). Effect of the area’s accessibility on population structure Figure 4B shows no clear pattern regarding the effect of accessibility on population structure. Kruskal–Wallis

tests indicated that this factor does not influence any of the population parameters taken into consideration: density of recruits (K = 0.307; p = 0.858), density of large–size individuals (K = 1.495; p = 0.749), average shell size (F = 0.833; p = 0.454) or maximum shell size (F = 0.974; p = 0.404). Estimated individual growth rates We considered two recently created artificial structures: Fuentecaballos (sector D) and the concrete block breakwater area of Dique de Poniente (Sector J). For the former, a density of 0.10 ind./m was recorded, and maximum shell length was 7 cm. Taking into account that this breakwater was finished in April 2005 (J. L. Ruiz, pers. com.), and that the census was carried out in March 2010, we could estimate that individuals had an average growth rate of 1.40 cm/year. The area where C. nigra subpopulations reached the maximum shell length recorded for the species, 13.3 cm (February 2010), was Sector J, where there is one of the largest populations of the species in Ceuta (34.6 ind./m). This


Rivera–Ingraham et al.

Frequency

324

150

A

B

C

D

E

F

G

H

I

J

K

L

M

N

100 50

Frequency

0 150 100 50

Frequency

0 150 100 50

Frequency

0 150

1 2 3 4 5 6 7

1 2 3 4 5 6 7

100 50 0

1 2 3 4 5 6 7

1 2 3 4 5 6 7 Size class (cm)

Fig. 3. Size frequencies for each of the sectors considered. Letters correspond to the code used in table 1 and figure 1: 1 (0–1), 2 (2–3), 3 (4–5), 4 (6–7), 5 (8–9), 6 (10–11), 7 (12–13). Fig. 3. Frecuencias de tamaño para cada uno de los sectores considerados. Las letras corresponden al código usado en la tabla 1 y la figura 1. (Para las abreviaturas de los tamaños de clase, ver arriba.)

area was constructed in the early months of 2004 (J. Medina, pers. com.), so we were able to estimate a growth rate of 2.66 cm/year. Discussion After inspecting 12.13% of the coast of Ceuta that could potentially present C. nigra individuals, we estimated that there could be about 48,473 individuals in this area. The methodology used to reach this estimate has been used successfully by other authors to calculate the total number of Patella ferruginea individuals in

places such as Habibas Islands (Espinosa, 2009), Zembra Island (Boudouresque & Laborel–Deguen, 1986) and Ceuta (Rivera–Ingraham et al., 2011b). Although it has been emphasised that it would be of interest to quantify endangered invertebrate populations such as P. ferruginea (Paracuellos et al., 2003), this is the first quantification of a C. nigra population. It should additionally be considered that our estimates may be conservative, as even though special attention was paid to locating recruits (< 20 mm), this fraction of the population can easily be underestimated (Rivera–Ingraham, 2010). The idea of this estimate being conservative is also supported by the recent location


Animal Biodiversity and Conservation 34.2 (2011)

325

Table 2. Summary statistics for each C. nigra population considered in the study: Sc. Sector; P. Population; D. Density (ind./m); As. Average shell size (cm); Ms. Maximum shell size (cm); K. Kurtosis († Kurtosis is considered significant when its absolute value is greater than 2*SE Kurtosis); Kt. Kurtosis type (Pk. Platikurtic; Lk. Leptokurtic ); Sk. Skewness (Skew is considered significant when its absolute value is greater than 2*SE Skew); Am. Asymmetry (P. Positive; N. Negative). Tabla 2. Estadística resumen para cada subpoblación de C. nigra considerada en el estudio. Sc. Sector; P. Población; D. Densidad (ind./m); As. Tamaño medio de la valva (cm); Ms. Tamaño máximo de la valva (cm); K. Curtosis († la curtosis se considera significativa cuando su valor absoluto es mayor que el doble de su EE); Kt. Tipo de curtosis (Pk. Platicúrtica; Lk. Leptocúrtica); Sk. Coeficiente de asimetría (la desviación se considera significativa cuando su valor absoluto es mayor que el doble de su EE); Am. Asimetría (P. Positiva; N. Negativa). Sc P

D

As

Ms

K

Kt

Sk

Am

A Frontier shoreline

0

A 'Pineo' islet

2

5.4

5.40

B 'Piedras Gordas' islets

0.12

4.0

6.60

–0.937

0.343

B 'Brazo' islet

0

C Chorrillo jetty

0.02

4.60

4.70

C Foso jetty

0.12

2.82

7.10

8.457†

Lk

2.784

D Fuentecaballos jetty

0.10

4.54

7.00

–0.878

–0.284

E Mellizos–Desnarigado

0.28

3.28

5.90

–0.749

0.560

F Desnarigado–Point Sirena

1.12

4.65

10.00

–0.519

0.334

G San Amaro

3.42

4.91

9.50

–0.103

0.493*

P

H Dique de Levante

0.84

3.42

6.00

–0.748

–0.134

I

Parque del Mediterráneo

2.96

6.67

11.10

–0.461

–0.407

J

D. Poniente (concrete cube section) 7.34

7.19

8.80

–0.571†

Pk

0.127*

P

J

D. Poniente (limestone section)

34.6

4.38

13.30 –0.468†

Pk

–0.275*

N

1.6

4.43

10.70

2.178†

Lk

0.938

K Benítez L

Desaladora–Point Bermeja

(natural substrate section)

L

Desaladora–Point Bermeja

(breakwater section)

1.16 2.32

4.10 3.44

8.20 7.50

–0.905 1.081†

0.058

Lk

0.838*

– P

M Point Bermeja–Point Blanca

0.84

3.15

5.90

0.179

–0.048

N Point Blanca–Benzú

2.40

4.09

10.40

1.675†

Lk

1.223*

P

of isolated subtidal rocks (1–5  m deep and with no intertidal regions) on Ceuta’s North Bay, within the limits of sectors L and M. These structures are around 150–250 m from the coast and are fully colonized by large C. nigra individuals. Our preliminary observations indicate that around 220 individuals could be living in these areas. However, further surveys are needed to determine the presence of similar sublitoral regions in the area. The density of C. nigra individuals on the coast of Ceuta is clearly influenced by the location of the population. The largest populations were located in North Bay, which is mainly influenced by Atlantic waters. It was in this area where a contagious distribution pattern was especially patent, agreeing with the results previously obtained by Rivera–Ingraham et al. (2011a). However,

these populations become scarcer as we moved southwest. This is a predictable pattern considering that C. nigra is a typically Atlantic species. This distribution contrasts with that shown by P. ferruginea, endemic to the Mediterranean, and mainly distributed in South Bay (which has a greater Mediterranean influence) (Rivera–Ingraham et al., 2011b). No significant differences were obtained between populations located on artificial or natural substrates. One of the main differences between natural and artificial surfaces is their heterogeneity (Bulleri & Chapman, 2010). The substrates presenting the highest irregularities showed coefficients higher than 1.30 (using the equation provided by Blanchard & Bourget, 1999) and around 1.017 (Rivera–Ingraham, 2010) using fractal dimensions (Mandelbrot, 1967),


Rivera–Ingraham et al.

326

A

Transform: Log (x + 1) Resemblance S17 Bray Curtis similarity

Bray Curtis similarity

20 40 60 80

100 A

I

J

C

D

B

B

M I J

C

G F Sectors

K

N

E

H

M

2D Stress: 0,01 Substrate heterogeneity A High

I C

L

F G KN D L B H E M

Medium Low Collection impact High Medium Low

Fig. 4. Multivariate analyses: A. Cluster analysis (continuous lines indicate significantly different groups, SIMPROF analysis, p < 0.05); B. Spatial representation of centroids (MDS) for each sector: circle diameter is positively correlated with the area's substrate roughness (THI) (Blanchard & Bourget, 1999): high (>  1.30), medium (1.15–1.30), low (< 1.15); colours are associated with the different grades of impact by collection suffered by individuals in the area: high (easily accessible areas, where is common to find people collecting intertidal macro–invertebrates), medium (areas of relatively easy access, although they do not present high impact by collection), low (areas with difficult or no access by land to the intertidal fringe, and where no people have been seen). Fig. 4. Análisis multivariante: A. Análisis de conglomerados (las líneas continuas indican grupos significativamente distintos, análisis SIMPROF de perfil de similitud, p < 0,05); B. Representación espacial de los centroides (MDS, escalamiento multidimensional) para cada sector: el diámetro de los círculos está correlacionado positivamente con la rugosidad del sustrato de la zona (THI, índice de heterogeneidad topográfica) (Blanchard & Bourget, 1999): alta (> 1,30), media (1,15–1,30), baja (< 1,15); los colores se asocian con los distintos grados de impacto debido a la recolección sufrida por los individuos de la zona: alta (áreas muy accesibles, donde es común ver gente recolectando macroinvertebrados intermareales), media (áreas de acceso relativamente fácil, aunque no presentan un gran impacto de recolección), baja (áreas sin acceso o con un acceso difícil desde tierra a la franja intermareal, donde no se ha visto a nadie recolectando).

according to an adapted method from Beck (2000) and using profile gauges to obtain rock profiles as in Frost et al. (2005). Irregularities in substrates are associated with high recruitment rates (as these

structures can enhance settlement and provide shelter for juvenile limpets) (e.g. Creese, 1982). This is also true for other limpet species such as P. ferruginea (Rivera–Ingraham et al., 2011b). However, no dif-


Animal Biodiversity and Conservation 34.2 (2011)

Average density (ind./m)

5

Size class < 25 mm 25–50 mm > 50 mm

b

4

3

2

b

1

a a

a

327

0

High

a

a

a

a

Medium Low Substrate roughness

Fig. 5. Average density values recorded for populations located on substrates of high (THI > 1.30), medium (THI = 1.30–1.15) and low (THI < 1.15) roughness (measured using equation provided by Blanchard & Bourget, 1999). Results are divided into three size classes (< 25 mm, 25–50 mm, > 50 mm). Values associated with the same letter (a, b) and colour belong to the same roughness subset based on a oneway ANOVA and a a posteriori multiple comparison test Student–Neuman–Keuls. Fig. 5. Valores promedio de densidad registrados para poblaciones localizadas sobre sustratos de alta (THI > 1,30), media (THI = 1,30–1,15) y baja (THI < 1,15) rugosidad (medida utilizando una ecuación proporcionada por Blanchard & Bourget, 1999). Los resultados se dividen en tres clases de tamaño (<  25 mm, 25–50 mm, > 50 mm). Los valores asociados con la misma letra (a, b) y color pertenecen al mismo subconjunto de rugosidad, basándose en un ANOVA de un solo factor, y un test de Student– Neuman–Keuls de comparación múltiple realizado a posteriori.

ferences in recruitment rates were detected among different substrates for C. nigra. Therefore, some other factor could influence recruitment (aside from the physical influence of the substrate’s irregularities). Previous studies have shown that the species is not homogenously distributed on the coast, and that is has an aggregated distribution (Rivera–Ingraham et al. 2011a). This was later attributed to the possibility that C. nigra larvae are attracted by adult conspecifics through chemical signalling (Rivera–Ingraham et al., 2011c), as occurs in other mollusc species (Morse et al. 1979; Hadfield, 1984; García–Lavandeira et al., 2005; Mesías–Gansbiller et al., 2008). Small scale topographical heterogeneity did influence the density of medium sized and large individuals, increasing as THI decreased. Based on these facts, it appears that a high topographical index may indeed favour recruitment, while on smooth surfaces recruitment may be equally important thanks to the influence of the presence of adult individuals.The area’s accessibility may highly influence population structures in intertidal

communities. It has been frequently observed that human collection of organisms reduces population density and preferentially affects the largest fraction of populations. This is true for giant limpets such as Lottia gigantea (Kido & Murray, 2003), and P. ferruginea (Espinosa et al., 2009; Rivera–Ingraham, 2010). It is true that the largest C. nigra individuals were found at Parque del Mediterráneo (11.10 cm) and at the concrete block breakwater area of ‘Dique de Poniente’ (13.30 cm). These areas show low THI and are hard to access, the former because it is private property, and the latter because it is physically difficult to reach. The low collection rate in these port areas can be explained by their inaccessibility and also by the fact that people are usually reluctant to consume organisms from port areas (presumably living in more polluted waters) (Doneddu & Manunza, 1992). This in turn favours the survival of large individuals (Espinosa et al., 2009). However, no statistical differences were recorded among the three groups of populations, and all of them showed similar mortality rates. This could


Rivera–Ingraham et al.

328

be partly due to the fact that C. nigra shells, regardless of the enormous sizes they can reach, lack radial ribs or vivid colours (see Rivera–Ingraham et al., 2011a), so they are not as conspicuous as limpet species like P. ferruginea. Moreover, the species shows a clear vertical segregation, with the largest individuals being located in the lower intertidal areas or even at subtidal levels (Rivera–Ingraham et al., 2011a). Because of this, shells are usually colonized by macroalgae tuffs such as Corallina elongata (Rivera–Ingraham, 2010), making individuals even less noticeable. Finally, two recently created artificial structures provide additional interesting information regarding individual growth rates: Fuentecaballos (sector D) and the new area of ‘Dique de Poniente’ (sector J). For the former, individuals should have an average growth rate of 1.40 cm/year, while for the latter, values would be around 2.66 cm/year. It has previously been observed that growth rates in the species highly depend on the age of the individuals, and smaller/ younger individuals show higher growth rates than larger/ older individuals (Rivera–Ingraham, 2010). The same study also determined that for well established populations (e.g. Parque del Mediterráneo or C. nigra populations located in Algeciras Bay), C. nigra individuals showed an average growth rate of 0.95 cm/ year (being 2.6 and 0.1 cm/year for individuals with initial sizes of < 3 and > 10 cm respectively). For both structures, values are considerably greater than those recorded in normal conditions. The same was observed for P. ferruginea (Rivera–Ingraham et al., 2011b) and these differences were attributed to the fact that the recorded individuals were probably the first to colonize these new structures (with abundant microalgae biofilm and almost no fauna). If this were the case, newly settled individuals would have abundant trophic resources at their disposal and this could result in higher growth rates than those for individuals that settle in areas with well developed subpopulations (where individuals would have to compete for space and food) (Branch, 1975). Furthermore, a shore may contain individuals in very different conditions of wave exposure or emersion. It may be inappropriate to average the population dynamics of such assemblages at the shore scale when demographic rates are likely to vary greatly within shores (Johnson, 2006). The present study provides the first quantitative data of a C. nigra population. We recommend, however, that this work be repeated in coming years in order to monitor the evolution of the species and to promptly implement adequate conservation measures if needed. Acknowledgements The authors express their gratitude to Jorge Francisco Marín Lora for help in the sampling process. Thanks also go to the Consejería de Medio Ambiente de Ceuta (OBIMASA) staff for their support, and to Dr. José Templado and one anonymous referee for their comments on the original manuscript. The present study was financed by a F. P. U. grant awarded to G. A. Rivera–Ingraham (AP–2006–04220).

References Beck, M. W., 2000. Separating the elements of habitat structure: independent effects of habitat complexity and structural components on rocky intertidal gastropods. Journal of Experimental Marine Biology and Ecology, 249: 29–49. Blanchard, D. & Bourget, E.,1999. Scales in coastal heterogeneity: influence on intertidal community structure. Marine Ecology Progress Series, 179: 163–173. Boudouresque, C. F. & Laborel–Deguen, F., 1986. Patella ferruginea. In: Le benthos marin de l’ile de Zembra (Parc National, Tunisie): 105–110 (C. F. Boudouresque, J. G. Harmelin, A. Jeudy de Grissac, Eds.). GIS Posidonie Publishers, Marseille. Branch, G. M., 1975. Intraspecific competition in Patella cochlear Born. Journal of Animal Ecology, 44: 263–282. Bray, J. R. & Curtis, J. T., 1957. An ordination of the upload forest communities of southern Wisconsin. Ecological Monographs, 27: 325–349. Bulleri, F. & Chapman, M. G., 2010. The introduction of coastal infrastructure as a driver of change in marine environments. Journal of Applied Ecology, 47: 26–35. Christiaens, J., 1974. Révision du genre Patella (Mollusca, Gastropoda). Bulletin du Museum National d’Histoire Naturelle, 182: 1305–1392. – 1983. The genus Patella along the coast of Malaga (Spain). Conchiglie, 15: 15–17. Creese, R. G., 1982. Distribution and abundance of the acmaeid limpet, Patelloida latistrigata, and its interaction with barnacles. Oecologia, 52: 85–96. Doneddu, M. & Manunza, B., 1992. Valutazione dell’impatto antropico relativo alla balneazione estiva su una populazione di Patella ferruginea Gmelin, 1791 del litorale di Aglientu. Bollino. Malacologico, 28: 161–168. Espinosa, F., 2009. Populational status of the endangered mollusc Patella ferruginea Gmelin, 1791 (Gastropoda: Patellidae) on Algerian islands (SW Mediterranean). Animal Biodiversity and Conservation, 32: 19–28. Espinosa, F., Guerra–García, J. M. & García–Gómez, J. C., 2007. Sewage pollution and extinction risk: an endangered limpet as a bioindicator? Biodiversity and Conservation, 16: 377–397. Espinosa, F., Nakano, T., Guerra–García, J. M. & García–Gómez, J. C., 2010. Population genetic structure of the endangered limpet Cymbula nigra in a temperate Northern hemisphere region: influence of paleoclimatic events? Marine Ecology, 32: 1–5. Espinosa, F., Rivera–Ingraham, G. A., Fa, D. A. & García–Gómez, J. C., 2009. Effect of human pressure on population size structures of the endangered ferruginean limpet: towards future management measures. Journal of Coastal Research, 25: 857–863. Frenkiel, L., 1975. Contribution a l’étude des cycles de reproduction des patellidae en Algérie. Pubblicazione de la Stazzione Zoologica di Napoli, 39: 153–189. Frost, N. J., Burrows, M. T., Johnson, M. P., Hanley,


Animal Biodiversity and Conservation 34.2 (2011)

M. E. & Hawkins, S. J., 2005. Measuring surface complexity in ecological studies. Limnology and Oceanography, 3: 203–210. García–Lavandeira, M., Silva, A., Abad, M., Pazos, A. J., Sánchez, J. L. & Parallé, L., 2005. Effects of GABA and epinephrine on the settlement and metamorphosis of the larvae of four species of bivalve molluscs. Journal of Experimental Marine Biology and Ecology, 316: 149–156. Grandfils, R. & Vega, R.,1982. Il genre Patella lungo la costa di Malaga (Spagna). Conchiglie, 14: 6–8. Guerra–García, J. M., Corzo, J., Espinosa, F. & García– Gómez, J. C., 2004. Assessing habitat use of the endangered marine mollusc Patella ferruginea (Gastropoda, Patellidae) in northern Africa: preliminary results and implications for conservation. Biological Conservation, 116: 319–326. Guerra–García, J. M., Maestre, M. J., González, A. R. & García–Gómez, J. C., 2006. Assessing a quick monitoring method using rocky intertidal communities as a bioindicator: a multivariate approach in Algeciras Bay. Environmental Monitoring and Assessment, 116: 345–361. Hadfield, M. G., 1984. Settlement requirements of molluscan larvae: new data on chemical and genetic roles. Aquaculture, 39: 283–298. Johnson, M. P., 2006. Rocky intertidal invertebrates: the potential for metapopulations within and among shores. In: Marine Metapopulations: the potential for metapopulations within and among shores: 247–270 (J. P. Kritzer & P. F. Sale, Eds.). Academic Press Inc. Kido, J. S. & Murray, S. N., 2003. Variation in the limpet Lottia gigantea population structures, growth rates, and gonadal production on Southern California rocky shores. Marine Ecology Progress Series, 257: 111–124. Koufopanou, V., Reid, D. G., Ridgway, S. A. & Thomas, R. H., 1999. A molecular phylogeny of the patellid limpet (Gastropoda: Patellidae) and its implications for the origin of their antitropical distribution. Molecular Phylogenetics and Evolution, 11: 138–156. Kruskal, J. B. & Wish, M., 1978. Multidimensional scaling. Sage Publications, Beverly Hills, California. Mandelbrot, B. B., 1967. How long is the coast of Britain? Statistical self–similarity and fractional dimension. Science, 155: 636–638. Martínez, J. L. & Peñas, A., 1996. Fauna malacológica de Caleta, Mijas costa, Málaga. Malakos, 5: 75–92. Mesías–Gansbiller, C., El Amine Bendimerad, M., Román, G., Pazos, A. J., Sánchez, J. L. & Pérez– Parallé, M. L., 2008. Settlement behavior of Black scallop larvae Chlamys varia L. in response to GABA, epinephrine and IBMX. Journal of Shellfish Research, 27: 261–264. Moreno, D., 2006. Tesoros sumergidos: la flora y fauna marinas. In: Entre África y Europa. Historia Natural de la Isla de Alborán: 67–85 (M. Paracuellos, J. C. Nevado & F. J. Mota, Eds.). RENPA, Consejería de Medio Ambiente, Junta de Andalucía, Sevilla. Moreno, D. & Arroyo, M. C., 2008. Cymbula nigra (da Costa, 1771). In: Libro rojo de los invertebrados de Andalucía: 301–309 (J. M. Barea–Azcon, E. Bal-

329

lesteros & D. Moreno, Eds.). Consejería de Medio Ambiente, Junta de Andalucía, Sevilla. Morse, D. E., Hooker, N., Duncan, H. & Jensen, L., 1979. Gamma–aminobutyric acid, a neurotransmitter induces planktonic larvae to settle and begin metamorphosis. Science, 204: 407–410. Nakano, T. & Espinosa, F., 2010. New alien species in the Atlantic Ocean? Marine Biodiversity Records. doi: 10.1017/ S1755267210000357. Paracuellos, M., Nevado, J. C., Moreno, D., Giménez, A. & Alesina, J. J., 2003. Conservational status and demographic characteristics of Patella ferrguinea Gmelin, 1791 (Mollusca, Gastropoda) on the Alboran Island (western mediterranean). Animal Biodiversity and Conservation, 26.2: 29–37. Pasteur–Humbert, C.,1962. Les mollusques marins testacés du Maroc. Catalogue non critique. I. Gasterópodes. Travaux de l’Institut Scientifique Chérifien. Série Zoologie, 23: 1–245. Peñas, A., Rolán, R., Luque, A. A., Templado, J., Moreno, D., Rubio, F., Salas, C., Sierra, A. & Gofas, S., 2006. Moluscos marinos de la Isla de Alborán. Iberus, 24: 23–151. Ridgway, S. A., Reid, D. G., Taylor, J. D., Branch, G. M. & Hodgson, A. N., 1998. A cladistic phylogeny of the family Patellidae (Mollusca: Gastropoda). Philosophical Transactions of the Royal Society of London B, 353: 1645–1671. Rivera–Ingraham, G. A., 2010. Biología de la conservación de especies de patélidos en el umbral Atlántico–Mediterráneo. Tesis doctoral, University of Seville. Rivera–Ingraham, G. A., Espinosa, F. & García–Gómez, J.C., 2011a. Ecological considerations and niche differentiation between juvenile and adult black limpets Cymbula nigra. Journal of the Marine Biological Association of the UK, 91: 191–198. – 2011b. Conservation status and updated census of Patella ferruginea (Gastropoda, Patellidae) in Ceuta: distribution patterns and new evidence of the effects of environmental parameters on population structure. Animal Biodiversity and Conservation, 34.1: 83–99. – 2011c. Effect of gamma–amino butyric acid on limpet populations: towards the future management and conservation of endangered patellid species. Journal of Chemical Ecology, 37: 1–9. Sá–Pinto, A., Branco, M., Harris, D. J. & Alexandrino, P., 2005. Phylogeny and phylogeography of the genus Patella based on mitochondrial DNA sequence data. Journal of Experimental Marine Biology and Ecology, 325: 95–110. Sagarin, R. D., Ambrose, R. F., Becker, B. J., Engle, J. M., Kido, J., Lee, S. F., Miner, C. F., Murray, S. N., Raimondi, P. T., Richards, D. V. & Roe, C., 2007. Ecological impacts on the limpet Lottia gigantea populations: human pressure over a broad scale on island and mainland intertidal zones. Marine Biology, 150: 399–413. Spada, G. & Maldonado Quiles, A., 1974. Nota preliminare sulle specie di molluschi a diffusione prevalentemente atlantica e presenti anche in Mediterraneo nel mare di Alboran. Quaderni della Civica Stazione


330

di Idrobiologia, Milano, 5: 51–69. Templado, J., Calvo, M., García, A., Luque, A. A., Maldonado, M. & Moro, L., 2004. Guía de invertebrados y peces marinos protegidos por la legislación nacional e internacional. Organismo Autónomo de Parques Nacionales, Ministerio de Medio Ambiente,

Rivera–Ingraham et al.

Madrid. Templado, J., Calvo, M., Moreno, D., Flores, A., Conde, F., Abad, R., Rubio, J., López–Fe, C. M. & Ortiz, M., 2006. Flora y fauna de la Reserva Marina y Reserva de Pesca de la Isla de Alborán. Secretaría General de Pesca Marítima, MAPA, Madrid.


Animal Biodiversity and Conservation 34.2 (2011)

331

Barrier effects on vertebrate distribution caused by a motorway crossing through fragmented forest landscape J. L. Tellería, J. A. Díaz, J. Pérez–Tris, E. de Juana, I. de la Hera, P. Iraeta, A. Salvador & T. Santos Tellería, J. L., Díaz, J. A., Pérez–Tris, J., De Juana, E., De la Hera, I., Iraeta, P., Salvador, A. & Santos, T., 2011. Barrier effects on vertebrate distribution caused by a motorway crossing through fragmented forest landscape. Animal Biodiversity and Conservation, 34.2: 331–340. Abstract Barrier effects on vertebrate distribution caused by a motorway crossing through fragmented forest landscape.— We analysed the effects of a 25–year–old motorway on the distribution of five vertebrates inhabiting a fragmented forest landscape and differing in their ability to move across linear infrastructures. We found clear evidence of barrier effects on the distribution of the forest lizard Psammodromus algirus. The roe deer (Capreolus capreolus) was also unequally distributed on both sides of the motorway, but this could also be due, at least in part, to fragmentation. The eyed lizard (Timon lepidus), that can move through open fields, showed no evidence of barrier effects. The distribution of two small birds (Erithacus rubecula and Phylloscopus bonelli) was unaffected by the motorway. Our results show that a motorway may severely restrict the distribution of species which can withstand high levels of forest fragmentation but show limited dispersal ability, highlighting the role of linear infrastructures in shaping species’ ranges at regional scales. Key words: Abundance patterns, Barrier effect, Dispersive ability, Lizard, Road ecology, Roe deer. Resumen Efecto barrera en la distribución de vertebrados causada por las autovías que cruzan un paisaje forestal fragmentado.— En este trabajo se analizan los efectos de una autovía construida hace 25 años sobre la distribución en un paisaje forestal fragmentado de cinco vertebrados que difieren en su capacidad de atravesar infraestructuras lineales. Se encontraron evidencias de efecto barrera en la distribución de la lagartija colilarga Psammodromus algirus. El corzo (Capreolus capreolus) presentó una distribución desigual a ambos lados de la carretera, aunque también atribuible, al menos en parte, a diferencias en el grado de fragmentación. El lagarto ocelado (Timon lepidus), que puede moverse a través de los cultivos, no mostró ninguna evidencia de efecto barrera. La distribución de dos pequeños paseriformes (Erithacus rubecula y Phylloscopus bonelli) no se vió afectada por la autovía. Estos resultados demuestran que una autovía puede restringir la distribución de especies capaces de soportar altos niveles de fragmentación del paisaje pero con escasa capacidad de dispersión, poniendo de manifiesto el efecto potencial de las infraestructuras lineales sobre la distribución de las especies a escala regional. Palabras clave: Patrones de abundancia, Efecto barrera, Capacidad de dispersión, Lagartijas, Ecología de las carreteras, Corzo. (Received: 6 IX 11; Conditional acceptance: 21 X 11; Final acceptance: 29 XI 11) José L. Tellería, José A. Díaz, Javier Pérez–Tris, Eduardo de Juana, Iván de la Hera, Pablo Iraeta & Tomás Santos, Depto. de Zoología y Antropología Física (Vertebrados), Fac. de Biología, Univ. Complutense, E–28040 Madrid, España (Spain).– Alfredo Salvador, Depto. de Ecología Evolutiva, Museo Nacional de Ciencias Naturales–CSIC, c./ José Gutiérrez Abascal 2, E–28006 Madrid, España (Spain). Corresponding author: Tomás Santos. E–mail: tsantos@bio.ucm.es ISSN: 1578–665X

© 2011 Museu de Ciències Naturals de Barcelona


Tellería et al.

332

Introduction Barrier effects of linear infrastructures may have different origins, such as occupation of natural dispersal pathways of populations, neo–phobic behaviour of individuals (which may alter dispersal patterns), or physical isolation caused by fences that run alongside roads and railways to prevent collision. As a consequence, linear infrastructures usually divide animal populations into sub–populations that may become isolated, thereby compromising their long–term persistence (Forman & Alexander, 1998; Trombulak & Frissell, 2000). Many studies have assessed the effectiveness of management actions to correct the barrier effect, such as building overpasses, underpasses or culverts to favour dispersal (Clevenger & Waltho, 2005; Grilo et al., 2008; Mader, 1984; Mata et al., 2005). Other studies have emphasized the need to determine the effects of interrupted gene flow between populations separated by linear infrastructures (e.g. Epps et al., 2005; Riley et al., 2006; Strasburg, 2006). However, less effort has been devoted to study the impact of linear infrastructures on the regional distribution of species (Baguette & Van Dyck, 2007; Roedenbeck et al., 2007; Underhill & Angold, 2000). Ranges of local species may undergo different dynamics on either side of a linear infrastructure when the barrier disconnects marginal sectors from core areas in the regional distribution of each species (Taylor et al., 1993). In these circumstances, the barrier effect will decrease the ability of local populations to avoid extinction through the rescue effect due to individuals coming from source areas (Brown & Kodric–Brown, 1977; Goodwin & Fahrig, 2002). Reduced connectivity may promote independent population trends on either side of the linear infrastructure. These trends are typically mediated by asymmetric source–sink dynamics that may eventually cause local extinctions in isolated habitat patches (Pulliam, 1988; Taylor et al., 1993, Shepard et al., 2008). We studied the effect of a motorway on the distribution of several forest species in a fragmented landscape in northern Spain (fig. 1). The forest patches in this area become scarcer and increasingly fragmented westwards. The motorway runs perpendicular to the westernmost tips of this fragmented area, isolating them from eastern habitat patches. These eastern habitat patches are better connected to large forests located further east in the ‘Sierra de la Demanda’ mountain range, and might act as source habitats. When the motorway was constructed 25 years ago, populations of forest vertebrates were expanding westwards from these mountains as a consequence of human abandonment of low productivity highlands (Sáez–Royuela & Tellería, 1986; Tellería & Sáez–Royuela, 1984). A barrier effect might therefore have caused asymmetric population dynamics on either side of the motorway, obstructing the recovery of local extinctions in western forest patches by individuals moving from mountain forests. Our aim was to assess whether there is any evidence of this barrier effect two and a half decades after the construction of this infrastructure. If the motorway

is a barrier against the dispersal of a species, we could expect different patterns of habitat occupancy and abundance on either side of the road, with reduced patch occupancy and/or abundance on the western side of the motorway. We can also expect asymmetric distribution of species with different susceptibility to barrier effects (Fahrig & Rytwinski, 2009). To identify such differences, we analysed the distribution of five vertebrate model species: a large mammal (the roe deer Capreolus capreolus), two small birds (the European robin Erithacus rubecula and the Bonelli’s warbler Phylloscopus bonelli), and two lacertid lizards (the eyed lizard Timon lepidus and the large psammodromus Psammodromus algirus). In principle, the motorway should not represent a barrier against dispersal of birds as they can fly over it. Material and methods Study area We studied the distribution of the selected species in an agricultural landscape located around Lerma, northern Spain (fig.1). In this area, forests cover ca. 10% of their former range and are dominated by evergreen Holm Oaks Quercus ilex and deciduous oaks Q. pyrenaica and Q. faginea (Díaz et al., 2005). The distribution of forest vertebrates in this fragmented landscape has been studied by the authors since the mid–1980s, when the highway A–1 (E–50 according to European nomenclature) was built to substitute an unfenced, single carriageway road (N–1). The motorway, which opened to traffic in 1985, is a dual–carriageway protected by fences throughout its length. It is one of the main transportation highways to connect Spain, Portugal and the Maghreb countries with the rest of Europe. We studied animal distributions in forest remnants located on either side of the motorway between Gumiel de Izán (41º  46'  24''  N, 3º  41'  17''  W; 854 m a.s.l.) and Saldaña de Burgos (42º 15' 34'' N, 3º 41' 48'' W; 863 m a.s.l.; fig. 1). The study area represents 40 km of the road, which are crossed by 27 culverts (0.7/km), 24 overpasses (0.6/km), and 29 underpasses (0.7/km) that may potentially be used by several vertebrates (Mata et al., 2005), meaning an average of 2 passes/km. The frequency of passes meets the Spanish regulation (Ministerio de Medio Ambiente, 2006) aimed to favour dispersal of large mammals (0.3–1 passes/ km) and small vertebrates (1–2 passes/km). A 4–km long segment crossing forested landscape to the north of Lerma (fig. 1) is critical for forest vertebrates to cross from one side of the motorway to the other. This segment is crossed by one overpass and five underpasses (1.5  passes/km) that are suitable for dispersal of large mammals. This is also in compliance with Spanish regulation for motorways crossing through forested areas (1–2 passes/km). We did not study the forest stretch to the south of Lerma (fig. 1) because it is occupied by a golf course surrounded by a metallic fence with a concrete base that prevents colonization by terrestrial animals.


Animal Biodiversity and Conservation 34.2 (2011)

333

Irún–Béhobie

CE

Burgos A–1

Study area

CW

N VW

VE

Madrid

ME

MW EW

A–1

EE

Saldaña

10 km

mE

mW

A–1 TNW

TSW

TSE

Lerma

TSE BW

BE A–1

A–1

'Sierra de la Demanda' mountains

Gumiel

Fig. 1. Geographical location of the study area. The location of the A–1 motorway is traced on the Iberian peninsula in the upper–right panel, with the distribution of highlands and mountains (areas over 1,000 m a.s.l.) shown in a darker shade. The lower–right panel shows a general view of the study area. The three small squares represent the locations of study sites (magnified on the left). Codes refer to forest shown in table 2, with the E and W suffix differentiating eastern and western locations, respectively. Fig. 1. Localización geográfica del area de estudio. En el panel superior derecho se encuentra el trazado de la autopista A–1 en la península ibérica, junto con la distribución de montañas y mesetas (zonas situadas a más de 1.000 m s.n.m.) en color más oscuro. El panel inferior de la derecha presenta una vista general del área de estudio. Los tres cuadrados pequeños representan la localización de las zonas de estudio (aumentados a la izquierda). Los códigos hacen referencia a los bosques de la tabla 2, con los sufijos E y W diferenciando las localizaciones oriental y occidental respectivamente.

Sampling protocol We selected eight forest fragments as closely located to the motorway as possible on its western side, all of which had a contiguous forest fragment on the other side of the motorway (fig. 1). By selecting our fragments according to a paired design, we aimed to study differences in the occurrence and abundance of animals between forest fragments separated by the road. These fragments were more closely located to each other than any pair of fragments located on the same side of the road. We constrained our selection of sites to avoid fragments that were too small (< 1.5 ha) as patch size rather than isolation might determine occurrence of the studied species.

However, the layout of fragment sizes and locations on either side of the road did not allow for a perfect design. Thus, we sometimes had to use adjacent fragments on the eastern side of the motorway to match western fragments suitable for our study (e.g., the EE and mE sites in fig. 1, which are more closely located to each other than to their western counterparts). In one case (TNE and TSE sites in fig. 1), we had to use two study sites located on the same large fragment on the eastern side of the motorway. Nevertheless, these sites were more distantly located than some neighbouring forests in our study area, and could be considered to function as different sources of colonization of western fragments if dispersal westwards involved crossing the road.


Tellería et al.

334

We scored presence of vertebrates and measured their abundance in each study site from May to June 2009. We assessed the abundance of roe deer by counting faecal groups detected within a 1 + 1 m wide band (Tellería & Virgós, 1997) during the same time intervals dedicated to sampling lizards (see below). Abundance was estimated as the number of faecal groups/ha. Birds were counted early in the morning along 25 + 25  m wide transects and their abundance was estimated as number of birds/ha. Lizards were counted by means of time–controlled search. To prevent the confounding effects of circadian and/or weather induced variations in the activity levels of lizards (Díaz, 1991), two two–person teams connected by telephone simultaneously sampled the two fragments of each pair on either side of the motorway. Both teams started sampling at the same time, when meteorological conditions were appropriate for lizard activity. In each fragment, lizards were searched for during three person–hours, which was sufficient to detect the presence of the large psammodromus (mean ± SE time required to detect the first lizard was 12 ± 2.8', range 1–28'). Abundance was expressed as the number of lizards detected per 10' search. For eyed lizards, we only noted their presence/absence during the other species’ counts, given their low densities and limited detectability (Díaz et al., 2006). To control for possible confounding effects of features other than location of study sites with respect to the motorway, we measured the size of the fragments (which is a major determinant of vertebrate distribution in the study area) and their minimum distance to the motorway. These variables were estimated by means of the SIGPAC facility (http://sigpac.mapa.es/fega/visor/). We also examined vegetation structure as this can affect the distribution of forest vertebrates. We sampled habitat features within 25 m–radius circles evenly distributed along the itineraries used to census animals. In these plots, we visually estimated the cover of leaf litter, grass, shrubs < 2 m high, shrubs and trees > 2 m high, and deciduous oaks (Quercus pyrenaica and Quercus faginea). We also assessed the mean height of the tree canopy and the number of tree and shrub species. These variables were chosen because shrub and leaf litter are used as refuges by forest lizards (Díaz & Carrascal, 1991), the presence of deciduous oaks is related to more moist and more productive conditions in these Mediterranean forests, and the diversity of tree and shrub species has been related to the presence of roe deer, which require a variety of plant species to feed on (Virgós & Tellería, 1998). Analyses We used principal components analysis (PCA) to transform the variables of vegetation structure into a reduced set of independent components (table 1). PC–1 reflected a gradient of increased forest development, higher cover of leaf litter and deciduous oaks, and higher numbers of shrub and tree species. PC–2 represented a gradient of increased cover of grasslands with a scarcity of low shrubs. Factor scores of forest fragments on these gradients were used as surrogates of vegetation structure.

Table 1. Results of principal components analysis of the variables measured to evaluate the vegetation structure of forest fragments. Figures show the factor loading of each variable on each component, the eigenvalues and the percentage of variance explained by each component. Significant factor loadings are shown in bold. Tabla 1. Resultado del análisis de componentes principales de las variables medidas para evaluar la estructura de la vegetación de los fragmentos forestales. Las cifras muestran la saturación factorial de cada variable sobre cada componente, los valores propios y el porcentaje de varianza explicado para cada componente. Las saturaciones factoriales significativas se indican en negrita. Variable Cover of grass Cover of leaf litter

PC–1

PC–2

–0.047

0.831

0.725 –0.272

Cover of shrubs < 2 m high 0.564 –0.629 Cover of shrubs and trees > 2 m-high 0.309 0.675 Cover of deciduous oaks 0.809 0.048 Number of shrub and tree species 0.685 -0.250 Height of the tree canopy Eigenvalue Explained variance (%)

0.819

0.202

2.74

1.72

39.07

24.56

Differences in fragment occupancy between species and sides of the motorway were tested for significance using Fisher exact tests applied to 2 x 2  contingency tables. We used within–subjects ANOVA to analyse pairwise variation between eastern and western forests in the abundance of roe deer, birds, and large psammodromus, as well as in vegetation structure (as estimated by PC scores), fragment size, and distance to the motorway. Whenever it was necessary, variables were arcsine– or log–transformed to fulfil the requirements of parametric tests. All analyses were done with Statistica v7.1. Results Our study sites (table 2, fig. 2) were located at comparable distances from the motorway (F1,7 = 1.99, P = 0.201) and showed similar vegetation structure on both sides of the motorway (PC1: F1,7 = 0.03, P = 0.864; PC2: F1,7 = 0.09, P = 0.763). However, the westward– directed ongoing forest fragmentation process made it impossible to find fragments of similar size on either side of the motorway (forest remnants were consistently


Animal Biodiversity and Conservation 34.2 (2011)

335

Table 2. Fragment size (area in ha), distance to the motorway (m), and abundance of four vertebrate species measured in study sites located on either side of the A–1 motorway (see text for details on the abundance indices used for each species: RD. Roe deer fecal groups/ha; LP. Number of large psammodromus/10'; ER. Number of European robins/ha; BW. Number of Bonelli’s warblers/ha). Forests are presented from northernmost to southernmost, and codes shown in figure 1 are in brackets. (a Robins were contacted outside the census band). Tabla 2. Tamaño de los fragmentos (área en ha), distancia a la autovía (m) y abundancia de las cuatro especies de vertebrados, medidas en las zonas de estudio situadas a ambos lados de la autovía A–1 (véase el texto para los detalles de los índices de abundancia utilizados para cada especie: RD. Grupos fecales de corzo/ha; LP. Número de lagartijas colilargas/10'; ER. Número de petirrojos/ha; BW. Número de mosquiteros papialbos/ha). Los bosques se presentan desde los más septentrionales a los más meridionales con los códigos de la figura 1 entre paréntesis. (ª Se encontraron petirrojos fuera de la franja de censo.)

Western study sites

Eastern study sites

Forest

Area

D RD LP ER BW Area

D RD

LP

Cogollos (C)

79.4 180 0.0 0.0 0.24 0.24 100.3 300

0.0 0.0 0.29 0.43

Valdorros (V)

10.4 100 0.0 0.0 0.86 0.86 20.1 270

0.0 1.54 0.22 0.89

Madrigal del Monte (M)

2.9 400 0.0 0.0 0.0 1.79

Encinillas (E)

4.9 400 0.0 0.0 0.0 0.74 39.5 40 25.0 1.61 0.17 0.83

Madrigalejo (m)

1.7 10 16.7 2.17

ER BW

22.7 200 0.0 0.0 0.47 0.94 18.6 50 14.3 1.22

Torrecilla del Monte N (TN)

0.0 2.94 0.0 0.86

8.0 100 17.9 0.0 0.0 0.24 76.5 30 35.7 1.21 1.43 1.19 a

Torrecilla del Monte S (TS)

14.1

40 17.8 0.0 0.48 0.71 76.5 30 33.0 1.55 0.88 1.10

Bahabón de Esgueva (B)

48.6

40 0.0 0.83 0.0a 0.48 463.0 90 65.9 1.25 0.22 1.10

Fragment size (ha)

170 130

Distance to motorway (m) 220

0.4

180

90

Vegetation structure (factor scores) PC–1

PC–2

0.2

140

50

0.0 100

–0.2 –0.4

10

W

E

60

W

E

W

E

W

E

Fig. 2. Main features of the study fragments. Characteristics of the forest fragments studied on the western (W) and eastern (E) sides of the motorway (means ± 1 SE). The Y axes were plotted on log scale for fragment size and distance to the motorway to meet the assumptions of the analysis. Fig. 2. Características principales de los fragmentos estudiados. Características de los fragmentos forestales estudiados en los lados occidental (W) y oriental (E) de la autovía (media ± DE). El eje de las ordenadas está a escala logarítmica, segun el tamaño del fragmento y la distancia a la autovía, para ajustarse a las suposiciones del análisis.


Tellería et al.

336

No. of occupied fragments

Western side of motorway 8 7 6

Eastern side of motorway P = 1.000

P = 0.005 P = 0.067

5

P = 0.715

P = 0.690

4 3 2 1

P. algirus C. capreolus T. epidus E. rubecula P. bonelli Fig. 3. Frequency of occupancy of forest fragments. Number of forest fragments occupied by each species of forest vertebrate on the western and eastern sides of the motorway. P–values indicate the results of Fisher exact tests. Fig. 3. Frecuencia de ocupación de los fragmentos forestales. Número de fragmentos forestales ocupados por cada especie de vertebrado en los lados occidental y oriental de la autovía. Los valores P indican los resultados del test exacto de Fisher.

larger on the eastern side: F1,7 = 6.74, P = 0.036; figs. 1, 2). As a consequence, within–subjects ANOVAs with abundance indices as the dependent variable were re–analysed using the difference in fragment size between eastern and western sites as a covariate. Also, distance to the road was larger, though not significantly, on the western side. Thus, we included the difference in distance to the road between eastern and western sites as a second covariate. A preliminary inspection of fragment occupancy on either side of the motorway revealed significant differences among our five model species (fig. 3, table 3). The large psammodromus and the roe deer were more frequent on eastern fragments than on western fragments (although the result was marginally non–significant in the case of the deer), whereas eyed lizards and birds occupied fragments with identical frequencies on both sides of the motorway (fig. 3). On the eastern side, where forest cover was higher and less fragmented, all but the most frequent species, Bonelli warbler, and the least frequent species, the eyed lizard, showed similar frequencies of occupancy (upper right half of table 3). However, on the western side, that had smaller and scarcer fragments interspersed among cereal fields (fig. 1), the large psammodromus occupied significantly fewer fragments than birds, with roe deer and eyed lizards showing intermediate frequencies between the large psammodromus and the birds (lower left half of table 3). The numbers of animals on each side of the motorway also showed different patterns in the four

species for which we could obtain accurate estimates of abundance (fig. 4). While psammodromus lizards (F1,7  = 6.74, P = 0.002), roe deer (F1,7  = 9.37, P  =  0.018), and Bonelli’s warblers (F1,7 = 7.87, P = 0.026) decreased abundance on the western side of the motorway, robins showed similar abundance on both sides (F1,7 = 0.38, P = 0.559). However, the difference vanished in the bird and the deer when fragment size and distance from the motorway were included as covariates in the within–subjects analyses (roe deer: F1,5 = 3.44, P = 0.123; Bonelli’s warblers: F1,5=  0.74, P  = 0.428). On the other hand, the abundance of lizards was still lower on the western forests after controlling for the effects of covariates (F1,5  =  18.99, P = 0.007). Including vegetation variables (PC–1 and PC–2 scores) as covariates did not change the above results qualitatively (results not shown). Discussion An overview of the effects of the A–1 motorway on vertebrate distribution A particular example of the negative consequences of roads for wildlife is the production of asymmetric effects on the surrounding landscape (Forman & Alexander, 1998; Taylor et al., 1993). Such an effect appears to be important in our study system where extensive eastern woodlands ('Sierra de la Demanda'; fig. 1) seem to operate as source areas for the


Animal Biodiversity and Conservation 34.2 (2011)

337

Table 3. Fisher exact tests comparing the frequency of occupancy of forest fragments by pairs of vertebrate species on the eastern (upper right half of the matrix) and western (lower left half) sides of the motorway. Tabla 3. Test exacto de Fisher, comparando la frecuencia y la ocupación de los fragmentos forestales por pares de especies de vertebrados en los lados oriental (mitad superior derecha de la matriz) y occidental (mitad inferior izquierda) de la autovía.

P. algirus

T. lepidus

C. capreolus

E. rubecula

P. bonelli

P. algirus

0.141

0.500

0.500

0.500

T. lepidus

0.141

0.304

0.304

0.039

C. capreolus

0.500

0.304

0.715

0.233

E. rubecula

0.020

0.304

0.067

0.233

< 0.001

0.039

0.003

0.233

P. bonelli

colonization of the fragmented landscape. Increased fragmentation westwards leads to a reduced cover of forests, which may reduce regional density of forest vertebrates around the motorway. As a consequence, and assuming a metapopulation dynamics scenario (Hanski, 1998), fragments on the western side of the motorway will have fewer opportunities to receive immigrants to compensate for local extinctions through a rescue effect (Brown & Kodric–Brown, 1977).

1.6

P. algirus (nº/10' of search)

On the other hand, our design may be criticized because the present distribution of animals on the western side of the motorway should be compared with their distribution before the motorway was constructed, but we lack the data needed to perform such a comparison. This limits our ability to determine to what extent the observed distribution pattern is due to the motorway–caused interruption of the directional flow of individuals or to the reduced potential of

C. capreolus (nº of faecal groups/ha) 40 6.0

E. rubecula (nº/10 ha) 16

30

14

20

1.2

4.5

10

0.8

P. bonelli (nº/10 ha)

12 10

3.0

8

0.4 1.5

0.0 W

E

W

E

6 W

E

W

E

Fig. 4. Abundance patterns in forest fragments. Mean (± 1 SE) values of abundance indices computed for large psammodromus, roe deer, robins and Bonelli’s warblers in forest fragments located on the western (W) and eastern (E) sides of the motorway. The Y axes were plotted on log scale when required to meet the assumptions of the analysis. Fig. 4. Patrones de abundancia en los fragmentos forestales. Valores medios (± DE) de los índices de abundancia calculados para la lagartija colilarga, el corzo, el petirrojo y el mosquitero papialbo, en fragmentos forestales localizados en los lados occidental (W) y oriental (E) de la autovía. Las ordenadas se consignaron en valores logarítmicos cuando fue necesario, para ajustarse a las suposiciones del análisis.


338

the western forest archipelago to maintain metapopulations by autonomous extinction–recolonization dynamics. However, both processes predict population declines on the western side of the motorway, and they may even interact to produce synergistic effects. For instance, local or regional extinctions in disconnected western fragments will be difficult to avoid if animals coming from eastern woodlands are unable to cross the motorway. That the motorway hampers westwards dispersal was supported by the fact that isolated forests on the eastern edge of the motorway sustain populations of all studied species, while some of these species are absent or show reduced abundance in neighbouring forests on the western side of the road. Thus, the idea that reduced connectivity is responsible for the observed pattern seems to be supported by our data because, at the local scale of our study, landscape structure does not differ between eastern and western fragments (fig. 1). In fact, all study sites should be regarded as part of the same fragmented landscape (the distance between pairs of fragments ranges from 70 to less than 500  m; see table 2). As a consequence, we should not expect to find any difference between paired eastern and western fragments if the motorway, that provides the very basic criterion to divide the landscape into western and eastern sides, had not acted as a barrier that disrupts the westwards movement of individuals Species–specific effects of the motorway The distribution of species on either side of the motorway reveals lower frequencies of occurrence in western fragments for the large psammodromus and the roe deer (fig. 3). However, the other species occurred with exactly the same frequency on both sides of the motorway, although they differed in overall frequency of occurrence. The two bird species showed high frequencies of occurrence on both sides of the motorway (fig. 3), which was expected from their high dispersal capability. Besides, the abundance of bird species did not differ between the two sides of the motorway. We found a lower abundance of Bonelli’s warblers on the western side, but the effect vanished when fragment size was included as a covariate in the analysis. Eyed lizards had a scattered distribution, occurring in 50% of the fragments on both sides of the motorway. This result suggests that the distribution of eyed lizards is not affected by the motorway. In fact, compared to the large psammodromus, which is a forest specialist, the eyed lizard is a habitat generalist that can use a wide variety of habitats (uncultivated cropland boundaries, slopes or hedges) as corridors in cultivated landscape (Santos & Tellería, 1989). Moreover, eyed lizards are often found basking on road verges (in fact, road–killed eyed lizards are not uncommon in our study area). Such behaviour brings about opportunities for individuals to cross roads. However, further data (e.g. radiotracking, genetic analyses, etc.) is necessary to confirm that the motorway is unimportant as a barrier for the dispersal of eyed lizards.

Tellería et al.

Roe deer were broadly distributed on the eastern side of the motorway where they were only absent from the two northernmost forests, perhaps due to the disturbing effects of human activity generated in the nearby town of Burgos (population 180.000; ca. 10 km) and in three small villages in the close vicinity (< 1 km) of these forests. In contrast, on the western side roe deer occupied only the two forests located close to 'Torrecilla del Monte' (table 2, fig. 1). In this sector, the motorway is crossed by one overpass and five underpasses that are suitable for this species (protective fences running aside the motorway keep roe deer from crossing over the road). However, it is important to note that although the motorway might act as a barrier against the dispersal of roe deer, the distribution pattern of this species in the study area seems to respond mainly to fragmentation and habitat deterioration effects. In a study on the expansion of roe deer in central Spain, Virgós & Tellería (1998) showed that their need for large forests to rest and to browse was the main limiting factor. From this perspective, the western sector of the motorway offered less suitable habitat than the large forest patches on the eastern side, and in fact the difference in abundance between eastern and western remnants disappeared after controlling for the effects of fragment size. On the other hand, the Torrecilla sector, besides offering suitable underpasses, also retains a high forest cover on both sides of the motorway (fig. 1). Finally, the large psammodromus offered the clearest example of the barrier effect caused by the motorway. This species was present in seven out of eight study sites on the eastern side, but only in one out of eight sites on the western side (fig. 2, table 2). Abundance was lower in western fragments, even after controlling for the effects of woodlot size and distance from the motorway. Previous studies suggest that the current size and habitat structure of forest fragments on the western side of the motorway make them suitable to host stable populations of these lizards (Santos et al., 2008); according to the logistic regression model used by these authors, the average probabilities of finding lizard populations in eastern and western fragments are 0.963 and 0.933, respectively, with no significant differences between the two sides of the motorway (P = 0.380). Therefore, it follows from our results that the large psammodromus faces a major barrier against dispersal, which has presumably stopped the westward expansion of the species. Moreover, a study of the distribution of the large psammodromus further to the west in our study area failed to detect populations in 17 forest fragments with suitable habitat for the species (Díaz et al., 2000). The absence of the species from the western part of this area was attributed to local extinction in forest remnants due to shrub cleaning produced by overgrazing and cultivation during the 1940s–1950s (Díaz et al., 2000). With the abandonment of such practice since the 1960s, forested patches have been regenerating in the study area, and we have found lizards in an increasing number of sites on the eastern side (Santos et al., 2008), an effect that has not been paralleled on the western side of the motorway.


Animal Biodiversity and Conservation 34.2 (2011)

Further research is needed to determine why the large psammodromus is so sensitive to the barrier effect, but some biological traits are candidate to explain its failure to cross the motorway. In the first place, the dispersal of this species through inhospitable habitat may be inhibited by its reluctance to move on bare ground. These lizards are positively associated with low shrub cover (Díaz & Carrascal, 1991) and they spend seasonal and daily periods of inactivity hidden in vegetated ground. Díaz (1992) showed that the choice of compass directions around oak shrubs allows sun–seeking lizards to minimize their escape distance towards the nearest shrub, thus reducing predation risk. It has also been shown that the approach and escape distances of lizards from a deciduous oak forest were larger at the times of year when oaks were unleaved (Martín & López, 1995). In the study area, although we have observed dispersal among nearby forest fragments across open fields (Santos et al., 2009), the species is rarely seen on bare ground. All these facts point towards the idea that the large psammodromus strictly avoids open surfaces such as paved roads and their ditches, which should make motorways formidable barriers for this species. Indeed, we have never found road–killed individuals of this species in our study area, although they are much more abundant on the eastern side of the motorway than eyed lizards. Conclusions Despite the well–known decline of animal populations in the proximity of linear infrastructures (Benítez–López et al., 2010), little work has been dedicated to explore the effects of roads on the spatial patterning of animal abundance. Although our study is geographically and taxonomically restricted, it supports the view that linear infrastructures are emerging determinants of the current configuration of species’ ranges. Despite the efforts made to increase the effectiveness of crossing structures, many animal populations will face the detrimental effects of reduced population connectivity. This may produce decreased dispersal ability, delayed or suppressed re–colonization and, in situations of long– term isolation, extinctions and/or micro–evolutionary processes (Saccheri et al., 1998; Shepard et al., 2008; Strasburg, 2006). In order to correct barrier effects caused by existing roads and highways and to prevent similar effects of the infrastructures to come, transportation planning and policy will require adding more proactive approaches to the already recommended construction of suitable passes. It will require monitoring the long–term effects of infrastructures on affected populations, and delineating specific management measures to preclude population decline or extinction. Options such as traffic management in some road sectors (e.g. traffic calming; Langevelde & Jaarsma, 2009), or translocation of individuals to reinforce isolated populations (Santos et al., 2009) appear to be suitable management approaches to prevent these negative effects on the distribution of terrestrial forest vertebrates.

339

Acknowledgements Projects CCG07–UCM/AMB–3010 (Comunidad de Madrid–Universidad Complutense), 910577–658 (Universidad Complutense–BSCH), and CGL2010–17928 (Ministerio de Ciencia e Innovación, Spain). References Baguette, M. & Van Dyck, H., 2007. Landscape connectivity and animal behavior: functional grain as a key determinant for dispersal. Landscape Ecology, 22: 1117–1129. Benítez–López, A., Alkemade, R. & Verweij, P. A., 2010. The impacts of roads and other infrastructure on mammal and bird populations: A meta–analysis. Biological Conservation, 143: 1307–1316. Brown, J. H. & Kodric–Brown, A., 1977. Turnover rates in insular biogeography: effect of immigration on extinction. Ecology, 58: 445–49. Clevenger, A. P. & Waltho, N., 2005. Performance indices to identify attributes of highway crossing structures facilitating movement of large mammals. Biological Conservation, 121: 453–464. Díaz, J. A., 1991. Temporal patterns of basking behaviour in a Mediterranean lacertid lizard. Behaviour, 118: 1–14. – 1992. Choice of compass directions around shrub patches by the heliothermic lizard Psammodromus algirus. Herpetologica, 48: 293–300. Díaz, J. A. & Carrascal, L. M., 1991. Regional distribution of a Mediterranean lizard: influence of habitat cues and prey abundance. Journal of Biogeography, 18: 291–297. Díaz, J. A., Carbonell, R., Virgós, E., Santos, T. & Tellería, J. L., 2000. Effects of forest fragmentation on the distribution of the lizard Psammodromus algirus. Animal Conservation, 3: 235–240. Díaz, J. A., Monasterio, C. & Salvador, A., 2006. Abundance, microhabitat selection, and conservation of eyed lizards Lacerta lepida: a radiotelemetric study. Journal of Zoology, 268: 295–301. Díaz, J. A., Pérez–Tris, J., Tellería, J. L., Carbonell, R. & Santos, T., 2005. Reproductive investment of a lacertid lizard in fragmented habitat. Conservation Biology, 19: 1578–1585. Epps, C. W., Palsbøll, P. J., Weyhausen, J. D., Roderick, G. K., Ramey, R. R. & McCullough, D. R., 2005. Highways block gene flow and cause a rapid decline in genetic diversity of desert bighorn sheep. Ecology Letters, 8: 1029–1038. Fahrig, L. & Rytwinski, T., 2009. Effects of roads on animal abundance: an empirical review and synthesis. Ecology and Society, 14(1): 21. http://www.ecologyandsociety.org/vol14/iss1/art21 Forman, R. T. T. & Alexander, L. E., 1998. Roads and their major ecological effects. Annual Review of Ecology and Systematics, 29: 207–231. Goodwin, B. J. & Fahrig, L., 2002. How does landscape structure influence landscape connectivity? Oikos, 99: 552–570. Grilo, C., Bissonette, J. A. & Santos–Reis, M., 2008.


340

Response of carnivores to existing highway culverts and underpasses: implications for road planning and mitigation. Biodiversity and Conservation, 17: 1685–1699. Hanski, I., 1998. Metapopulation dynamics. Nature, 396: 41–49. Langevelde, F. & Jaarsma, C. F., 2009. Modeling the effect of traffic calming on local animal population persistence. Ecology and Society, 14: 39. http://www.ecologyandsociety.org/vol14/iss2/art39 Mader, H. J., 1984. Animal habitat isolation by roads and agricultural fields. Biological Conservation, 29: 81–96. Martín, J. & López, P., 1995. Influence of habitat structure on escape tactics of the lizard Psammodromus algirus. Canadian Journal of Zoology, 73: 129–132. Mata, C., Hervás, I., Herranz, J., Suárez, F. & Malo, J. E., 2005. Complementary use by vertebrates of crossing structures along a fenced Spanish motorway. Biological Conservation, 124: 397–405. Ministerio de Medio Ambiente, 2006. Prescripciones técnicas para el diseño de pasos de fauna y vallados perimetrales. Documentos para la reducción de la fragmentación de hábitats causada por infraestructuras de transporte. O. A. Parques Nacionales. Ministerio de Medio Ambiente, Madrid. [In Spanish.] Pulliam, H. R., 1988. Sources, sinks, and population regulation. American Naturalist, 132: 652–669. Riley, S. P. D., Pollinger, J. P., Sauvajot, R. M., York, E. C., Bromley, C., Fuller, T. K. & Wayne, R. K., 2006. A southern California freeway is a physical and social barrier to gene flow in carnivores. Molecular Ecology, 15: 1733–1741. Roedenbeck, I. A., Fahrig, L., Findlay, C. S., Houlahan, J. E., Jaeger, J. A. G., Klar, N. Kramer–Schadt, S. & Van der Grift, E. A., 2007. The Rauischholzhausen agenda for road ecology. Ecology & Society, 12: 11. http://www.ecologyandsociety.org/vol12/iss1/art11/ Saccheri, I., Kuussaari, M., Kankare, M., Vikman, P., Fortelius, W. & Hanski, I., 1998. Inbreeding and extinction in a butterfly metapopulation. Nature, 392: 491–494. Sáez–Royuela, C. & Tellería, J. L., 1986. The in-

Tellería et al.

creased population of the wild boar (Sus scrofa L) in Europe. Mammal Review, 16: 97–101. Santos, T., Díaz, J. A., Pérez–Tris, J., Carbonell, R. & Tellería, J. L., 2008. Habitat quality predicts the distribution of a lizard in fragmented woodlands better than habitat fragmentation. Animal Conservation, 11: 46–56. Santos, T., Pérez–Tris, J., Carbonell, R., Tellería, J. L. & Díaz, J. A., 2009. Monitoring the performance of wild–born and introduced lizards in a fragmented landscape: Implications for ex situ conservation programmes. Biological Conservation, 142: 2923–2930. Santos, T. & Tellería, J. L., 1989. Preferencias de hábitat y perspectivas de conservación en una comunidad de lacértidos en medios cerealistas del centro de España. Revista Española Herpetología, 3: 259–272. [In Spanish.] Shepard, D. B., Kuhns, A. R., Dreslik, M. J. & Pillips, C. A., 2008. Roads as barriers to animal movement in fragmented landscapes. Animal Conservation, 11: 288–296. Strasburg, J. L., 2006. Roads and genetic connectivity. Nature, 440: 875–876. Taylor, P. D., Fahrig, L., Henein, K. & Merriam, G., 1993. Connectivity is a vital element of landscape structure. Oikos, 68: 571–573. Tellería, J. L. & Sáez–Royuela, C., 1984. The large mammals of central Spain –an introductory view. Mammal Review, 14: 51–56. Tellería, J. L. & Virgós, E., 1997. Distribution of an increasing roe deer population in a fragmented Mediterranean landscape. Ecography, 20: 247–252. Trombulak, S. C. & Frissell, C. A., 2000. Review of ecological effects of roads on terrestrial and aquatic communities. Conservation Biology, 14: 18–30. Underhill, J. E. & Angold, P. G., 2000. Effects of roads on wildlife in an intensively modified landscape. Environmental Reviews, 8: 21–39. Virgós, E. & Tellería, J. L., 1998. Roe deer habitat selection in the southwestern edge of its range (Spain). Canadian Journal of Zoology, 76: 1294–1299.


Animal Biodiversity and Conservation 34.2 (2011)

341 Review

Conservation genetics of Iberian raptors B. Martinez–Cruz

Martinez–Cruz, B., 2011. Conservation genetics of Iberian raptors. Animal Biodiversity and Conservation, 34.2: 341–353. Abstract Conservation genetics of Iberian raptors.— In this paper I provide an overview of conservation genetics and describe the management actions in the wild that can benefit from conservation genetic studies. I describe the genetic factors of risk for the survival of wild species, the consequences of loss of genetic diversity, inbreeding and outbreeding depression, and the use of genetic tools to delimitate units of conservation. Then I introduce the most common applications of conservation genetics in the management of wild populations. In a second part of the paper I review the conservation genetic studies carried on the Iberian raptors. I introduce several studies on the Spanish imperial eagle, the bearded vulture, the black vulture and the red kite that were carried out using autosomal microsatellite markers and mitochondrial DNA (mtDNA) sequencing. I describe studies on the lesser kestrel and Egyptian vulture that additionally applied major histocompatibility complex (MHC) markers, with the purpose of incorporating the study of non–neutral variation. For every species I explain how these studies can be and/or are applied in the strategy of conservation in the wild. Key words: Conservation genetics, Conservation genomics, Molecular markers, Iberian raptors, Management of threatened populations. Resumen Genética de la conservación de rapaces ibéricas.— En este artículo se da una visión global de lo que es la genética de la conservación y cuáles son las acciones de manejo en la naturaleza que pueden beneficiarse de los estudios genéticos. Se presentan en primer lugar los factores genéticos de riesgo para la supervivencia de las especies y cuáles son las consecuencias de la pérdida de diversidad genética y de la depresión tanto por endogamia como por exogamia. Se explica el uso de las herramientas genéticas en la delimitación de las unidades de conservación. Tras ello se explica cuáles son las aplicaciones más comunes de la genética de la conservación en el manejo de poblaciones silvestres. En una segunda parte del artículo se hace una revisión de los estudios en genética de la conservación llevados a cabo en rapaces ibéricas. Se explican varios estudios llevados a cabo sobre el águila imperial ibérica, el quebrantahuesos, el buitre negro y el milano real usando marcadores en microsatélites autosomales y secuencias de ADN mitocondrial (mtDNA). Se describen estudios sobre el cernícalo primilla y el alimoche que han utilizado adicionalmente marcadores en el complejo de histocompatibilidad mayor (MHC) con el propósito de incorporar el estudio de variación no neutral. Para cada una de las especies se explica cómo estos estudios se pueden aplicar y/o se aplican en las estrategias de conservación de dichas especies en la naturaleza. Palabras clave: Genética de la conservación, Genómica de la conservación, Marcadores moleculares, Rapaces ibéricas, Manejo de poblaciones amenazadas. (Received: 27 VII 11; Conditional acceptance: 19 XI 11; Final acceptance: 29 XI 11) Begoña Martinez–Cruz, Inst. de Biologia Evolutive (CSIC–UPF), Dept. de Ciències Experimentals i de la Salut, Univ. Pompeu Fabra, c./ Doctor Aigüader 88, 08003 Barcelona, Espanya (Spain).

ISSN: 1578–665X

© 2011 Museu de Ciències Naturals de Barcelona


342

What is conservation genetics? Population genetics theory predicts that when a population goes through a steep and maintained decline drift will become the preponderant evolutionary force in detriment of selection. Under genetic drift, allele frequencies will change arbitrarily from one generation to another, with some alleles getting fixed and others getting lost, thus decreasing genetic diversity and increasing inbreeding in the population. Conservation genetics is a relatively novel discipline dealing with these issues by describing genetic patterns and assessing evolutionary processes in endangered species. A loss of genetic diversity could undermine the adaptive potential of the population, and it has been shown that the adaptive diversity of a population is most of the times already present in the population and not created anew when it is exposed to an environmental change or a new environment (Barrett & Schluter, 2008). The consequences of a limited genetic diversity when facing an environmental perturbation are clearly exemplified in the response of hosts to novel pathogen infections (O’Brien & Evermann, 1988). Otherwise, some long–lived species show intrinsic low levels of diversity and perform as well as their sister species with higher diversity levels (e.g. Milot et al., 2007), but this seems to be more the exception than the rule. In addition, from generation to generation, the inheritance of identical copies of the same allele will increase even under a random mating system. At the same time, recessive deleterious alleles that purifying selection maintains at low frequencies, and mostly in heterozygosity, might evolve under drift as neutral variants and eventually increase in frequency. Both factors contribute to the inbreeding depression of the population, compromising the persistence of the species or population in the long term. This has been documented extensively in the wild and for multiple taxa (Crnokrak & Roff, 1999; Keller & Waller, 2002). For instance, inbreeding depression has been invoked as the origin of the lethal coronary disease and low sperm quality in the Florida panther (Roelke et al., 1993), the mortality due to viruses in dolphins (Valsecchi et al., 2004), and the decrease of sperm quality in rabbits (Gage et al., 2006). Although desirable, predicting whether and to what degree a population will suffer from inbreeding depression is almost impossible due to the stochastic nature of drift, but some patterns have been recognized experimentally (Reed & Frankham, 2003; Reed et al., 2003; Armbruster & Reed, 2005).

Martinez–Cruz

The combination of reduced genetic diversity and inbreeding depression are expected to precipitate the extinction of endangered populations through their effects on fitness components. Even though some authors have argued that the extinction of a population will happen long before the genetic factors become sufficiently important to provoke it (Lande & Shannon, 1996), the existence of lower levels of genetic diversity in endangered species than in their non–endangered sister species argues against this view (e.g. Frankham, 1996). Furthermore, inbreeding depression has been shown to severely affect extinction probabilities in simulations (Tanaka, 2000; O’Grady et al., 2006) and empirical studies (Saccheri et al., 1998; Vilas et al., 2006; Reed et al., 2007). The introduction of new alleles from external populations can ameliorate the negative effects of inbreeding, a phenomenon that has been termed 'genetic rescue'. For instance, the arrival of a single individual increased the genetic diversity and population size of a wolf population in Scandinavia (Vila et al., 2003), and the introduction of eight puma females into the Florida panther population restored genetic diversity and reversed the negative demographic trend of the population (Johnson et al., 2010). On the other hand, the mix of divergent gene pools may lead to the opposite problem, termed outbreeding depression, in which offspring or e second generation hybrids have a lower fitness than parents due to phenotypes not adapted to local conditions or the disruption of coadapted gene complexes by recombination (Lynch & Walsh, 1998). As in the case of inbreeding depression, evaluating the risks for outbreeding depression is very difficult. In addition to measuring genetic diversity, molecular tools have been applied in conservation genetics to define conservation units (e.g., subpopulation, population, species) through the study of the evolutionary history of the population. The definition of these units relies on any genetic discontinuity requiring independent management and tries to answer the question of which populations to manage conjunctly or independently. Two main units of conservation have been proposed: ESUs (Evolutionary Significant Units) and MUs (Management Units, inside the ESUs). Although there are different definitions of ESU in the literature (e.g. Ryder, 1986; Moritz, 1994; Crandall et al., 2000), the main criteria for a synthetic definition encompass reproductive isolation, adaptive differentiation and concordance across genetic, morphological,

Fig. 1. Maps with the distribution ranges of the Spanish imperial eagle (A), bearded vulture (B), lesser kestrel (C), Egyptian vulture (D), black vulture (E), and red kite (F). For A, C, D, E, and F maps were taken from the IUCN (2011, IUCN Red List of Threatened Species, version 2011.1 <www.iucnredlist. org>), downloaded on 10 June 2011. Fig. 1. Mapas con los rangos de distribución de águila imperial ibérica (A), quebrantahuesos (B), cernícalo primilla (C), alimoche (D), buitre negro (E) y milano real (F). Los mapas A, C, D, E y F se han tomado de IUCN (2011, Lista roja IUCN de las especies amenazadas, versión 2011.1 <www.iucnredlist.org>), descargados el 10 de junio 2011.


Animal Biodiversity and Conservation 34.2 (2011)

343

A

B

Spain Portugal

Algeria Morocco

Russian Federation

C Poland Ukraine

France Spain

Kazakhastan

D France

Mongolia

Spain

Turkey

Mali

Egypt

Niger

Saudí Arabia

Kazakhastan

Turkey

China Iraq

Algeria

India

Egypt

Saudí Arabia

Mauritania

Chad Sudan

Nigeria

Russian Federation

China

Iraq Algeria

PolandBelarus Ukraine

Niger

Mali

Yemen

Chad Sudan

Ethiopia Nigeria

India

Ethiopia

Kenya Angola Angola Zambia Namibia

Range type Native (resident) Native (breeding) Native (non breeding)

E

F

Norway

Finland Sweden

Germany

Russian Federation

Poland Ukraine

Spain

Kazakhastan

Ukraine

Mongolia Spain

Turkey

Mali

Egypt Chad

Nigeria

Saudí Arabia

Turkey

China

Iraq Algeria

Russian Federation

Belarus

Hungary Romania

France France

Poland

Iraq

Morocco India

Algeria

Libyan Arab Jamahiriya

Sudan

Egypy

Ethiopia Mauritania

Mali

Niger Chad

Angola Australia

Guinea

Nigeria

Sudan

Saudi Arabia

Yemen Ethiopia


344

behavioural, geographical data types (Allendorf & Luikart, 2007). Information from sets of molecular markers evolving at different rates provides a good base for assessing the main criteria for the delimitation of ESUs. This is usually based on neutral variation only, but the new genomic approaches promise to overcome this limitation (e.g. Primmer, 2009). Every ESU might be sub–divided into one or more Management Units (MUs), described as a population or group of populations that are demographically independent (Moritz, 1994, 2002). Genetic markers do help delimitate MUs through the analysis of panmixia, gene flow and genetic structure (Waples & Gaggiotti, 2006; Palsboll et al., 2007). Application of conservation genetics in the management of wild populations The most commonly practised actions in management for conservation of endangered populations that can benefit from genetic studies are captive breeding, reintroduction of individuals in the wild and translocations. All three kinds of actions are complementary to any other management action and may try to maximize the global genetic diversity while preserving local adaptations. Any captive breeding program should aim at incorporating a good number of healthy founders representing the genetic diversity remaining in the wild and at minimizing effects of the unavoidable drift. Genetic management of captive breeding populations should be conducted after compiling basic pedigree and demographic data on the population (Ballou & Foose, 1996). In the usual situation where information on the relatedness of the founders is lacking, molecular markers could provide these estimates to optimize genetic management (Gautschi et al., 2003a; Russello & Amato, 2004). Deleterious alleles in captive populations might stochastically drift to high frequencies and their elimination may then enter into conflict with the preservation of overall genetic diversity. For instance, the hereditary condrodistrophy, a Mendelian disease due to a single recessive allele, could be prevented in the Californian condor captive population by avoiding crossing breeders that would produce affected individuals, but the reproduction of carriers could not be prevented without a major impact on global genetic diversity (Ralls et al., 2000). Nonetheless, the success of the reintroduction of Californian condor to the wild attests to the success of the application of genetic criteria in the captive breeding program (Ralls & Ballou, 2004). The second common management action considers the reintroduction of populations in the wild. Genetic tools may assist in the careful selection of the origin, number, and sex of the released individuals, and in the monitoring of the reintroduced population. Whenever possible, the best candidates to be reintroduced in a determined area would be individuals belonging to the same lineages of the extinct populations or from nearby populations in order to maximize adaptation (Frankham et al., 2002).

Martinez–Cruz

In the short term, inbreeding depression and loss of genetic diversity can be counteracted by the translocation of individuals from a different population (the genetic rescue effect mentioned above). In this case, and in order to minimize the possible effects of outbreeding depression, individuals should come from nearby populations and the number of them should be the minimum needed to revert to the normal situation (Kleiman & Stanley, 1994). Conservation genetics on Iberian raptors Avian biodiversity in the Iberian peninsula is one of the highest in Europe. Due to its geographical location it is an important stopover for migratory birds on their way from northern Europe to Africa or viceversa. But Iberia is not only a crossing point as millions of birds migrate directly to the Iberian estuaries and wetlands to winter. Additionally, the high diversity of biotopes in Iberia harbours a large number of resident species. Despite their ecological relevance, many birds of prey are threatened to different degrees due to a long history of direct human persecution and anthropogenic disturbances, such as electrocution on power lines, decreases in prey populations, poisoning, direct shooting, and landscape fragmentation. Many raptor populations have recently declined and become fragmented, some of them being currently confined to small and geographically isolated patches (see below). The genetic consequences of such processes and how genetics can assist raptor conservation are issues that have received much attention during the last decades. We present here what has been done in the field of conservation genetics as it applies to Iberian raptors. The vast majority of works on this subject were carried out at the Molecular Ecology Laboratory (LEM) of Doñana Biological Station in Sevilla, Spain (on Spanish imperial eagles,Egyptian vultures, some of the bearded vulture studies, works on lesser kestrel as well as red kites ) or elsewhere in collaboration with Doñana researchers (studies on bearded vultures by Gautschi et al., 2003). The creation of LEM in the mid 1990s boosted conservation–genetics–driven projects not only in raptors but also in many other non–model organisms, disentangling the genetic processes and patterns occurring in the wild and how they affect and can be used in the conservation of wild species. The Spanish imperial eagle, Aquila adalberti The Spanish imperial eagle is one of the most iconic species of the Iberian fauna. With around 200 pairs in the wild distributed in several breeding nuclei in the southern quadrant of the Iberian peninsula (fig.  1A) (Birdlife international), the Spanish imperial eagle is one of the most endangered raptors in the world (Collar & Andrew, 1988; Birdlife International, 2004). The species must have been relatively abundant until the mid–20th century with a distribution area that comprised most of the Iberian peninsula and


Animal Biodiversity and Conservation 34.2 (2011)

the north of Africa (Gonzalez et al., 1989). By the end of the 19th century and during most of the 20th century, the species went through a steep decline and population fragmentation due to human pressures (Collar & Andrew, 1988), mainly direct prosecution and collection (Gonzalez et al., 1989), poisoning (Ferrer, 2001), electrocution on power lines (Ferrer & Hiraldo, 1992), and two severe viral epidemics of its main prey, the rabbit Oryctologus cuniculus (Villafuerte et al., 1995). Many traditional breeding areas were lost and the population that had previously been settled in humanized areas such as forested plains and fluvial valleys moved to non–humanized mountainous areas, probably as a response to human prosecution (Ferrer, 2001). The 1974 breeding survey revealed only 38 couples in the wild, although the populations started to recover in the 1980s. By the 1990s the species reached around 130 couples (Ferrer, 2001), and thanks to the conservation actions the population has been increasing since 2000, reaching 282 pairs in the wild in 2010 (Spanish Ministry of Environment, unpublished report). However, the population is still classified as vulnerable by the IUCN (International Union for Conservation of Nature) due to its small population size and because it still relies on human actions for its survival (IUCN, 2011). To understand whether the demographic history of decline and fragmentation of the Spanish imperial eagle had affected the genetic variation of the species, a study was carried out using an extensive sampling of the present population covering the whole distribution range, and of the historical population, for both mitochondrial sequences and nuclear microsatellite markers (Martinez–Cruz et al., 2004, 2007). The historical population was represented by museum individuals covering the historical distribution range from 1853 to 1904 when the population was abundant and not endangered. The origin and evolution of the Spanish imperial eagle were further investigated by analysing its divergence from its sister species, the Eastern imperial eagle (Martinez–Cruz & Godoy, 2007), that numbers over 5000 breeding pairs (Ferrer & Negro, 2004) and is not globally threatened. Only three different mitochondrial control region haplotypes differing in one or two base pairs were found in the Spanish imperial eagle, resulting in very low levels of contemporary haplotype diversity (H = 0.322), although levels were already low in the historical population (H = 0.498) in comparison with the higher diversity found in the Eastern imperial eagle (H = 0.779). In contrast, diversity in the nuclear genome seemed to have been unaffected by the demographic decline, and levels of diversity measured by heterozygosity between either current vs. historical populations (0.603 and 0.627, respectively) or between Spanish vs. Eastern imperial eagle populations (0.549 and 0.627, respectively) were similar. Nonetheless, the breeding nucleus of Doñana, in South Western Andalusia, showed lower levels of diversity and significant genetic differentiation from the other breeding nuclei. These effects were observed even though genetic and field monitoring data suggested some gene flow and dispersal from and into the population, which indicates extremely low effective

345

population sizes. Indeed, the number of breeding pairs in Doñana National Park is very low, there being only nine in 2009 (López–López et al., 2011), and it is the most geographically isolated nuclei. Historical genetic data indicate a globally panmictic population in the past, so the currently observed genetic structure is the direct consequence of the demographic decline and the fragmentation of the population. Under these circumstances, the authors indicated that conservation management actions should aim to restore the historical situation of the Spanish imperial eagle by managing the population as a whole as a single ESU and aimed to connect the genetically isolated nucleus of Doñana to the rest of the population. This could be done in the short term by the translocation of individuals, and in the long term by the recovery of populations that could serve as stepping stones to connect Doñana with the remaining nuclei. Since 2002, a reintroduction project in Cádiz has released about 45 individuals into the wild by hacking. In 2010, the project finally succeeded with the first reproduction attempt of a reintroduced couple (Junta de Andalucía, Counsel of Environment, www. juntadeandalucia.es/medioambiente). On the other hand, the continuous efforts to breed the Spanish imperial eagle in captivity have been unsuccessful so far, precluding the development of an ex situ conservation program for the species. Nonetheless, thanks to assisted reproduction techniques, the first imperial eagle chick born in captivity hatched in May 2011 (http://www.jccm.es/web/en/CastillaLaMancha/index/ notaPrensa1212703217441np/index.html), but there is still a long way to go to develop a successful captive breeding program for the species. Based on a recent divergence date and lack of complete isolation in the past estimated through coalescent–based analyses of genetic data (Martinez– Cruz & Godoy, 2007), the possibility of introducing individuals from the sister Eastern imperial eagle might not be completely ruled out if the situation turned critical. The bearded vulture, Gypaetus barbatus The bearded vulture is a large scavenger that mainly feeds on bones of medium–sized ungulates (Hiraldo et al., 1979). The species inhabits high–altitude mountain ranges in Eurasia and Africa; it is relatively abundant in Asia, from Anatolia to the Tian Shan and the Himalaya Mountains, and in Africa, with populations in South Africa, Lesotho, the Rift Valley and Morocco (Fig. 1B) (IUCN 2011), although very few individuals have been observed in Moroccoin recent years (Godino et al., 2003, 2004). While in the past the species was widely distributed in Southern Europe, during the 20th century the population declined dramatically, becoming extinct in the Alps, the Balkans, Southern Spain and continental Greece. By the end of last century the European population was confined to the Pyrenees (with around 100 breeding pairs, Heredia & Heredia, 1991), and two small and highly endangered populations in Crete and Corsica (with less than 10 breeding couples on each island, Thibault et al., 1992; Xirouchakis & Nikolakakis, 2002; Schaub et al., 2009).


346

In Central Europe a captive breeding program was established to serve as source population for the reintroduction of the bearded vulture in the Alps (Frey et al., 1995). The origin of founders was vaguely known, the majority coming from Central and South Asia and the Caucasus, and some of them from Europe (Frey et al., 1995). By 2005, 137 individuals had been released in the wild (Schaub et al., 2009) and at present there are almost 20 breeding territories and up to 15 incubating couples per year (Zink, 2009). After such success, a plan started for the reintroduction of the bearded vulture in the Cazorla Mountains, where the population had become extinct by the end of the 1970s (Hiraldo et al., 1979), and the first releases took place in 2006 (Simón et al., 2007) and continued until 2010. Additionally, many European countries started in situ conservation actions to preserve the remaining populations. Although management actions were already on their way, it was recognized that application of molecular tools could help to improve the short and long term success of such actions (Negro & Torres, 1999; Gautschi et al., 2003b; Godoy et al., 2004; Simón et al., 2007). Additionally, the knowledge of the evolutionary history of the species would be crucial in the delimitation of conservation units for the optimal management of genetic variation in the species and to decide whether the design of the breeding program in the Alps, which included the mixing of geographically separated populations, might have inadvertently and inappropriately mixed divergent evolutionary lineages (Gautschi et al., 2003a; Godoy et al., 2004). Gautschi et al. (2003a) used 14 nuclear microsatellite loci and sequence information on 268 bp of the mitochondrial control region to estimate kinship in the artificially built captive population, to test close kinship between founder pairs and to determine the unknown population of origin of some founders. After assessing the power and reliability of their dataset and the Queller & Goodnight (1989) kinship estimator to estimate known kinship within the captive population, and with the help of the scarce information on the origins of some individuals, the authors were able to exclude the existence of full–sib and parent–offspring relationships among most founders, but conservatively cautioned against the pairing of some founders. They concluded that unless the strategy of avoiding crosses among close relatives was abandoned and as long as the size of the captive population remained small, inbreeding was not a potential problem in the bearded vulture captive population. Subsequently, Godoy et al. (2004) performed a study on the mitochondrial DNA variation in both the current and the historical population (before the process of decline of the European population) to assess patterns in population diversity and structure and to make an inference on the evolutionary or ecological processes that originated them. The study of the historical population was possible through the sampling of museum specimens covering the whole historical distribution range. Their analysis of 500 bp and 273 bp of the mitochondrial control region in the present and in the historical population, respectively, showed low levels of current diversity in Europe (H = 0.40–0.90)

Martinez–Cruz

in contrast with Asian levels (H = 0.94). The diversity in the Pyrenees was lower in the current population than in the historical population (H = 0.40 vs. H = 0.72). This decrease in diversity was due to the loss of some haplotypes and the increase in the frequency of the most frequent haplotype. Besides, historical bearded vulture populations showed a high level of genetic differentiation, with FST values higher than 0.25 even among close populations. This could be due to low rates of dispersal in both sexes as the same pattern has been found in nuclear microsatellite markers (Gautschi, 2001), or it could be the consequence of the recent isolation of populations and the consequent increase in genetic drift. Field observations support the highly philopatric dispersal behaviour hypothesis that is also consistent with the clear phylogeographic pattern observed. In any case, the marked genetic structure of bearded vulture populations indicated that the extinction of European populations during the last century resulted in the loss of a high proportion of the global genetic diversity of the species. Moreover, this process may continue in the small populations of Crete and Corsica as they approach extinction. The phylogeographic analysis of the mitochondrial variation revealed the existence of two mitochondrial lineages with a mostly disjoint geographical distribution in Eastern (Asian and East African) and Western (European and North African) populations, but with some European populations exhibiting similar frequencies of both lineages (Godoy et al., 2004). The observed patterns suggested that the evolutionary history of the bearded vulture have been marked by a period of isolation in allopatry and a subsequent secondary admixture forming a hybrid zone in Europe and possibly in North Africa following the range expansion of the Easter/African lineage. Furthermore, these results were not concordant with the delimitation of two subspecies, G. b. barbatus in Eurasia and North Africa and G. b. meridionalis in South and East Africa, previously proposed on the basis of morphological differences (Hiraldo et al., 1984). The absence of reciprocal monophyly between the European and the Central Asian populations impeding the rejection of the hypothesis of genetic exchangeability, and the lack of evidence to reject ecological exchangeability argue against their delimitation as different ESU. The authors (Godoy et al., 2004) recommend managing all the population as a whole, validating the strategy of admixing Central Asian and European founders in the current captive breeding for posterior reintroduction programs in the Alps. In fact, the eventual dispersal of reintroduced individuals or their descendants into the remnant European populations could help counteract the effects of drift in these threatened populations. The lesser kestrel Falco naumanni The lesser kestrel Falco naumanni is a small falcon that inhabits steppe and pseudo–steppe ecosystems (Cramp & Simmons, 1980). It is widely distributed in the western Palaearctic and mainly winters in sub– Saharan Africa (Fig. 1C; IUCN, 2011). Although the


Animal Biodiversity and Conservation 34.2 (2011)

global population outnumbers 100,000 individuals, the population was been considered vulnerable until 2011 by the IUCN after a decline of up to 95% in the Palearctic since 1950, mainly due to agriculture intensification, afforestation and urbanisation (Birdlife International, 2011). Many populations in Europe have since become extinct (Biber, 1990), resulting in the current patchy distribution. Recent surveys estimate 8,000 breeding pairs in Spain (IUCN, 2011). Various studies have shown the high natal and breeding philopatric behaviour of this species, as indicated by a steep negative relationship between effective dispersal and geographical distance (Negro et al., 1997; Serrano et al., 2001; Serrano & Tella, 2003; Serrano et al., 2003). Several studies have analysed populations in Spain, France, Italy, Greece, Israel and Kazakhstan in order to understand how the decline and fragmentation of the lesser kestrel population, together with the phylopatric behaviour of the species, has affected the genetics of European populations. For this purpose they have analysed both the neutral –autosomal microsatellite markers and mitochondrial control region sequences–, and the functional variation –the second exon of an MHC class II B locus (Alcaide et al., 2008a, 2008b, 2009a), that is involved in the immune response against parasites and bacterial infections.The contrast of both neutral and non–neutral systems was expected to reveal the extent of local adaptations in a species with low gene flow levels. Both neutral and non–neutral markers showed that populations had a pattern of isolation by distance, probably increased by the fragmentation of the populations, as suggested by the authors (Alcaide et al., 2009a). However, the allelic distribution for the highly diverse MHC system across the distribution range (average heterozygosity = 0.98) was less uniform and many private alleles were observed, in contrast to neutral nuclear diversity, which may have been homogenised by relatively rare long–distance natal dispersal events (Prugnolle et al., 2003; Alcaide et al., 2009b). The MHC geographic pattern was more similar to that found for mtDNA, and the authors attributed the lower effects of drift observed at the MHC locus, in comparison to mtDNA, to balancing selection. They concluded that given that all lesser kestrel populations inhabit similar ecosystems their results would be explained by adaptation to local pathogen communities favouring the most locally efficient alleles although preserving a lot of genetic diversity to deal with a high array of pathogens. In contrast, the sister species, the Eurasian kestrel F. tinnunculus (that contrary to the lesser kestrel is panmictic, cosmopolitan, generalist and a long distance disperser) showed a higher level of MHC diversity in correlation with a higher burden of parasites and bacterial infections when populations of both species were compared, despite showing no higher levels of neutral microsatellite diversity in the continent (Ho = 0.66 for both species). Among the management actions taken to recover European lesser kestrel populations, there are several ongoing captive breeding and reintroduction programs. The captive and some reintroduced populations were

347

genetically monitored through the study of genetic diversity and paternity analyses using the same eight microsatellites markers from the studies above (Alcaide et al., 2010). The genetic diversity in the captive breeding populations was similar to that in the wild and rejected a low genetic diversity as the cause of low hatching rates documented in captivity (Colás et al., 2002; Pomarol et al., 2004). However, levels in reintroduced populations were slightly lower than those of their source captive populations. This may have been caused by large variances in breeding performance of the reintroduced individuals: 28% of the breeders fathered 56% of the offspring, which would negatively affect the effective population size (Ne). Neither captive population differed genetically from its source populations. Neither did reintroduced populations differ from natural populations, suggesting that the captive breeding programs are not compromised by genetic factors, even in the absence of explicit genetic management. However, genetic monitoring was prudently advised in order to maximize the diversity of any reintroduced population in the wild, especially if this population was to be geographically isolated. It would also be very useful to monitor the evolution of the breeding success of the reintroduced individuals over time, as a main determinant of Ne and thus genetic diversity and inbreeding, and this could include not only neutral markers but also fitness–related markers as genes of the MHC. In a nice example on how adaptive loci can help in conservation and management of wild populations, Rodríguez et al. (2011) applied MHC markers to infer the breeding origin of lesser kestrels in wintering quarters in sub–Saharan Africa. The genetic distances estimated confirmed the genetic connectivity of the western European populations to the wintering quarters in Senegal and rejected their connectivity to the South African populations, giving valuable information on the migration patterns of an endangered species. The Egyptian vulture Neophron percnopterus The Egyptian vulture is a long–lived, highly philopatric scavenger widely distributed around the Mediterranean Sea, the Middle East, Central/Southern Asia, the Indian sub–continent and sub–Saharan Africa (Fig. 1D). Considered as 'endangered' by the IUCN due to anthropogenic disturbances (IUCN, 2011), in the last decades the Egyptian vulture experienced a severe decline throughout its European distribution range (Tucker & Heath, 1994). Currently, the largest European population lives in the Iberian peninsula, with around 1,500 breeding pairs (Birdlife International) distributed throughout Navarra, Aragón, Cordillera Central, in the North, and Extremadura and Andalucía in the South (Perea et al., 1990), although the decrease during the 1980s reached 70% in some of these regions (IUCN, 2011). Additionally, there are two island populations in Spain, one in Menorca in the Balearic archipelago of the Mediterranean Sea, with 41 breeding pairs (De Pablo, 2000), and one in the Canary Islands in the Atlantic, where most of the population concentrates in Fuerteventura (around 30 breeding territories during the last decade, Agudo et


348

al., 2011b) and it is considered a different subspecies N. p. majorensis (Donázar et al., 2002b). Island populations have also suffered strong declines in the recent past (De Pablo, 2000; Palacios, 2000; Donázar et al., 2002b). Whereas the Iberian population is migratory (Brown and Amadon 1968), both island populations are non–migratory (De Pablo, 2000; Palacios, 2000). Under this circumstance, populations in Menorca and in the Canary Islands could be genetically differentiated from each other and from the mainland, and could be evolving in isolation, thus exacerbating potential genetic problems in such small populations. Whether these island populations are actually genetically differentiated and whether there is gene flow among them has been the object of several studies (Kretzmann et al., 2003; Agudo et al., 2010, 2011b). Kretzman et al. (2003) genotyped nine microsatellite nuclear markers and Agudo et al. (2010, 2011a) used twenty–two microsatellites in an extensive sampling from all the Spanish distribution range from both mainland and inland populations. Consistently, the population from the Canaries was found to have less genetic variability than Menorca and Iberian populations, with Menorca showing levels of diversity similar to those of the mainland, and both inland populations showing signals of a recent bottleneck (Kretzmann et al., 2003; Agudo et al., 2011b). Effective population sizes on the islands were low, with 38 individuals in Fuerteventura and 34 individuals in Menorca (Agudo et al., 2010), whereas they were higher but still low for the Iberian populations (45 individuals for the southern and 128 individuals for the northern). The analysis of population structure showed that both inland populations are genetically differentiated from the mainland populations (FST = 0.078–0.095 for Menorca and FST= 0.103–0.114 for Fuerteventura respectively), and that the level of differentiation between the two inland populations is even larger (FST = 0.144; Agudo et al., 2010). Still, some migrants from the peninsula and even from further populations have been found to reach the island populations and manage to reproduce (Agudo et al., 2010, 2011b). Nonetheless, as this level of gene flow is insufficient to prevent the primary role that drift is having over selection it might compromise the ability of the island populations to cope with new infectious diseases and thus may have important consequences for those populations in the long term (Agudo et al., 2011a). The study of MHC class II genes (Agudo et al., 2011a) in both mainland and inland populations suggests that even if drift is the main acting force and that diversity is reduced in inland and bottlenecked populations, the co–evolution of duplicate genes is counteracting the losses of genetic diversity in the MHC, thus maximizing the capacities of antigen recognition to face infections, which would be of crucial importance overall in these insular populations. Although geography did not explain any genetic difference, mainland populations also showed moderate levels of structure, indicating that the population in Andalucía differs genetically from the rest, probably as a consequence of its recent fragmentation (Agudo et al., 2010). This Iberian structure was not found by

Martinez–Cruz

Kretzman et al. (2003) but this may be caused by their lower sampling size and the low number and heterospecific origin of the microsatellites used. The authors recommended that special attention should be paid to this southern population in order to prevent losses of genetic diversity and to facilitate conservation of the Iberian populations as a whole. The black vulture Aegypius monachus The distribution area of the black vulture has been severely reduced due to anthropogenic disturbances in the last century (Hiraldo, 1974; Donázar et al., 2002a) when many breeding areas across the European continent were lost (Cramp & Simmons, 1980; Meyburg & Meyburg, 1984). Currently, European populations are restricted to the Iberian peninsula, the Balkan and the Caucasus, extending to the less known Mongolian, Pakistan and Kazakhstan populations further east (fig. 1E) (IUCN, 2011). Field observations have failed to provide any evidence of contact between the Balkan and the Iberian populations since the disappearance of intermediate populations (Poulakakis et al., 2008). Poulakakis et al. (2008) carried out a genetic study across the whole distribution range in order to evaluate the current genetic status of the vulnerable Balkan and Iberian populations. By sequencing a fragment of the mitochondrial cytochrome b gene (Cyt b) and genotyping eight nuclear microsatellites in three European and the Mongolian populations they aimed to gain insights into the current and past processes shaping the genetic diversity in the species, and also to delineate management units for conservation. While a lower mitochondrial diversity was found in both European populations than in Mongolia, no signals of genetic erosion were found at the nuclear level. The seven mitochondrial haplotypes found clustered in four deeply divergent and geographically separated lineages: A– Balkans, B–Iberia, C–Caucasus, D–Mongolia. Microsatellite analyses supported this phylogeographic distribution. The authors suggested a low level of gene flow due to the highly philopatric dispersal behaviour of the species and claim for the delimitation of different ESUs, based on reciprocal monophyly of the mtDNA clades and a high differentiation at microsatellite allele frequencies (Moritz, 1994, 1999). Nonetheless, when looking at the structure plot (figure 3 in Poulakakis et al., 2008) one can suspect the admixed ancestry of some individuals, at least in the Caucasus and the Mongolian populations, as well as the existence of one migrant in the Mongolian population. Given the low sample size of the population in the Caucasus, a better sampling could eventually reveal higher levels of migration in both directions. In that case, the delimitation as separate ESUs might not be warranted. Moreover, additional criteria should be taken into account in order to delimitate ESUs (Allendorf & Luikart, 2007). More information could be obtained through the analysis of the historical population using museum specimens, as well as a more informative set of genetic markers. In the case that the historical genetic and ecological exchangeability could not be rejected


Animal Biodiversity and Conservation 34.2 (2011)

349

(Crandall et al., 2000), the reintroduction of populations in Europe and the translocation of individuals could be envisaged as actions to increase the genetic diversity of the species.

relationship. For the impoverished Spanish populations, they suggest the central Spanish one as the source population of choice.

The red kite Milvus milvus

Conclusions and perspectives

With a geographical distribution limited to the western Palearctic (fig. 1F) the red kite is considered as near–threatened by the IUCN (IUCN, 2011). During the 20th century there was a strong decline in most of its distribution range. Some populations went extinct, as in the Canary Islands (Cramp & Simmons, 1980; Evans & Pienkowski, 1991), and the range became fragmented. This process of decline left two disjointed populations in Spain, Central and Southern Spain, and the process is still going on in southwest and eastern Europe (Viñuela et al., 1999). At present, the northwestern and central European populations are stable and recovering whereas the southern populations are not. Roques & Negro (2005) made a genetic study covering the Central European (Germany, Switzerland, Luxemburg, and France) and the Southern European (Central Spain, Southern Spain, Italy, Mallorca and Menorca) populations by sequencing 357 bp of the mitochondrial control region. For comparison, they also typed one population of its successful sister species, the black kite M. migrans from Southern Spain. They tested the hypothesis that bottlenecks in the last century could have negatively affected the diversity of the species. Indeed they found very low levels of diversity, with two mitochondrial haplotypes dominating (82% in global) distributed in almost all populations. The lowest diversities were found in Southern Spain, Menorca and Mallorca (with a single haplotype; H  =  0.000–0.529; π  =  0.0000–0.0016), populations that are peripheral and in decline, and the highest diversities were found in central Europe (where the bulk of the population is; H = 0.800; π = 0.0062) and in central Spain (H = 0.905; π = 0.0045). The exception was Italy, with high levels of diversity (H = 0.667; π = 0.0037) despite a low breeding size that may still be reflecting pre–bottleneck levels of diversity. When comparing the southern Spanish red and black kite populations, levels of diversity were around half for the former. Both species depict two phylogenetic clades that would have diverged during the last glaciations of the Pleistocene, between 129 and 650 ky ago. Although no clear structure is found among the red kite populations probably due to the high dispersion capabilities of the species, a shallow separation exists between an eastern group (Germany, Central Europe and Italy) and a western group (Central Spain, Menorca and South Spain). The authors suggested that the genetic patterns found may have been shaped by the evolutionary history of the species and some contemporary events, including northward post–glacial expansions and recent successive bottlenecks and range contractions affecting more the peripheral southern populations. For these reasons they advise the reinforcement of those populations with individuals from healthier populations showing the closest genetic

Many species of Palearctic raptors have common recent histories of decline and fragmentation due to anthropogenic perturbations. Although Iberian raptors are capable of flying long distances, in most cases studies have shown common patterns of high genetic structure, in part due to low gene flow and low genetic diversity, which is exacerbated in the case of small and isolated populations. Studies on the populations predating the most recent anthropogenic disturbances show indeed that some of these patterns were already present historically, so one can hypothesize that the decline of the species could have started long age. In other cases, however, the study of the historical population has revealed genetic erosion that may be directly attributed to the recent decline. Similar processes may have different consequences for different species, implying that different management actions are needed. Genetics can help in adopting the adequate conservation measures, but most notably so when adequate markers are used and historical and contemporary patterns and processes are conjunctly investigated. To date genetic tools have been successfully applied to conservation and a huge amount of insightful works have been performed in the very recent field of conservation genetics. However, many questions remain unresolved, in many cases due to technical limitations. The markers used have some limitations as the neutrality of mtDNA loci is not widely accepted (see Ballard & Witlock, 2004, for a review on the biology of this organelle) and microsatellite markers suffer from high homoplasy and back mutation rates (but see Estoup et al., 2002). Additionally, the use of a low number of markers has some limitations regarding accurateness in estimating parameters of interest to population genetics. Non–neutral markers have also been applied, but such studies are strongly biased towards MHC markers. Fortunately, with the recent development in genomic techniques such as next generation sequencing, whole genome scans and gene–expression pattern analyses, together with the decrease in costs of such techniques it is becoming possible to apply genomic approaches to non–model species of ecological and conservation relevance to overcome these eventualities. Such techniques allow us to tackle many of the unresolved issues in conservation genetics that could not be properly addressed before now. On one hand, the use of thousands of neutral genetic markers is increasing the accuracy in estimating population genetic parameters that are important for managing populations (e.g. kin relationships, inbreeding coefficients, effective population size, demographic bottlenecks, migration rates, etc), by identifying and eliminating loci under selection (outlier loci) that bias these estimates. On the other hand, using genomic tools might help in the identification of


350

variation with potential functional importance to study the genetic basis of local adaptation and inbreeding depression, highly relevant processes in conservation. In the last few years there have been some very interesting reviews on the high number of possibilities that genomics is opening to conservation studies (e.g. Primmer, 2009; Allendorf et al., 2010; Ouborg et al., 2010a, 2010b), and there is no doubt that genomic approaches will become routine in conservation in the very near future. Acknowledgements I would like to sincerely thank the editors for their invitation to write this review and the manuscript editor and an anonymous referee for their helpful comments. I specially wish to thank J. A. Godoy for all the very useful discussions and constructive comments on this manuscript and conservation genetics in general. References Agudo, R., Alcaide, M., Rico, C., Lemus, J. A., Blanco, G., Hiraldo, F. & Donázar, J. A., 2011a. Major histocompatibility complex variation in insular populations of the Egyptian vulture: inferences about the roles of genetic drift and selection. Mol. Ecol., 20: 2329–2340. Agudo, R., Rico, C., Hiraldo, F. & Donázar, J. A., 2011b. Evidence of connectivity between continental and differentiated insular populations in a highly mobile species. Divers. Distrib., 17: 1–12. Agudo, R., Rico, C., Vila, C., Hiraldo, F. & Donázar, J. A., 2010. The role of humans in the diversification of a threatened island raptor. BMC Evol. Biol., 10: 384. Alcaide, M., Edwards, S. V., Negro, J. J., Serrano, D. & Tella, J. L., 2008a. Extensive polymorphism and geographical variation at a positively selected MHC class IIB gene of the lesser kestrel (Falco naumanni). Mol. Ecol., 17: 2652–2665. Alcaide, M., Negro, J. J., Serrano, D., Antolin, J. L., Casado, S. & Pomarol, M., 2010. Captive breeding and reintroduction of the lesser kestrel Falco naumanni: a genetic analysis using microsatellites. Conserv. Genet., 11: 331–338. Alcaide, M., Negro, J. J., Serrano, D. & Rodriguez, A., 2008b. A genetic assessment of captive breeding and reintroduction programs of the lesser kestrel (Falco naumanni) using neutral and adaptive loci. Avian Biol. Res., 1: 40–40. Alcaide, M., Serrano, D., Negro, J. J., Tella, J. L., Laaksonen, T., Muller, C., Gal, A. & Korpimaki, E., 2009a. Population fragmentation leads to isolation by distance but not genetic impoverishment in the philopatric Lesser Kestrel: a comparison with the widespread and sympatric Eurasian Kestrel. Heredity, 102: 190–198. Alcaide, M., Serrano, D., Tella, J. L. & Negro, J. J., 2009b. Strong philopatry derived from capture–recapture records does not lead to fine–scale

Martinez–Cruz

genetic differentiation in lesser kestrels. J. Anim. Ecol., 78: 468–475. Allendorf, F. W., Hohenlohe, P. A. & Luikart, G., 2010. Genomics and the future of conservation genetics. Nat. Rev., 11: 697–709. Allendorf, F. W. & Luikart, G., 2007. Conservation and the genetics of populations. Blackwell Publishing, Malden, USA. Armbruster, P. & Reed, D. H., 2005. Inbreeding depression in benign and stressful environments. Heredity, 95: 235–242. Ballard, J. W. & Whitlock, M. C., 2004. The incomplete natural history of mitochondria. Mol. Ecol., 13: 729–744. Ballou, J. D. & Foose, T. J., 1996. Demographic and genetic management of captive populations. In: Wild mammals in captivity, principles and techniques: The University of Chicago Press,Chicago, IL. Barrett, R. D. H. & Schluter, D., 2008. Adaptation from standing genetic variation. Trends Ecol. Evol., 23: 38–44. Biber, J. P., 1990. Action plan for the conservation of western lesser kestrel (Falco naumanni) populations International Council for Bird Preservation, Cambridge, UK. BirdLife International, 2004. Threatened Birds of the World 2004. CD–ROM. BirdLife International, Cambridge, UK. – 2011. Falco naumanni. In: IUCN 2011. IUCN Red List of Threatened Species. Version 2011.2. <www. iucnredlist.org>. Downloaded on 10 January 2012. Brown, L. H. & Amadon, D., 1968. Eagles, hawks and falcons of the world. 2 vols. McGraw–Hill, New York. Colás, J., Corroto, M., Garcia Brea, A., Gough, R. & Jimenez Gomez, P., 2002. Dramatic infertility and embryo mortality in a lesser kestrel (Falco naumanni) captive breeding program in Spain. J. Wildlife Dis., 38: 1. Collar, N. J. & Andrew, P., 1988. Birds to watch: The ICBP world checklist of threatened birds. Smithsonian Institution Press, Washington, D. C. and Cambridge, England. Cramp, S. & Simmons, K. E. L., 1980. The Birds of the Western Palearctic. Oxford University Press, Oxford. Crandall, K. A., Bininda–Emonds, O. R. P., Mace, G. M. & Wayne, R. K., 2000. Considering evolutionary processes in conservation biology. Trends Ecol. Evol., 15: 290–295. Crnokrak, P. & Roff, D. A., 1999. Inbreeding depression in the wild. Heredity, 83: 260–270. De Pablo, F., 2000. Estatus del alimoche (Neophron percnopterus) en Menorca (Islas Baleares). Anuari Ornitològic de les Balears, 15: 3–9. Donázar, J. A., Blanco, G., Hiraldo, F., Soto–Largo, E. & Oria, J., 2002a. Effects of forestry and other land–use practices on the conservation of cinereous vultures. Ecol. Appl., 12: 1445–1456. Donázar, J. A., Negro, J. J., Palacios, C. J., Gangoso, L., Godoy, J. A., Ceballos, O., Hiraldo, F. N. & Capote, N., 2002b. Description of a new subspecies of the Egyptian Vulture (Accipitridae: Neophron percnopterus) from the Canary Islands. J. Raptor


Animal Biodiversity and Conservation 34.2 (2011)

Res., 36: 17–23. Estoup, A., Jarne, P. & Cornuet, J.–M., 2002. Homoplasy and mutation model at microsatellite loci and their consequences for population genetic analysis. Mol. Ecol., 11: 1591–1604. Evans, I. & Pienkowski, M., 1991. A background to the experimental reintroduction to England and Scotland. British Birds, 84: 171–187. Ferrer, M., 2001. The Spanish Imperial Eagle. Lynx Edicions, Barcelona. Ferrer, M. & Hiraldo, F., 1992. Man–Induced Sex–Biased Mortality in the Spanish Imperial Eagle. Biol. Conserv., 60: 57–60. Ferrer, M. & Negro, J. J., 2004. The near extinction of two large European predators: Super specialists pay a price. Conserv. Biol., 18: 344–349. Frankham, R., 1996. Relationship of genetic variation to population size in wildlife. Conserv. Biol., 10: 1500–1508. Frankham, R., Ballou, J. D. & Briscoe, D. A., 2002. Introduction to conservation genetics. Cambridge University Press, Cambridge. Frey, H., Knotzinger, O. & Llopis, A., 1995. Bearded vulture reintroduction into the Alps, annual report. Foundation for the Conservation of the Bearded Vulture, Wein. Gage, M. J. G., Surridge, A. K., Tomkins, J. L., Green, E., Wiskin, L., Bell, D. J. & Hewitt, G. M., 2006. Reduced heterozygosity depresses sperm quality in wild rabbits, Oryctolagus cuniculus. Curr. Biol., 16: 612–617. Gautschi, B., 2001. Conservation genetics of the Bearded vulture. PhD Thesis, Universität Zürich. Gautschi, B., Jacob, G., Negro, J. J., Godoy, J. A., Muller, J. P. & Schmid, B., 2003a. Analysis of relatedness and determination of the source of founders in the captive bearded vulture, Gypaetus barbatus, population. Conserv. Genet., 4: 479–490. Gautschi, B., Muller, J. P., Schmid, B. & Shykoff, J. A., 2003b. Effective number of breeders and maintenance of genetic diversity in the captive bearded vulture population. Heredity, 91: 9–16. Godino, A., Paz, J. L. & Simón, M. A., 2003. Naturalistas españoles localizan en Marruecos 5 quebrantahuesos. Quercus, 205: 46–47. – 2004. Localizan más quebrantahuesos adultos y jóvenes en el Alto Atlas. Quercus, 127: 12–13. Godoy, J. A., Negro, J. J., Hiraldo, F. & Donázar, J. A., 2004. Phylogeography, genetic structure and diversity in the endangered bearded vulture (Gypaetus barbatus, L.) as revealed by mitochondrial DNA. Mol. Ecol., 13: 371–390. Gonzalez, L. M., Hiraldo, F., Delibes, M. & Calderon, J., 1989. Reduction in the range of the Spanish imperial eagle (Aquila adalberti Brehm, 1861) since Ad 1850. J. Biogeogr., 16: 305–315. Heredia, R. & Heredia, B., 1991. El quebrantahuesos (Gypaetus barbatus) en los Pirineos. Ministerio de Agricultura, Pesca y Alimentacion, Madrid. Hiraldo, F., 1974. Colonias de cría y censo de los buitres negros (Aegypius monachus) en España. Naturalia Hispanica, 2: 1–31. Hiraldo, F., Delibes, M. & Calderon, J., 1979. El Que-

351

brantahuesos Gypaetus barbatus L. Monografia No 22. ICONA, Madrid. – 1984. Comments on the taxonomy of the bearded vulture Gypaetus barbatus (Linnaeus, 1758). Bonner Zoologische Beitrage, 35: 91–95. IUCN 2011. IUCN Red List of Threatened Species. Version 2011.1. <www.iucnredlist.org>. Downloaded on 10 June 2011. Johnson, W. E., Onorato, D. P., Roelke, M. E., Land, E. D., Cunningham, M., Belden, R. C., McBride, R., Jansen, D., Lotz, M., Shindle, D., Howard, J., Wildt, D. E., Penfold, L. M., Hostetler, J. A., Oli, M. K. & O’Brien, S. J., 2010. Genetic Restoration of the Florida Panther. Science, 329: 1641–1645. Keller, L. F. & Waller, D. M., 2002. Inbreeding effects in wild populations. Trends Ecol. Evol., 17: 230–241. Kleiman, D. G. & Stanley, P. M. R., 1994. Criteria for reintroductions. In: Creative Conservation–Interactive management of wild and captive animals: 287–303 Chapman & Hall, London. Kretzmann, M. B., Capote, N., Gautschi, B., Godoy, J. A., Donázar, J. A. & Negro, J. J., 2003. Genetically distinct island populations of the Egyptian vulture (Neophron percnopterus). Conserv. Genet., 4: 697–706. Lande, R. & Shannon, S., 1996. The role of genetic variation in adaptation and population persistence in a changing environment. Evolution, 50: 434–437. Lopez–Lopez, P., Ferrer, M., Madero, A., Casado, E. & McGrady, M., 2011. Solving man–induced large– scale conservation problems: the Spanish Imperial eagle and power lines. Plos One, 6(3): e17196. Lynch, M. & Walsh, J. B., 1998. Genetics and Analysis of Quantitative Traits. Sinauer Assocs., Inc., Sunderland, MA. Martinez–Cruz, B. & Godoy, J. A., 2007. Genetic evidence for a recent divergence and subsequent gene flow between Spanish and Eastern imperial eagles. BMC Evol. Biol., 7: 170. Martinez–Cruz, B., Godoy, J. A. & Negro, J. J., 2004. Population genetics after fragmentation: the case of the endangered Spanish imperial eagle (Aquila adalberti). Mol. Ecol., 13: 2243–2255. – 2007. Population fragmentation leads to spatial and temporal genetic structure in the endangered Spanish imperial eagle. Mol. Ecol., 16: 477–486. Meyburg, B. U. & Meyburg, C., 1984. Distribution et statut actuels du vautour moine Aegypius monachus. Rapinyaires Mediterranis, 2: 26–31. Milot, E., Weimerskirch, H., Duchesne, P. & Bernatchez, L., 2007. Surviving with low genetic diversity: the case of albatrosses. Proc. R. Soc. B–Biol. Sci., 274: 779–787. Moritz, C., 1994. Defining Evolutionarily–Significant– Units for Conservation. Trends Ecol. Evol., 9: 373–375. – 1999. Conservation units and translocations: strategies for conserving evolutionary processes. Hereditas, 130: 217–228. – 2002. Strategies to protect biological diversity and the evolutionary processes that sustain it. Syst. Biol., 51: 238–254. Negro, J. J., Hiraldo, F. & Donázar, J. A., 1997. Causes


352

of natal dispersal in the lesser kestrel: inbreeding avoidance or resource competition? J. Anim. Ecol., 66: 640–648. Negro, J. J. & Torres, M. J., 1999. Genetic variability and differentiation of two bearded vulture Gypaetus barbatus populations and implications for reintroduction projects. Biol. Conserv., 87: 249–254. O’Brien, S. J. & Evermann, J. F., 1988. Interactive influence of infectious–disease and genetic diversity in natural–populations. Trends Ecol. Evol., 3: 254–259. O’Grady, J. J., Brook, B. W., Reed, D. H., Ballou, J. D., Tonkyn, D. W. & Frankham, R., 2006. Realistic levels of inbreeding depression strongly affect extinction risk in wild populations. Biol. Conserv., 133: 42–51. Ouborg, N. J., Pertorldi C., Loeschcke, V., Bijlsma, R. (Kuke) & Hedrick, P. W., 2010a. Conservation genetics in transition to conservation genomics. Trends Ecol. Evol., 26: 177–187. Ouborg, N.J., Angeloni, F. & Vergeer, P., 2010b. An essay on the necessity and feasibility of conservation genomics. Conserv. Genet., 11: 643–653. Palacios, C. J., 2000. Decline of the Egyptian vulture (Neophron percnopterus) in the Canary Islands. J. Raptor Res., 34: 61. Palsboll, P. J., Berube, M. & Allendorf, F. W., 2007. Identification of management units using population genetic data. Trends Ecol. Evol., 22: 11–16. Perea, J. L., Morales, M. & Velasco, J., 1990. Programa de seguimiento de las poblaciones de alimoche (Neophron percnopterus) en España y primera encuesta sobre su estado de conservación. Evora, Portugal. Pomarol, M., Carbonell, F., Heredia, G., Valbuena, E., Alonso, M. & Serrano, D., 2004. Cría en cautividad y reintroducción. In: El cernícalo primilla en Andalucía. Bases ecológicas para su conservación: Consejeria de Medio Ambiente, Junta de Andalucia, Sevilla. Poulakakis, N., Antoniou, A., Mantziou, G., Parmakelis, A., Skartsi, T., Vasilakis, D., Elorriaga, J., De La Puente, J., Gavashelishvili, A., Ghasabyan, M., Katzner, T., McGrady, M., Batbayar, N., Fuller, M. & Natsagdorj, T., 2008. Population structure, diversity, and phylogeography in the near–threatened Eurasian black vultures Aegypius monachus (Falconiformes, Accipitridae) in Europe: insights from microsatellite and mitochondrial DNA variation. Biol. J. Linnean Soc., 95: 859–872. Primmer, C. R., 2009. From Conservation Genetics to Conservation Genomics. Annals of the New York Academy of Sciences, 1162: 357–368. Prugnolle, F., Pilard, P., Brun, L. & Tavecchia, G., 2003. First–year and adult survival of the endangered Lesser Kestrel Falco naumanni in southern France. Bird Study, 50: 68–72. Queller, D. C. & Goodnight, K. F., 1989. Estimating relatedness using genetic–markers. Evolution, 43: 258–275. Ralls, K. & Ballou, J. D., 2004. Genetic status and management of California Condors. Condor, 106: 215–228.

Martinez–Cruz

Ralls, K., Ballou, J. D., Rideout, B. A. & Frankham, R., 2000. Genetic management of chondrodystrophy in California condors. Anim. Conserv., 3: 145–153. Reed, D. H. & Frankham, R., 2003. Correlation between fitness and genetic diversity. Conserv. Biol., 17: 230–237. Reed, D. H., Lowe, E. H., Briscoe, D. A. & Frankham, R., 2003. Fitness and adaptation in a novel environment: Effect of inbreeding, prior environment, and lineage. Evolution, 57: 1822–1828. Reed, D. H., Nicholas, A. C. & Stratton, G. E., 2007. Inbreeding levels and prey abundance interact to determine fecundity in natural populations of two species of wolf spider. Conserv. Genet., 8: 1061–1071. Rodriguez, A., Alcaide, M., Negro, J. J. & Pilard, P., 2011. Using major histocompatibility complex markers to assign the geographic origin of migratory birds: examples from the threatened lesser kestrel. Anim. Cons., 14: 306–313. Roelke, M. E., Martenson, J. S. & Obrien, S. J., 1993. The consequences of demographic reduction and genetic depletion in the endangered Florida panther. Curr. Biol., 3: 340–350. Roques, S. & Negro, J. J., 2005. MtDNA genetic diversity and population history of a dwindling raptorial bird, the red kite (Milvus milvus). Biol. Conserv., 126: 41–50. Russello, M. A. & Amato, G., 2004. Ex situ population management in the absence of pedigree information. Mol. Ecol., 13: 2829–2840. Ryder, O. A., 1986. Species conservation and systematics–the dilemma of subspecies. Trends Ecol. Evol., 1: 9–10. Saccheri, I., Kuussaari, M., Kankare, M., Vikman, P., Fortelius, W. & Hanski, I., 1998. Inbreeding and extinction in a butterfly metapopulation. Nature, 392: 491–494. Schaub, M., Zink, R., Beissmann, H., Sarrazin, F. & Arlettaz, R., 2009. When to end releases in reintroduction programmes: demographic rates and population viability analysis of bearded vultures in the Alps. J. Appl. Ecol., 46: 92–100. Serrano, D. & Tella, J. L., 2003. Dispersal within a spatially structured population of lesser kestrels: the role of spatial isolation and conspecific attraction. J. Anim. Ecol., 72: 400–410. Serrano, D., Tella, J. L., Donázar, J. A. & Pomarol, M., 2003. Social and individual features affecting natal dispersal in the colonial Lesser Kestrel. Ecology, 84: 3044–3054. Serrano, D., Tella, J. L., Forero, M. G. & Donázar, J. A., 2001. Factors affecting breeding dispersal in the facultatively colonial lesser kestrel: individual experience vs. conspecific cues. J. Anim. Ecol., 70: 568–578. Simón, M. A., Couto, S., Lucio Carrasco, A., García– Baquero, M. J., Godino, A., Gutiérrez, J. E., Hernández, F. J., Jiménez, E., Liñán, M., López, M., Padial, J. M., Salamanca, J. C., Bautista, F., del Barco, M., Cabrera, L. & Macías, E., 2007. The reintroduction of the Bearded vulture Gypaetus barbatus in Andalusia, southern Spain. Vulture


Animal Biodiversity and Conservation 34.2 (2011)

News, 56: 29–40. Tanaka, Y., 2000. Extinction of populations by inbreeding depression under stochastic environments. Popul. Ecol., 42: 55–62. Thibault, J. C., Vigne, J. D. & Torre, J., 1992. The diet of young Lammergeiers Gypaetus barbatus in Corsica: its dependence on extensive grazing. Ibis, 135: 203–214. Tucker, G. M. & Heath, M. F., 1994. Birds in Europe: Their Conservation Status. BirdLife International, Cambridge, UK. Valsecchi, E., Amos, W., Raga, J. A., Podesta, M. & Sherwin, W., 2004. The effects of inbreeding on mortality during a morbillivirus outbreak in the Mediterranean striped dolphin (Stenella coeruleoalba). Anim. Conserv., 7: 139–146. Vila, C., Sundqvist, A. K., Flagstad, O., Seddon, J., Bjornerfeldt, S., Kojola, I., Casulli, A., Sand, H., Wabakken, P. & Ellegren, H., 2003. Rescue of a severely bottlenecked wolf (Canis lupus) population by a single immigrant. Proc. R. Soc. B–Biol. Sci., 270: 91–97. Vilas, C., San Miguel, E., Amaro, R. & Garcia, C., 2006. Relative contribution of inbreeding depres-

353

sion and eroded adaptive diversify to extinction risk in small populations of shore Campion. Conserv. Biol., 20: 229–238. Villafuerte, R., Calvete, C., Blanco, J. C. & Lucientes, J., 1995. Incidence of viral hemorrhagic disease in wild rabbit populations in Spain. Mammalia, 59: 651–659. Viñuela, J., Marti, R. & Ruiz, A., 1999. El Milano Real en España. Monografía No. 6, SEO/BirdLife, Madrid. Waples, R. S. & Gaggiotti, O., 2006. What is a population? An empirical evaluation of some genetic methods for identifying the number of gene pools and their degree of connectivity. Mol. Ecol., 15: 1419–1439. Xirouchakis, S. & Nikolakakis, M., 2002. Conservation implications of the temporal and spatial distribution of bearded vulture Gypaetus barbatus in Crete. Bird Conservation International, 12: 269–280. Zink, R., 2009. International Bearded vulture monitoring annual report. Hohe Tauern National Park/Owl and Raptor Centre Haringsee c/o Research Institute of Wildlife Ecology, University of Veterinary Medicine, Vienna.


366


Animal Biodiversity and Conservation 34.2 (2011)

355

The diet of Great Tit Parus major nestlings in a Mediterranean Iberian forest: the important role of spiders E. Pagani–Núñez, Í. Ruiz, J. Quesada, J. J. Negro & J. C. Senar

Pagani–Núñez, E., Ruiz, Í, Quesada, J., Negro, J. J. & Senar, J. C., 2011. The diet of Great Tit Parus major

nestlings in a Mediterranean Iberian forest: the important role of spiders. Animal Biodiversity and Conservation, 34.2: 355–361. Abstract The diet of Great Tit Parus major nestlings in a Mediterranean Iberian forest: the important role of spiders.— The diet of the Great Tit Parus major when rearing chicks has been described in many studies. However, data from the Mediterranean area is scarce. Here we describe the diet of nestlings in a population of Great Tits in a Mediterranean forest in Barcelona (north–east Spain) during two breeding seasons using two methods: neck–collars and video recording. The main prey were caterpillars (44% from neck–collar data and 62% from video–recorded data), but in our latitudes spiders also seemed to be an important food resource (24% from neck–collar data and 42% from video–recorded data). We did not find any significant differences in the quantity of spiders collected by parents in relation to stage of chick development, main vegetation surrounding nest boxes, size of the brood, or year. Our results stress the importance of spiders as a food source in Mediterranean habitats. Key words: Great Tits, Nestlings, Diet, Spiders. Resumen Dieta de los pollos de Carbonero Común Parus major en un bosque Ibérico Mediterraneo: la importáncia de las arañas.— La dieta del Carbonero Común Parus major cuando alimenta a los pollos ha sido descrita en muchos artículos. Sin embargo, la información sobre el área mediterránea es bastante escasa. Aquí describimos la dieta de los pollos en una población de Carbonero Común en un bosque Mediterráneo de Barcelona (nordeste de España) en dos temporadas de cría a través de dos métodos (collares y grabaciones de video). Las principales presas cebadas fueron las orugas (44% a partir de datos de collares y 62% a partir de datos de grabaciones), pero en nuestras latitudes las arañas parecen ser un importante recurso (24% [datos collares] y 42% [datos de grabaciones]). No encontramos diferencias significativas en la cantidad de arañas recolectadas por los padres en relación con el estado de desarrollo de los pollos, vegetación alrededor de las cajas nido, tamaño de puesta y año. Nuestros resultados subrayan la importancia de las arañas como recurso trófico en ambientes mediterráneos. Palabras clave: Carbonero Común, Pollos, Dieta, Arañas. (Received: 10 XI 11; Final acceptance: 28 XII 11) Emilio Pagani–Núñez, Íker Ruiz, Javier Quesada & Juan Carlos Senar, Evolutionary Ecology Associate Research Unit (CSIC), Natural History Museum of Barcelona, Psg. Picasso s/n., 08003 Barcelona, Espanya (Spain).– Juan José Negro, Estación Biológica de Doñana (CSIC), Avda. Americo Vespucio s/n., 41092 Sevilla, España (Spain).

ISSN: 1578–665X

© 2011 Museu de Ciències Naturals de Barcelona


Pagani–Núñez et al.

356

Introduction The Great Tit Parus major has often been used as a model species in studies of feeding behaviour and diet (Royama, 1970; Wansink & Tinbergen,1994; Biard et al., 2006; Senar et al., 2010). Although some variation appears between areas and periods, the species has been reported to focus highly on caterpillars when rearing chicks (Gibb & Betts, 1963; Balen, 1973; Gosler, 1993; Wilkin et al., 2009). Nevertheless, the species has also been recorded to provide other animal content such as butterflies, Phasmidae or Orthoptera. Interestingly, high proportions of spiders have been reported early in the chicks’ diet, with a peak around 5–6 days of age (see Ramsay & Houston [2003] for references). These results are independent of date and habitat, which stresses that Tits specifically select spiders in their diet during that period. This preference for arachnids has also been recorded for other similar species such as Blue Tits Cyanistes caeruleus (Banbura et al., 1999; Arnold et al., 2007; Arnold et al., 2011; Garcia–Navas & Sanz, 2011), Pied Flycatchers Ficedula hipoleuca (Sanz, 1998) and Red–Breasted Flycatchers Ficedula parva (Mitrus et al., 2010). Earlier work found that the proportion of spiders increased markedly in Mediterranean habitats, with this food resource being used right throughout the breeding period. Blondel et al. (1991) reported that although the percentage of spiders in Blue Tit diets was about 16% in the mainland, it rose to 26% in the islands. Naef–Daenzer et al. (2000) found that the percentage of spiders in very earlier stages of Great Tit chick development could rise to 75%, although it rapidly decreased again to 5%. The aim of this work was to analyse in detail the relative contribution of spiders to the diet of Great Tit nestlings in a typical Mediterranean forest. We analysed the diet in relation to spider availability and Great Tit breeding phenology. Specifically, we aimed 1) to assess whether Great Tits actively selected prey to feed their brood; 2) to determine if there were any differences in the quantity of spiders provided to nestlings by parents in relation to the age of the nestlings, brood size and forest structure; and 3) to determine if there were any differences in the quantity of spiders provided to nestlings according to the sex of the parents. Material and methods Great Tits were studied over two breeding seasons in 2001 and 2004, in a mixed forest dominated by oaks and pines in the field station of Can Catà, within the Park of Collserola (Cerdanyola, Barcelona, NE of the Iberian Peninsula, latitude 45º 27' N, length 2º 8' E). Nest boxes were distributed throughout the whole area (80 ha). They were located on the trunks of oaks, at an approximate height of 1.30 m. Birds entered the nest boxes through a cylindrical tube of 10 cm in length and 5 cm in diameter designed to protect the box from predators.

The study area was highly varied. Altitude ranged from 80 to 225 m a.s.l. At the bottom of the valley vegetation coverage was dominated by Holm Oaks (Quercus ilex) and Oaks (Quercus cerrioides) and had a highly developed Mediterranean understory. On the slopes the Aleppo Pine (Pinus halepensis) was the predominant tree species. We considered the nest boxes in accordance with the main vegetation surrounding them. We considered two zones, the valley, where Quercus trees represented > 70% of the arboreal vegetation, and the slopes, where pines showed higher presence the higher the altitude and Quercus trees represented < 70% of the arboreal vegetation. Recording diet: filming The diet provided by Great Tit parents to their chicks was studied in spring 2001 by filming the parents’ entries and exits from the nest box. Filming was undertaken using a domestic video camera that was camouflaged by means of a net of cryptic colours and vegetation at five meters‘distance from the box. The tapes had an hour of duration. Three days before filming we placed a tripod on the ground in the exact position where we later filmed to get the birds used to the setting. The height of the camera with the tripod did not surpass 50 cm above ground level. Wire netting was placed over the entrance of the nest–box to make it more difficult for the birds to enter the box. This delayed their entry, allowing a clear view of each prey item (Currie et al., 1996; Atienzar et al., 2009). The nest boxes were checked twice each week to gather breeding data, including laying date, clutch size, hatching date, numbers of nestlings and fledglings. We eliminated tapes that recorded fewer than five visits by the male, since we considered they could be biased in some way. This left us with a sample size of 25 nest boxes. Tape recordings allowed us to determine the sex of the parents and the exact time of each feeding. The method provides a photographic record of the prey items for later identification and is not biased by the size of the prey. Prey are sometimes difficult to identify, however (Barba & Monrós, 1999). Neck collars Diet was analysed in spring 2004 using neck collars (Barba & Gil–Delgado, 1990). Neck collars allow simultaneous gathering of samples from several nests. Collars were made from a wire cable, and the loop was carefully laid around the neck of the chick, allowing it to breathe unhindered, but unable to swallow food (Poulsen & Aebischer, 1995). Collars remained fitted for two hours. We sampled a total of 37 nest boxes. The food was carefully extracted from the mouth and oesophagus of the chick, and the neck collar was removed. The food was suitably stored in individual vials and the number on the box and date were noted. The mass of each sample was later measured in the lab. We recorded the diet of chicks using this method when they were


Animal Biodiversity and Conservation 34.2 (2011)

357

Table 1. Arthropod abundance in leaves of trees in the study area. We surveyed fifteen trees of each the tree main species (Holm Oaks, Quercus ilex; Oaks, Quercus cerrioides; and Aleppo Pine, Pinus halepensis) on the slopes and at the bottom of the valley. Census made during springs in 2002 and in 2003. Table 1. Abundancia de artrópodos en las hojas de los árboles de la zona de estudio. Se testaron quince árboles de cada una de las tres espécies principales (encinas, Quercus ilex; robles, Quercus cerrioides y pino carrasco, Pinus halepensis) en las laderas y los fondos del valle. El censo se realizó durante las primaveras de 2002 y 2003. Prey Lepidoptera larvae Symphita larvae

N

N / tree

St. Dev

% N

V (mm3)

% Mass

156

0.06

0.29

6.35

600

42.90

89

0.03

0.31

3.62

450

18.36

168

0.07

0.84

6.84

150

11.55

Orthoptera

71

0.03

0.2

2.89

300

9.76

Coleoptera

448

0.18

0.55

18.23

20

4.11

Lepidoptera adults

Arachnida

397

0.16

0.53

16.16

20

3.64

Chrysalis

10

0

0.07

0.41

750

3.44

Tipulae

53

0.02

0.17

2.16

120

2.92

119

0.05

0.28

4.84

20

1.09

Homoptera Phasmida

3

0

0.03

0.12

600

0.83

Neuroptera

25

0.01

0.1

1.02

50

0.57

Formicidae

481

0.19

1.42

19.58

2

0.44

Mantidae

1

0

0.02

0.04

Heteroptera

1

0

0.02

0.04

Myriapoda

1

0

0.02

0.04

Gasteropoda

1

0

0.02

0.04

both five and twelve days old, to analyse changes in the feeding behaviour in relation to the age of the nestlings. Using this method it is not possible to determine which of the parents deliver every item (Barba & Monrós, 1999). We used both video filming and neck collar data to characterize the nestling diet and to determine which variables could affect the quantity of spiders provided to nestlings by Great Tit parents.

(2, 20, 120, 150, 300, 450, 600 and 750 mm3). The mass equivalence was obtained by multiplying volume by abundance. This allowed us to evaluate the abundance of the different groups of arthropods present in the forest and to assess whether there was any active selection of prey by the parents.

Phenology of prey

Arthropod census on trees

During the breeding seasons of the years 2002 and 2003 we made a census of arthropods present in the leaves of the three main tree species at Can Catà: Holm Oaks (Quercus ilex), Oaks (Quercus cerrioides), and Aleppo Pine (Pinus halepensis). We randomly selected fifteen trees of each species for each year, ten located on the slopes and five at the bottom of the valley; this allowed us to take altitudinal variations into account. Census lasted four minutes per tree. We recorded all insects seen on external branches and leaves (Carrascal et al., 1998). Arthropod volume was estimated and classified into one of eight sizes

Census of arthropods on tree leaves showed that the groups most frequently recorded were Formicidae (20%), Coleoptera (18%), and Arachnidae (16%). Lepidoptera represented 13% of the records (larvae 6%, adults 7%). When we considered the volume (mm3) as a proxy for the biomass of the prey delivered, however, their relative importance changed: Lepidoptera became the main prey available (larvae 43%, adults 12%), followed by Symphyta (larvae 18%, adults were not found) and Orthoptera (10%); many other taxa were recorded but their frequencies were very low (table 1).

Results


Pagani–Núñez et al.

358

70%

Collected prey

60%

Filming Collars

50% 40% 30% 20% 10% 0

Caterpillars

Spiders

Others

Fig. 1. Percentage of prey collected by Great Tit parents, according to the recording method: video (2001) or neck collars (2004). Fig. 1. Porcentaje de presas recolectadas por los padres de Carbonero Común, de acuerdo con el método de grabación: vídeo (2001) o collares (2004).

Prey brought by parents to nestlings Great Tits brought a variety of insects to the nest: caterpillars (both Lepidoptera and Symphyta), spiders (including also eggs), butterflies, Phasmidae, Orthoptera and formless remains. Caterpillars included Symphyta such as Diprion sp.; Noctuidae such as Orthosia sp., Catocala sp. and Spodoptera sp.; Geometridae such as Lycia sp. and Idaea sp.; hairy caterpillars such as Lymantria sp. and Orgyia sp. Spiders included mainly Zoropsis sp., Olios sp., Gibbaranea sp., Scotophaeus sp., Chiracantium sp., Philodromus sp., Synema sp. and Thomisius sp. We grouped prey in three groups: caterpillars, spiders and 'others'. Data obtained from filming nests showed that caterpillars were the main prey provided to nestlings (44%), followed by spiders (42%) (fig. 1). Frequencies did not correspond to availability, since caterpillars and spiders were consumed more than expected when compared to other insects (x2 = 360.55; df = 2; P  <  0.005). When we compared consumption to availability only for caterpillars and spiders we found that caterpillars were consumed more than expected (x2 = 13.73; df = 1; P [ 0.005). Data obtained from neck collars showed caterpillars and butterflies were the main (62%) groups of prey brought by the parents to the nestlings. In the second place we found spiders (24%). The remaining arthropods were present at very low frequencies (fig.  1). Again, frequencies did not correspond to availability, since caterpillars and spiders were also more frequently consumed than expected (x2 = 375.55; df = 2;

Table 2. MANOVA of quantity of spiders provided by parents to their nestlings according to their age (5 and 12 days old). We included also as factors the size of the brood and the percentage of Quercus surrounding (25 m) nest boxes (measured in a qualitative scale: > 70% and < 70% of Quercus). Tabla 2. MANOVA de la cantidad de arañas proporcionada por los padres a sus crías en función de su edad (5 y 12 días). Se incluyeron también como factores el tamaño de puesta y el porcentaje de Quercus cercanos a las cajas nido (25 m) (medido en una escala cualitativa: > 70% y < 70% de Quercus).

F1,16

P

% Quercus

0.3

0.57

Brood size

0.5

0.48

Age of the nestlings

3.0

0.10

% Quercus x brood size

0.4

0.52

Age nestlings x % Quercus

1.0

0.33

Age nestlings x brood size

1.8

0.20

Age nestlings x brood size x x % Quercus 0.4 0.51


Animal Biodiversity and Conservation 34.2 (2011)

80% 70% % spiders

P < 0.005). When we considered only caterpillars and spiders, we found that caterpillars were ingested more often than expected according to availability (x2  = 63.64; df = 1; P < 0.005). Collar data showed that the quantity of spiders provided to nestlings by Great Tit parents did not vary with the age of the nestlings, brood size or forest structure, measured as a percentage of Quercus around the nest (table 2). Video data showed that females provided young with more spiders than males (fig. 2), and that males provided more caterpillars than females (tables 3, 4). However, in absolute terms, males provided more prey items than females, including spiders (table 3). Repeated measures ANOVA within each pair stressed that the males collected more caterpillars whereas females tended to deliver more spiders (table 4). When comparing abundance and selection of caterpillars and spiders in function of sex, we found that males seemed to actively select caterpillars (x2  =  15.18; df  = 1; P < 0.005), but females did not (x2  = 0.08; df = 1; P = 0.77). Analyses of pooled data from both recording methods showed no relationship between the quantity of spiders fed by the parents and the factors 'method/year', 'size of brood' or 'habitat structure' measured as percentage of Quercus around the nest (table 5).

359

60% 50% 40% 30% 20% Males

Females

Fig. 2. Mean percentage and 95% confidence intervals of spiders fed to young by males and females during the 2001 breeding season, according to video data (see table 4). Fig. 2. Porcentaje medio e intérvalos de confianza del 95% de las arañas que alimentan a los jóvenes recolectadas por los machos y hembras durante la temporada de cría de 2001, según grabaciones de vídeo (ver tabla 4).

Discussion

Table 3. Percentage and absolute values of different preys collected by Great Tit males (n = 25) and females (n = 22) during the 2001 breeding season (filming data). Tabla 3. Porcentaje y valores absolutos de las distintas presas recolectadas por machos (n = 25) y hembras (n = 22) de Carbonero Común durante la temporada de cría 2001 (grabaciones de video). Percentage

N

Caterpillars

48.96%

94

Spiders

40.63%

78

Others

10.42%

20

100%

192

Males

Sum

Females Caterpillars

35,56%

32

Spiders

55.56%

50

Others

8.89%

8

Sum

100%

90

Caterpillars are generally the main food resource used by Tits to feed their chicks (Sillanpaa et al., 2008; Wilkin et al., 2009). Nevertheless, several studies have pointed out the important role of spiders as a key resource during the early stages of chick development (Ramsay & Houston, 2003), when spiders may

Table 4. RM ANOVA comparing males and females within each nest–box in relation to the number of caterpillars and spiders collected. Data recorded during the 2001 breeding season (filming data). Tabla 4. RM ANOVA comparando machos y hembras dentro de cada caja nido en relación con el número de orugas y arañas recolectados. Datos registrados durante la temporada de cría 2001 (grabaciones de vídeo). Caterpillars Spiders

F1,17

P

11.1

< 0.001

3.5

0.08


Pagani–Núñez et al.

360

Table 5. ANOVA on the variation in the abundance of spiders collected by Great Tit parents in relation to the recording method (collars and filming data), percentage of Quercus surrounding nest boxes (categorical variable with two classes: > 70% and < 70% of Quercus) and size of the brood. Tabla 5. ANOVA de la variación en la abundancia de las arañas recolectadas por los padres de Carbonero Común en relación con el método de grabación (collares y grabaciones de video), porcentaje de Quercus que rodean los nidos (variable categórica con dos clases: > 70% y < 70% de Quercus) y tamaño de puesta.

F1,57

P

Method

9.0

0.20

Brood size

3.1

0.33

% Quercus

09

0.52

Method x brood size

3.4

0.32

Method x % Quercus

1.5

0.43

Brood size x % Quercus

0.6

0.59

Method x nº brood x x % Quercus 0.2 0.69

constitute 25% of the nestlings’ diet (Tinbergen, 1960; Royama, 1970; Balen, 1973; Cowie & Hinsley, 1988, Woodburn, 1997). Our data stress that the overall contribution of spiders to the diet of the nestling Tits is even higher in the Mediterranean ecosystems, where figures can rise to 25 or even 40%, independently of the age of the chicks (see also Blondel et al., 1991; Naef–Daenzer, 2000). It could be argued that the higher use of spiders as a food source in this area is a collateral result of the generalized lack of food and the harder conditions of the Mediterranean forests (Royama, 1970; Blondel et al., 1991). However, the high proportion of spiders fed to the nestlings in our study area, independently of location and habitat structure, suggests that spiders may also be selected as a main food source. Additionally, if we consider the volume of prey rather than frequency, the availability of caterpillar increases greatly, implying that spiders are a more favoured food resource. Further, more detailed data on the size of the prey brought to nestlings is needed, however, to confirm this consideration. Variations found over the two years of the study may be due to yearly variations or, more probably, to differences in the recording method used. Small spiders may easily be ingested by chicks when collars are used, and videotaping is probably a less biased recording method (Barba & Monrós, 1999). Another important pattern found was that males

and females provided different quantities of spiders to their nestlings, with females capturing a higher percentage of spiders. Sexual differences in the use of spiders has not been previously documented in Great Tits (Atienzar et al., 2009; Garcia–Navas & Sanz, 2010; Mitrus et al., 2010), although the pattern is consistent with results found by Grieco (2001), who stated that female Blue Tits were more flexible in relation to their feeding behaviour, while males maintained a constant proportion of food components between years. Differences between sexes may be determined by sex–differential strategies in feeding, or even, foraging behaviour (Tschirren et al., 2005). Nevertheless, we should point out that even though females provided a higher proportion of spiders than males, figures reversed when we analysed absolute values, so that males provided in general higher quantities of food, including spiders. Our results contrast with Wright & Cuthill (1990) and Mitrus et al. (2010) who reported higher feeding rates for females. This means that males are responsible for an important part of chick provisioning and development in our population. Future studies should assess whether the active selection of spiders in the Mediterranean is constant through the different stages within the life–history of the species (e.g., when moulting or breeding) or with the age of the birds. Finally, given the nutritional value of spiders due to their high content in taurine (Ramsay & Houston, 2003), it would be interesting to determine the extent to which their preferential selection in the Mediterranean area can have physiological consequences for the Great Tits. Acknowledgements The present study was funded by CGL2009–10652 research project to JCS and JJN, and by FPI grants BES2010–040359 to EPN, from the Ministry of Science and Innovation, Spanish Research Council. Birds were handled with the permission of the Catalan Ringing Office (ICO) and the Departament de Medi Ambient, Generalitat de Catalunya. We thank the owners of Can Catà for kindly allowing us to use their facilities. We also thank J. A. Barrientos for his help with the identification of Arthropods. References Arnold, K. E., Ramsay, S. L., Donaldson, C. & Adam, A., 2007. Parental prey selection affects risk–taking behaviour and spatial learning in avian offspring. Proceedings of the Royal Society B–Biological Sciences, 274: 2563–2569. Arnold, K. E., Ramsay, S. L., Henderson, L. & Larcombe, S., 2011. Seasonal variation in diet quality: antioxidants,invertebrates and Blue Tits Cyanistes caeruleus. Biol. J. Linn. Soc., 99: 708–717. Atienzar, F., Andreu, J., Alvarez, E. & Barba, E., 2009. An improved type of wire cage for the study of parental feeding behaviour in hole–nesting pas-


Animal Biodiversity and Conservation 34.2 (2011)

serines. Revista Catalana d’Ornitologia, 25: 26–31. Balen, J. H. V., 1973. A comparative study of the breeding ecology of the Great Tit Parus major in different habitats. Ardea, 61: 1–93. Banbura, J., Lambrechts, M., Blondel, J., Perret, P. & Cartanson, M., 1999. Food handling time of Blue Tits chicks: constraints and adaptations to different prey tipes. J. Avian Biol., 30: 263–270. Barba, E. & Gil–Delgado, J. A., 1990. Seasonal variation in nestling diet of the Great Tit Parus major in orange groves in eastern spain. Ornis Scand., 21: 296–298. Barba, E. & Monrós, J. S., 1999. Métodos de estudio de la alimentaciónen pollos de paseriformes: Una revisión. EtoloGuía, 17: 31–52. Biard, C., Surai, P. F. & Moller, A. P., 2006. Carotenoid availability in diet and phenotype of Blue and Great Tit nestlings. J. Exp. Biol., 209: 1004–1015. Blondel, J., Dervieux, A., Maistre, M. & Perret, P., 1991. Feeding ecology and life history variation of the Blue Tit in Mediterranean deciduous and sclerophyllous habitats. Oecologia, 88: 9–14. Carrascal, L. M., Senar, J. C., Mozetich, I., Uribe, F. & Domènech, J., 1998. Interactions among environmental stress, body condition, nutritional status, and dominance in Great Tits. Auk, 115: 727–738. Cowie, R. J. & Hinsley, S. A., 1988. Feeding ecology of Great Tits Parus major and Blue Tits Parus caeruleus breeding in suburban areas. J. Anim. Ecol., 57: 611–626. Currie, D., Nour, N. & Adriaensen, F., 1996. A new technique for filming prey delivered to nestlings, making minimal alterations to the nest box. Bird Study, 43: 380–382. Garcia–Navas, V. & Sanz, J. J., 2010. Flexibility in the Foraging Behavior of Blue Tits in Response to Short–Term Manipulations of Brood Size. Ethology, 116: 744–754. – 2011. The importance of a main dish: nestling diet and foraging behaviour in Mediterranean Blue Tits in relation to prey phenology. Oecologia, 165: 639–649. Gibb, J. A. & Betts, M. M., 1963. Food and Food Supply of Nestling Tits (Paridae) in Breckland Pine. J. Anim. Ecol., 32: 489–533. Gosler, A. G., 1993. The Great Tit. Hamlyn, London. Grieco, F., 2001. Short–term regulation of food–provisioning rate and effect on prey size in Blue Tits,

361

Parus caeruleus. Anim. Behav., 62: 107–116. Mitrus, C., Mitrus, J. & Sikora, M., 2010. Changes in nestling diet composition of the Red–Breasted Flycatcher Ficedula parva in relation to chick age and parental sex. Animal Biology, 60: 319–328. Naef–Daenzer, L., Naef–Daenzer, B. & Nager, R. G., 2000. Prey selection and foraging performance of breeding Great Tits Parus major in relation to food availability. J. Avian Biol., 31: 206–214. Poulsen, J. G. & Aebischer, N. J., 1995. Quantitative Comparison of Two Methods of Assessing Diet of Nestling Skylarks Alauda arvensis. The Auk, 112(4): 1070–1073. Ramsay, S. L. & Houston, D. C., 2003. Amino acid composition of some woodland arthropods and its implications for breeding Tits and other passerines. Ibis, 145: 227–232. Royama, T., 1970. Factors governing hunting behaviour and selection of food by Great Tit Parus major. J. Anim. Ecol., 39: 619–668. Sanz, J. J., 1998. Effect of habitat and latitude on nestling diet of Pied Flycatchers Ficedula hypoleuca. Ardea, 86: 81–88. Senar, J. C., Moller, A. P., Ruiz, I., Negro, J. J., Broggi, J. & Hohtola, E., 2010. Specific appetite for carotenoids in a colorful bird. Plos ONE 5: e10716 Sillanpaa, S., Salminen, J. P., Lehikoinen, E., Toivonen, E. & Eeva, T., 2008. Carotenoids in a food chain along a pollution gradient. Science of the Total Environment, 406: 247–255. Tinbergen, L., 1960. The natural control of insects in pinewoods. Arch. Néerl. Zool., 13: 259–379. Tschirren, B., Fitze, P. S. & Richner, H., 2005. Carotenoid–based nestling colouration and parental favouritismin the Great Tit. Oecologia, 143: 477–482. Wansink, D. & Tinbergen, J. M., 1994. The influence of ambient temperature on diet in the Great Tit. Journal of Avian Biology, 25: 261–267. Wilkin, T. A., King, L. E. & Sheldon, B. C., 2009. Habitat quality, nestling diet, and provisioning behaviour in Great Tits Parus major. J. Avian Biol., 40: 135–145. Woodburn, R., 1997. Breeding ecology of the blue tit and great tit and the possible effects of climate change. Ph. D. Thesis, Oxford University. Wright, J. & Cuthill, I., 1990. Manipulation of sex differences in parental care: the effect of brood size. Anim. Behav., 40: 462–471.


366


Animal Biodiversity and Conservation 34.2 (2011)

363

Actes del Cinquè Simposi Internacional del Mussol Comú, 4–6 febrer 2011, Vic, Catalunya (Spain) Actas del Quinto Simposio Internacional del Mochuelo Europeo, 4–6 febrero 2011, Vic, Cataluña (Spain) Proceedings of the Fifth International Little Owl Symposium, 4–6 February 2011, Vic, Catalonia (Spain)

Comitè organitzador / Comité organizador / Organizing Committee Jordi Baucells Grup de Natrualistes d'Osona, Barcelona, Spain Hugo Framis Barcelona, Spain Dries Van Nieuwenhuyse Natuurpunt, Herzele, Belgium Íñigo Zuberogoitia Estudios Medioambientales Icarus, La Rioja, Spain

Comitè científic / Comité científico / Scientífic committee Hugo Framis Barcelona, Spain Juan Carles Senar Museu de Ciències Naturals, Barcelona, Spain Íñigo Zuberogoitia Estudios Medioambientales Icarus, La Rioja, Spain Assessors dels articles / Asesores de los artículos / Referees of papers Martin Grüebler, University of Bern, Switzerland Grzegorz Grzywaczewski University of Life Sciences, Lublin, Poland Geoffrey L. Holroyd, Canadian Wildlife Services, Environment Canada, Edmonton, Canada David Johnson Global Owl Project, Alexandria, Virginia, USA José Enrique Martínez, Bonelli´s Eagle Study and Conservation Group, Murcia, Spain Jose Antonio Martínez Alicante, Spain Bernd–Ulrich Meyburg Berlin, Germany Enrique Murgui Grupo para el Estudio de las Aves, Valencia, Spain Jerry Olsen Institute for Applied Ecology, University of Canberra, Australia Diego Pavón University of Helsinki, Finland Vincenzo Penteriani Estación Biológica de Doñana–CSIC, Sevilla, Spain Maurizio Sara Universitá degli Studi di Palermo, Italy Michael Schaub, University of Bern, Switzerland Peter Sunde National Environmental Research Institute, Aarhus University, Denmark Jabi Zabala IHOBE–Sociedad Pública de Gestión Ambiental, Vizcaiya, Spain


364

Participants / Ponentes / Speakers Raúl Alonso Centro de Recuperación de Rapaces Nocturnas–BRINZAL, Madrid, Spain Jordi Baucells Grup de Naturalistes d'Osona, Barcelona, Spain Hugo Framis Barcelona, Spain Grzegorz Grzywaczewski University of Life Sciences in Lublin, Poland Paul Haustraete Regional Landschap Vlaamse Ardennen, Belgium Geoffrey L. Holroyd Canadian Wildlife Service, Environment Canada, Edmonton, Canada Emily Z. K. Joáchim School of Biological Sciences, University of Reading, Reading, UK Patrick Lecomte Etudes Chevêche, Onservatoire Scientifiquede l'Avifaune d'Ille de France, France Guillem Mas Grup de Naturalistes d'Osona, Barcelona, Spain Rafael Molina–López Centre de Fauna Salvatge de Torreferrussa, Barcelona, Spain Ronal Van Harxen STONE, The Netherlands Dries Van Nieuwenhuyse Natuurpunt, Herzele, Belgium Íñigo Zuberogoitia Estudios Medioambientales Icarus, La Rioja, Spain


Animal Biodiversity and Conservation 34.2 (2011)

365

Índex / Índice / Contents 367 H. Framis, J. Baucells, I. Zuberogoitia & D. Van Nieuwenhuyse The Fifth International Little Owl Symposium, 4–6 February 2011, Vic, Catalonia (Spain) 369–378 H. Framis, G. L. Holroyd & S. Mañosa Home range and habitat use of little owl (Athene noctua) in an agricultural landscape in coastal Catalonia, Spain 379–387 G. L Holroyd & H. E. Trefry Tracking movements of Athene owls: the application of North American experiences to Europe

389–393 R. Alonso, P. Orejas, F. Lopes & C. Sanz Pre–release training of juvenile little owls Athene noctua to avoid predation 395–400 I. Zuberogoitia, J. Zabala & J. E. Martínez Bias in little owl population estimates using playback techniques during surveys 401–405 R. A. Molina–López & L. Darwich Causes of admission of little owl (Athene noctua) at a wildlife rehabilitation centre in Catalonia (Spain) from 1995 to 2010


366


Animal Biodiversity and Conservation 34.2 (2011)

367

The Fifth International Little Owl Symposium, 4–6 February 2011, Vic, Catalonia (Spain) H. Framis, J. Baucells, I. Zuberogoitia & D. Van Nieuwenhuyse

Framis, H., Baucells, J., Zuberogoitia, I. & Van Nieuwenhuyse, D., 2011. The Fifth International Little Owl Symposium, 4–6 February 2011, Vic, Catalonia (Spain). Animal Biodiversity and Conservation, 34.2: 367. The First International Little Owl Symposium took place in Champ–sur–Marne (France) in November 2000, organized by the International little owl Working Group (ILOWG). It was the first international meeting regarding a sole owl species in Europe. Since then, efforts have been made to improve our knowledge and develop management plans for little owl conservation, and every new advance has been shared among all of us, owl researchers and owl friends, at regular international meetings. The Fifth Symposium gave continuity to the series of previous meetings held in France, England and Belgium. The most recent meeting was held in Herzele, Flanders, and the highlight was the presentation of the latest little owl monograph (Van Nieuwenhuyse et al., 2008). This meeting demonstrated the need to promote future gatherings of researchers. This year´s symposium therefore aimed to serve a dual purpose. On one hand, it was a call to stimulate studies of the species from consolidated groups from central and northern Europe, and on the other, it aimed to promote the less known initiatives from southern European countries where there is less tradition in the study of little owl. Researchers from Poland, The Netherlands, England, France, Belgium, Canada and Spain attended the call and offered 14 presentations to an audience of 60 little owl enthusiasts. It was an opportunity to catch up with recently completed studies as well as ongoing projects. Topics ranged from population studies and citizen science to the use of technology in tracking owl movements, measuring the effectiveness of broadcast surveys and reviewing results from rehabilitation centers. This issue of the journal includes five peer reviewed papers from among these presentations, and others are expected in future publications. The recent publication of an owl conservation book in Catalonia (Baucells, 2010) favoured this symposium being held in Vic, and the support of several sponsors contributed greatly to the meeting becoming a reality. We are extremely thankful to the Departament de Medi Ambient de la Generalitat de Catalunya, Grup Naturalista d’Osona, Institut Català d’Ornitologia (ICO), Patronat d’Estudis Osonencs, Escola Sant Miquel, Grup d’Anellament Calldetenes Osona (GACO), Caty Lorton, and last but not least to the Institut de Cultura: Museu de Ciències Naturals de Barcelona for their readiness to publish the papers presented at the Fifth Symposium. As the organizing committee, we hope that these proceedings provide a valuable step towards advancing our knowledge of the little owl and contribute to its conservation. References Baucells, J., 2010. Els rapinyaires nocturns de Catalunya. Biologia, gestió i conservación de les vuit espècies de rapinyaires nocturns catalans i els seus hàbitats. I. G. Sta. Eulàlia–Santa Eulàlia de Ronçana, Barcelona. Van Nieuwenhuyse, D., Génot, J.–C. & Johnson, D. H., 2008. The little owl: conservation, ecology and behavior of Athene noctua. Cambridge University Press, U.K. Hugo Framis, c/ Sant Joaquim 12 1, 08302 Mataró, Barcelona, Espanya (Spain).– J. Baucells, Grup de Naturalistes d'Osona & Grup d'Anellament de Calldetenes Osona (GACO), Molí de Torrellebreta s/n., 08519 Malla, Espanya, Spain.– Iñigo Zuberogoitia, Estudios Medioambientales Icarus S.L., Pintor Sorolla 6 1º C, 26007 Logroño, España (Spain).– Dries Van Nieuwenhuyse, Natuurpunt, Herzele, Belgium.


366


Animal Biodiversity and Conservation 34.2 (2011)

369

Home range and habitat use of little owl (Athene noctua) in an agricultural landscape in coastal Catalonia, Spain H. Framis, G. L. Holroyd & S. Mañosa

Framis, H., Holroyd, G. L. & Mañosa, S., 2011. Home range and habitat use of little owl (Athene noctua) in an agricultural landscape in coastal Catalonia, Spain. Animal Biodiversity and Conservation, 34.2: 369–378. Abstract Home range and habitat use of little owl (Athene noctua) in an agricultural landscape in coastal Catalonia, Spain.— In recent decades agricultural landscapes in Catalonia have undergone a profound transformation as in most of Europe. Reforestation and urban development have reduced farmland and therefore the availability of suitable habitat for some bird species such as the little owl (Athene noctua). The outskirts of the city of Mataró by the Mediterranean Sea exemplify this landscape change, but still support a population of little owl where agriculture is carried out. Three resident little owls were monitored with telemetry weekly from November 2007 until the beginning of August 2008 in this suburban agricultural landscape. Mean home range ± SD was 10.9  ± 5.5 ha for minimum convex polygon (MCP100) and 7.4 ± 3.8 ha for Kernel 95% probability function (K95). Home ranges of contiguous neighboring pairs overlapped 18.4% (MCP100) or 6% (K95). Home range varied among seasons reaching a maximum between March and early August but always included the nesting site. Small forested patches were associated with roosting and nesting areas where cavities in carob trees (Ceratonia siliqua) were important. When foraging in crop fields, the owls typically fed where crops had recently been harvested and replanted. All three owls bred successfully. Key words: Little owl, Athene noctua, Telemetry, Conservation, Home range, Habitat use, Agricultural landscape. Resumen Área de campeo y uso del hábitat del mochuelo europeo (Athene noctua) en un paisaje agrícola de la costa de Cataluña, España.— El paisaje agrícola en Cataluña ha sufrido una profunda transformación en las últimas décadas, tal y como ha ocurrido en gran parte de Europa. La reforestación y especialmente el desarrollo urbanístico han reducido las tierras agrícolas y con ello se ha perdido hábitat adecuado para especies como el mochuelo europeo (Athene noctua). Los alrededores de la ciudad de Mataró, a orillas del mar Mediterráneo, son un buen ejemplo de este cambio, pero todavía acogen una población de mochuelos allí donde se da actividad agrícola. Entre noviembre de 2007 y principios de agosto de 2008 se siguieron semanalmente con telemetría tres mochuelos residentes en este entorno agrícola periurbano. La media del área de campeo ± DE estimada con el polígono convexo mínimo (MCP100) fue de 10,9 ± 5,5 ha, y de 7,4 ± 3,8 ha, con el estimador de Kernel 95% (K95). Las áreas de campeo de las parejas vecinas se solapaban un 18,4% (MCP100) o un 6% (K95). Las áreas de campeo entre temporadas variaron a lo largo del seguimiento y llegaron a un máximo entre marzo y principios de agosto, aunque éstas siempre albergaron la zona del nido. Las pequeñas manchas arboladas se asociaron a áreas de reposo y nidificación, donde las cavidades naturales de los algarrobos (Ceratonia siliqua) eran importantes. Cuando los mochuelos se detectaron en los campos, fue en cultivos recién cosechados o replantados. Los tres mochuelos criaron con éxito. Palabras clave: Mochuelo europeo, Athene noctua, Telemetría, Conservación, Área vital, Área de campeo, Uso del hábitat, Paisaje agrícola. (Received: 7 IV 11; Conditional acceptance: 26 VI 11; Final acceptance: 10 X 11) H. Framis, c/ Sant Joaquim 12 1, 08302 Mataró, Espanya (Spain).– G. L. Holroyd, Environment Canada, Room 200, 4999–98 Ave., Edmonton, AB, T6J 1Z1, Canada.– S. Mañosa, Dept. de Biologia Animal, Univ. de Barcelona, Av. Diagonal 643, 08028 Barcelona, Espanya (Spain). Corresponding author: H. Framis. E–mail: hugoframis@yahoo.com ISSN: 1578–665X

© 2011 Museu de Ciències Naturals de Barcelona


370

Introduction Landscape transformation and particularly changes in agricultural areas have resulted in reductions of little owl (Athene noctua) populations throughout Europe (Van Nieuwenhuyse et al., 2008). Similarly, in Spain where the species is still broadly distributed (Martí & del Moral, 2003), populations are in decline (Blas García & Muñoz Gallego, 2003). The owl’s distribution in Catalonia, in northeast Spain, was reduced in 61 10  x  10 km quadrates (21%) in the 20–year period from 1982 to 2002. The most recent breeding population estimate in Catalonia was between 9,000 and 14,500 breeding pairs (Framis, 2004). The little owl, a species that primarily inhabits open, unforested areas, has occupied agricultural landscapes and has been influenced by the prevailing transformation of agricultural environments (Van Nieuwenhuyse et al., 2008). During the last decades, changes in agricultural landscapes have been characterized by two independent processes: a) agricultural intensification, and b) land abandonment with subsequent reforestation or urbanization. Intensification of agriculture is characterized by the elimination of interstitial elements such as hedgerows between arable fields, isolated trees and stone buildings, accompanied by the introduction of new farming techniques such as the use of fertilizers and pesticides and the introduction of extensive monocultures (Sanderson et al., 2005; Onrubia & Andrés, 2005). Intensification has been correlated with the decline of populations of other bird species that specialize in farmland environments (Krebs et al., 1999; Donald et al., 2001; Donald et al., 2002; Sanderson et al., 2005). Land abandonment results in the replacement of farmland with urban areas or shrubland and forest. In Catalonia, a region of slightly over 32,000 km2, 61% of farmland was lost in 50 years, a reduction of 16 km2 annually (Montasell, 2010). Total forested surface in Catalonia increased from 36% in 1970 to almost 61% in 2005 (Montasell, 2010). City enlargement has also transformed the landscape, especially throughout the Barcelona Metropolitan Region (Catalán et al., 2008). These changes in land uses and traditional agricultural practices have reduced little owl habitat availability (Baucells, 2010; Andino, 2005; Framis, 2004). Little owls occupy small territories year round in Catalonia (Muntaner et al., 1983; Calvet et al., 2004; Aymí & Tomás, 2003). However, no attempt had yet been made to investigate their home range and habitat use in this region. The objectives of this study were to measure the home range of little owls within an intensive market–garden agricultural landscape on the Mediterranean coast, to analyze habitat use related to little owl’s main activities, and to propose some conservation priorities to conserve the local population of little owls within the agricultural landscape. Methods The study area is located north–east of the city of Mataró, a town of 120,000 inhabitants in the Maresme county on the Mediterranean Sea (N 41º 33' E 2º 28', fig. 1). The study area is within the 263 ha fertile gentle

Framis et al.

slope called the Cinc Sènies, where the main activity is intensive market–garden agriculture (Montasell, 2006). Main crops are celery, parsley, spinach, onion and potatoes. Small forest patches of carob trees (Ceratonia siliqua), littoral oak (Quercetum ilex) and stone pine trees (Pinus pinea) occur mainly around the hill named Turó d’Onofre Arnau (131 m a.s.l.) at the centre of the Cinc Sènies. The area extends longitudinally, between the coast and the so–called littoral mountain range, from 10–131 m a.s.l. The agricultural and surrounding open area is categorized as peri–urban environment due to its proximity to a medium size city (Montasell, 2006). Industrial sites, railroad tracks, roads, highways, and suburbs have an obvious impact on this landscape whose infrastructure is still under intensive development. The local Mediterranean climate is characterized by an annual average temperature between 15–16ºC. In the summer, temperatures as high as 30ºC occur but are not normal. However, the annual relative humidity of 72% makes the weather sultry (Andino et al., 2005). In winter, temperatures decline to 5–7ºC (Atles Climàtic Digital de Catalunya) but frost is rare and it does not snow due to the nearby sea. Precipitation is between 550–600 mm annually (Andino et al., 2005). The climate is mild enough to support continuous cropping of the market garden agriculture. Intraguild predation is known to influence little owl densities and behavior (Zabala et al., 2006; Zuberogoitia et al., 2008). Barn owl (Tyto alba) and scops owl (Otus scops) are present in the study area; tawny owl (Strix aluco) is absent, but distributed in the nearby areas where woods are continuous (Framis, 2008). The effect of predators was not taken into account for this study. Radio telemetry Knowledge of owl territories from a previous census in 2007 (Framis, 2008) allowed us to decide where to trap owls, for telemetry. Little owl call playback was used to attract individuals into mist nets just after dark in early November 2007. The nets were checked for trapped owls every 15’. Eleven owls were trapped in a week and handled following the Catalan Ornithological Institute ringing standards (1/03 edition; ICO, 2003). Age was assigned following Martínez et al. (2002). We attached transmitters to five of the 11 trapped owls using the backpack style harnesses with 4 mm wide Teflon® webbing (Bally Ribbon Mills Inc, Pennsylvania, USA). Battery life of the transmitters was 9 months and frequencies were between 172.097 and 172.659 MHz (Holohil Systems Ltd, Ontario, Canada). The mass of the transmitters was 6.3 g and together with the attaching material, the total mass was ~4.3% of the owl’s mass which did not exceed the maximum of 5% recommended by Kenward (2001). An R–1000 telemetry receiver (Communication Specialists Inc., Orange, California, USA) and a three element Yagi antenna were used to triangulate the locations of the transmitters. The effective range of the transmitters was limited by line–of–sight and was 1–3 km depending upon terrain and location of the owls. Telemetry started in November 2007 and lasted until the first week of August 2008. Intensive work at day and


Animal Biodiversity and Conservation 34.2 (2011)

371

night took place in the first 10 days after captures to find the roosting areas. As soon as owls’ roosts were located, each owl’s movements were determined one or two days a week, starting with a different owl each afternoon just before sunset to confirm daily roost sites and continuing with the other owls alternatively through the night to determine foraging locations (from 1 to 4 hours/session). Intensive individual owl tracking was conducted for one of the owls in each session, and at least one random location was secured for the other two owls. Each location was considered to be independent from the others and they were analyzed regardless of the time spent at each point. The telemetry schedule after sunset and before sunrise coincided with the high–activity foraging time as described for the little owl (Van Nieuwenhuyse et al., 2008). The mosaic of arable plots and access roads, at different elevations, prevented us from taking compass bearings from specific vantage points, so we had to move regularly to get bearings on owls. All owl locations were plotted on a map (1:3,400) during each session, generating a different map each day. Locations were frequently confirmed by spontaneous calls of the owls and by direct observations, made possible by the light pollution from the city and from the scattered houses in the area. When adverse conditions of wind or rain occurred, telemetry sessions were cancelled to avoid poor reception. Locations were recorded with a GPS device (Garmin Etrex, Garmin Ltd. 2000–2003). Whenever possible, for each location, we recorded time, landscape type, owl activity and the structural feature where it was detected (table 1). When appropriate, crop variety, height and density were also recorded. Transmitters were removed from two of the owls in August and October 2008 and the other three were not retrapped. Monitored pairs bred successfully (2.3  fledglings/pair), two in naturally occurring chambers in carob trees and one most probably in an old concrete irrigation canal.

N

To France (140 km)

Mataro

Urban area To Barcelona (30 km)

it

ed

M

r er

an

e

an

a

Se

Little owl 1 ♀ Little owl 2 ♂ Little owl 3 ♂ Municipality limit

0

2000 m

Fig. 1. Location of the study area and minimum convex polygons 100% (MCP100) of the three monitored little owls within the limits of the Mataró municipality. Fig. 1. Localización del área de estudio y mínimos polígonos convexos 100% (MCP100) de los tres mochuelos europeos monitorizados dentro de los límites del municipio de Mataró.

Table 1. Distribution (%) of the little owls’ telemetry locations according to landscape (U. Urban; W. Woodland; C. Crop), activity (F. Foraging; R. Roosting; O. Other; Nk. Not known) and landscape feature (B. Building; P. Pole; G. Ground; T. Tree; Nk. Not known). (Number of locations for each owl in brackets, total = 469). Tabla 1. Distribución (%) de las localizaciones por telemetría de los mochuelos europeos según las características del paisaje (U. Urbano; W. Bosque; C. Cultivo), la actividad llevada a cabo (F. Alimentación; R. Dormidero; O. Otras; Nk. No conocida) y las características del mismo (B. Edificio; P. Poste; G. Suelo; T. Árbol; Nk. No conocido). (Número de localizaciones para cada mochuelo entre paréntesis, total = 469). Little owl 1♀

Landscape U 2(3)

W

Activity C

F

R

O

Landscape feature Nk

B 1(1)

P

G

T

Nk

49(70) 49(70)

43(62) 24(35) 16(23) 16(23)

3(5) 43(61) 50(72) 3(4)

2♂

10(16) 62(102) 29(48)

41(69) 15(26) 12(21) 30(50)

1(2) 20(33) 15(25) 60(99) 4(7)

3♂

52(83) 16(26) 32(51)

52(83) 12(19) 11(18) 25(40)

33(53) 17(27) 10(16) 33(52) 8(12)


372

Framis et al.

Table 2. Total home range estimates for each owl with minimum convex polygon 100% (MCP100), Kernel 95% (K95) and 50% (K50) (mean ± SD) in hectares. Seasonal variation of little owl home ranges estimated with MCP100 and K95. Study period 16 XI 07–6 VIII 08: W. Winter season (16 XI 07–28 I 08); B. Breeding season (1 III 08–6 VIII 08). (Number of locations for each owl in brackets, total = 469). Tabla 2. Estimas del área de campeo para cada mochuelo con un mínimo polígono convexo 100% (MCP100) y un estimador de Kernel 95% (K95) y 50% (K50) (media ± DE) en hectáreas. La variación estacional de las áreas vitales de los mochuelos europeos se estimó mediante MCP100 y K95. Período de estudio 16 XI 07–6 VII 08: W. Estación invernal (16 XI 07–28 II 08); B. Estación de cría (1 III 08–6 VIII 08). (El número de localidades para cada mochuelo se incluye entre paréntesis, total = 469). Little owl

MCP 100

K95

W

B

1♀

3.1(81)

3.7(62)

2.8

2♂

7.5(72)

13.4(94)

3♂

9.0(64)

Mean ± SD 6.6 ± 3.1

10.9(96) 9.3 ± 5.1

W

B

MCP100

K95

K50

2.3

4.8

3.2

0.6

7.2

10.9

15.5

10.5

2.7

5.9

8.5

5.28 ± 2.3 7.2 ± 4.4

Home range analysis Home range sizes were calculated using two estimators: MCP (minimum convex polygon) 100% and fixed Kernel 95% contours. The MCP created a polygon by connecting the most outer locations, (White & Garrot, 1990); MCP has been commonly used in similar studies of the species (summary in Van Nieuwenhuyse et al., 2008; Grzywaczewski, 2009). The Kernel method estimated the distribution of the locations creating contours of probability of the individual’s presence (Sissons, 2003; Zuberogoitia et al., 2007; Sunde et al., 2009). Calculations were done using the location analysis application from Ranges 7 v0.67 software (South et al., 2005). Little owls’ basic habitat requirements are known to vary throughout the year to meet different life history needs, and home range dimensions change accordingly, particularly in the breeding season (Finck, 1990). In order to find variation in home range through the nine month monitoring period, which embraced most of the annual activity of the species, two main seasonal periods were defined: winter, November–February, as in Van Nieuwenhuyse et al. (2008), and breeding: March–August which included courtship,nesting, and early fledgling (Fink, 1990; Van Nieuwenhuyse et al., 2008; Grzywaczewski, 2009). Monitoring of individuals within the same population allowed us to determine their interaction, particularly interesting for a territorial species with a well developed social activity (Hardouin et al., 2006; Zuberogoitia et al., 2007). The overlap application from Ranges 7 v0.67 software (South et al., 2005) was used to quantify shared space using the telemetry locations. Habitat and activity data analysis Habitat use was analyzed by means of Chi–square test, comparing the frequencies of telemetry locations

12.5

8.4

2.0

10.9 ± 5.5

7.4 ± 3.8

1.8 ± 1.1

in each land cover type with the expected frequencies according to the proportion of habitats within the 100% MCP range of each owl. Land use availability for each MCP was determined from the 2007 regional land cover map (Ibáñez & Burriel, 2008; CREAF, 2009). We calculated the proportions of different land cover types within the owls’ home ranges and the distribution of owls’ nocturnal locations (15 minutes after sunset to 15 before sunrise) (n = 343) within land cover types using MiraMon GIS (Pons, 2004) and ArcGis (v.9.2) software. Associations between activity variables and landscape or landscape feature variables were performed by cross tabulation analysis and Chi–square test. For this, we categorized owl locations into a simplified three principal landscape types (assigned from field data) to facilitate further analysis. The available land cover types from the regional map within the study area were distributed into the three categories as follows: crop (other irrigated herbaceous crops, greenhouse, greenhouse & market garden, fields & grassland); forest (shrubland, pine woods, nonbuild urban land, riparian shrubland); urban (housing development, farmland under transformation, greenhouse in a chicken farm). Significant differences between observed and expected frequencies on a given cell were evaluated by means of the standardized residuals. Statistical analyses were conducted with SPSS statistics software 15.0 for Windows (SPSS Institute, Chicago, Illinois, USA). Results Home ranges Two of the five tagged owls were hatch–year age and three were adults (two males and one female). The hatch–year owls were not detected again in the study area despite random telemetry searches to


Animal Biodiversity and Conservation 34.2 (2011)

373

Table 3. Hectares of each land use type within each MCP range based on land cover maps, number of fixes within each category only at night (15' after sunset or 15' before sunrise) (n = 343), and significance of the Chi–square test results for owl night locations distribution in relation to habitat availability. (+) Land cover type is used more often than expected from its availability; (–) Land cover type is used less than expected; Sf. Surface; F. Fixes. S. Significance; NS. Non–significant difference between observed and expected frequencies; – Land cover not available for that owl; ** p ≤ 0.05 and *** p ≤ 0.01. Tabla 3. Hectáreas de cada uso del paisaje dentro de cada rango MCP, basándose en mapas de cubierta del suelo, número de fijos dentro de cada categoría sólo durante la noche (15' después de la puesta de sol o 15' antes del alba) (n = 343), y significación de los resultados del test de la ji–cuadrado para las distribuciones de las localizaciones nocturnas de los mochuelos en relación con la disponibilidad del hábitat: (+) El tipo de cubierta se usa con mayor frecuencia de la esperada por su disponibilidad; (–) El tipo de cubierta se usa menos de lo esperado; Sf. Superficie; F. Corrección; S. Significación; NS. Diferencia no significativa entre las frecuencias observadas y esperadas; – Cubierta del suelo no disponible para ese mochuelo; ** p ≤ 0,05 y *** p ≤ 0,01.

Little owl 1♀

Little owl 2♂

Land cover type

Sf

F

S

Sf

Other irrigated herbaceous crops

2.7

68

NS

6.8

40 **(–)

5.6

19 ***(–)

Shrubland

0.5

9

NS

1.3

23 ***(+)

0.3

7

NS

Pinewood (Pinus pinea) (> 20%)

0.5

34

***(+)

2.5

39 ***(+)

0.84

0

NS

Housing development and isolated houses 0.2

0

2.4

11 NS

2.26 27 NS

Non–built urban land

F

S

Little owl 3♂ Sf

0.4

2

***(–)

0.02

0

0

1.7

1 ***(–)

Farmland under transformation

0

0

0.1

0

0

0

Riparian shrubland

0

0

0.3

2

NS

0.1

2

NS

Fields and grassland

0

0

0.1

0

0.1

1

NS

Warehouse (Chicken farm)

0

0

0.03

0

0.8

32 ***(+)

Total area / total number of fixes

4.34 113

try to locate them through the study period (within a maximum radius of 1.1 km from the study area). The three adults stayed on the study area throughout the nine months of monitoring. Telemetry effort totaled 212.5 hours over 81 days, producing 469 locations for the three monitored owls (mean 156.3 ±11.9 SD locations/owl), yielding 5.7 ± 3.8 total locations/session. Monitoring effort was relatively evenly distributed between November 2007 and July 2008, 23.0 ± 5.7 hours/month and was terminated in early August 2008. Average MCP 100% home range was 10.9 ± 5.5 ha and mean home range estimated with Kernel 95% was 7.4 ± 3.8 ha. Estimates of home range size are dependent, in part, on the number of telemetry points used to determine the home range (Kenward, 2001). As the number of points increased, the estimated home range size reached an asymptote. For the three owls the total numbers of locations over the monitoring period reached the asymptote at around 100 locations, for both MCP and Kernel analysis. Home ranges were calculated for the winter and the breeding periods using both MCP and Kernel analysis

15.25 116

2.4

0

S

Greenhouse market garden agriculture 0.04

0

F

26 NS

12.4 114

(table 2). For both these periods a minimum of 30 locations was used to define home range (Kenward, 1987). Seasonal home ranges of each individual were nested within each other. While the size of the home ranges varied seasonally, these areas largely overlapped and always included the nest site for all three owls. The three owls held contiguous territories. Overlap of home ranges was low, accounting for 6 and 18.4% (22.2 ha K 95% and 32.79 ha MCP 100% respectively) of the total home range used by all three owls. Habitat use Crop fields (see methods section for land cover typology) accounted for 45–63% of the total available home range of the owls. Each owl had a particular habitat use pattern. Forest areas were visited more than expected by two owls, and urban landscape by the third owl (table  3). Activity was not independent of habitat type (table 4). Woodland was preferred for roosting for two owls, while crops were preferred for foraging by the same owls. The third owl showed no significant relationship between the variables landscape


374

Framis et al.

Lo–2

Land cover type Farmland under transformation Fields and grasslands Greenhouse market garden agriculture Housing development and isolated houses Other irrigated herbaceous crops Woodpine (Pinus pinea) (> 20%) Non-built urban land Riparian shrubland Greenhouse Warehose

Lo–1

Shrubland

Lo–3

N 0

200 m

Fig. 2. Home ranges of little owl (Lo) 1, 2 and 3, MCP100 and land cover typology within them. Fig. 2. Áreas de campeo de los mochuelos europeos (Lo) 1, 2 y 3, MCP100 y usos del suelo dentro de ellas.

and activity, since it nested in the chicken farm and roosted in trees both in and outside the farm. When owls foraged at night in open ground they avoided dense crop cover; 93% (85–97%) of the foraging locations were in fields with < 60% crop cover and 82% (71–90%) were in fields with crops ≤ 15 cm height. Locations (n = 82) were recorded in up to eight field varieties; bare soil accounted for 61% of fields and parsley accounted for the highest percentage of crop visits (31%), mainly by the female. Most mature crops in this intensive market garden area were very dense, with little space between crop stems. The proportion of mature crop cover increased up to 83.3% for celery, 96.1% for spinach and 100% for parsley. Within these crop fields the pattern of owl foraging was quite distinct, as will be discussed. Discussion Home range analysis and interactions These little owls were permanent residents of the agricultural area of Cinc Sènies in coastal Catalonia. Mean value of the home ranges of the three owls was smaller than elsewhere (average MCP 100% home range was 10.9 ± 5.5 ha and with Kernel 95% was 7.4 ± 3.7 ha) and showed large variation between them. In four studies in Germany and France annual home ranges were 14.5 ha, 14.6 ha, 27.4 ha and 31 ha (n = 12, n = 19, n = 4, n = 8, respectively; Van

Nieuwenhuyse et al. 2008). In northern Spain, annual home range was 15.1 ± 2.4 ha (range 10.3–18.6 ha, n = 9) estimated in MCP 95% (Zuberogoitia et al., 2007). Large variation in home range size was also shown in other studies (Grzywaczewski, 2009; Sunde et al., 2009). Even smaller home ranges have been found elsewhere in northern Europe (Van Nieuwenhuyse et al., 2008). Nonetheless, our study sample size is smaller than that used in other similar approaches (Grzywaczewski, 2009; Sunde et al., 2009; Van Nieuwenhuyse et al., 2008; Zuberogoitia et al., 2007), and must be taken into account when making generalizations (Hebblewhite & Hidon, 2010). Small home ranges in our study may be due to several factors. Mild Mediterranean weather conditions throughout the year with average temperatures of 15–16ºC and humid summers of almost 30ºC (Andino et al., 2005) may facilitate the foraging of owls compared to areas with snow cover in winter at higher latitudes (Finck, 1990). Additionally, the monitored owls would have had the advantage of good feeding opportunities throughout the year due the mosaic of intensive market–garden agriculture that is continuously cropped. The home range was larger in the breeding season than in winter, as opposed to findings in other studies (Finck, 1990; Zuberogoitia et al., 2007; Van Nieuwenhuyse et al., 2008). Winter home range is usually the largest because this is the time of year when owls are not related to a specific area, and therefore enjoy greater mobility, in contrast to the breeding season


Animal Biodiversity and Conservation 34.2 (2011)

375

Table 4. Association between habitat type and activity. Each cell indicates the significance and sign of the association between habitat categories and activities for each owl (Chi–square tests for contingence tables between activity and habitat): (+) The combination occurs more frequently than expected; (–) The combination occurs less frequently than expected; Blanks indicate a non–significant difference between observed and expected frequencies. Results for urban habitat and activity are lacking since no combination showed significance: C&R. Crop and roosting; C&F. Crop and foraging; W&R. Woodland and roosting; W&F. Woodland and foraging; W&O. Woodland and other activity; E(< 5). Percentage of frequencies under 5. (Significance level: ** p ≤ 0.05; *** p ≤ 0.01). Tabla 4. Asociación entre tipo de hábitat y actividad. Cada celda indica la significación y el signo de la asociación entre las categorías de hábitat y las actividades para cada mochuelo (test de la ji–cuadrado para las tablas de contingencia entre actividad y hábitat): (+) La combinación se da con mayor frecuencia de lo esperado; (–) La combinación se da con menor frecuencia de lo esperado. Las celdas en blanco indican que no existe una diferencia significativa entre lo observado y lo esperado. No se representan el hábitat urbano y la actividad urbana, dado que ninguna combinación presentó significación: C&R. Cultivo y descanso; C&F. Cultivo y alimentación; W&R. Bosque y descanso; W&F. Bosque y alimentación; W&O. Bosque y otras actividdes; E(< 5). Procentaje de frecuencias por debajo de 5. (Nivel de significación: ** p ≤ 0,05; *** p ≤ 0,01). Little owl

C&R

C&F

W&R

1♀

***(–)

***(+)

***(+)

2♂

***(–)

***(+)

***(+)

3♂

**(–)

when they are linked to their nest site. Accessibility and availability of resources in the study area might have favored this uncommon situation. Monitoring of three neighboring owls showed the territoriality of this species, since they had little overlap in home range all year round. The overlap of total home ranges of all three individuals was only 18.4% (MCP100) and 6% (K95). Social activity and interaction has been shown for little owl during the winter, especially in February (Zuberogoitia et al., 2007). This was not the case here, since the maximum home range overlap occurred between March and August; 5.4 ha (19%) with MCP 100%. However, overlap always occurred in the feeding grounds and at the boundaries of their territories and away from the core nesting and roosting areas. The breeding success was within the average of the county (Andino et al., 2005) and those of the same area for the previous year (Framis, 2008). This suggests that transmitters did not adversely affect the normal activity of the owls outfitted with them. Little owl habitat use As expected, the owls embraced a high percentage of cropland within their home ranges as well as a large proportion of an adjacent housing development. However, the analysis of the telemetry locations showed a higher than expected use of woodland by two owls, where both had their main roosts and nests. Similarly, the third owl showed a higher than expected use of buildings in the chicken farm where

W&F

W&O

***(–)

**(+)

E(< 5) 0% 0% 16.7%

he roosted and nested. In this case, holding a territory in a farm presumably offered shelter and plenty of feeding opportunities from the constant manure management. In contrast, the other two owls were frequently detected foraging in the crops. Both showed positive associations for crops as feeding grounds. While feeding in crops, owls were always on the ground associated with recently harvested and planted portions of fields that were actively irrigated and had 70–100% bare soil. Hedgerows of bushes or trees were extremely scarce between plots. Arundo donax, an invasive species of cane (Andreu & Vila, 2009), is the only common vegetation between them, but it did not offer support to perch on. Owls made use of wooden electricity poles as stopping perches or simply as vantage spots for surveillance at the edge of the agricultural area when flying from forest patches into the fields. In other Mediterranean areas of treeless pseudo–steppe, low piles of stones make equivalent foraging perches (Tomé et al., 2011). Access to ground prey has been negatively correlated with the height of vegetation (Van Nieuwenhuyse et al., 2008; Grzywaczewski, 2009). Parsley, the most visited crop, was not harvested all at once, but was collected at irregular frequencies a few rows each time, allowing plant regrowth and at the same time making new foraging opportunities accessible to the owls. Foraging on the ground after harvesting was common all year round; it might have been based on availability of insects, and especially earthworms (Lumbricidae), common in the little owl diet (Finck, 1990; Hounsome et al., 2004; Van Nieuwenhuyse et


376

al., 2008). Moreover, the sprinkle irrigation system in those fields might favor access to earthworms that come to the surface of very wet fields. Other resources were also occasionally used, such as manure piles, demonstrating the foraging plasticity of the owl to exploit new resources when they become available (Finck, 1990). Conservation implications The arable area of Cinc Sènies is the most likely site to find little owls in the landscapes surrounding the city of Mataró (Framis, 2008). The cluster of little owl territories found in surveys and by telemetry suggests that the 263 ha study area gives refuge to a relatively high density of litte owls. With home ranges of 7.4–10.9 ha, as many as 35–24 pairs of little owls could live in this agricultural area. However, regional farmland has been reduced in the past fifty years due to agricultural abandonment, and the growth of continuous forests and the city (Sabater et al., 1997, 2008). New urban plans are waiting to be approved for the remaining open space. A right–of–way for an orbital train and extension of the existing highway will isolate the agricultural area from other open spaces even more. Abundant roads have a negative effect on owl occupancy (Zabala et al., 2006). Thus, the continuity of farmland connected with the remaining open areas to the north of the city and beyond should be promoted to preserve little owl habitat (Framis, 2008). Any landscape restoration would also be a contribution towards the welfare of other farmland birds, undergoing negative trends countrywide (Herrando, 2008). Results from this telemetry study show the need to protect patches of forest or traditional tree crops within farmland, especially carob trees, which provide natural chambers that the owls use as nesting and roosting sites. Restoration of hedgerows with trees would increase the chance to preserve nest sites in the centre of the plain, which now lacks much tree cover. Currently, some nest sites depend on the old irrigation canals which remain at risk of disappearing. It has also been suggested that the lack of perches is a limiting factor for foraging in open spaces (Van Nieuwenhuyse et al., 2008). Agricultural management should also preserve the mosaic character of the agriculture carried out in the area and promote patches of bare ground as foraging areas. Agricultural activity is not only an essential cultural and economical asset for the city of Mataró but is also a key element in the conservation of little owls (Sabater et al., 2008). Acknowledgements We are very grateful to the Beaverhill Bird Observatory, Alberta, Canada for providing travel funds for Dr. Holroyd and for the transmitters, to Environment Canada for the loan of the telemetry receiver equipment; to Jordi Tejero and Beatriu Jiménez–Barrera for their dedication to the GIS analysis; to Raül Aymí,

Framis et al.

from the Catalan Ornithological Institute (ICO) for providing the required permits and valuable references; to Marta Comerma and Jordi Corbera from the Secció de Ciències Naturals del Museu de Mataró for their encouragement; to the farmers in the study area, especially the Riera–Lorton, Masjoan and Vinyals families for letting us wander around their fields, day and night; and to four anonymous referees for the comments and suggestions on an earlier version of the manuscript. And finally to the friends and colleagues who gave their support and advice along the way: Gemma Fors, Jose Manuel de los Reyes and Rafel Rocaspana. References Andino, H., Badosa, E., Clarabuch, O. & Lleberia, C. (Eds.), 2005. Atles dels ocells nidificants del Maresme. Barcelona. Andreu, J. & Vilà, M., 2009. Gestió de les invasions vegetals a Catalunya. L’Atzavara, 18: 67–75. Atles Climàtic de Catalunya. http://opengis.uab.es/ wms/acdc/index.htm Aymí, R. & Tomàs, X. 2003. Bird–ringing report of the Catalan Ornithological Institute (ICO) for the period 2000–2002. Revista Catalana d’Ornitologia, 20: 28–115. Baucells, J., Camprodon, J., Cerdeira, J. & Vila, P., 2003. Guía de las cajas nido y comederos para aves y otros vertebrados. Lynx Edicions, Barcelona. Baucells, J., 2010. Els rapinyaires nocturns de Catalunya. Biologia, gestió i conservación de les vuit espècies de rapinyaires nocturns catalans i els seus hàbitats. I. G. Sta. Eulàlia–Santa Eulàlia de Ronçana, Barcelona. Blas García, J. & Muñoz Gallego, A. R., 2003. Atlas de las aves reproductoras de España: 318–319 (R. Martí & J. C. Del Moral, Eds.). Madrid, Ministerio de Medio Ambiente y SEO/BirdLife. Calvet, J., Estrada, J., Mañosa, S., Moncasí, F., Solans, J. (Ed.), 2004. Els ocells de la plana de Lleida. Pagès Editors, Lleida. Catalán, B., Saurí, D. & Serra, P., 2008. Urban sprawl in the Mediterranean? Patterns of growth and change in the Barcelona Metropolitan Region 1993–2000. Landscape and Urban Planning, 85: 174–184. Clavell, J., 2002. Catàleg dels ocells dels Països Catalans. Lynx Edicions, Barcelona. CREAF, 2009. Mapa de Cobertes del Sòl de Catalunya (2005/2007). Centre de Recerca Ecològica i Aplicacions Forestals. Donald, P. F., Green, R. E. & Heath, M. F., 2001. Agricultural intensification and the collapse of Europe’s farmland bird populations. Proceedings of the Royal Society of London Series B–Biological Sciences, 268: 25–29. Donald, P. F., Pisano, G., Rayment, M. D. & Pain, D. J., 2002. The Common Agricultural Policy, EU enlargement and the conservation of Europe’s farmland birds. Agriculture Ecosystems & Environment, 89: 167–182.


Animal Biodiversity and Conservation 34.2 (2011)

Finck, P., 1990. Seasonal variation of territory size with Little Owl (Athene noctua). Oecologia, 83: 68–75. Framis, H., 2004. Mussol comú (Athene noctua). In: Atles dels ocells nidificants de Catalunya 1999–2002: 290–291 (J. Estrada, V. Pedrocchi, L. Brotons & S. Herrando, Eds Institut Català d’Ornitologia (ICO)/Lynx Edicions, Barcelona. – 2008. Approach to the distribution and abundance of little owl (Athene noctua) in periurban areas and arable land around the city of Mataró (Barcelona). L’Atzavara, 17: 43–50. Grzywaczewski, G., 2009. Home range size and habitat use of the Little Owl Athene noctua in East Poland. Ardea, 97: 541–545. Hardouin, L. A., Tabel, P. & Bretagnolle, V., 2006. Neighbour–stranger discrimination in the little owl, Athene noctua. Animal Behaviour, 72: 105–112. Hebblewhite, M. & Hidon, D. T., 2010. Distinguishing technology from biology: a critical review of the use of GPS telemetry data in ecology. Philosophical Transactions of the Royal Society B, 365: 2303–2312. Herrando, S., 2008. The state of farmland birds in Catalonia and the specific case of Les Cinc Senies. L’Atzavara, 17: 37–41. Hounsome, T., O’Mahony, D. & Delahay, R., 2004. The diet of Little Owls Athene noctua in Gloucestershire, England. Bird Study, 51: 282–284. ICO, 2003. Catalan Ornithological Institute ringing standards (1/03 edition). Barcelona. Ibáñez, J. J. & Burriel, J. A., 2008. Mapa de cubiertas del suelo de Catalunya: metodología de la tercera edición. In: Tecnologías de la Información Geográfica para el Desarrollo Territorial: 70–75 (L. Hernández & J. M. Parreño, Eds.). Servicio de Publicaciones y Difusión Científica de la ULPGC, Las Palmas de Gran Canaria. Kenward, R. E., 1987. Wildlife Radio Tagging. Equipment, Field Techniques and Data Analysis. London: Academic Press. – 2001. A manual for wildlife radio tagging. Academic Press, London, San Diego. Krebs, J. R., Wilson, J. D., Bradbury, R. B. & Siriwardena, G. M., 1999. The second silent spring? Nature, 400: 611–612. Martí, R. & Del Moral, J. C. (Eds.) 2003. Atlas de las aves reproductoras de España. Direción General de Conservación de la Naturaleza–Sociedad Española de Ornitología. Madrid. Martínez, J. A., Zuberogoitia, I. & Alonso, R., 2002. Rapaces Nocturnas. Guía para la determinación de la edad y el sexo en las Estrigiformes ibéricas. Monticola Ed., Madrid. Montasell, J., 2006. Els espais agraris de la regió metropolitana de Barcelona. L’Atzavara, 14: 73–89. – 2010. Els espais agraris. In: Els rapinyaires nocturns de Catalunya. Biologia, gestió i conservación de les vuit espècies de rapinyaires nocturns catalans i els seus hàbitats: 163–178 (Baucells, J., Ed.). I. G. Sta. Eulàlia, Santa Eulàlia de Ronçana, Barcelona. Muntaner, J., Ferrer, X & Martínez–Vilalta, A., 1983. Atlas dels ocells nidificants de Catalunya i Andorra.

377

Ketres, Barcelona. Onrubia, A. & Andrés, T., 2005. Impact of human activities on steppic–land birds: a review in the context of the western Palearctic. In: Ecology and conservation of steppe–land birds: 185–209 (G. Bota, M. B. Morales, S. Mañosa & J. Camprodon, Eds.). Lynx Edicions & Centre Tecnològic Forestal de Catalunya, Barcelona. Pons, X., 2004. Miramon. Sistema d’Informació Geogràfica i software de teledetecció. Centre de Recerca Ecològica i Aplicacions Forestals–CREAF, Bellaterra. Sabater, F., Benaiges, N. & Valls, I., 1997. La recent transformació del paisatge a la conca de la riera d’Argentona: Anàlisi de l’evolució del paisatge dècada a dècada des de l’any 1967 fins el 1994. L’Atzavara, 7: 29–37. Sabater, F., Corbera, J., Valls, I., Benaiges, N., Guardiola, M., Basagaña, J., Guardiola, M., Marfà, V., Gallés, A., Buscà, Ll., Comerma, M., Campeny, R., Parera, J. M. & Triadó, S., 2008. Els valors dels espais agraris periurbans: el cas de les Cinc Sènies–Mata–Valldeix de Mataró. L’Atzavara, 17: 51–60. Sanderson, F. J., Donald, P. F. & Burfield, I. J., 2005. Farmland birds in Europe: from policy change to population decline and back again. In Ecology and conservation of steppe–land birds: 211–236 (G. Bota, M. B. Morales, S. Mañosa & J. Camprodon, Eds.). Lynx Edicions & Centre Tecnològic Forestal de Catalunya, Barcelona. Sissons, R. A., 2003. Food and Habitat Selection of Male Burrowing Owls (Athene cunicularia) on Southern Alberta Grasslands. Master’s of Science in Wildlife Ecology Thesis, Dept. of Renewable Resources, Univ. of Alberta, Canada. South, A. B., Kenward, R. E. &, Walls, S. S., 2005. Ranges7 v1.0: For the analysis of tracking and location data. Online manual. Anatrack Ltd. Wareham, UK. Sunde, P., Thorup, K., Jacobsen, L. B., Holsegard– Rasmussen, M. H., Ottessen, N., Svenne, S. & Rahbek, C., 2009. Spatial behaviour of little owls (Athene noctua) in a declining low–density population in Denmark. Journal of Ornithology, 150: 537–548. Tomé, R., Dias, M. P., Chumbinho, A. C. & Bloise, C., 2011. Influence of perch height and vegetation structure on the foraging behaviour of Little Owls Athene noctua: how to achieve the same success in two distinct habitats. Ardea, 99(1): 17–26. Van Nieuwenhuyse, D., Génot, J.–C. & Johnson, D. H., 2008. The little owl: conservation, ecology and behavior of Athene noctua. Cambridge University Press, U.K. White, G. C. & Garrott, R. A., 1990. Analysis of Wildlife Radio–tracking Data. Academic Press, New York, USA. Zabala, J., Zuberogoitia, I., Martínez–Climent, J. A., Martínez, J. E., Azkona, A., Hidalgo, S. & Iraeta, A., 2006. Occupancy and abundance of Little Owl Athene noctua in an intensively managed forest area in Biscay. Ornis Fennica, 83: 97–107.


378

Zuberogoitia, I., Zabala, J., Martínez, J. A., Hidalgo, S., Martínez, J. E., Azkona, A. & Castillo, I., 2007. Seasonal dynamics in social behaviour and spacing patterns of the Little Owl Athene noctua. Ornis Fennica, 84: 173–180.

Framis et al.

Zuberogoitia, I., Martínez, J. E., Zabala, J., Martínez, J. A., Azkona, A., Castillo, I. & Hidalgo, S., 2008. Social interactions between two owl species sometimes associated with intraguild predation. Ardea, 96(1): 109–113.


Animal Biodiversity and Conservation 34.2 (2011)

379

Tracking movements of Athene owls: the application of North American experiences to Europe G. L. Holroyd & H. E. Trefry

Holroyd, G. L. & Trefry, H. E., 2011. Tracking movements of Athene owls: the application of North American experiences to Europe. Animal Biodiversity and Conservation, 34.2: 379–387. Abstract Tracking movements of Athene owls: the application of North American experiences to Europe.— Migration and dispersal are important ecological and evolutionary processes and understanding them is a requirement for species conservation efforts. Burrowing owl, Athene cunicularia, the North American equivalent of little owl, A. noctua, is migratory in the northern parts of its range. In Canada their populations have declined dramatically and are classified as endangered. Movements of burrowing owls have been studied using banding (ringing), VHF telemetry, stable isotopes, genetics (DNA), geolocators and satellite transmitters. Geolocators and satellite transmitters provide the most reliable information about migrations but to operate successfully they are both dependent upon exposure to sunlight, which can be limited for nocturnal owls. Ringing encounters and winter influxes of little owls into Spain, including the Balearic Islands, indicate that some migration movement may be occurring. A stable isotope study could determine if wintering owls in southern Europe includes owls originating in northern Europe. Key words: Athene, Movements, Migration, Dispersal, Techniques. Resumen Seguimiento de los desplazamiento de los mochuelos del género Athene: aplicación de las experiencias norteamericanas a Europa.— Migración y dispersión son procesos importantes desde el punto de vista de la ecología y la evolución, y entenderlos es un requisito importante para los programas de conservación de las especies. El mochuelo de madriguera, Athene cunicularia, el equivalente norteamericano del mochuelo europeo, A. noctua, es migratorio en las zonas septentrionales de su área de acción. En Canadá sus poblaciones han disminuido de forma notoria, y se han clasificado como amenazadas. Se han estudiado los desplazamientos del mochuelo de madriguera utilizando el anillado, la telemetría VHF, los isótopos estables, la genética (ADN), los geolocalizadores y los transmisores por satélite. Los geolocalizadores y los transmisores por satélite proporcionan la información más fiable sobre las migraciones, pero su buen funcionamiento depende de la exposición a la luz solar, que es limitada en el caso de las rapaces nocturnas. Los hallazgos de animales anillados, y los flujos migratorios invernales del mochuelo común hacia España, incluyendo las Islas Baleares, indican que pueden estarse dando desplazamientos migratorios. Un estudio mediante isótopos estables podría determinar si entre los mochuelos que invernan en el sur de Europa, se incluyen mochuelos del norte de Europa. Palabras clave: Athene, Desplazamientos, Migración, Dispersión, Técnicas. (Recieved: 25 IV 11; Conditional acceptance: 2 IX 11; Final acceptance: 10 X 11) G. L. Holroyd & H. E. Trefry, Environment Canada, Room 200, 4999–98 Ave, Edmonton, Alberta, T6B 2X3, Canada. Corresponding author: G. L. Holroyd. E–mail: geoffrey.holroyd@ec.gc.ca

ISSN: 1578–665X

© 2011 Museu de Ciències Naturals de Barcelona


380

Introduction Migration and dispersal are important ecological and evolutionary processes and understanding them is a requirement for species conservation efforts (Faaborg et al., 2010). Additionally, species specific information on winter range, winter habitat use and survival are needed to fully develop and implement conservation action (Terborgh, 1989; Stutchbury, 2007). Despite the importance of movement issues on the conservation of avian species, knowledge about avian migration and dispersal is still patchy (Faaborg et al., 2010). Burrowing owl (Athene cunicularia) is the Western Hemisphere equivalent of little owl (A. noctua). The western burrowing owl (A. cunicularia hypugaea), which occurs in western North America, is totally migratory in the northern parts of its range and partially in the southern parts (Haug et al., 1993). In Europe, most researchers treat the little owl as a non–migrant species (Van Nieuwenhuyse et al., 2008), but some records indicate this premise should be examined more closely. Clavell (2002) documented 23 records of little owl on the Balearic Islands in the western Mediterranean Sea in autumn, which is consistent with a southward migration. Ferrer et al. (1994) stated the little owl was accidental on these islands and that one owl ringed in Germany was encountered on Menorca. Thus, the possibility that little owls are migratory in at least portions of its range should not be dismissed. These records are similar to autumn and winter ringing records of burrowing owls from Canada that indicated where this migratory species was wintering. Migration movements may be more pronounced in certain age or sex groups. Ogonowski & Conway (2009) found the probability of burrowing owls in Arizona migrating decreased with age and was less likely for males than females. Also the migratory tendency could change from one year to the next. Local natal and breeding dispersal has been demonstrated in little owls (Van Nieuwenhuyse et al., 2008). If little owls are found to migrate longer distances, then evidence of longer range dispersal should also be considered. Dispersal of one–year old owls, natal dispersal, is particularly likely since these birds are setting up nests for the first time. Breeding dispersal, adult movements between years, may be less likely since adults have established a nest once and may have become resident at that site. However, short–distance breeding dispersal has been documented in non–migrant populations of burrowing owls in California (Rosier et al., 2006) and long distance breeding dispersal has been documented in Canadian burrowing owls (Duxbury, 2004; Holroyd et al., 2011). In Canada, burrowing owls are migratory because of the severe weather conditions across their range in winter. The Canadian population has declined dramatically and the species is classified as Endangered (Wiggins, 2006). The prairie population in Canada declined 95% in the 1990s (Holroyd et al., 2001) and has remained below 5% of the 1990 numbers since then. In the past twenty years, studies in Canada have focused on productivity, diet, and foraging behaviours in an effort to determine limit-

Holroyd & Trefry

ing factors that could be implicated in the declines (Wellicome, 2000; Todd, 2001; Sissons, 2003; Shyry, 2005). However, mortality away from the breeding season (Terborgh, 1980) and inter–year dispersal could also be implicated in the declines (Macdonald & Johnson, 2001). Thus, studies of burrowing owls after they depart the breeding grounds were undertaken. Until 1990, the winter destinations of burrowing owls from Canada were unknown. Ringing recoveries on migration indicated they were headed south–easterly across the Great Plains of the USA (James, 1992). In order to undertake studies of burrowing owls’ winter survival and ecology the movements and destinations of owls after the breeding season had to be determined. This article reviews the techniques used in Canada to track migrant and dispersing burrowing owls, applications that could be useful to determine if little owls are migratory. Movements of burrowing owls have been studied using ringing (James, 1992; Holroyd et al., 2011), VHF telemetry (Holroyd et al., 2010), stable isotopes (Duxbury, 2004), genetics (DNA; Macias–Duarte, 2011), geolocators and satellite transmitters. This article reviews the results of these various techniques, their logistical issues and expenses, and their possible applications to study little owl movements. The information regarding the geolocators and satellite transmitters is published here for the first time. Ringing The attachment of metal rings on the legs of birds has a long history and the resulting encounters have been analyzed for many species. Encounters are defined by the North American Bird Banding Office as any handling of a banded bird, dead or alive subsequent to the initial banding (http://www.pwrc.usgs.gov/bbl/ manual/glossary.cfm). As of the end of 2009, a total of 38,242 burrowing owls had been ringed in North America, of which 11,611 were ringed in Canada (L. Laurin, Canadian Bird Banding Office). Reported encounters of ringed owls have been few (Holroyd et al., 2010), but the number of encounters has increased in recent years due to three factors: extensive use of alpha–numeric coloured, anodized aluminum, rivet rings (Acraft Sign and Nameplate Co., Edmonton, AB, Canada), digital photography, which enables the reading of the alpha–numeric codes of color rings, and internet communication. Until recently, there was not an easy way for Mexicans to report rings but the inclusion of Mexico in a 1–800 free call reporting system and the wider use of the internet may help. By 1987 only three encounters with burrowing owls ringed in prairie Canada had been recorded in the USA and none in Mexico, and none in winter; the encounters were in spring and autumn and all in the US Great Plains (James, 1992). At that time, 1.0% of 2,512 owls ringed in the US had been encountered between November–February, i.e. winter. By 2003, some 20,597 burrowing owls had been ringed in Canada, US and Mexico but they resulted in only


Animal Biodiversity and Conservation 34.2 (2011)

260 (1.3%) encounters (Harman & Barclay, 2007). Of these encounters, 102 were in California (2.2% of 4,708 ringed). Thus only 158 encounters occurred from 15,889 owls ringed in US and Canada outside California for a return rate of 0.99%. The low encounter rate of ringed burrowing owls provided limited information on their seasonal movements. The lack of winter recoveries led James (1992) to hypothesize that Canadian owls were exhibiting leap–frog migration over US wintering grounds to winter in Mexico. By 2008, 10 ring encounters from the Great Plains of Canada had defined a migration route south into Texas but still none in Mexico (Holroyd et al., 2010). Sixteen encounters from British Columbia showed a western migration route into the three western US states (ibid). Only six interstate movements had been documented in California by 2003 (Harman & Barclay, 2007). Despite a major effort to ring over 20,000 owls, little was known about their wintering grounds. Ring encounters provided some clues regarding natal and breeding dispersal. In 2002, encounter data from the US Bird Banding Office provided records of 43 summer encounters of burrowing owls. Of these, nine were restricted to the non–migratory Florida population, and 23 were in the same year as ringing. The 11 remaining summer encounters were one to three years after ringing with an average movement of 127 km. If these owls were breeding at each site, this distance provided a first estimate of inter–annual dispersal. This estimate does not include zero dispersal distances since many researchers do not report owls returning to their ringing site to the Bird Banding Office. In 2003, a unique burrowing owl ring encounter provided an unusual breeding dispersal record. A female with a vascularized brood patch, a mate and one young ringed in Tucson, Arizona, USA on 30 April 2003 was encountered 1,860 km north in southern Saskatchewan, Canada, in July of the same year with a new male and seven nestlings (Holroyd et al., 2011). Two intensive studies, the use of colour rings and an observant biologist led to this record long–distance nesting dispersal. Ring encounters provide additional information, such as causes of death. For example, in California, 7.4% were caught due to an injury, 4.6% were shot, 4.6% were found dead in buildings, 2.8% were hit by vehicles, and 1.9% were hit by aircraft, but 48% were found dead due to unknown causes (Harman & Barclay, 2007). These causes of death are biased due to the likelihood of the public finding an owl. Any owl that dies underground will not be found and if they are eaten by a raptor they are less likely to be found than if they hit a building or aircraft. The known causes of death cannot therefore be accepted to represent the true mortality factors. In conclusion, ringing of owls in Canada failed to identify the owls’ wintering grounds, gave a biased view of the causes of death, and provided very limited information on dispersal. The ringing of over 38,000 owls did provide some information on migration routes from Canada, and also on wintering locations of British Columbia and USA owls (Holroyd et al., 2010).

381

VHF telemetry VHF telemetry was used to locate wintering burrowing owls that had transmitters attached in Canada the previous summer (Holroyd et al., 2010). The transmitters were initially used for 2–month foraging studies of burrowing owls in Canada (Todd, 2001; Sissons, 2003; Shyry, 2005). With slightly larger batteries the transmitters were manufactured to last 6–9 months (Holohil Systems Ltd., Newmarket, Ontario, Canada). A fixed wing aircraft flew for a total of 167.7 hours over southern Texas and north– eastern and central Mexico during three winters (1997–2000), resulting in nine of 125 transmitters being located. Three of these encounters were in Mexico, the first indication that burrowing owls from Canada wintered south of the US (see Holroyd et al., 2010 for more details). These nine records were effectively ring encounters but they had the added advantage that the owls were located while they were still present at their wintering sites. It was therefore possible to gather specific information about habitat and winter movements of the owls (Holroyd et al., 2010). The habitat information indicated burrowing owls utilize a much broader range of habitats in winter than they do in summer. They wintered in shrubland with access to grassland or cropland where they forage; they were not limited to open grassland as they are in Canada in summer. Indeed, the telemetry offered an unbiased search method that explored all habitat types in an area. This aerial telemetry provided valuable insights into the distribution and habitats of burrowing owls from Canada in winter. However, it was costly and the encounter rate was low, likely hindered by the nocturnal, burrow dwelling habits of the species. Genetics (DNA) The analysis of the genetic structure of burrowing owl populations in western North America has led to the conclusion that the populations are panmictic, with no structure to the populations’ genetics (Korfanta et al., 2005; Desmond et al., 2001). Canadian samples were included in a broader and more intensive analysis of burrowing owl genetics that reached a similar conclusion (Macias–Duarte, 2011). This new study confirmed that western burrowing owls were genetically similar. However, the apparently isolated population around Mexico City was genetically distinct, and confirmed that the population on Isla Clarion, 700 km west of the tip of Baja California Sur, Mexico in the Pacific Ocean had been isolated for about 200,000 years (Macias– Duarte, 2011). Although genetics will not identify current movements of owls, it is a powerful tool to interpret past movements and isolation of burrowing owls. All three studies have shown there is no genetic isolation in the western burrowing owl from northern Mexico, through the western USA to south–western Canada. Genetic markers have been used in other species to show migratory pathways (see Boulet & Norris, 2006 review), a technique which may apply to little owls.


382

Stable isotopes Stable isotopes analysis (SIA) of feathers has been used to track bird movements (Chamberlain et al., 1997); both migration and dispersal (Hobson & Wassenaar, 2008). This aspect of SIA determines the location where feathers were grown. Determining movements with SIA has several advantages but also limitations. The approximate origin of every feather collected can be determined through this technique. Tail feathers were collected from burrowing owls trapped during winter and summer studies. The stable isotope ratios of H, C and N in the feathers were compared to a map of stable isotope ratios in nestling feathers (Duxbury, 2004). Feathers collected in winter provided an approximate location where the owls grew the tail feather the previous summer. Feathers collected in the breeding season before molt provided an approximate location where the owls grew the feathers the previous breeding season, thus an indication of breeding and natal dispersal. The technique is most easily interpreted with feathers since they are a stable tissue but an often unproven assumption is that adult owls grow feathers where they breed, which may not always be true. The moult cycle of the bird must be fully understood before deciding which feathers should be used. One large advantage of SIA is that the moult location for each owl handled can be determined through its feathers. The disadvantages are that the locations are approximate with errors of ± 125 km (Duxbury, 2004), and recent studies have demonstrated further factors that introduce error (Meehan et al., 2003; Smith et al., 2009). Despite these issues, SIA provided the origin of owls trapped in the winter in Texas and Mexico (Duxbury, 2004). Seven of 105 owl feather samples collected in winter in central Mexico were grown in Canada the previous summer. The majority of feather samples showed that most owls wintering in south Texas and central Mexico were short distant migrants from northern Mexico and southern USA. In addition, SIA was used to study breeding dispersal by obtaining estimations of where breeding owls had grown their feathers the previous season (Duxbury, 2004). Feather samples collected in summer from breeding owls in Canada showed that burrowing owls relocated an average 400 km between breeding seasons. Over half of the breeding owls in prairie Canada each year were in the USA or Mexico the previous year. Duxbury (2004) estimated that the net loss of owls to the northern US Great Plains approximately equaled the 20% decline in the owl population from prairie Canada. The high rate of dispersal had a marked impact on the population dynamics of owls in prairie Canada. Geolocators The advent of small geolocators or light–level data loggers provides new opportunities to learn more about small species’ migrations (Stutchbury et al., 2009). Geolocators provide information about the

Holroyd & Trefry

departure dates, stop–overs, migration route and arrival at wintering grounds, and the units are small enough to be carried by small birds. Fourteen ~7 gm geolocators were attached to adult breeding burrowing owls in July 2005 at the Onefour Agricultural Research Sub–Station in extreme south–eastern Alberta, Canada (49.04oN 110.38oW). The geolocators determined the light levels every minute and recorded the time, GMT. One unit was retrieved from an owl in June 2006. The fate of the other 13 owls is unknown. The unit operated from August 1to December 25 in 2005. The movements of the owl were interpreted from dawn and dusk determined from the unit’s records of light levels and time. Each 4' change in dawn and dusk from GMT = 1 degree of longitude. Day length was used to determine latitude with the aid of the US Navy website (http://aa.usno.navy.mil/data/docs/ RS_OneYear.php). A geolocator that remained at the capture site at Onefour was used to determine that the combined error in estimation of dawn, dusk and day length resulted in errors of 55.5 km (SD ± 83.9) longitude and 25.0 km (SD ± 194.8) latitude in August. These error estimates are within the blanket error ± 300 km quoted in Stutchbury et al. (2009). The owl followed the migration route shown from ring encounters in the Great Plains to Texas (Holroyd et al., 2010), before turning west into central Mexico (fig. 1). The owl appears to have remained in the high plains until it reached northern Texas when it went to the coastal Gulf lowlands, then into the bajio of central Mexico. When the geolocator quit on December 25, the owl may not have completed its migration since it moved in the weeks prior to this date, being stationary in central Mexico for only about 10 days. However, VHF telemetry of burrowing owls indicated that owls do remain on a wintering location starting in December and do not wander throughout the winter (unpubl. data). The winter location was near where other owls from Canada have been tracked (Holroyd et al., 2010). Previously, the length of time that burrowing owls spend on migration was estimated from known arrival and departure dates. The new information provided by this geolocator was the dates of start and end of migration, a minimum length of the migration of at least 3 months, and the dates and location of two staging sites. A significant disadvantage is that only one of 14 geolocators was recovered. Units must be recovered before any data can be extracted since they do not transmit any data. Thus any owls that dispersed to breed outside a study area are not likely to be detected and their movements not recorded. The return rate of the male owls with data–loggers was lower than the anticipated return rate for ringed adult male owls. Other authors report low recovery rates (e.g. 10% for purple martin, Stutchbury et al.,, 2009; 13% female, but 66% of male wood thrush, Stutchbury et al., 2010). In addition, the locations have a large error estimate as shown by the irregular directions of the apparent movement. The line was not ‘smoothed’ as done by Stutchbury et al. (2009, 2010) since smoothing masks how variable the apparent movement can be without manipulation.


Animal Biodiversity and Conservation 34.2 (2011)

British Columbia

383

Manitoba

Alberta

Saskatchewan Washington North Dakota Oregon

Nebraska Colorado

California

Oklahoma

Texas

Louisiana

N W

E S

Satellite transmitter Geolocator Michoacan

Fall & spring Winter 0

250

500

1,000

1,500

2,000 km

Fig. 1. The migration of burrowing owls from Canada as indicated by a geolocator and satellite transmitter superimposed on ring recoveries and VHF telemetry (Holroyd et al., 2010). Fig. 1. Migraciones del mochuelo de madriguera desde el Canadá, tal como indican los estudios con geolocalizadores y transmisores por satélite, superpuestos a la recuperación de animales anillados y a la telemetría VHF (Holroyd et al., 2010).

Satellite transmitters Satellite transmitters (PTTs) have been used to track larger birds for two decades (e.g. Fuller et al., 1998; Kochert et al., 2011). In January 2010, Microwave Telemetry Inc (Maryland, USA) introduced a 5 gm

solar satellite transmitter. The transmitters send signals on a predetermined duty cycle that are received by polar orbiting NOAA/ARGOS satellites which triangulate the location of the transmitter using Doppler shift algorithms of the Argos satellite positioning system.


384

In 2010, 5–5 gm PTTs were attached to adult breeding burrowing owls in southern Alberta and Saskatchewan using Teflon back–pack harnesses. The project’s objective were to study the owls’ migration routes and timing, and to determine wintering areas, non–breeding habitats and breeding dispersal. These transmitters operated with a duty cycle of 8 hours on, 48 hours off. Accuracy varied from a minimum error of 150 m to over 1 km, depending upon the ARGOS satellites’ view of the transmitter. The signal from one PTT on a female owl ended in August at the nest site, with evidence (a freshly eaten skull and digging) suggesting the female died as a result of predation by a swift fox, Vulpes velox. Three PTTs, on one female and two male owls, stopped sending signals in September. The voltage of the units declined prior to the end of messages, presumably from declining light levels and possibly due to owls spending more time in burrows after the completion of the post–fledging foraging period. One of these three PTTs was recovered from a male in July 2011 near its original capture site. The PTT was not working, possibly due to damage to the casing. The second PTT, on a female, started transmitting the following summer on July 4, 2011 after being silent all winter. A site visit confirmed she was nesting with at least four young about 55 km east of the original capture site. The third of three PTTs has not been reported since it went silent in September 2010. The fifth transmitter is described below. Thus of five transmitters, one owl was depredated, one operated for a year, and three quit after two months but one was recovered, one restarted transmissions in the follow summer, and the fate of the other was unknown. The fifth PTT operated through the autumn and winter of 2010 (fig. 1). It was attached to a female burrowing owl with young at the Onefour Agricultural Research Sub–Station in extreme south–eastern Alberta, Canada (49.04o N 110.38o W) on 24 June, 2010. She stayed in the vicinity of her nest through July. After the young fledged, she moved 5 km south to Montana, spending two months in cultivated fields adjacent to native prairie. Since micro–PTTs, unlike geolocators, provide locations in real time, researchers could visit her current pre–migration roosting and foraging sites. On October 21 she was in north–eastern New Mexico, 1,400 km from her Montana roost. Six days later, she was 470 km further south in south–east New Mexico. The next transmission on November 10 was 1,046 km west on the Baja California peninsula, Mexico (fig. 1). This burrowing owl followed a new migration route previously not described. While ringing recoveries from prairie Canada have been between the Great Plains and Texas, this owl followed the foothills of the Rockies before heading west to a previously unknown wintering location. The owl spent the winter on the west coast of Baja California, Mexico, where researchers were able to study her habitat due to the transmission of current PTT signals. Another major advantage of PTTs is that breeding and natal dispersal can be determined if they operate for a full year, identifying two sequential

Holroyd & Trefry

breeding sites. While breeding and natal dispersal records can result from ringing, they are rare for burrowing owls, as discussed above. Meyburg et al. (2011) were able to track a Eurasian hobby (Falco subbuteo) for two years and determine sequential breeding efforts; similar data have been collected for up to three breeding seasons for peregrine falcons, Falco peregrinus, (Holroyd & Trefry, unpubl. data). The locations can be geo–referenced with habitat data to determine the types of habitats used post–breeding, on migrations and in winter. Satellite transmitters are therefore the only option if the goal is to follow birds over more than one season. Like the geolocators, the small PTTs also require light to recharge the battery, and this can be problematic due to the lifestyle of the burrowing owl and feather coverage. However, techniques to overcome these problems, such as back pads to reduce feather coverage and programming the PTTs to receive less frequent signals, are options being tested. Discussion The combined studies of burrowing owls over the past 20 years in Canada provide data showing they winter in south Texas and across the full width of central Mexico from Veracruz to Baja California Sur. This leads to the conclusion that Canadian burrowing owls have weak or no migratory connectivity with specific wintering areas (as illustrated in figure 1 in Boulet & Norris, 2006). While the specific proportions of Canadian owls that migrate to various parts of the winter range are unknown, determining the exact proportions would be very expensive and time consuming with questionable conservation benefits. The wide geographic range of wintering areas and habitats also makes their protection a trinational conservation challenge (Holroyd, 2005) All of the techniques reviewed here have advantages and disadvantages that include the type of data collected, cost, and delays in accessing the data (table 1). A few general statements can be made about the data collected. Knowledge of burrowing owl migration and dispersal was very limited from ringing encounters. Recovery data can be geographically biased if reporting of rings is much more likely to occur in some areas. For example, ring recoveries from Texas and coastal California are more likely due to denser human populations, public education about rings, and access to easy methods of reporting. On the other hand, Mexico lacked a method of reporting rings easily until recently, and phone and mail access are still limited in rural areas. This led to the assumption that Canadian owls were going to Mexico. This initial analysis was in fact incorrect in that burrowing owls do not appear to demonstrate leap–frog migration as hypothesized by James (1992). Rather, aerial telemetry and stable isotope analysis indicate that Canadian owls winter from south Texas into central México mixed with owls from lower latitudes. Newer models of geolocators appear more reliable and should function throughout an entire year, unlike the 2005 prototype reported here. However, stable isotope


Animal Biodiversity and Conservation 34.2 (2011)

385

Table 1. Advantages and disadvantages of techniques used by the authors to track burrowing owls. Tabla 1. Ventajas y desventajas de las tĂŠcnicas utilizadas por los autores, para el seguimiento de los mochuelos de madriguera. Technique Numbered rings Advantages 1. 2. Disadvantages 1. 2. 3. Colour rings Advantages 1. 2. Disadvantages 1. 2. Genetics Advantages 1. 2.

Relative cost 1 Inexpensive Exact location at start and end of movement known Low return rate Route and timing of movement unknown Delay in accessing data 2 Relatively inexpensive Identification possible with optical equipment Low return rate Route and timing of movement unknown 3 Relatively inexpensive Provides historical evidence of movements

Disadvantages 1. No precision of movements because of lack of genetic differentiation of most populations Geolocators 5 Advantages 1. Provides data on timing and location of seasonal migration 2. Light weight (~1.5 gm) Disadvantages 1. Low return rate 2. Bird must be recaptured to retrieve geolocator 3. Natal and breeding dispersals not recorded. Imprecise (~250km) 4. Delay in receiving locations. Aerial radio transmitters 6 Advantages 1. Provides real time location Disadvantages 1. Relatively expensive 2. Owls must be above ground to detect transmitter Stable isotopes 4 Advantages 1. Applicable to all trapped birds 2. Multiple isotopes can increase accuracy estimate 3. Relatively inexpensive Disadvantages 1. Mostly limited to feathers 2. Route and timing of movement unknown 3. Location of molt may differ from breeding site 4. Imprecise (~250 km) 5. Isoscape map required for each element 6. Requires clear view of sky at dawn and dusk Satellite telemetry 7 Advantages Disadvantages

1. Provides relatively exact data on timing and location of seasonal movements and dispersal 2. Provides real time location 1. Most costly 2. Weight (6 gm with harness) at maximum for 150 gm owl 3. Relatively long antennae 4. Requires clear view of sky most of the time


386

analysis predicts high dispersal in this species and ring returns for adult males are 30%, so the probability of retrieving the units is low. The actual recovery rate of the units (1/14 = 7%) indicates caution and geolocators are not recommended for use with burrowing owls. Also, the errors associated with the units are large, and the owl’s behavior of occupying burrows at dawn and dusk resulted in many days with no estimate of dawn and dusk. The satellite transmitter provided the highest quality data on movements of one owl, and it is the only technology that did so in ‘real’ time, such that the owl’s staging areas in Montana and wintering area in Mexico were visited while the owl was present. PTT data could be used to estimate breeding, staging and wintering home ranges, although with some error due to the estimates of location based on Doppler shift algorithms. The applicability of these techniques to determine movements of little owls is of course dependent on the scale and goals of the study and available funds. The encounters of rings have led many authors to conclude that the little owl is sedentary (Van Nieuwenhuyse et al., 2008). However, one ringing encounter from Germany to Menorca and influxes of owls in winter into the Balearic Islands and Spain indicate that some migration–like movement is undertaken by some little owls. In addition, little owls that experience more rigorous winters, such as in Kazakhstan, are nomadic or migratory (Gavrin, 1962 in Van Nieuwenhuyse et al., 2008: p. 302). Long distance breeding dispersal events have been documented with rings up to 600 km (reviewed in Van Nieuwenhuyse et al., 2008). The possibility that migration in little owls varies with age and sex, as occurs in burrowing owls (Ogonowski & Conway, 2009), should also be considered. Such dispersal has important implications for population dynamics and viability (Macdonald & Johnson, 2001). The extent of migration and dispersal of little owls within Europe should therefore be explored. The little owl appears to spend more daytime hours in the dark, potentially limiting the usefulness of geolocators and solar satellite transmitters, but only a trial study would determine this. Stable isotope analysis would show large scale movements of little owls if they were sampled at the extremes of their range, and particularly on the Iberian Peninsula in winter. The moult is well documented for the species, an important consideration when determining which feathers to collect. Since the species retains some secondaries, after–second–year birds could provide feathers with data for more than one year. The variation in stable isotopes from northern to southern Europe for H and N are not as great as in North America (Bowen & West, 2008), but should be sufficient to show if owls are migrants from northern into southern Europe. Based on the experiences tracking migrating burrowing owls, movements of little owls using stable isotope analysis should be explored in Europe. Acknowledgements Many people contributed to our studies of movements of burrowing owls. We thank them all and wish that we

Holroyd & Trefry

could list all their names. Our studies in Mexico benefited from the cooperation with E. Valdez–Gomez and R. and V. Peacock. We thank H. Framis for assistance with literature searches and editing, and G. Turney for producing the map. The above studies were funded by Environment Canada, US National Fish and Wildlife Foundation, World Wildlife Fund Canada, Beaverhill Bird Observatory and Alberta Sustainable Resource Development. Burrowing owls were trapped under federal, provincial and Parks Canada Agency banding and research permits, and approval by Animal Care Committee, Prairie and Northern Region, Environment Canada, Edmonton, Alberta, Canada. This article benefited from the comments by D. Pavón Jordán and I. Zuberogoitia. References Bowen, G. J. & West, J. B., 2008. Isotope Landscapes for terrestrial migration research. In: Tracking animal migration with stable isotopes: 79–105 (K. A. Hobson & L. I. Wassenaar, Eds.). Elsevier Inc., Oxford, United Kingdom. Boulet, M. & Norris, D. R., 2006. The past and present of migratory connectivity. Ornithological Monographs, 61: 1–13. Chamberlain, C. P., Blum, J. D., Holmes, R. T., Feng, X. H., Sherry, T. W. & Graves, G. R., 1997. The use of isotope tracers for identifying populations of migratory birds. Oecologia, 109: 132–141. Clavell, J., 2002. Cataleg dels ocells dels Paisos Catalans. Lynx Edicions, Barcelona, Spain. Desmond, M. J., Parsons, T., Powers, T. O. & Savidge, J. A., 2001. An initial examination of mitochondrial DNA structure in burrowing owl populations. Journal of Raptor Research, 35: 274–281. Duxbury, J. M., 2004. Stable isotope analysis and the investigation of the migrations and dispersal of peregrine falcons (Falco peregrinus) and burrowing owls (Athene cunicularia hypugaea). Ph. D. Thesis, University of Alberta, Edmonton, Canada. Faaborg, J., Holmes, R. T., Anders, A. D., Bildstein, K. L., Dugger, K. M., Gauthreaux, S. A. Jr., Heglund, P., Hobson, K. A., Jahn, A. E., Johnson, D. H., Latta, S. C., Levey, D. J., Marra, P. P., Merkord, C. L., Nol, E., Rothstein, S. I., Sherry, T. W., Sillett, T. S., Thompson, F. R. III & Warnock N., 2010. Recent advances in understanding migration systems of New World land birds. Ecological Monographs, 80: 3–48. Ferrer, X., Martínez i Vilalta, A. & Muntaner, J., 1994. Fauna dels Països Catalans. Enciclopèdia Catalana, Barcelona, Spain. Fuller, M. R., Seegar, W. S. & Schueck, L. S., 1998. Routes and travel rates of migrating peregrine falcons Falco peregrinus and Swainson’s hawks Buteo swainsoni in the western hemisphere. Journal of Avian Biology, 29: 433–440. Harman, L. M. & Barclay, J. H., 2007. A summary of California burrowing owl banding records. In: Proceedings of the California burrowing owl Symposium, November 2003: 123–131. Bird Popula-


Animal Biodiversity and Conservation 34.2 (2011)

tions Monograph Number 1. The Institute for Bird Populations and Albion Environmental Inc., Point Reyes Station, California, USA. Haug, E. A., Millsap, B. A. & Martell, M. S., 1993. Burrowing owl (Athene cunicularia). In: The birds of North America, 61 (A. Poole & F. Gill, Eds.). The Academy of Natural Sciences, Philadelphia, & American Ornithologists’ Union, Washington, DC, U.S.A. Hobson, K. A. & Wassenaar, L. I., 2008. Tracking animal migration with stable isotopes. Elsevier Inc., Oxford, United Kingdom. Holroyd, G. L. (Ed.), 2005. North American conservation action plan for the burrowing owl. Commission for Environmental Cooperation, Montreal, Quebec, Canada. Holroyd, G. L., Conway, C. J. & Trefry, H. E., 2011. Breeding Dispersal of a burrowing owl from Arizona to Saskatchewan. Wilson Journal of Ornithology, 123: 378–381. Holroyd, G. L, Rodriguez–Estrella, R. & Sheffield, S. R., 2001. Conservation of the burrowing owl in western North America – issues, challenges and recommendations. Journal of Raptor Research, 35: 399–407. Holroyd, G. L., Trefry, H. E. & Duxbury, J. M., 2010. Winter destinations and habitats of ‘Canadian’ burrowing owls. Journal of Raptor Research, 44: 294–299. James, P. C., 1992. Where do Canadian burrowing owls spend the winter? Blue Jay, 50: 93–95. Kochert, M. N., Fuller, M. R., Schueck, L. S., Bechard, M. J., Woodbridge, B., Holroyd, G. L., Bond, L. & Banasch, U., 2011. Migration patterns, use of stopover areas, and austral summer movements of Swainson’s Hawks. Condor, 113: 89–106. Korfanta, N. M., McDonald, D. B. & Glenn, T. C., 2005. Burrowing owl (Athene cunicularia) population genetics: a comparison of North American forms and migratory habits. Auk, 122: 464–478. Macias–Duarte, A., 2011. Change in migratory behavior as a possible explanation for population declines of burrowing owls in northern latitudes. Ph. D. Thesis, University of Arizona, USA. Macdonald, D. W., & Johnson, D. D. P., 2001. Dispersal in theory and practice: consequences for conservation biology. In: Dispersal: 358–372 (J. Clobert, A. A. Danchin, A. A. Dhondt & J. D. Nichols, Eds.). Oxford University Press, United Kingdom. Meehan, T. D., Rosenfield, R. N., Atudorei, V. N., Bielefeldt, J., Rosenfield, L. J., Stewart, A. C., Stout, W. E. & Bozek, M. A., 2003. Variation in hydrogen stable–isotope ratios between adult and nestling Cooper’s Hawks. Condor, 105: 567–572. Meyburg, B., Howey, P. W., Meyburg, C. & Fiuczynski, K. D., 2011. Two complete migration cycles of an adult Hobby tracked by satellite. British Birds, 104: 2–15. Ogonowski, M. S. & Conway, C. J.. 2009. Migratory

387

decisions in birds: extent of genetic vs. environmental control. Oecologia, 161:199–207. Rosier, J. R., Ronan, N. A. & Rosenberg, D. K., 2006. Post–breeding dispersal of burrowing owls in an extensive California grassland. American Midland Naturalist, 155: 162–167. Shyry, D., 2005. Juvenile burrowing owl survival, prey use and availability, and nocturnal foraging habitat selection, over the post–fledging period. M. Sc. Thesis, University of Alberta, Edmonton, Canada. Sissons, R. A., 2003. Food and habitat selection of male burrowing owls (Athene cunicularia) on southern Alberta grasslands. M. Sc. Thesis, University of Alberta, Edmonton, Canada. Smith, A. D., Lott, C. A., Smith, J. P., Donohue, K. C., Wittenberg, S., Smith, K. G. & Goodrich, L., 2009. Deuterium measurements of raptor feathers: does a lack of reproducibility compromise geographic assignment? Auk, 126: 41–46. Stutchbury, B., 2007. Silence of the songbirds. HarperCollins Publishers, Toronto. Stutchbury, B. J. M., Gow, E. A., Done, T., MacPherson, M., Fox, J. W. & Afanasyev, V., 2010. Effects of post–breeding moult and energetic condition on timing of songbird migration into the tropics. Proceedings of the Royal Society published online 21 July 2010 doi: 10.1098/rspb.2010.1220. Stutchbury, B. J. M., Tarof, S. A., Done, T., Gow, E., Kramer, P. M., Tautin, J., Fox, J. W. & Afanasyev, V., 2009. Tracking long–distance songbird migration by using geolocators. Science, 323: 896. Terborgh, J. W., 1980. The conservation status of neotropical migrants: present and future. In Migrant birds in the neotropics: ecology, behavior, distribution, and conservation: 21–30 (A. Keast & E. S. Morton, Eds.). Smithsonian Institution Press, Washington, D.C., USA. Terborgh, J., 1989. Where have all the birds gone? Princeton University Press, Princeton, New Jersey, USA. Todd, L. D., 2001. Survival and dispersal of juvenile burrowing owls (Athene cunicularia) during the post–fledging, pre–migratory period. M. Sc. Thesis, University of Regina, Saskatchewan, Canada. Van Nieuwenhuyse, D., Genot, J.–C. & Johnson, D. H., 2008. The little owl: conservation, ecology and behavior of Athene noctua. Cambridge University Press, UK. Wellicome, T. I., 2000. Effects of food on reproduction in burrowing owls (Athene cunicularia) during three stages of the breeding season. Ph. D. Thesis, University of Alberta, Edmonton, Canada. Wiggins, D. A., 2006. Update COSEWIC status report on burrowing owl Athene cunicularia. Prepared for Committee on the Status of Endangered Wildlife in Canada, Environment Canada, Ottawa, Ontario, Canada.


366


Animal Biodiversity and Conservation 34.2 (2011)

389

Pre–release training of juvenile little owls Athene noctua to avoid predation R. Alonso, P. Orejas, F. Lopes & C. Sanz

Alonso, R., Orejas, P., Lopes, F. & Sanz, C., 2011. Pre–release training of juvenile little owls Athene noctua to avoid predation. Animal Biodiversity and Conservation, 34.2: 389–393. Abstract Pre–release training of juvenile little owls Athene noctua to avoid predation.— Anti–predator training of juvenile little owls was tested in a sample of recovered owls raised in captivity in Brinzal Owl Rescue Center (Madrid, Spain). Mortality caused by predators has been described previously in released individuals. Nine little owls were conditioned during their development to a naturalized goshawk and a large live rat, whose presence was paired to the owl’s alarm call. All nine owls and seven non–trained individuals were then released during the late summer and autumn and radio–tracked for six weeks to test their survival. In total 71.4% of the trained owls survived while only the 33.3% of the untrained group were alive at the end of week six. The only cause of death that was detected was predation. Antipredator training, therefore, seems to be beneficial in maximizing survival after the release of juvenile little owls. Key words: Little owl, Athene noctua, Reintroduction, Release, Survival, Antipredator training. Resumen Entrenamiento antes de la liberación en mochuelos europeos Athene noctua para evitar su depredación.— Un entrenamiento sobre mochuelos juveniles para evitar la depredación, se ha testado en una muestra de ejemplares recuperados y criados en el Centro de Recuperación de Rapaces Nocturnas Brinzal (Madrid, España). Previamente se ha descrito una alta mortalidad en ejemplares liberados, causada por los depredadores. Se condicionaron nueve ejemplares durante su desarrollo, frente a un azor naturalizado y a una rata viva de gran tamaño, cuya presencia se había asociado a una llamada de alarma del mochuelo. Estos nueve ejemplares, junto a siete más no entrenados, se liberaron durante la última parte del verano y el otoño y fueron radiomonitorizados durante seis semanas con objeto de comprobar su supervivencia. En total sobrevivió el 71,4% de los mochuelos entrenados, mientras que sólo el 33,3% de los no entrenados sobrevivía a las seis semanas. No se registró ninguna otra causa de mortalidad que no fuera la depredación. El entrenamiento antidepredación parece ser beneficioso para la liberación de juveniles de mochuelo, de cara a maximizar su supervivencia. Palabras clave: Mochuelo europeo, Athene noctua, Reintroducción, Liberación, Supervivencia, Entrenamiento antidepredación. (Received: 10 VI 11; Conditional acceptance: 8 VII 11; Final acceptance: 10 X 11) Raúl Alonso, Patricia Orejas, Francisca Lopes & Carmen Sanz, Centro de Recuperación de Rapaces Nocturnas Brinzal, Albergue Juvenil Richard Schirrmann, Casa de Campo s/n., Madrid, España (Spain). Corresponding author: R. Alonso. E–mail: raulalonso@brinzal.org

ISSN: 1578–665X

© 2011 Museu de Ciències Naturals de Barcelona


390

Introduction In recent decades, governments and private organizations have increasingly used captive bred animals in reintroduction attempts to reestablish or restock populations for conservation or hunting (Seddon et al., 2007). Post–release predation has been documented as one of the main cause of failure in release projects for mammals and birds (e.g. Kleiman, 1989; Beck el al., 1991; Short, 1992; Miller, 1994; Parish & Sotherton, 2007). As a result of captivity, animals can show variation in behavior, especially when the number of captive generations increases. This could result in decreased survival upon reintroduction to the wild (McPhee, 2003). Artificial rearing of birds tends to disrupt the normal development in recognizing innate predators (Curio, 1998) and may decrease opportunities for animals to acquire essential learned behaviors, such as predator recognition (Kleiman, 1989; Griffin et al., 2000). As adequate responses imply the presence of recognition processes (Curio, 1993), it has been proposed that animals that have been isolated from predators may loose antipredator behavior (Berguer, 1998; Lima & Dill, 1990). Understanding and mitigating these factors is essential for ethical reasons and to ensure successful reintroduction (Seddon et al., 2007). Animals should therefore be given the opportunity to acquire the necessary information to enable survival in the wild through training in their captive environment (IUCN/SSC). Fearless animals can be trained to respond to predators (e.g. Ellis et al., 1977; Curio, 1988; Maloney & McLean, 1995), and in recent decades interest has grown in training animals to avoid post–release predation (e.g. Curio, 1998; Griffin et al., 2000). Antipredator training is usually done through the use of Pavlovian or classical conditioning. In this process, trials of conditioned stimulus (the predator figure) are paired to an unconditioned stimulus (a frightening, alarming or pain–inducing stimulus). Such training can be used not only to illicit a fear response but also to improve the efficiency of responses (Griffin et al., 2000) and to enhance initially low–level antipredator responses (Miller et al., 1990; McLean et al., 1996). Moreover, this type of training increases vigilance behaviour (McLean et al., 1996). The feasibility of antipredator training depends on the type of isolation (evolutionary –over generations– or ontogenetic –during an animal´s life) and on the specific components of antipredator behavior (avoidance, recognition, and response) that have been lost (Griffin et al., 2000). Animals that have been isolated only ontogenetically (e.g. bred in captivity) may have the capacity to express appropriate antipredator behaviour, but this might not occur without specific experience (Griffin et al., 2000). In the case of little owl Athene noctua, several release projects have been attempted in Europe with little or no success (Stahl, 1982; Mohr, 1989; Leicht, 1992; Möller, 1993; Génot & Sturm, 2003) primarily due to predation (Van Nieuwenhuyse et al., 2008). Implementation of animal behaviour experiments has been proposed to determine why reintroductions fail, to shed light on what can be done to improve

Alonso et al.

the chances of success, and to evaluate the impact of interventions such as environmental enrichment (Mathews et al., 2005). Although the evidence suggests that a large percentage of released juvenile little owls do not survive in the wild, nothing has been done to decrease the main cause of death, predation. Here, our main aim was to determine whether antipredator training is effective in improving survival of released juvenile little owls. Material and methods Little owls entered the Brinzal Owl Rescue Center (Madrid, Spain) as eggs, chicks, fledglings or fully fledged individuals. Eggs were artificially incubated and owl chicks were hand–fed only during their first week of life. Chicks and fledglings received medical treatment when necessary. After the first week of life, special care was taken to keep the owls isolated from humans and to assure establishment of species imprinting, which promotes adequate reproductive behavior. A relation between misimprinting and predator avoidance has been described (Curio, 1998). As imprinting occurs not only in parents but also in siblings (Fox, 1995), chicks were exposed 24 h/day to a square angle mirror placed inside their intensive care cages. When medical treatment was given(fractures, large wounds, etc.), a non–releasable adult accompanied them. Chicks were transferred to adult foster parents in outdoor cages as soon as possible after the first week of life. Antipredator training was performed when the owls were in outdoor cages (foster parent cages and aviaries). Classical or Pavlovian conditioning trials were conducted using conditioned stimuli as follows: a stuffed goshawk in flight position was moved along an iron cable above the cages and a live rat was trained to cross a mesh corridor connecting two boxes with one–way doors. Since alarm calls are closely associated with predatory events and can potentially favour species–specific learning mechanisms (Griffin et al., 2000), we used a digital recorded alarm call of the little owl as unconditioned stimulus. The call was manually activated by remote control during the goshawk 'flight', and automatically activated by the rat when it triggered a switch while running through the mesh corridor. Goshawk trials occurred at any time during the day, while rat trials were performed at night. Although learning about predators occurs after just one or two trials in controlled conditions (Maloney & McLean, 1995; McLean et al., 1999), we conducted two to four trials per week from fledgling until the bird was released. The trials were randomly spaced and we made sure that no noises or activity occurred before these trials to avoid establishing clues regarding their onset. To avoid habituation, the appearance of the predator was very short in time (a few seconds only) and it was always associated with the alarm call. To allow appropriate responses (e.g., hiding, freezing, flying or staying high enough to escape) to the predators,


Animal Biodiversity and Conservation 34.2 (2011)

391

Table 1. Results of the radio–monitoring of little owls. Tabla 1. Resultados del seguimiento por radio de mochuelos. Group

Release date

Predation

Signal lost

Alive at week 6

Control

7

2007

4

1

2

Experimental

9

2010

2

2

5

cages and aviaries were equipped with natural vegetation, nest–boxes and high perches. The experimental group consisted of nine individuals that were trained to avoid predation while the control group was made up of seven individuals that had no antipredator training. For logistical reasons, the control group of owls was reared and released in 2007, while the trained experimental group was reared and released in 2010. Before their release into the wild, all the individuals had at least 15 days of hunting training. Finally, they were ringed and radio–tagged with backpack tags (TW–4 from Biotrack Ltd.) attached with Teflon ribbon (Bally Ribbon, Bally, PA) harness (Kenward, 1987). These were switched on, on the day before release to check function and attachment system. Because the typical maximum allowable tag to body weight ratio is 3%, the total weight of each tag was 3.8 g. (2.5% of the average little owl weight). The receivers were a TRX–1000 (Wildlife Materials, Inc.) and a VR–500 (Vertex Standard Co., Ltd.) connected to a three–element Yagi antenna (Biotrack Ltd.). Birds were released during the late summer or autumn (August to October) in an appropriate habitat (olive orchards, open holm oak forests, or ash trees in meadows) in the province of Madrid where we knew little owls occurred. After release, owls were monitored a minimum of four times per week during the day or at night. Since most predation events in artificially reared birds occur in the first weeks after release (e.g. Sokos et al., 2008), little owls were monitored for six weeks, the minimum time considered necessary to evaluate the success or failure of released rehabilitated raptors (Duke et al., 1981). Results Table 1 shows the results of the monitoring. Signals from three individuals were lost in the first five days after release due to unknown reasons. Two of these were from birds in the trained group. Four of six controls (66.6%) were preyed upon in the first twenty–five days (mean 14.25) after release (survival rate = 33.3%). In the experimental group, two of the seven individuals (28.5%) were depredated on the third and eighth day (mean 5.5) after release (survival rate = 71.4%). Predators were a variety of raptors and mammals. Probable predators (judged on the condition of the carcasses)

included two tawny owls (Strix aluco), a sparrowhawk (Accipiter nisus), a goshawk (Accipiter gentilis), a least weasel (Mustela nivalis) and a genet (Genetta genetta). No other causes of death cause were found. Discussion Although our sample was relatively small, predation was two–fold higher in the control group than in the experimental group, suggesting that pre–release training of juvenile little owls could improve the efficacy of release projects. Nevertheless, further testing is needed to determine whether these results are sustained over time, as has been shown in other species (e.g. Azevedo & Young, 2006). A release program carried out in Toledo, Spain in 2001 and 2009 showed that 15 out of 23 (65%) radio–monitored little owls were depredated during the first four weeks after release (P. Cervera, pers. comm.). We should emphasize that the predation rates could have been be affected by the fact that the control and experimental releases occurred in different years. It should also be kept in mind that our sample sizes were small. Nevertheless, our results suggest the benefits of training and the technique merits further testing and adoption when birds raised in captivity are released into the wild. As knowledge acquired via training is supposedly transmitted culturally (Curio, 1998), in a similar way to conditioning, antipredator training in captive–bred birds should not be overlooked. When trained animals are reintroduced into the wild, they could potentially serve as models for predator–naive individuals, including their offspring and other adults (Griffin et al., 2000). To conclude, we suggest that the extent of natural predation in recovered juvenile raptors should be studied in greater depth. Results to date seem to indicate that predation on birds reared in captivity could explain reintroduction/restocking failures more than any other factor. Acknowledgements We would like to thank Fundación Biodiversidad (Ministerio de Medio Ambiente y Medio Rural y Marino), Comunidad de Madrid (Consejería de Medio Ambiente y


392

Dirección General de Juventud), Ayuntamiento de Madrid, VOLCAM–Caja Mediterráneo, Manimals, S.L. and Imaginarte, S.L., for their economic support and for their encouragement with this project. This work was possible thanks to the generous collaboration of Antonio Agudo, Sara Agudo, Carolina Battistoni, José Mª Blázquez, Mª José Caballero, Marta Conejo, Manuel Galán, Emilio Escudero, Iván García, Marta García, Lucas González, José Manuel Hebrero, Abel Herrero, Fernando Jiménez, Alfonso Mamán, Rebeca Urquía y Cristina Ruiz. We would specially like to thank Rocío Blanco, Lidia Lopez and Verena Valenzuela for their dedication and tenacity spending countless hours radiotracking. Thanks too to Pilar Cervera (CERI) for sharing with us the results of little owl radiotracking in Toledo, Cristina González–Onandía and Victoria Pérez for helping us beyond their own professional responsibilities sharing with us the enthusiasm for these exciting big–eyed animals. This article benefited from comments on an earlier draft by I. Zuberogoitia, G. L. Holroyd and V. Penteriani. References Azevedo, C. S. & Young, R. J., 2006. Do captive–born greater rheas Rhea Americana remember antipredator training? Revista brasileira de zoologia, 23(1): 194–201. Beck, B. B., Kleiman, D. G., Dietz, J. M., Castro, I., Carvalho, C., Martins, A. & Rettberg–Beck, B., 1991. Losses and reproduction in reintroduced golden lion tamarins Leontopithecus rosalia. Dodo: Journal of the Jersey Wildlife Preservation Trust, 27: 50–61. Berguer, J., 1998. Future prey: some consequences of the loss and restoration of large carnivores. In: Behavioral ecology and conservation biology: 80–100 (T. M. Caro, Ed.). Oxford University Press, New York. Curio, E., 1988. Cultural transmission of enemy recognition by birds. In: Social learning: psychological and biological perspectives: 75–97 (T. R. Zentall & B. G. Galef Jr., Eds.). Lawrence Eribaum Associates, Hillsdale, New Jersey. – 1993. Proximate and developmental aspects of antipredator behaviour. Advances in the study of behaviour, 22: 135–238. – 1998. Behavior as a tool for management: intervention for birds. In: Behavioral ecology and conservation biology: 80–100 (T. M. Caro, Ed.). Oxford University Press, New York. Duke, G. E., Redig, P. T. & Jones, W., 1981. Recoveries and Resightings of Released Rehabilitated Raptors. Journal of Raptor Research, 15(4): 97–107. Ellis, D. H., Dobrott, S. J. & Goodwin, J. G., 1977. Reintroduction techniques for Masked Bobwhites. In: Endangered birds: management techniques for preserving threatened species: 345–354 (S. A. Temple, Ed.). University of Wisconsin Press, Madison. Fox, N., 1995. Understanding the bird of prey. Hancock House, Surrey

Alonso et al.

Génot, J. C. & Sturm, F., 2003. Bilan de l´expérience de reinforcement des populations de Cheveche d´Athena Athene noctua dans le Parc naturel regional des Vosges du Nord. Alauda, 71: 175–178. Griffin, A. S., Blumstein, D. T. & Evans, C. S., 2000. Training captive–bred or traslocated animals to avoid predators. Conservation biology, 14(5): 1317–1326. IUCN/SSC. Guidelines for Re–Introductions. http://intranet.iucn.org/webfiles/doc/SSC/SSCwebsite/Policy_statements/Reintroduction_guidelines. pdf (consulted on May 10th of 2011). Kenward, R., 1987. Wildlife radio tagging: Equipment, field techniques, and data analysis. Academic Press, London. Kleiman, D. G., 1989. Reintroduction of captive mammals for conservation: guidelines for reintroducing endangered species into the wild. BioScience, 39: 152–161. Leicht, U., 1992. Erfahrungen mit der steinkauzzucht und der Auswilderung. Naturschutzzentrum Wasserschloss Midwitz, 2: 35. Lima, S. L. & Dill, L. M., 1990. Behavioral decisions made under the risk of predation: a review and prospectus. Canadian Journal of Zoology, 68: 619–640. Maloney, R. F. & McLean, I. G., 1995. Historical and experimental learned predator recognition in free living New Zealand robins. Animal Behaviour, 50: 1193–1201. Mathews, F., Orros, M., McLaren, G. Gelling, M. & Foster, R., 2005. Keeping fit on the ark: assessing the suitability of captive–bred animals for release. Biological Conservation, 121: 569–577. McLean, I. G., Hölzer, C. & Strudholme, B. J. S., 1999. Teaching predator–recognition to a naive bird: implications for management. Biological Conservation, 87: 123–130. McLean, I. G., Lundie–Jenkisns, G. & Jarman, P. J., 1996. Teaching an endangered mammal to recognise predators. Biological Conservation, 75: 51–62. McPhee, M. E., 2003. Generations in captivity increases behavioral variance: considerations for captive breeding and reintroduction programmes. Biological Conservation, 115: 71–77. Miller, B., Biggins, D., Hanebury, L. & Vargas, A., 1994. Reintroduction of the black–footed ferret (Mustela nigripes). In: Creative Conservation: interactive management of wild and captive animals: 455–464 (P. J. S. Olney, G. M. Mace & A. T. C. Feistner, Eds.). Chapman and Hall, London. Miller, B., Biggins, D., Wemmer, C., Powell, L., Hanebury, L. & Wharton, T., 1990. Development of survival skills in captive raised Siberian polecats (Mustela eversmanni) II. Predator avoidance. Journal of Ethology, 8: 95–104. Mohr, H., 1989. Steinkäuze brüten wieder in Schwaben. Gefierdete Welt, 113: 308–309. Möller, B., 1993. Erste Ergebnisse zur Wiedereinbürgerung des Steinkauzes (Athene noctua) in den Landkreisen Hildesheim und Peine. Beiträge zur Naturkunde Niedersachsens, 46: 72–81. Parish, D. M. B. & Sotherton, N. W., 2007. The fate of released captive–reared grey partridges Perdix


Animal Biodiversity and Conservation 34.2 (2011)

perdix: implications for reintroduction programmes. Wildlife Biology, 13(2): 140–149. Seddon, P. J., Armstrong, D. P. & Maloney, R. F., 2007. Developing the science of reintroduction biology. Conservation Biology, 21(2): 303–312. Short, J., Bradshaw, S. D., Giles, J., Prince, R. I. T. & Wilson, G. R., 1992. Reintroduction of macropods (Marsupialia: Macropoidea) in Australia: a review. Biological Conservation, 62: 189–204.

393

Sokos, C. K., Birtsas, P. K. & Tsachalidis, E. P., 2008. The aims of galliforms release and choice of techniques. Wildlife Biology, 14(4): 412–422. Stahl, D., 1982. Zucht und Auswilderung des Steinkauzes (Athene noctua). Voliere, 5: 178–180. Van Nieuwenhuyse, D., Génot, J. C. & Johnson, D. H., 2008. The Little Owl. Conservation, Ecology and Behavior of Athene noctua. Cambridge University Press, Cambridge.


366


Animal Biodiversity and Conservation 34.2 (2011)

395

Bias in little owl population estimates using playback techniques during surveys I. Zuberogoitia, J. Zabala & J. E. Martínez

Zuberogoitia, I., Zabala, J. & Martínez, J. E., 2011. Bias in little owl population estimates using playback techniques during surveys. Animal Biodiversity and Conservation, 34.2: 395–400. Summary Bias in little owl population estimates using playback techniques during surveys.— To test the efficiency of playback methods to survey little owl (Athene noctua) populations we carried out two studies: (1) we recorded the replies of radio–tagged little owls to calls in a small area; (2) we recorded call broadcasts to estimate the effectiveness of the method to detect the presence of little owl. In the first study, we detected an average of 8.12 owls in the 30' survey period, a number that is close to the real population; we also detected significant little owl movements from the initial location (before the playback) to the next locations during the survey period. However, we only detected an average of 2.25 and 5.37 little owls in the first 5' and 10', respectively, of the survey time. In the second study, we detected 137 little owl territories in 105 positive sample units. The occupation rate was 0.35, the estimated occupancy was 0.393, and the probability of detection was 0.439. The estimated cumulative probability of detection suggests that a minimum of four sampling times would be needed in an extensive survey to detect 95% of the areas occupied by little owls. Key words: Little owl, Survey methods, Presence Program, Detection efficiency, Vocal activity. Resumen Problemas en las estimas poblacionales de mochuelos cuando se realizan censos con reclamos.— Se desarrollaron dos estudios diferentes para probar la eficiencia de los censos por medio de reclamos de mochuelos (Athene noctua): (1) un seguimiento intensivo de las respuestas a los reclamos de mochuelos radio–marcados en una pequeña área, (2) un estudio extensivo utilizando reclamos para estimar la eficiencia del método como herramienta para detectar la presencia de mochuelos. En el primer caso, se detectaron 8,12 mochuelos de media en un periodo de censo de 30', número cercano al tamaño de población real; además, se detectaron desplazamientos significativos de los mochuelos desde la posición inicial (antes de conectar el reclamo) a las posiciones siguientes durante la ejecución del reclamo. Sin embargo, tan sólo se detectó una media de 2,25 y 5,37 mochuelos en los primeros 5' y 10' respectivamente. En el segundo caso, se detectaron 137 territorios de mochuelos en 105 unidades de muestreo positivas. La tasa de ocupación fue de 0,35, la ocupación estimada de 0,393 y la probabilidad de detección de 0,439. La probabilidad acumulada estimada de detección sugiere que se precisarían de al menos cuatro muestreos en un estudio extensivo si se pretende detectar el 95% de las áreas ocupadas por mochuelos. Palabras clave: Mochuelo común, Métodos de censo, Programa Presence, Eficacia de detección, Actividad vocal. (Received: 25 IV 11; Conditional acceptance: 7 X 11; Final acceptance: 13 X 11) Iñigo Zuberogoitia, Estudios Medioambientales Icarus S. L., Pintor Sorolla 6 1º C, 26007 Logroño, España (Spain).– Jabi Zabala, Ihobe, Climate change and Biodiversity Area, Alameda Urquijo 36 6º, 48011 Bilbao, España (Spain).– José Enrique Martínez, Bonelli´s Eagle Study and Conservation Group, Apdo. 4009, E–30080, Murcia, España (Spain). Corresponding author: I. Zuberogoitia. E–mail: zuberogoitia@icarus.es ISSN: 1578–665X

© 2011 Museu de Ciències Naturals de Barcelona


Zuberogoitia et al.

396

Introduction Owls are notoriously difficult to count because they are secretive and primarily nocturnal, and roost in concealed locations during the day (Mikkola, 1983). Monitoring of owl populations to date has generally taken long periods and the use of several tools to achieve accurate results. In current practice, population monitoring is most likely achieved in small areas where monitoring programmes are developed over long time periods. However, if we need to know the status of an owl guild in large areas, a balance must be made between time, effort, budget and the desired accuracy of the results we expect to achieve. Among the most effective methods to estimate owl populations in terms of cost and results is listening to owl calls during their activity period (see i.e. Zuberogoitia & Campos, 1997; Galeotti & Sacchi, 2001; Zuberogoitia, 2002; Martínez et al., 2007; Van Nieuwenhuyse et al., 2008). Although this is no simple task, it is worth the effort because knowledge of this aspect of bird behaviour (their responsiveness to survey methods) is necessary to optimize returns. Playback of tape–recorded calls has been widely used to survey owl species worldwide and several studies rely on the use of elicited calls to systematically obtain data on the relative abundance of owls (e.g. Zuberogoitia & Campos, 1997; Martínez & Zuberogoitia, 2002; Escandell, 2005; Navarro et al., 2005; Crewe & Badzinski, 2006; Conway et al., 2008). However, whether this technique is effective in detecting some owl species is still a controversial issue (Martínez et al., 2002), and there are some aspects of this technique that require further research to determine applicability and possible shortcomings of the results. For the last 20 years we have been studying owls in Bizkaia (Northern Spain), focusing our attention on owl distribution and status, survey methodologies and ecology (see Zuberogoitia, 2002). One of the target species, little owl (Athene noctua), was monitored to detect and document any possible population decline. Our main goal was to develop a replicable survey method to evaluate population trends. Based on our experience and a literature review we detected two main drawbacks in current practice and little owl survey methodologies. On one hand, abundance estimates obtained simply by listening to spontaneous calls or the response to broadcast calls are suspected of being affected by issues related with the studied population (Martínez & Zuberogoitia, 2004), and by methodological aspects such as the duration of the listening period. On the other hand, previous studies (Zuberogoitia & Campos, 1998; Johnson et al., 2009) and our own field experience suggest that presence/absence surveys based on broadcast calls generate an unknown number of omission errors (false absences) that could have important consequences for management. These errors could be overcome, or at least controlled, if the detection capability of the methods could be estimated. We conducted two studies to throw light on these aspects. First, we carried out an intensive study on answers to calls of radio–tagged little owls in a small area

(approx. 1 km²) to determine individual and seasonal variations in owl detection and to assess how broadcast and listening times affect the number of owls estimated. Second, we conducted an extensive survey using call broadcasts to estimate the effectiveness of the method to detect the presence of little owl in the surveyed area. Study area This study was carried out in two nested areas: the extensive survey covered the whole of the Basque Country (approx. 7,200 km²), whereas the intensive study was carried out in the Mungia valley in an area of approximately 1 km². The Autonomous Community of the Basque Country (lying between 42º and 43º N and 3º and 1º W) has two clearly defined areas (roughly north and south). The northern area runs along the coast of the Bay of Biscay, with its Atlantic climate and mild temperatures and an annual rainfall of 1,200–2,000 mm. The land there is mountainous and densely populated, with extensive urban and industrial areas, mainly located in valleys and on the gentler slopes. Forestry plantations (Pinus radiata and Eucaliptus spp.) have become widespread in the last 80 years, gradually replacing grazing land for extensively–reared livestock, traditional agricultural activities, and a few remnants of native forest. The second large area, of some 2,500 km2, lies to the south and is situated in an area of transition to the Mediterranean climatic region. The climate is Mediterranean and the landscape is dominated by arable lands, vineyards, Mediterranean scrub and holm–oak woods in the sloping areas. The little owl population in the Atlantic area is divided into several small patches, which vary in density depending on the prevailing vegetation types, with areas of open fields and meadows harbouring the highest densities (see Zuberogoitia & Campos, 1997, 1998; Zabala et al., 2006). For the intensive radio–tracking study we selected one of these small population patches according to the following characteristics: 1) a high density of little owls: in the study area some fields contained seven pairs/km2 (Zuberogoitia & Campos, 1998), 2) knowledge of the little owl population size and distribution, which had been surveyed previously (Zuberogoitia & Campos, 1997, 1998). This population was located in the Mungia valley, an area of 10 km2 dominated by pastures for cattle, and small–holdings where the 1 km2 intensive study area was located. The climate is rainy oceanic, with annual rainfall of around 1,500 mm, and annual average temperatures varying from 13.8ºC to 12ºC. Winters are mild and there is no summer drought. Methods Intensive study: changes in owl detection We captured nine little owls using mist nets; they were then radio–tagged and radio–monitored between January and September 2004. All the little owls sur-


Animal Biodiversity and Conservation 34.2 (2011)

397

Fig. 1. Sample units (○) and detected territories of little owl (●) in the Basque Country in the breeding season of 2009. Fig. 1. Unidades de muestreo (○) y territorios de mochuelo detectados (●) en el País Vasco durante el periodo reproductor de 2009.

vived throughout the study period (see Zuberogoitia et al., 2007, 2008). We radio–tracked these owls and listened to their calls simultaneously for a total of 250 hours, as described in more detail below. The tape recorder was set randomly in eight different points in the study area between February and June 2004, and it played three different sources of recordings of little owls (see Hardouin et al., 2004). The recording was always played for 30' in the first two hours after sunset. Simultaneously, we monitored the nine radio– tracked little owls to know their position before and during the 30–minute period, thereby following their movements during the experiment. This allowed us to test whether the radio–tagged little owls, if present, responded to the playback recording, and if so, when. Extensive study: effectiveness of call broadcasts to detect little owls In this case, we ran a large scale presence/absence survey encompassing all the Basque Country area using the knowledge obtained in the previous test and other studies on little owl (Zuberogoitia & Campos, 1997, 1998; Zuberogoitia, 2002; Zabala et al., 2006; Zuberogoitia et al., 2005, 2007, 2008). The area was divided into patches considered ecologically suitable for little owl (open areas) (Zabala et al., 2006). We considered as suitable three different types of area: (1) Coastal countryside: open areas located close to the coast dominated by grass–fields for small cattle farms, urban parks, coastal heather lands, dunes and marshes; (2) Atlantic countryside: orchards and grass–fields for cattle in narrow valleys surrounded by mountains which covered by timber plantations and oak patches; and (3) Mediterranean agricultural areas: dry–farmed and irrigated crops (cereal, potatoes, beetroot and vineyards).

We distributed 300 sample points in 30 areas (10  sample points/area), 80% of them located in areas of good quality for little owls and 20% in low quality areas (open areas surrounded by large forests). We needed some low quality areas to obtain zeros to run the statistics (see MacKenzie et al., 2006). Sample points inside the sample areas should be independent and so the units were at least one km apart. Every sample unit was censused between 2 and 8 times (an average of 5 times per unit, see Meredith, 2008), obtaining a total of 1,500 censuses. Different censuses in the same sample unit were conducted on different days, or in a different location on the same day and at least half an hour after the first census. In this way, the effects and variations caused by sampling site location were included as well as temporal variation. Starting at dusk, we broadcasted little owl calls for 5' and waited another 5' for answers in every sample unit. In this 10' period, we recorded every response and the point of the first detection. The study was carried out between the 1st of June and the 8th of July 2009, coinciding with the stage in which adults are raising their offspring and when their home range around the nest is minimal, i.e. just in the period in which adults respond to the voices of conspecifics mainly within their core areas (Zuberogoitia et al., 2007; Sunde et al., 2009). Outside this period little owls reduce their territorial behaviour and can be located in communal feeding places, sometimes far from the nesting areas (Zuberogoitia et al., 2007). Spontaneous vocal activity is low during the breeding period (Zuberogoitia et al., 2007) but the little owls detected were mainly breeders, answering close to the nesting sites, and therefore representing true occupation of the areas in question.


Zuberogoitia et al.

398

To estimate the effectiveness of the method to detect little owls, we analysed results using maximum likelihood estimators (MLE). The results of the survey were considered to be the combination of two unknown probabilities: (a) the probability of the little owl being present at the sample site; and (b) the probability of the species being detected with the method applied. The Presence 2.4. Program (MacKenzie et al., 2006) was used to obtain the MLE of these probabilities from the raw field data. For this analysis we assumed that occupancy of little owl populations in the area to be homogeneous (no strong differences in presence among sub–areas) and that the probability of detection was the same throughout the sampling period (1st June to 8th July 2009). To estimate the effect of repeated sampling we estimated the cumulative probability of detection using the following formula:

The coefficient of variation of the number of little owls detected in the area and calling in the 30–minute period (CV = 0.30) increased when the 10–minute period was considered (CV = 0.43), and was even higher for the 5–minute period (CV = 0.57). Extensive study: effectiveness of call broadcasts to detect little owls

where q stands for the probability of going undetected if present, p for the probability of detection if present and n the number of times the survey is conducted in the area. Presence 2.4. was run using the 300 sample units (1,500 surveys) to obtain the occupancy (Psi) and the probability of detection (P).

Overall, 137 little owl territories were detected in 105 positive sample units (table 1, fig. 1). This suggests an apparent occupation rate of 0.35 (i.e. 35% of the sampled areas was occupied). On the other hand, the software Presence produced a model that estimated occupancy at 0.393 (CV 0.052) and probability of detection at 0.439 (CV 0.080). Therefore, a single 5– minute broadcast with 5' of subsequent listening would only have detected little owls in approximately half (44%) of the occupied areas. A survey of the study area based on a single broadcast per sample unit would have estimated an occupancy rate of 0.017 (estimated occupancy, 0.393, multiplied by probability of detection with a single sampling occasion, 0.439). The estimator for the cumulative probability of detection (fig. 2) suggests that in an extensive survey a minimum of four repetitions of censuses in the same sample points in each sample area is required to detect 95% of the areas occupied by little owls.

Results

Discussion

Intensive study: changes in owl detection

Intensive study: changes in owl detection

The radio–tracked owls were located at an average distance of 264.78 m (SD = 125.18, range  =  30– 553  m) just before broadcasting the calls. After the broadcast they moved closer to the radio tape from the initial distance, to an average distance of 145.03 m (SD = 135.56, range = 0–530). Differences between pre– and post–broadcast distances were statistically significant (t–Student test for matched samples, t = 4.991, P = 0.000, df = 34). There was a positive correlation between the distance at the beginning of the broadcast period and the answer time (R2 = 0.366, P = 0.006, n = 55). The average number of detected owls in the 30– minute period was 8.12 (SD = 2.47), ranging between 4 and 11 individuals. However, if only the first 5' of playback were considered, the number of detected owls would be 2.25 (SD = 1.28, range 1–4), and if 10' were considered, the number of detected birds would be 5.37 (SD = 2.33, range 2–9). Differences were statistically significant among groups (Kruskal–Wallis test, H = 18.77, P = 0.000). The number of little owls calling in the 30–minute period was similar to the number of little owls detected in the study area (Mann–Whitney test, U = 27.5, P = 0.645), but the number of little owls calling in the first 10' was significantly lower than the number of owls known to be present in the surroundings (Mann–Whitney test, U = 13, P = 0.04) and, obviously, the number of detected owls in the first 5' was much lower (Mann–Whitney test, U = 1, P = 0.001).

The results of the censuses varied substantially depending on whether we considered playback and listening periods of 5' or 10'; they would improve, though not by much, if the 10–minute period is considered. The number of little owls detected was close to the real number of territories using the 30–minute periods. However, even in this case, the results were erroneous due to the high mobility of little owls. In this sense, we knew the identity of every little owl tracked and we observed how one owl calling 500 m away could be calling again 2 m close to the broadcast 1' later and fly several hundred meters away a few seconds later. During the study period, we noted that the home range of little owls varied over the months, being larger during winter, decreasing during the pre–courtship period, and increasing again just after the breeding period (Zuberogoitia et al., 2007). We also detected a high degree of social interactions among paired and unpaired owls. Almost all the tracked little owls shared the same fields during the winter period, and even the unpaired owls and those that had lost clutches used communal fields and foraged in the home ranges of the neighbours. Likewise, we observed that the vocal activity was higher during the courtship period (March and April), decreasing during the breeding season. However, when owls were incubating or hatching owlets they called close to the nests and responded to broadcasts from their breeding home

qn= (1–p)n and pn=1–qn


Animal Biodiversity and Conservation 34.2 (2011)

ranges. Only non–paired owls and those which had failed the incubation approached the playback and produced calls from everywhere. Had we developed typical censuses of little owl in the area, the results would have been different. Also, by slightly changing the study period, the results would change between years, depending on the number of successful breeding pairs. However, it was impossible to ascertain the real number of territories using a typical playback survey methodology. In such a case, knowledge of the animals’ behaviour is needed to interpret the results.

399

Table 1. Number of times each sample unit was surveyed, and percentage of positive sample units and number of points with 100%, 75%, 50% and 25% positive results. Tabla 1. Número de ocasiones en las que se censa cada unidad de muestreo y el porcentaje de unidades de muestreo positivas, además del número de puntos con el 100%, 75%, 50% y 25% de los resultados positivos.

Extensive study: effectiveness of call broadcasts to detect little owls

Number of times censused 2

39

As stated above, although the probability of detection is lower during the breeding period due to the reduction in vocal activity, the positive responses to the playback during this time represent breeding units and not dispersers. Under these circumstances, the detection probability was close to 0.44. This means that, in only one survey, we could hope that more than half of the sample points could be null, even considering that little owls were in the sample points (false negatives). In fact, if we wanted to develop a survey methodology capable of detecting the species with statistical confidence (P > 95%; see fig. 2) we would need at least four censuses in every sampling point to be able to accept or reject the presence of little owls. The implications of these results, however, are of high practical relevance. The methodological error derived from low intensity surveys (only one visit per site for example) affects estimations of distribution area. This error is usually associated with a low detectability rate and large confidence intervals, most times unknown. Results of such surveys are often used to investigate habitat issues of the species in question, changes in distribution, population trends and other issues. In the case of an annual large scale survey, the inter–annual

3

103

4

18

5

46

6

37

7

17

8

40

Positive sample units (%)

35

100% positive results

4.3

75% positive results

3.3

50% positive results

9.3

25% positive results

12.3

variability due to the changes in the probability of detection could be higher than the real variation in the population trends, and this would lead to erroneous interpretations of the real population trends. Similarly, in habitat studies based on the presence/absence of the target species, according to our results there would

Basque Country

0.7 0.6

1–P

0.5 0.4 0.3 0.2 0.1

0

1

2

3 4 5 6 Number of censuses

7

8

Fig. 2. Probability of false negatives (probability of non–detection knowing the presence) depending on the number of censuses, considering the average value of P. Fig. 2. Probabilidad de falsos negativos (probabilidad de no detección a sabiendas que hay mochuelo) dependiendo del número de censos, considerando el valor medio de P.


400

be more false absences than valid presence points, seriously questioning the validity of the conclusions that could be drawn. These results should clearly be treated with caution due to the high variability in little owl behaviour in different regions and countries. Nevertheless, they reveal biases associated with broadcasting methodologies, and these could cause serious errors when attempting to ascertain the population dynamics of a given species. We therefore suggest methods should be standardized, beginning with the period of censuses, the minimum length of the monitoring area and number of censuses for large–scale surveys. Acknowledgements We thank L. Astorkia, A. Azkona, A. Iraeta, J. Ituarte, I. Castillo, C. González, S. Hidalgo, J. Fernández and I. Palacios for their field assistance in the first part of the study, Gorka Belamendia, Carlos González de Buitrago, Juan José Torres and Bruno Iglesias for their field assistance in the second part of this study, and an anonymous referee who made valuable suggestions that improved an early version of the manuscript. We are also grateful to IHOBE S. A., for allowing the use of the results for this paper. References Conway, C. J., García, V., Smith, M. D. & Highes, K., 2008. Factors affecting detection of Burrowing Owl nests during standardized surveys. Journal of Wildlife Management, 72: 688–696. Crewe, T. & Badzinski, D., 2006. Ontario Nocturnal Owl Survey. 2005 Final Report. Ontario Ministry of Natural Resources–Terrestrial Assessment Unit. Ontario. Escandell, V. (Ed)., 2005. Programas de seguimiento de SEO/Birdlife en 2005. SEO/Birdlife, Madrid. Galeotti, P. & Sacchi, R., 2001. Turnover of territorial Scops owls as estimated by sperctrographic analyses of male hoots. Journal of Avian Biology, 32: 256–262. Hardouin, L. A., Tabel, P. & Bretagnolle, V., 2004. Neighbour–stranger discrimination in the Little Owl, Athene noctua. Animal Behaviour, 72: 105–112. Johnson, D. H., Van Nieuwenhuyse, D. & Genot, J.–C., 2009. Survey Protocol for the Little Owl Athene noctua. In: Owls–ambassadors for the protection of natura in their changing landscapes (D. H. Johnson, D. Vam Nieuwenhuyse & J. R. Duncan, Eds.). Ardea, 97(4): 403–412. MacKenzie, D. I., Nichols, J. D., Royle, A. J., Pollock, K. H., Bailey, L. L. & Hines, J. E., 2006. Occupancy Estimation and Modeling; Inferring Patterns and Dynamics of Species Occurrence. Elsevier Publishing. Martínez, J. A. & Zuberogoitia, I., 2002. Factors affect-

Zuberogoitia et al.

ing the vocal behaviour of Eagle Owl Bubo bubo: effects of sex and territorial status. Ardeola, 49: 1–10. – 2004. Effects of habitat loss on perceived actual abundance of the Little Owl Athene noctua. Ardeola, 51(1): 215–219. Martínez, J. A., Zuberogoitia, I., Colás, J. & Macía, J., 2002. Use of recorder calls for detecting Long– eared Owls Asio otus. Ardeola, 49: 97–101. Martínez, J. A., Zuberogoitia, I., Martínez, J. E., Zabala, J. & Calvo, J. F., 2007. Patterns of territory settlements by Eurasian Scops Owl (Otus scops) in altered semi–arid lanscape. Journal of Arid Environments, 69: 400–409. Meredith, M. E., 2008. Analyzing Camera Trap Data with PRESENCE. In: Problem–solving in conservation biology and wildlife management: 105–124 (J. P. Gibbs, M. J. Hunter & E. J. Sterling, Eds.). Blackwell Publishing, Malden, USA. Mikkola, H., 1983. Owls of Europe. T & A.D. Poyser. Calton. Navarro, J., Mínguez, E., García, D., Villacorta, C., Botella, F., Sánchez–Zapata, J. A., Carrete, M. & Giménez, A., 2005. Differential effectiveness of playbacks for Little Owls (Athene noctua) surveys before and after sunset. J. Raptor Res., 39(4): 457–461. Sunde, P., Thorup, K., Jacobsen, L. B., Holsegard, M. H., Ottenssen, N., Svenné, S. & Rahbek, C., 2009. Spatial behaviour of Little Owls (Athene noctua) in a declining low–density population in Denmark. J. Ornithol, 150: 537–548. Van Nieuwenhuyse, D., Génot, J.–C. & Johnson, D. H., 2008. The Little Owl. Conservation, Ecology and Behaviour of Athene noctua. Cambridge University Press, Cambridge. Zabala, J., Zuberogoitia, I., Martínez, J. A., Martínez, J. E., Azkona, A., Hidalgo, S. & Iraeta, A., 2006. Occupancy and abundance of Little Owl (Athene noctua) in an intensively managed forest area in Biscay. Ornis Fennica, 83: 97–107. Zuberogoitia, I., 2002. Eco–etología de las rapaces nocturnas de Bizkaia. Ph. D. Thesis, Basque Country University, Leoia. Zuberogoitia, I. & Campos, L. F., 1997. Intensive census of nocturnal raptors in Biscay. Munibe, 49: 117–127. – 1998. Censusing owls in large areas: a comparison between methods. Ardeola, 45: 47–53. Zuberogoitia, I., Martínez, J. A., Zabala, J. & Martínez, J. E., 2005. Interspecific aggression and nest–site competition in a European owl community. Journal of Raptor Research, 39: 156–159. Zuberogoitia, I., Martínez, J. E., Zabala, J., Martínez, J. A., Azkona, A., Castillo, I. & Hidalgo, S., 2008. Social interactions between two owl species sometimes associated with intraguild predation. Ardea, 96: 109–113. Zuberogoitia, I., Zabala, J., Martínez, J. A., Hidalgo, S., Martínez, J. E., Azkona, A. & Castillo, I., 2007. Seasonal dynamics in social behaviour and spacing patterns of the Little Owl Athene noctua. Ornis Fennica, 84: 173–180.


Animal Biodiversity and Conservation 34.2 (2011)

401

Causes of admission of little owl (Athene noctua) at a wildlife rehabilitation centre in Catalonia (Spain) from 1995 to 2010 R. A. Molina–López & L. Darwich

Molina–López, R. A. & Darwich, L., 2011. Causes of admission of little owl (Athene noctua) at a wildlife rehabilitation centre in Catalonia (Spain) from 1995 to 2010. Animal Biodiversity and Conservation, 34.2: 401–405. Abstract Causes of admission of little owl (Athene noctua) at a wildlife rehabilitation centre in Catalonia (Spain) from 1995 to 2010.— This retrospective study analyzes the causes of morbidity of little owl (Athene noctua) admitted to the Wildlife Rehabilitation Centre of Torreferrussa from 1995 to 2010. A total of 1,427 little owls were included in the study, with an average of 89 cases per year (range: 73–116). As regards the sex category, 80.7% animals (1,152/1,427) were classified as undetermined gender, 9.1% (130/1,427) were sexed as females and 10.2% (145/1,427) as males. The overall age distribution according to the calendar year showed that 66.6% (951/1,427) of birds were '1st calendar year and 16.6% (237/1,427) were '> 1 calendar year'. Age could not be determined in 16.7% of birds. Primary causes of admission were orphaned young (53.2%), unknown trauma (24.7%), impact with motor vehicles (5.6%), other cause (5.5%), undetermined (3.7%), illegally captive (2.1%), natural diseases (2.1%), and gunshot (1.1%). Within the breeding season the frequency of admissions due to traumas –unknown trauma (x2 = 147.108; p < 0.001)– and impact with motor vehicles (x2 = 28.528; p < 0.001) and other cause (x2  = 11.420; p = 0.003) were the most important causes. In winter, admissions were mainly related to unknown trauma and gunshot. Over the fifteen years we observed a significant increase in the orphaned young category. Key words: Little owl, Rehabilitation centres, Morbidity, Epidemiology. Resumen Causas de la admisión de mochuelos comunes (Athene noctua) en un centro de rehabilitación de animales salvajes de Cataluña (España) desde el 1995 al 2010.— Este estudio retrospectivo analiza las causas de morbilidad de los mochuelos comunes (Athene noctua) admitidos en el Centro de Recuperación de Fauna Salvaje de Torreferrussa desde 1995 a 2010. En este estudio se incluyeron un total de 1.427 mochuelos comunes, con un promedio de 89 casos por año (rango: 73–116). Con referencia a la categoría sexual, el 80,7% de los animales (1.152/1.427) se clasificaron como de género indeterminado, un 9,1% (130/1.427) como hembras, y un 10,20% (145/1.427) como machos. La distribución general de edades, calculadas en años naturales, mostraba que un 66,6% de las aves (951/1.427) tenían un año natural, y que el 16,6% (2.37/1.427) eran menores de un año. En el 16,7% de las aves no se pudo determinar la edad. Las principales causas de admisión fueron jóvenes huérfanos (53,2%), trauma desconocido (24,7%), impacto de vehículos a motor (5,6%), otras causas (5,5%), por causas indeterminadas (3,7%), cautividad ilegal (2,1%), enfermedades naturales (2,1%), y disparos (1,1%). Durante la estación de cría, la frecuencia de admisiones debidas a traumas –trauma desconocido (x2 = 147,108; p < 0,001) e impacto por vehículo a motor (x2 = 28,528; p < 0,001)– y otras causas (x2 = 11,420; p = 0,003), fueron las causas más importantes. En invierno, las admisiones se producían principalmente en relación con traumas desconocidos y disparos. Durante el periodo de 15 años observamos un aumento significativo en la categoría de jóvenes huérfanos. Palabras clave: Mochuelo común, Centros de Recuperación, Morbilidad, Epidemiología. (Received: 8 VII 11; Conditional acceptance: 21 IX 11; Final acceptance: 21 X 11) Rafael A. MolinacLópez, Centre de Fauna Salvatge de Torreferrussa, Catalan Wildlife Service, Forestal Catalana, 08130 Santa Perpètua de la Mogoda, Espanya (Spain).– Laila Darwich, Dept de Sanitat i Anatomia Animals, Fac. of Veterinary, Univ. Autònoma de Barcelona, 08193 Bellatera, Espanya (Spain). Corresponding author: Rafael A. Molina–López ISSN: 1578–665X

© 2011 Museu de Ciències Naturals de Barcelona


Molina–López & Darwich

402

Introduction Birds of prey and owls have become valuable sentinels of environmental changes because of their position on the ecological food chain (Kovács et al., 2008; Sergio et al., 2006). Moreover, identifying and understanding causes of the variation or decline of wildlife population is essential in order to implement conservation measures (Salafsky et al., 2008). Morbidity and mortality studies, including those based on data from wildlife rehabilitation centres (WWC), have complemented the understanding of menaces posed to raptors by identifying the underlying natural and anthropogenic factors. WWCs have therefore become a key data source, providing valuable information concerning the health of ecosystems (Sleeman, 2008). The population of free–living little owls (Athene noctua) in Europe is considered to be in moderate decline (Burfield, 2008). The decrease has been related to changes in their habitat, resulting in a fragmented and isolated breeding population (Exo, 2005). Catalonia is an Autonomous Community in Spain located in the Mediterranean subregion of the western Palaearctic (3º 19'–0º 9' E and 42º 51'–40º  1'  N). Eight different owl species have been observed in this area, most of them being breeding species (Estrada et al., 2004). In Catalonia, a decrease in their distribution area has also been reported and the little owl is considered near threatened (Framis, 2011). Epidemiological studies of little owl focusing on morbidity and mortality are scarce (Hernández, 1988), especially those covering a long period of time. The objective of the present study was to analyze the main causes of morbidity and mortality of the little owl population admitted to the Wildlife Rehabilitation Centre of Torreferrussa (Catalonia) over a fifteen–year period. Methods Study design and animals A retrospective unicentric study was performed using the original medical records of wild little owls hospitalized at the Wildlife Rehabilitation Centre of Torreferrussa (Catalonia, North–East Spain) from 1995 to 2010. Non– wild born individuals and cases with no epidemiological information were excluded from the analysis. Definition of variables The following variables were included in the analyses: species, sex, age, date and primary cause of admission. Sex was determined when possible by gonadal inspection during clinical diagnostic procedures or at necropsy. Age was categorized as '1st calendar year' and '> 1 calendar year' according to Martínez et al. (2001). The year was divided into three seasons: breeding (from March to July), post–nuptial migration (from August to October) and wintering (from November to February), according to Herrando et al. (2011). General classification of primary morbidity and mortality causes was adapted from the categories defined

by different authors (Morishita et al., 1998; Wendell et al., 2002; Naldo & Samour, 2004) and were based on definitive diagnoses, as follows: trauma, toxicosis, infectious and parasitic diseases, metabolic or nutritional diseases, orphaned young birds, and undetermined. The metabolic, nutritional, infectious, and parasitic categories were grouped as natural disease. The overall trauma category was subdivided into specific causes as follows: collision, elec