Page 1

SURG Studies by Undergraduate Researchers at Guelph

Fall 2012 • Volume 6 • Issue 1

ISSN: 2291-1367


About SURG

The SURG Team

Support

Editor-in-Chief Anita Acai Associate Editor (Arts) Bethany Philpott Associate Editor (Layout) Moez Valliani Assistant Editors Conor Hedley Jessa Letargo Sara Wilmshurst Director, Research Communication Owen Roberts Project Management Liz Snyder Scholarly Communication Librarians Jane Burpee Wayne Johnston

SURG recognizes the ongoing support for research infrastructure – equipment, materials and personnel – at the University of Guelph, from the Natural Sciences and Engineering Research Council, the Canadian Institutes of Health Research, the Social Sciences and Humanities Research Council and the Ontario Ministry of Agriculture, Food and Rural Affairs, among others.

The Journal Studies by Undergraduate Researchers at Guelph (SURG, pronounced "surge") is a refereed, multi-disciplinary online journal that publishes research and review articles by University of Guelph undergraduate students. It was developed through co-operation of the University’s Office of Research, the McLaughlin Library and the Office of Graduate Studies. SURG is open-access – available for free to everyone. ISSN: 2291-1367.

Call for Papers Eligibility Undergraduate students from all disciplines at the University of Guelph are eligible to submit a paper to SURG. Most submissions will be based on summer research or fourth-year research courses. However, we encourage submissions from all undergraduates engaging in research. Requirements SURG accepts both original articles and reviews. All submissions must be approved by a faculty supervisor. Please visit our website www.uoguelph.ca/~surg for author guidelines. Pre-submission inquiries may be sent to surg@uoguelph.ca.

Call for Editorial Staff Open Access

Cover Photo (Contest Winner)

Contact Anita Acai (SURG Editor-in-Chief) Research Communications Office of Research, University Centre University of Guelph, Guelph, ON N1G 2X1 www.uoguelph.ca/~surg surg@uoguelph.ca

Jordan Raycroft

Volume 6 • Issue 1 • Fall 2012

This journal provides open access to all of its content on the principle that making research freely available to the public supports a greater global exchange of knowledge. Such access is associated with increased readership and increased citation of an author's work. For more information on this approach, see the Public Knowledge Project (www.pkp.ubc.ca), which has designed this system to improve the scholarly and public quality of research, and which freely distributes the journal system as well as other software to support the open access publishing of scholarly resources. Student papers are licensed under a Creative Commons Attribution-Noncommercial 2.5 Canada License.

SURG is looking for students interested in research and publishing to volunteer their time as editorial staff. Duties will include copy editing, layout editing, and assisting with day-to-day operations of the journal. Requirements include superior literacy skills and knowledge of Microsoft Office and/or HTML. Previous research experience is a plus. Interested students should contact surg@uoguelph.ca.

Studies by Undergraduate Researchers at Guelph (SURG) 2


Table of Contents

Welcome 4

Welcome from the SURG editor Anita Acai

Research Articles Effects of volunteerism and relationship status on empathy Melissa Milanovic Analyzing a Pentecostal “revolution”: Reflections on research methodology Lucas Burton National parks and protected areas in African countries: A free market environmentalism approach for social and environmental sustainability Zuzanna Drewniak, Kaitlyn Finnegan, Charlotte Miles, and Meredith Miller Tourism in Kenya’s national parks: A cost-benefit analysis Hubert Cheung

5

14

23

31

Educational assortative matching in the United States and the effect on income inequality by households from 1960 to 2005 Kathryn Swierzewski

42

Differences in composition of territories in relation to behaviour, stage, and depth of the three-spot damselfish, Stegastes planifrons, in Caribbean coral reefs Maria J. Arroyo Gerez

52

In silico modelling of Arp1 and Arp2 as targets of Photox ADP-ribosylation Lianne P. Davidge and Kristen V.L. Daly

58

Review Articles Iron in the brain: Heavy metal mismanagement Erin L. Stephenson

72

Volume 6 • Issue 1 • Fall 2012

Studies by Undergraduate Researchers at Guelph (SURG) 3


Editor’s Welcome

Welcome to SURG’s Fall 2012 edition! As SURG’s new editor-in-chief, I am delighted to present to you the 11th issue of Studies by Undergraduate Researchers at Guelph (SURG)! As always, we continue our tradition of highlighting undergraduate research from throughout the University of Guelph. Our current issue brings together a compelling mix of articles that highlight the diverse research interests that exist at our university. Our issue begins with a study on the link between volunteerism, empathy, and relationship status and a student’s candid reflection of his experiences as a summer research assistant in the Department of History. After this, it takes a more global perspective through an examination of free market environmentalism as a strategy for social and environmental sustainability in managing African national parks and a cost-benefit SURG Editor-in-Chief Anita Acai analysis of national park tourism in Kenya. Our next Bachelor of Science (Biochemistry) Department of Molecular & Cellular Biology article takes a look at the relationship between marriage based on education level and income inequality in the United States. Finally, you’ll get a chance to explore the world of science with an assessment of the effects of Damselfish behaviours on coral reef and algal diversity in the Caribbean, a study that models interactions between a bacterial toxin and two of its potential binding partners, and a review of multiple sclerosis including potential causes and therapies.

Happy reading and may you have a prosperous 2013,

Anita Acai SURG Editor-in-Chief

Volume 6 • Issue 1 • Fall 2012

I am incredibly grateful to the SURG editorial and advisory teams, faculty reviewers at the University of Guelph, all student authors whose work is featured in this edition of SURG, and of course, our readers – we could not have produced such a successful issue without your generous contributions. I also owe thanks to (Dr.) Rob Fieldhouse – my predecessor and SURG’s longest-serving editor-in-chief – who has made significant contributions to bringing SURG to the outstanding publication that it is today. I look forward to continuing to work with SURG to feature undergraduate research at the University of Guelph.

Studies by Undergraduate Researchers at Guelph (SURG) 4


Research Article

Effects of volunteerism and relationship status on empathy Melissa Milanovic Department of Psychology, College of Social and Applied Human Sciences, University of Guelph, Guelph, ON Canada. Faculty supervisor: Jeffrey Spence. For correspondence, please email: milanovmelissa@gmail.com.

Abstract Volunteerism and relationship status differences were examined for their relatedness to empathy level, in the context of a psychological instrument for measuring empathy, the Baron-Cohen and Wheelwright Empathy Quotient (EQ). This Likerttype scale consisting of 60 items was completed by each of 100 participants, who were categorized based on self-declared volunteerism (a volunteer or not) and relationship status (in a relationship or not). Volunteerism was found to be a statistically significant factor in empathy level: those who volunteered exhibited higher empathy levels than those who did not. Relationship status was additionally statistically significant, such that those in a relationship had higher empathy levels than those who were not. When analyzed together, the factors of volunteerism and relationship status showed significant interaction in their influence on the dependent variable, empathy. Keywords: volunteerism; relationship status; empathy

Introduction Behaviours related to empathy Certain circumstances or experiences may be related to the extent to which an individual is empathic. The present study was developed with the intent of investigating two such important factors, relationship status and volunteerism, and how they correlate with the way that individuals behave towards others. Relationship status is expected to be correlated with empathic behaviour, because being in a committed relationship with another individual requires consideration for their thoughts and feelings and a desire to care for their wellbeing. Similarly, given that volunteers are individuals who willingly devote their time to aiding others without monetary compensation, those who elect to be volunteers may demonstrate more empathic behaviour. Thus, both volunteerism and relationship status are hypothesized correlates of empathy. Background and literature review There are varying degrees of volunteerism, but research has found some concrete results pertaining to the connectedness of empathy and volunteerism. Wilhelm and Bekkers discovered that individuals who express high levels of empathic concern have particularly high levels of helping behaviour [4]. In line with this conclusion is evidence presented by Roberts and Strayer, which states that empathy

Volume 6 • Issue 1 • Fall 2012

By definition, empathy is the ability to understand and share another person’s emotion while being sensitive to their thoughts and feelings [1]. It additionally entails an appropriate affective response by the observer to the other individual’s mental state. In a cognitive sense, being capable of empathic expression involves being able to understand how another person is feeling, and responding to this nonegocentrically [2,3]. The capacity to be empathic, or to express empathy, ultimately allows one to understand the behaviours and feelings of others and predict their behaviour. Hence, it allows effective interaction within a social context. By nature, humans are social beings. Our lives are primarily influenced by interactions with others, and the concern that we hold for others plays an important role in the course of our interactions and relationships with them. Empathy is a behaviour exhibited consistently by few individuals because it requires looking outside of oneself and focusing on what another person is experiencing. In a world of frequent competition for money, jobs, and opportunities, it is often difficult to forget about our own needs and instead focus our efforts on another. Thus, it may be useful to delve into the roots of empathy, and determine factors that are correlated with higher levels of empathy. With this knowledge, individuals might be better able to predict how their actions might affect others and work towards better understanding the thoughts and feelings of those around them.

Studies by Undergraduate Researchers at Guelph (SURG) 5


Effects of volunteerism and relationship status on empathy (Milanovic) is an important contributor to prosocial behaviour, defined as voluntary behaviour intended to benefit another with no goal other than to help fellow humans [5]. It thus appears that higher levels of empathy are correlated with one’s likelihood of engaging in activities, such as volunteering, that seek to aid other individuals without material gain. It is further postulated that being in a relationship affects empathic expression. One particular investigation found that people who are married are more likely to volunteer than their single counterparts [6]. Additional research has demonstrated that individuals are generally more likely to volunteer for a cause if they are in a committed relationship, regardless of their sex [7]. This, in addition to findings by Roberts and Strayer suggests that individuals in relationships exhibit more empathic and prosocial behaviour and are thus driven towards volunteer roles [5]. This postulation could be probable, were it not for contradicting results presented by Stolinski and colleagues [8]. In particular, they discovered that participants, who were volunteers, were driven to engage in volunteer roles because of the challenge it presented and the importance that they felt that role possessed. This leads to the question of whether individuals engage in volunteer work for the challenge is presents, or for reasons associated with empathy. Perhaps individuals in relationships do show more empathic concern, but whether this shows correlation with greater volunteerism among such individuals remains ambiguous. Thus, uncertainty remains in regard to whether being in a relationship and partaking in volunteerism are significantly related to one’s empathy level. Objectives

Participants Empathy level was measured for a set of participants (N = 100, Mage = 20.880, SDage = 1.328, age range: 18 – 26), including both males (n = 46) and females (n = 54). All participants came from the population of students attending the University of Guelph. Thus, the participant sample was a convenience sample, recruited directly by the researcher. All participants completed the Empathy Quotient in its entirety. Materials The materials consisted of a booklet containing the Empathy Quotient. After completing the EQ via pen and paper, participants declared their volunteer status (whether they were a volunteer or not) and relationship status (whether they were in a relationship or not) by check box format. Additionally, demographic information at the end of the booklet asked participants to provide their age and sex. As such, participants assigned themselves into one of four categories: in a relationship and a volunteer, in a relationship and not a volunteer, not in a relationship and a volunteer, and not in a relationship and not a volunteer. The Empathy Quotient The Empathy Quotient (EQ), created by BaronCohen and Wheelwright was used as the measurement tool to assess empathy for each participant [1]. The EQ consisted of 60 questions, 40 of which measured empathy, and 20 of which were filler items. Examples of items that assessed empathy were, “I really enjoy caring for other people”, and “I am good at predicting how someone will feel”. An example of a filler question was “I am at my best first thing in the morning”. The filler items were simply included to avoid a continuous focus on empathy-related items, and to reduce the testing threat to internal validity. The EQ was a Likert-formatted scale, and thus respondents had the option to respond to each item with one of four responses, including “strongly agree”, “slightly agree”, “slightly disagree”, and “strongly disagree”. Approximately half of the empathy-based items were worded to produce a “disagree” response, and half to produce an “agree” response towards empathy, to avoid response bias. To account for this, reverse scoring was used on half of the items. The Empathy Quotient was selected primarily for its reliability in assessing empathy. Previous literature provided a reliability of r = 0.97 for the EQ [1]. Additionally, it was selected for its straightforward scoring scheme since the individual EQs were hand-scored (refer to Appendix 2 for scoring rules).

Volume 6 • Issue 1 • Fall 2012

The present study was developed with the intention of adding to previous literature in the area of empathy and the factors that are related to its exhibition. A comparison of empathy levels between those in a relationship and those not, as well as between those who are volunteers and those who are not, is needed to see whether empathy is related to each factor. In the present research, empathy is measured using the Baron-Cohen and Wheelwright Empathy Quotient (EQ) [1] (Appendix 1). Three hypotheses are postulated. Primarily, it is hypothesized that there will be a correlation between volunteerism and empathy, such that individuals who are volunteers will exhibit higher empathy levels than those who are not volunteers. Further, it is hypothesized that there will be a correlation between relationship status and empathy, such that individuals who are in a relationship will exhibit higher levels of empathy than those who are not. Finally, an interaction effect is hypothesized, such that those who are both in a relationship and are volunteers will have the highest overall level of empathy.

Methods

Studies by Undergraduate Researchers at Guelph (SURG) 6


Effects of volunteerism and relationship status on empathy (Milanovic) Procedure Data collection occurred over a period of approximately two weeks in calm, relatively private areas with minimal distractions. A standardized description of the research was presented to each participant to ensure consistent and standardized interaction between the participant and researcher. Upon receiving informed consent from each participant, the researcher gave all of them a package containing the EQ scale, demographic information, and self-declaration check boxes for relationship status and volunteerism. Since participants were self-assigned into the conditions of relationship status and volunteerism, the design takes the classification of non-experimental. Upon completion of the package, participants were each debriefed, and any concluding questions were answered. Participants did not receive any form of compensation for participation. Completed surveys were maintained in a safe, private location until all data were collected. Once data was collected from all 100 participants, all EQs were hand-scored, and the data were entered into a master SPSS file for appropriate analyses.

Results

Relationship status and empathy The second hypothesis (H2) stated that individuals who were in a relationship would exhibit higher levels of empathy than those who were not in a relationship, thus indicating a correlation between being in a relationship and empathy. Data analysis showed a significant correlation between relationship status and empathy, such that those who reported being in a relationship had significantly higher empathy scores than those who were not in a relationship, F(1, 94) = 4.795, p = 0.031, ηp2 = 0.049. Observed power was 0.582. For those in a relationship, M = 46.328, SE = 2.028 and for those not, M = 39.895, SE = 2.017. Thus, the second proposed hypothesis (H2) was not rejected. Volunteerism and relationship status on empathy The third hypothesis (H3) predicted an interaction effect between volunteerism and relationship status, such that those who were volunteers and in a relationship would have a higher overall level of empathy than all other groups. A significant interaction effect between volunteerism and relationship status was found, F(1, 94) = 4.448, p = 0.038, ηp2 = 0.045. Observed power was 0.551. Specifically, the empathy level of those who were volunteers and in a relationship (M = 53.554, SE = 3.012) was significantly higher than those who were volunteers and not in a relationship (M = 41.185, SE = 2.943), those who were not volunteers and were in a relationship (M = 39.101, SE = 2.813), and those who were not volunteers and were not in a relationship (M = 38.606, SE = 3.004) (Figure 1, Appendix 3). Thus, the third proposed hypothesis (H3) was not rejected.

Volunteering and empathy The first hypothesis (H1) stated that individuals who were volunteers would exhibit higher levels of empathy than those who were not volunteers, thus indicating a correlation between volunteerism and empathy. Data analysis showed a

Figure 1. Plot detailing volunteerism by relationship status. Note the absence of parallelism between both interaction slopes, which clearly displays the significant interaction between the factors.

Volume 6 • Issue 1 • Fall 2012

Analyses focused on each participant’s overall empathy level score. A two-way between subjects factorial Analysis of Variance (ANOVA) was used. Although it was hoped that the ANOVA assumption of groups having the same sample size would be met, numbers were slightly unequal in two of the four conditions. Refer to Appendix 3 for the sample size and descriptive statistics for each condition. A Levene’s Test of Equality of Error Variances was conducted to assess the assumption that the error variances of the dependent variable (empathy score) was equal across groups. The Levene’s Test value was significant, and as such, the null hypothesis of equality of error variances was not accepted, F(3, 96) = 3.884, p = 0.011. Since the resulting significance value of Levene’s Test was below the critical value of 0.05, the obtained differences in sample variances was unlikely to have occurred based on random sampling. As such, a difference exists between the variances in the population. This finding is not unreasonable considering a non-experimental design was used, and by definition, that entails that samples were not randomly selected. Since the equality of error variances is an assumption of ANOVA, caution must be exercised in interpreting the results obtained.

significant correlation between volunteerism and empathy, such that those who reported being volunteers had significantly higher empathy scores than those who were not volunteers, F(1,94) = 6.988, p = 0.010, ηp2 = 0.069. Observed power was 0.744. For volunteers, M= 47.369, SE = 2.178 and for non-volunteers, M = 38.854, SE = 2.077. Therefore, the first proposed hypothesis (H1) was not rejected.

Studies by Undergraduate Researchers at Guelph (SURG) 7


Effects of volunteerism and relationship status on empathy (Milanovic) Item analyses Item analyses were conducted on the data obtained from the 100 Empathy Quotients as completed by participants. Specifically, the 40 items that measured empathy were analyzed for internal consistency, and the resulting reliability was strong (α = 0.944). As such, it is fair to conclude that the 40 items assessed the same construct as the entire test to a high degree. The distribution of total empathy scores obtained by each participant approached a normal distribution. The maximum possible empathy score an individual could receive on the EQ was 80, and the minimum possible score was zero. The sample results indicated a mean total empathy score of 43.060, and standard deviation of 16.549. As such, the majority of total empathy scores clustered around the central range of possible scores and the standard deviation provided indication of an acceptable variance in the scores (Figure 2). It should again be noted that Levene’s Test of Equality of Error Variances was significant, and as such, the ANOVA assumption of equality of error variances between groups was not met. Although ANOVA is a relatively robust measure of analysis and all proposed hypotheses attained statistical significance, one should not uncritically assume their correctness without noting the lack of equality of error variance.

Discussion In the present study, volunteerism and relationship status were found to be correlated with empathy level, both individually and when considered together. In line with previous research, it was discovered that those who volunteer have higher levels of empathy than those who do not [4,5].

Practical implications In a world where empathic expression can sometimes be lacking, learning from those who are more likely to express it is an important tool in increasing our general knowledge on how to better demonstrate empathy. Since this research has determined sub-groups of people who show enhanced empathy, we can look towards further studying them. In particular, learning about their motives and underlying intentions for being attentive to the feelings of others might be a useful extension of this study. In terms of real world examples of who would benefit from those with higher levels of empathy, one should consider individuals with abusive tendencies. Abusers, or those who show abusive tendencies at an early age, might learn from the proper demonstration of empathy, and from the teachings of those who have been found to possess higher levels of empathy. In addition, encouraging these individuals to partake in activities that are correlated with higher levels of empathy, for example volunteering, could serve as a beneficial tool in remediation programs or correctional facilities. Thus, a better understanding of who in our society might be more empathic would aid in helping those who show consistent low regard for the thoughts and feelings of others. Additional real life implications of this research could be to use the findings to aid in selecting the best-suited individuals for tasks or jobs that require high levels of empathy (e.g. helping professions such as doctors or nurses). Research by Loran and colleagues revealed that people tend to underestimate the severity of others’ social pain, and the effect that ostracism has on others [9]. Further analysis of the sub-groups (that were determined in the present study to be substantially high in empathy level) could

Volume 6 • Issue 1 • Fall 2012

Figure 2. Histogram detailing the frequency of participants who obtained each of the possible scores, erected over discrete range intervals pertaining to the measure of empathy score, the Empathy Quotient (EQ).

Additionally, the finding that those in relationships exhibit higher levels of empathy than those who are not is also consistent with the literature [6,7]. The findings of Stolinski and colleagues brought upon a debate of whether the choice to volunteer is caused by higher levels of empathy, or if it is perhaps based on a separate factor such as the motivation to work through the challenges brought on by volunteering [8]. The present study found compelling evidence that there is indeed a link between volunteerism, relationship status, and empathy. However, this evidence does not preclude the possibility that other variables may be involved in the relationship between volunteerism and empathy, such as those discussed by Stolinski and colleagues [8]. Rather, it has contributed to the growing body of literature in this field. Specifically it has provided further evidence in addition to previous research that empathy likely plays a role in one’s decision to volunteer, or that volunteering instills empathy. Further, results of the present research provide evidence in line with previous investigation that people in relationships express higher levels of empathy, or that empathic people tend to be in relationships, which could result in their greater seeking of volunteer roles.

Studies by Undergraduate Researchers at Guelph (SURG) 8


Effects of volunteerism and relationship status on empathy (Milanovic) provide more insight into whether empathy is a learned trait or if it is in part hereditary. In the event that empathy is acquired at least in part through learned behaviours, studies could examine what methods are the most effective in teaching individuals to be considerate of others. This is particularly relevant in settings where bullying is of concern; for example, in middle schools and high schools, students could benefit from workshops and extra discussion aimed at developing their consideration for the feelings of others. Theoretical implications In a theoretical sense, extending our knowledge about the basis of empathy will help in our understanding of why humans are motivated to help and care for others. Psychological altruism is a motivation state with the ultimate goal of increasing the welfare of others [10]. As such, altruism is an important component for enhancing the survival of another. Having pointed out particular groups of individuals who appear to be significantly more empathetic, the world of altruism is at our fingertips to further explore. While examining the precise evolutionary implications of altruistic behaviour might be a stretch beyond what is possible within the scope of this research, it is ultimately the improvement in our understanding of empathy and our ability to express it and care for others that can in the long run help us enhance survival of others. Strengths

Like most non-experimental designs, limitation lies in the lack of random assignment of individuals to conditions. As such, a selection threat to internal validity results and it is difficult to prove equality among groups that are self-assigned. This research specifically lacks proof for the assumption of equal error variances between the groups and as such, all further results needed cautionary treatment. In order to truly assess whether the effects are statistically significant, a more robust analysis such as the BrownForsythe test (which is recommended as a robust measure for non-normal data) might be considered, in order to determine if statistical significance remains. Further limitation results from the restricting effect that a small sample size of one hundred participants places on the power and effect size. In order to have the most effective analysis and obtain the best results possible, a large sample size is of great importance. Additionally, instrumentation effects (a threat to internal validity) might have had an effect on results. In particular, the EQ is rather long, and instrumentation effects pertaining to being tired might have affected the respondents’ performances. The present study used a method of self-report data. This introduces the possibility of response bias, which could consequently alter proper observation of effects. Further, the declarations of relationship status and volunteerism were rather ridged, allowing for only one of two possible declarations: in a relationship/not in a relationship and volunteer/not a volunteer. As such, it does not allow for more specific relationship- and volunteer-based analysis. Perhaps individuals in differing types of volunteer positions differ in their empathy levels. Or, perhaps there is a difference in empathy between those in more casual versus long-term relationships. Thus, the ridged definitions of relationship status and volunteerism used in this research prohibit the investigation of more in-depth analyses of varying levels of each factor. Future research Future research along the same lines as the present study should consider increasing sample size, as it would increase the relevancy of the results obtained. In addition, further research might take the effects of relationship type, length or commitment level on empathy into consideration. Perhaps individuals in marital versus casual relationships would have significant differences in empathy level, and research into such specifics would improve knowledge regarding the factors that underlie empathy. Furthermore, in this research, it was decided not to assess the effect of sex on empathy due to compelling results presented by Hill and colleagues that sex does not have particular effects on empathy [7]. Despite this, additional research that

Volume 6 • Issue 1 • Fall 2012

The study provided evidence in favour of the three proposed hypotheses, and as such, an improved general understanding of situations in which individuals might be more likely to show empathy has resulted. Despite using a non-experimental design, the resulting sample size (determined through self-declaration) of each of the four conditions was almost a perfect split of the one hundred participants. Having as close to equal number of participants per condition as possible enhances power, or the ability to detect effects when they do exist. Effect size is a measure of the difference that exists between groups. Cohen describes a general rule that one does not want an effect size below 0.2 [11]. The present study managed to obtain an effect size greater than this in all tests of hypotheses; a fair extent of the variability between the groups of the study can be explained by their true differences. Furthermore, the power of each test of between subjects is acceptable. Each test’s power surpasses the level of chance, and therefore the likelihood of detecting an effect when there truly is one is greater than the likelihood of committing a type II error (otherwise stated as finding an effect when it actually does not exist) in all three tests.

Limitations

Studies by Undergraduate Researchers at Guelph (SURG) 9


Effects of volunteerism and relationship status on empathy (Milanovic) incorporates the effect of sex on empathy might be of use in further confirming or debating Hill and colleagues’ conclusion. Finally, on a neurobiological level, there may be certain areas of the brain or brain pathways that are implicated in the existence, development, or maintenance of empathic behaviour. Research into the neurological aspects of empathy would provide a more empirical analysis. Such investigation might further the research in the present paper by investigating whether there are areas of the brain that are particularly active in individuals who are in relationships or are volunteers, and whether these areas are linked to regions of the brain known to be involved in the empathetic response. To conclude, the present study has determined two subgroups of individuals who exhibit higher levels of empathy: individuals who are volunteers, and individuals who are in a relationship. As such, the potential to gain more insight into empathy through further analysis of these individuals has been revealed. The discoveries made up to this point regarding empathy should be acknowledged, and future attempts made to expand upon the understanding of factors and mechanisms that underlie empathy.

References Baron-Cohen, S., & Wheelwright, S. (2004). The Empathy Quotient: An investigation of adults with Asperger’s Syndrome or high functioning autism, and normal sex differences. Journal of Autism and Developmental Disorders, 34(2), 163–175.

2.

Kohler, W. (1929). Gestalt psychology. New York: Liveright.

3.

Piaget, J. (1932). The moral judgment of the child. London: Routledge & Kegan Paul.

4.

Wilhelm, M. O., & Bekkers, R. (2010). Helping behaviour, dispositional empathic concern, and the principle of care. Social Psychology Quarterly, 73(1), 11–32.

5.

Roberts, W., & Strayer, J. (1996). Empathy, emotional expressiveness, and prosocial behaviour. Child Development, 67, 449–470.

6.

Sundeen, R. A. (1990). Family life course status and volunteer behaviour: implications for the single parent. Sociological Perspectives, 33(4), 483–500.

7.

Hill, C. T., Rubin, Z., Peplau, L. A., & Willard, S. G. (1979). The volunteer couple: sex differences, couple commitment, and participation in research on interpersonal relationships. Social Psychology Quarterly, 42(4), 415 – 420.

8.

Stolinski, A. M., Ryan, C. S., Hausmann, L. R. M., & Wernli, M. A. (2004). Empathy, guilt, volunteer experiences, and intentions to continue volunteering among buddy volunteers in an AIDS organization. Journal of Applied Biobehavioural Research, 9(1), 1– 22.

9.

Loran, F. N., Banas, K., & MacDonald, G. (2011). Empathy gaps for social pain: why people underestimate the pain of social suffering. Journal of Personality and Social Psychology, 100(1), 120– 128.

10. Batson, C. D., & Moran, T. (1999). Empathyinduced altruism in a prisoner’s dilemma. European Journal of Social Psychology, 29(7), 909–924. 11. Cohen, J. (1988). Statistical Power Analysis for the Behavioural Sciences. Hillsdale, NJ: Erlbaum.

Volume 6 • Issue 1 • Fall 2012

1.

Studies by Undergraduate Researchers at Guelph (SURG) 10


Supplementary Information

Effects of volunteerism and relationship status on empathy (Milanovic)

Appendix 1 The Baron-Cohen and Wheelwright Empathy Quotient (EQ) Below is a list of statements. Please read each statement carefully and rate how strongly you agree or disagree with it by circling your answer. There are no right or wrong answers, or trick questions. strongly agree

slightly agree

slightly disagree

2. I prefer animals to humans.

strongly agree

slightly agree

slightly disagree

3. I try to keep up with the current trends and fashions.

strongly agree

slightly agree

slightly disagree

4. I find it difficult to explain to others things that I understand easily, when they don’t understand it the first time. 5. I dream most nights.

strongly agree

slightly agree

slightly disagree

strongly agree

slightly agree

slightly disagree

6. I really enjoy caring for other people.

strongly agree

slightly agree

slightly disagree

7. I try to solve my own problems rather than discussing them with others.

strongly agree

slightly agree

slightly disagree

8. I find it hard to know what to do in a social situation.

strongly agree

slightly agree

slightly disagree

9. I am at my best first thing in the morning.

strongly agree

slightly agree

slightly disagree

10. People often tell me that I went too far in driving my point home in a discussion. 11. It doesn’t bother me too much if I am late meeting a friend.

strongly agree

slightly agree

slightly disagree

strongly agree

slightly agree

slightly disagree

12. Friendships and relationships are just too difficult, so I tend not to bother with them. 13. I would never break a law, no matter how minor.

strongly agree

slightly agree

slightly disagree

strongly agree

slightly agree

slightly disagree

14. I often find it difficult to judge if something is rude or polite.

strongly agree

slightly agree

slightly disagree

15. In a conversation, I tend to focus on my own thoughts rather than on what my listener might be thinking. 16. I prefer practical jokes to verbal humor.

strongly agree

slightly agree

slightly disagree

strongly agree

slightly agree

slightly disagree

17. I live life for today rather than the future.

strongly agree

slightly agree

slightly disagree

18. When I was a child, I enjoyed cutting up worms to see what would happen.

strongly agree

slightly agree

slightly disagree

19. I can pick up quickly if someone says one thing but means another.

strongly agree

slightly agree

slightly disagree

20. I tend to have very strong opinions about morality.

strongly agree

slightly agree

slightly disagree

21. It is hard for me to see why some things upset people so much.

strongly agree

slightly agree

slightly disagree

22. I find it easy to put myself in somebody else’s shoes.

strongly agree

slightly agree

slightly disagree

23. I think that good manners are the most important thing a parent can teach their child 24. I like to do things on the spur of the moment.

strongly agree

slightly agree

slightly disagree

strongly agree

slightly agree

slightly disagree

25. I am good at predicting how someone will feel.

strongly agree

slightly agree

slightly disagree

26. I am quick to spot when someone in a group is feeling awkward or uncomfortable. 27. If I say something that someone else is offended by, I think that that’s their problem, not mine. 28. If anyone asked me if I liked their haircut, I would reply truthfully, even if I didn’t like it. 29. I can’t always see why someone should have felt offended by a remark.

strongly agree

slightly agree

slightly disagree

strongly agree

slightly agree

slightly disagree

strongly agree

slightly agree

slightly disagree

strongly agree

slightly agree

slightly disagree

30. People often tell me that I am very unpredictable.

strongly agree

slightly agree

slightly disagree

31. I enjoy being the center of attention at any social gathering.

strongly agree

slightly agree

slightly disagree

32. Seeing people cry doesn’t really upset me.

strongly agree

slightly agree

slightly disagree

strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree

Volume 6 • Issue 1 • Fall 2012

1. I can easily tell if someone else wants to enter a conversation.

Studies by Undergraduate Researchers at Guelph (SURG) 11


Effects of volunteerism and relationship status on empathy (Milanovic) 33. I enjoy having discussions about politics.

strongly agree

slightly agree

slightly disagree

34. I am very blunt, which some people take to be rudeness, even though this is unintentional. 35. I don’t tend to find social situations confusing.

strongly agree

slightly agree

slightly disagree

strongly agree

slightly agree

slightly disagree

36. Other people tell me I am good at understanding how they are feeling and what they are thinking. 37. When I talk to people, I tend to talk about their experiences rather than my own. 38. It upsets me to see an animal in pain.

strongly agree

slightly agree

slightly disagree

strongly agree

slightly agree

slightly disagree

strongly agree

slightly agree

slightly disagree

39. I am able to make decisions without being influenced by people’s feelings.

strongly agree

slightly agree

slightly disagree

40. I can’t relax until I have done everything I had planned to do that day.

strongly agree

slightly agree

slightly disagree

41. I can easily tell if someone else is interested or bored with what I am saying.

strongly agree

slightly agree

slightly disagree

42. I get upset if I see people suffering on news programs.

strongly agree

slightly agree

slightly disagree

43. Friends usually talk to me about their problems as they say that I am very understanding. 44. I can sense if I am intruding, even if the other person doesn’t tell me.

strongly agree

slightly agree

slightly disagree

strongly agree

slightly agree

slightly disagree

45. I often start new hobbies but quickly become bored with them and move on to something else.

strongly agree

slightly agree

slightly disagree

46. People sometimes tell me that I have gone too far with teasing.

strongly agree

slightly agree

slightly disagree

47. I would be too nervous to go on a big rollercoaster.

strongly agree

slightly agree

slightly disagree

48. Other people often say that I am insensitive, though I don’t always see why.

strongly agree

slightly agree

slightly disagree

49. If I see a stranger in a group, I think that it is up to them to make an effort to join in. 50. I usually stay emotionally detached when watching a film.

strongly agree

slightly agree

slightly disagree

strongly agree

slightly agree

slightly disagree

51. I like to be very organized in day-to-day life and often make lists of the chores I have to do. 52. I can tune into how someone else feels rapidly and intuitively.

strongly agree

slightly agree

slightly disagree

strongly agree

slightly agree

slightly disagree

53. I don’t like to take risks.

strongly agree

slightly agree

slightly disagree

54. I can easily work out what another person might want to talk about.

strongly agree

slightly agree

slightly disagree

55. I can tell if someone is masking their true emotion.

strongly agree

slightly agree

slightly disagree

56. Before making a decision I always weigh up the pros and cons.

strongly agree

slightly agree

slightly disagree

57. I don’t consciously work out the rules of social situations.

strongly agree

slightly agree

slightly disagree

58. I am good at predicting what someone will do.

strongly agree

slightly agree

slightly disagree

59. I tend to get emotionally involved with a friend’s problems.

strongly agree

slightly agree

slightly disagree

60. I can usually appreciate the other person’s viewpoint, even if I don’t agree with it.

strongly agree

slightly agree

slightly disagree

strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree strongly disagree

Volume 6 • Issue 1 • Fall 2012

Studies by Undergraduate Researchers at Guelph (SURG) 12


Information SupplementaryInformation Supplementary

Effects of volunteerism and relationship status on empathy (Milanovic)

Appendix 2 The Empathy Quotient Scoring Scheme Positively Worded Items These items are stated in such a way that responding towards the “agree” end of the response scale indicates a higher level of empathy. “Strongly agree” responses score 2 points and “Slightly agree” responses score 1 point on these items: 1, 6, 19, 22, 25, 35, 36, 37, 38, 41, 42, 43, 44, 52, 54, 55, 57, 58, 59, 60. These items receive a score of zero if responded with “Strongly disagree” or “Slightly disagree” by the participant. Negatively Worded Items These items are stated in such a way that responding towards the “disagree” end of the response scale indicates a higher level of empathy. “Strongly disagree” responses score 2 points and “Slightly disagree” responses score 1 point for items: 4, 8, 10, 11, 12, 14, 15, 18, 21, 27, 28, 29, 32, 34, 39, 46, 48, 49, 50. These items receive a score of zero if responded with “Strongly agree” or “Slightly agree” by the participant. Filler Items Filler items (items 2, 3, 5, 7, 9, 13, 16, 17, 20, 23, 24, 30, 31, 33, 40, 45, 47, 51, 53, and 56) are not scored, and thus do not contribute to overall empathy score. Range of Total Scores On each item, a participant can score a 0, 1, or 2. This results in a maximum possible empathy score of 80, and a minimum possible empathy score of zero.

Appendix 3 Participant between Subject Factors (Relationship Status and Volunteerism) Statistical Values 95% CI Relationship Status

Volunteerism

N

M

SE

LL

UL

Not in a Relationship

Not a Volunteer

27

38.606

3.004

32.642

44.570

Is a Volunteer

23

41.185

2.943

35.342

47.028

Not a Volunteer

25

39.101

2.813

33.516

44.687

Is a Volunteer

25

53.554

3.012

47.574

59.534

In a Relationship

Note: CI = confidence interval; LL = lower limit, UL = upper limit

Volume 6 • Issue 1 • Fall 2012

Studies by Undergraduate Researchers at Guelph (SURG) 13


Research Article

Analyzing a Pentecostal “revolution”: Reflections on research methodology Lucas Burton Department of History, College of Arts, University of Guelph, Guelph, ON Canada. Faculty supervisor: Femi Kolapo. For correspondence, please email: 1lucas.burton@gmail.com.

Abstract This study evaluates a research questionnaire and responses in a recent study on the influence of the Pentecostal movement in Nigeria. It observes areas of potential confusion within the questionnaire due to its wording, and suggests revisions and additional questions that might improve the relevance of the questionnaire. Through further analysis, it notes questions that may be unintentionally written with bias, thereby influencing responses, and discusses what are suggested to be the advantages and disadvantages of formatting questions in a specific manner. Finally, this study explains the process of using the aforementioned questionnaire to create an electronic database in which to organize responses. It evaluates the use of the database as a research tool and comments on the place of quantitative analysis as a whole within the study of history. Analysis of the questionnaire and database concludes that both were largely successful in accomplishing the goals of the study for which they were created. The database is found to be a valuable tool as it provides an efficient way to establish trends and focus further analysis of the information collected. The importance of this tool to the research project suggests that quantitative analysis can be of great use in historical research, though it should not replace traditional research methodology in this field. Keywords: mixed research methodologies (evaluation of); survey research; religion and politics; Pentecostal revival; Nigeria

Introduction which was distributed to a number of Pentecostal churches in Nigeria, as a sampling of what was hoped to become a much larger and more intensive study; 54 questionnaires were filled out by Pentecostal Christians in Nigeria. My first task was to render each hand written response to the questionnaire into an electronic format for easy access and information storage. At times, this proved more challenging than I initially expected, as I often dealt with illegible hand writing in addition to spelling and grammatical errors. My first instinct was to correct the grammar and spelling of the responses as I typed them, but Dr. Kolapo explained that this was tantamount to tampering with evidence as doing so would modify the respondent’s explanation or response. Having taken his advice, I soon found this process of transcribing to be incredibly helpful to my understanding of the study as a whole. After spending approximately one week reading each questionnaire as I transcribed them, I became intimately aware of the varying opinions and perspectives of the respondents surveyed, and already began noticing patterns and trends in their responses. Because the questionnaire did not ask for the names of respondents, before I began transcribing, I labelled each questionnaire with a unique identification code so we could

Volume 6 • Issue 1 • Fall 2012

Following the end of the 2010-2011 school year, I began a full-time research assistantship with Dr. Femi Kolapo, of the History Department at the University of Guelph, in which I worked with him to explore the relationship between the religious and the political domains in Nigeria. The goal of our study was to establish how or whether an acclaimed Pentecostal revival, commonly referred to as the Born Again movement, has impacted the political realm of Nigeria, and whether Pentecostal Nigerians view themselves as “a force with the concrete ability to effect change by impacting policies and governance” [a]. This research tests out Dr. Ruth Marshall’s principal position in her book, Political Spiritualities: the Pentecostal Revolution in Nigeria. Dr. Marshall employs Foucauldian understandings of history, power, and subjectivation to argue that Pentecostal Christianity is not only the “single most important sociocultural force in southern Nigeria today,” but also a movement with significant impact on the political realm [1]. Our study investigated the same relationship between religion and governance in Nigeria. There were a number of elements involved in my position as a research assistant to Dr. Kolapo. Prior to my arrival, Dr. Kolapo had created a four-page questionnaire,

Studies by Undergraduate Researchers at Guelph (SURG) 14


Analyzing a Pentecostal “revolution” (Burton)

Research Questionnaire The questionnaire used in this study is relatively simple. After asking for some basic personal information from the respondent, including their church affiliation, native town/village/city and state, profession, and age bracket, the remainder of the questionnaire is divided into six sections. The first of these, “Identity”, includes questions on what defines a Pentecostal Christian and what differentiates them from other Christians. The second section, “Process”, deals with the respondent’s experience in becoming a Pentecostal Christian, while the third, “Rupture”, details how the respondent’s conversion to Christianity has affected their relationships with family, friends, co-workers etc., and their lifestyle as a whole. Section four, “Personal Development”, discusses the impact that the respondent’s faith has had on their career progress, as well as the professions they believe to be out of bounds to the Pentecostal Christian. Section five, “National Development”, is by far the most detailed and extensive of the six sections, and asks questions on the role that Pentecostal Christians should have in the political domain, as well as the impact the Pentecostal movement currently has in Nigeria. It also includes questions that allow the respondent to express their opinion regarding how to address the issue of bad governance in Nigeria, as well as on the godly response to a corrupt government and to inter-faith conflict. Finally, section six, “Doctrine”, asks respondents to detail the Biblical position they accept as a guide for their life on a variety of topics ranging from sin and prosperity, to ill-health and what God allows Pentecostal Christians to do in regards to addressing bad governance. This section also includes a few other questions such as how often the respondent has changed churches or visited a church other than their own in the last several years and the last twelve months respectively. In reading through this questionnaire, it becomes evident that the overall organization of the questions it includes is coherent and well thought out. After beginning with a section on the respondent’s understanding of the basic theology of the faith, it makes a logical progression into the respondent’s experience of conversion, followed by the influences of such conversion on their personal life. Thereafter, it delves into a more specific line of questioning related to national development, thereby shifting the focus from the personal to a larger social level, before finishing with questions related to doctrine. However, a careful analysis of each question in addition to the pattern of responses reveals flaws within the construction of the questionnaire. Some questions proved to be poorly worded and unclear, while others, I feel, might have had a predisposing influence on the responses received. Moreover, in view of the responses received, it becomes evident that various other questions should be added, particularly in the National Development section, in order to ensure the reliability—and the reality—of the views expressed therein.

Volume 6 • Issue 1 • Fall 2012

distinguish them from one another. I labelled each questionnaire “PER_0XX”, where “XX” represents a number from 01 to 54. This proved to be extremely helpful throughout the research project, particularly when it came to completing my next task: entering the information into a database. My next task was to use Microsoft Excel and Microsoft Access to create an electronic database in which the information we collected could be organized. The larger purpose for creating this database, a process I will later describe in greater detail, was to have the ability to run queries on the information collected through the questionnaires in order to establish patterns and connections between respondents and their perspectives. Once this database was complete, I began running the queries and analysing our information for trends and inconsistencies. At the same time, I also began reading Dr. Ruth Marshall’s book on the Pentecostal “revolution” in Nigeria, in order to both understand and analyse her perspective on what defines political agency [1]. Each of these tasks provided me with a number of learning experiences, as most of them were unlike anything I have ever completed for class projects or papers. That is not to say, however, that my experience was easy or straightforward every step of the way. For instance, reading Dr. Marshall’s book was at times very difficult as her analysis relies heavily on Foucauldian philosophy, which I was unfamiliar with at the time. Likewise, constructing the electronic database was entirely new to me as it marked my first experience using Microsoft Access. Nonetheless, I appreciated the opportunity to be so closely involved within the research project. In this article, I detail the various aspects of my project with Dr. Kolapo in five sections, this introduction inclusive. In the second section I provide a description of the questionnaire used in this study and identify problems with respondents’ answers resulting from some questions that are confusing. In the third section, I discuss questions that may unduly influence the answers of respondents and suggest additional questions that can be added to the questionnaire to make it more relevant. The section also includes a discussion of what I hazard to be the advantages and disadvantages of employing checklists in questionnaires such as the one used in this study. In the fourth section, I discuss the process involved in my rendering the information that I transcribed into a database using Microsoft Access. I also comment on the advantages of creating and using this database and the questionnaire as a whole for this study. The fifth and final section summarizes the major conclusions that were determined from my project.

Studies by Undergraduate Researchers at Guelph (SURG) 15


Analyzing a Pentecostal “revolution” (Burton)

Presumptions and Question Format This exercise has alerted me to how the construction of a questionnaire and its questions might influence the responses it receives. Overall, it seems to me that the majority of the questions in this questionnaire do not appear have any bias associated with them, and the responses do not indicate otherwise. However, it can be argued that there are a few questions under the National Development section of the questionnaire that presuppose the underlying opinions of the respondents surveyed, thereby possibly influencing their response. There is also one question, though, that may have had a significant impact on responses, simply based on its construction. The first question under the National Development section of the survey asks, "Why do you think Nigeria is in the backward and underdeveloped state it is [?]" This question assumes that the respondent believes that Nigeria is indeed in a backward and underdeveloped state, when this may not be true. This is evidenced by the responses of two respondents: “Nigeria is not in the backward and underdeveloped state!” stated the first, and, “Nigeria is making progress, Niger[ia] of ‘today’ cannot be compared with Nigeria of ‘yesterday’,” emphasised the second. Although the responses of all other respondents surveyed indicate their support for the basic premise of this question (with exception of five respondents who did not respond to the question), it is possible that if the question was written from a neutral perspective, responses would change and possibly become more balanced as a whole. Perhaps the question would have been clearer if it was divided into two parts. The first part might ask, “In your opinion, is Nigeria in a backward and underdeveloped state?” while a second part, such as, “Why or why not?” would allow the respondent to support their opinion with relevant information. A similar issue is raised with the second question in this section, which asks, “In what area of political life does the Pentecostal movement in Nigeria already have an impact [?]” To a lesser extent, it can be argued that this question too assumes that the respondent believes that the Pentecostal movement is having a political impact in Nigeria, and might therefore influence them to answer in favour of that perspective. With that in mind, the responses we received do not indicate such an influence. Eighteen respondents answered by either expressing that the Pentecostal movement is having little if any impact on political life, or by indicating that they were unaware of any impact it might be having. This suggests that the respondents surveyed were more than able to express their opinion regardless of how the question was worded. Although it appears that this question is simply testing one of Marshall’s conclusions regarding the impact of the Pentecostal movement on the respondent, it might be better to make an effort to eliminate this presupposition from the question entirely - this could be achieved rather easily by rewording the question or dividing it into two parts as suggested above with the previous question.

Volume 6 • Issue 1 • Fall 2012

Two questions within the questionnaire stand out as being potentially confusing and, in the case of one of them, as being misunderstood by the vast majority of respondents surveyed. The first of these questions under the Personal Development section reads, “What professions or careers do you consider to be out of bounds for Pentecostal Christians [?]” Below the question is a list of ten professions or professional categories (e.g. “business”) in checklist format. Above the checklist reads, “tick, the appropriate ones.” Although the responses we received seem to indicate that the vast majority of respondents understood that the question asks them to tick off the professions that they considered out of bounds to Pentecostal Christians – in effect the inappropriate ones – it was evident that a few respondents who checked off every profession or category listed may have been confused by the wording of the question. Although the intention was that they check off the professions they believe to be inappropriate, it is likely that these respondents thought the question asked to them to check of those they believe to be appropriate for Pentecostal Christians. Later, I will discuss in greater detail the advantages and disadvantages of using a checklist of answers within such a questionnaire. The question that seems to have caused the most confusion is found under the Doctrine section. This particular question is a part of an overarching enquiry that asks the respondent to detail the “Biblical position that you accept as guide for your life in the following areas . . .” The area that caused the most confusion, as evidenced by the question being largely unanswered, was entitled “Managing versus changing reality.” The question hoped to address whether the respondent believes that God calls them to manage their reality (including, for example, the reality of living under a corrupt government), or to take action to change that reality. The majority of respondents did not answer this question at all, and although a few respondents attempted to provide an answer, almost all of those who did seemed to have misunderstood the question’s intention. Two respondents in particular wrote, “I don’t quite understand this,” and “I don’t know you’re [sic] intention.” I find it unfortunate that the confusion over this question lead to it being largely unanswered, as it speaks directly to the issue at the heart of this study – the Pentecostal Christian’s perspective on their power and freedom to change their reality. Having identified these and a few more minor areas of confusion within the questionnaire, it became evident that it might be helpful to reword some questions and explain others in greater depth to ensure that the future responses we receive are relevant, and fully address what is being asked. Nevertheless, this has certainly opened my eyes to the challenges involved in creating questionnaires and surveys for use in such a research project. Though I once imagined doing so to be a relatively straightforward task, I now see the various difficulties that can arise through the process. The reality of these difficulties became all the more apparent as I began to analyse the questionnaire in terms of how its nature may have influenced the responses we received.

Studies by Undergraduate Researchers at Guelph (SURG) 16


Analyzing a Pentecostal “revolution” (Burton) When reading through the completed questionnaires and trying to establish patterns and relationships in responses, I was intrigued by the consistency in responses to one particular question under the National Development section. The question reads, “What should guide who Pentecostal Christians should vote for in Nigerian elections (e.g. records of performance; being B.A [born-again], or Christian, or being popular) [?]” Of the forty-eight respondents who replied to this question, an astounding forty-one listed records of performance as one of the necessary guides to the Pentecostal Christian vote. This volume of consistent responses led me to question whether the same pattern would be present were a list of potential guidelines not included with the question. In reading the questionnaires again, I noticed that multiple respondents were using the examples listed in brackets as a checklist. Instead of writing out a response, many respondents underlined or circled one or more of the guidelines listed. Having recognized this, I do not believe it is unreasonable to suggest that were these suggestions not listed, not as many respondents would have mentioned records of performance when answering. There was not as much consistency in the responses concerning other potential guidelines (i.e. being B.A, being a Christian, and being popular). This suggests that including a list of potential guidelines may not have influenced the responses we received as much as I initially supposed; many respondents may have genuinely believed records of performance to be an important guideline, and may have included it even if it were not listed. It is also worth noting that the note written at the end of the survey may have had influenced the responses to the questionnaire overall. At the end of the questionnaire, Dr. Kolapo wrote a brief note that outlines the major goals of his research project [b]: . . . to map out, however subtle it may be, how or whether Pentecostal Christianity . . . has impacted the Pentecostal Christian and the larger Nigerian society, especially, its politics . . . and whether the Pentecostal Christians see themselves as a force with concrete ability to effect change by impacting policies and governance.

Volume 6 • Issue 1 • Fall 2012

Including this note with the questionnaire may have had one of two effects. On one hand, it may have brought greater context to the purpose of the questions asked and thus fostered more relevant answers. On the other hand, upon realizing the implication of the study on views of the Pentecostal movement and its role in Nigerian society, it may have led respondents to write more favourably of the movement and its impact on the political realm. When I raised this issue with Dr. Kolapo, his response was to explain that scholars are increasingly cautious not to imply that even the most disadvantaged or ignorant of people are unable to honestly express their opinions when given the opportunity. He also noted the need for researchers to inform their

interviewees of what their intentions are as a matter of respect. They cannot simply begin interviewing respondents without providing an explanation as to why they are demanding that the respondents give them their attention. While I still feel it is important to consider the potential influence this note may have on responses in order to draw the correct conclusions within the larger scope of this study, I appreciate that Dr. Kolapo informed me of some ethical considerations that researchers must deal with in the field. Whatever the case, one of the things I have learnt discussing this with Dr. Kolapo and working on the materials, is the critical importance of ensuring that one does not subtly influence responses by the way the questions are worded or phrased, or by what is included or excluded. After analysing the completed questionnaires using the electronic database, Dr. Kolapo and I began to discuss a common shortcoming we both agree is present within the questionnaire’s construction. Although the questionnaire contains multiple questions related to the respondent’s political perspective and beliefs on political activity, it lacks questions that ask the respondent to provide evidence regarding how they have personally acted on those beliefs. Consequently, a situation arises in which almost all respondents indicated their support of voting in elections, but we have no means of verifying that they actually did vote, for example, in the recent April 2011 federal election. The design of the questionnaire enables the respondents to indicate their support for a number of other political activities including party politics, contesting for votes, protest, revolution, etc., without providing a history of practice that confirms their support for these. This issue is significant when viewed in context of the goal of this study, as it seeks to determine not only what Pentecostal Christians believe they should do with respect to political agency, but what they actually do. Having said that, Dr. Kolapo and I agree that the addition of a few questions regarding the respondent’s record of political participation will help to address this issue. Examples of such questions might include, “Did you vote in the April 2011 election?”, “Are you currently a registered member of a political party?”, “Have you ever participated in a political protest?”, etc. Responses to questions such as these will help clarify to what extent Nigerian Pentecostal Christian philosophy on political agency is actively put into practice, thereby making the information collected through the questionnaires more relevant to the goals of the research project. Earlier in this section and in the previous section, I discussed two instances in which the employment of a checklist question format (or what was used as a checklist) had a potential influence on the responses we received. Since there is a total of eight questions within the questionnaire that make use of a checklist to some extent, as I analysed the information we collected, I began to wonder whether or not the advantages of using a checklist in a survey such as this one outweigh the potential disadvantages. Using a checklist of potential answers following a given question can be extremely useful when looking for a definite answer on a

Studies by Undergraduate Researchers at Guelph (SURG) 17


Analyzing a Pentecostal “revolution” (Burton) subject from respondents. For example, one question in the Doctrine section of our questionnaire asks, “. . . what does God allow Pentecostal Christians to do to change bad government to good [?]” The question is followed by a checklist that includes a number of forms of political activity (e.g. voting, protest, party politics, revolution, etc.), in addition to an “other” field. Such a checklist enabled us to directly measure the beliefs of respondents on specific methods of political activity, which may have otherwise not been commented on were the question left open for any answer. This proved to be highly advantageous when we constructed a database with information gathered from the questionnaire as it made quantifying the information much easier. Instead of having to make inferences as to whether or not a given respondent was in support of a certain form of political activity based on their written response, their opinion was more clearly represented. With that in mind, using a checklist in general may mean missing the opportunity to foster answers with greater depth and detail. This is especially significant when dealing with very broad ideas. For example, although we included “protest” and “revolution” as activities in our above checklist, the definition and understanding of these activities can vary considerably. This is particularly true in the case of this study, as some Pentecostal Christians will refer to acts of spiritual discipline, such as prayer and evangelism, as having “revolutionary” influence in and of themselves. Consequently, without constructing the question in a way that leads the respondent to provide more detail, it may become easier to misinterpret the true opinions and ideas of the respondents and thus the larger population they represent. It seems to me that having a balance of questions containing checklists and others left open for more detailed answers might help achieve a survey that is useful for collecting information on specific ideas and perspectives while also providing the context and detail needed to better understand those perspectives.

Constructing an Electronic Database

Volume 6 • Issue 1 • Fall 2012

Another important task accomplished is my creation of an electronic database of the information collected through the questionnaires. At first, this task seemed somewhat daunting, as Dr. Kolapo wanted to construct the database using Microsoft Access – a program that neither of us has used before. Nevertheless, the prospect of acquiring a new skill set and becoming familiar with Access was motivation enough. We began by outlining our vision for the database and its function. In addition to providing the means of organizing information such as the age, church affiliation, profession, and political views of each of the respondents surveyed, the database gave the ability to run searches, or “queries”, on the information collected in order to discern trends and relationships. The database was to be constructed in such a way that, within seconds, one could create a table

that displayed, for example, the church affiliation of each respondent in addition to their opinion on whether or not Christians should vote, protest, contest for votes. etc. As I began to research Microsoft Access, however, I noticed that the use of the program was largely for the purposes of organizing more numeric, and quantitative data. The first question I faced, therefore, was how I would be able to quantify what was evidently qualitative information so that it could be used in the database. Before addressing this challenge, Dr. Kolapo and I discussed exactly what information from the questionnaire we wanted to have represented in the database. We decided on a total of seven categories: Individual, Church Affiliation, Professions, Political Action, Explanation for the Nigerian Crisis, Christian Voting Guidelines, and Response to Interfaith Conflict. The “Individual” category includes information such as the age of the respondent, whether they are a member of the clergy or a layman, whether they live in an urban or rural region and, most importantly, whether their responses to the questions as a whole indicate that they have a “clear” or “unclear” political view. Both the “Church Affiliation” and “Professions” categories are selfexplanatory, while the “Political Action” category includes the respondent’s perspectives on Christians’ participation in voting, protests, party politics, contesting for elective positions, revolutions, etc. The “Explanation for Nigerian Crisis” category contains the respondent’s explanation for why Nigeria is in a “backward and underdeveloped state”, while the “Christian Voting Guidelines” category includes perspectives on what should guide the Pentecostal Christian vote. Finally, as the name suggests, “Response to Inter-faith Conflict” details the respondent’s beliefs on the response that Pentecostal Christians should have to the ongoing conflict between Christians and Muslims in northern Nigeria. Having established these categories as a framework, I began constructing a preliminary version of the database using Microsoft Excel. Because I have far more experience with this program, my goal was to set the structure and organization of the database using Excel before attempting to move the information into Microsoft Access. Using a separate spreadsheet for each of the seven categories, I created tables in which the information under each category could be organized. Since our data was by no means numeric (with the exception of ‘age’), I used headings to represent various questions from the questionnaire. For example, the headings included in the Political Action table are “Vote”, “Protest”, “Contest”, “Revolutions”, “Preach/Speak Out”, “Prayer”, and “Other”. The majority of these headings came directly from three different questions within the questionnaire, which were formatted as checklists. An additional heading, which is present in the table under each category, is “Person”. Under this heading, the ID codes for each respondent, PER_001 to PER_054, was listed. If, for example, when responding to the questionnaire, PER_001 indicated their support for voting and protesting, but not revolutions (by checking

Studies by Undergraduate Researchers at Guelph (SURG) 18


Analyzing a Pentecostal “revolution” (Burton) immediate use to me. Instead, I began watching YouTube videos of Access tutorials that Dr. Kolapo found online. These videos were incredibly helpful and after only a few hours, I was confident enough in my understanding of the program to begin constructing the database. In many regards, Access is quite similar to Excel, and at first glance the organization of our data looks nearly identical in both programs. Each category has its own table that is filled out just as it was in Excel. The significant difference between these databases is the ability to create “relationships” between tables in Access which allow execution of queries on the data. With each table in our database containing the “Person” heading, under which the unique ID codes of each respondent were listed, this heading became what is known in Access as the “Primary Key” that connects every table and their respective categories together; after setting up these relationships, the first version of our Access Database was complete. I immediately ran a few test queries on the data to ensure that things were operating smoothly and was extremely pleased to find it working successfully. When I first began working with Dr. Kolapo, we discussed the research project as a whole including the goals of the study and the desire to incorporate quantitative information using an electronic database. At the time, Dr. Kolapo mentioned that some students and scholars within the discipline of history might look with disfavour on the use of such methodology in a study that deals with people. There seems to be concern that the use of quantitative information can lead researchers to dehumanize the nature of a given study and ignore the reality faced by the people behind the numbers. Indeed, the extent to which quantified and numeric data has its place within a study such as ours may be debateable, but to reject entirely the opportunity to incorporate such information through a database such as we produced, is, in my opinion, to neglect what can be a highly useful tool. The amount of information organized and represented in our database is already immense, and I cannot imagine attempting to analyse so much information without the use of this tool. The value of this database is seen all the more when one considers that it represents the responses of only fifty-four respondents. As I mentioned earlier, Dr. Kolapo has plans to dramatically increase the scope of this study to include many hundreds, if not thousands, of respondents. Furthermore, the database perfectly accomplishes the goals we outlined for it, enabling us to efficiently and almost instantly tally the number of respondents surveyed who, for example, are affiliated with a specific church or believe that prayer is a necessary response to inter-faith conflict. Likewise, we can just as easily establish correlations between these categories and analyse how church affiliation is related to perspectives on inter-faith conflict. There is virtually no limit to the number of queries that can be run on the data and therefore the number of ideas that can be tested with respect to what influences

Volume 6 • Issue 1 • Fall 2012

off or not checking off these activities respectively), I simply wrote “Vote” under the “Vote” heading and “Protest” under the “Protest” heading, while leaving the cell under the “Revolutions” heading blank in the row dedicated to PER_001. Although simply entering “Yes” or “No” would be appropriate, I soon found that rewriting the heading name made it easier to read the information in the table. Using this method, I created headings for the remaining categories, and began rereading each questionnaire and filling out each category’s table for each respondent. Filling out some tables proved more difficult than others, as there were no “checklist-style” questions directly associated with the information they included. In such cases, I created headings based on the most consistent responses we received. For example, the information included in the “Explanation for Nigerian Crisis” category, is derived from a single question in the questionnaire. As I read through each response I soon noticed that “corruption”, “greed”, “poor leadership”, and “leaders who don’t fear God” were among the most consistent responses. Logically, I used these as the headings for the table under this category, and included an “Other” category to accommodate responses that do not fit under these headings. By far the hardest heading to fill out was “Political Idea” found under the Individual category. Filling out this heading required me to make an overarching judgement on whether or not each respondent was found to have a “clear” on an “unclear” political view based on their responses to the questionnaire. Dr. Kolapo and I discussed this process and defined a “clear” political view as one that exhibits a definite perspective on the political realm in Nigeria, and provides evidence of the respondent’s understanding of how to address issues related to bad governance. I understand that this definition is highly contestable, and although entire books could be devoted to arguing either for or against it, this article will not enter into the debate. While the responses of some respondents were very easily identified as relating to a clear or unclear political view, others were more difficult to immediately determine and required more thought and analysis. Although this made for a long process, it once again enabled me to become very familiar with the responses to the questionnaires themselves, which proved helpful when it came to analysing the data as a whole. With the creation of the Microsoft Excel database complete, it was time to attempt to recreate the database in Microsoft Access. Prior to beginning my work on the Excel database, I spent some time researching Access in order to become more familiar with the program. Initially, I borrowed from the library what can only be described as the Access manifesto, which boasted and astounding 734 pages of detailed explanations for every function of the program. Although I thought reading this book would be helpful, I soon realized that it was far too dense and detailed to be of any

Studies by Undergraduate Researchers at Guelph (SURG) 19


Analyzing a Pentecostal “revolution” (Burton) Pentecostal Christians and what perspectives they hold. The database also validates the interdisciplinary nature o f the current research, and especially highlights the sociological and ethnographic aspects of it that are clearly amenable to useable quantification. Having said that, the true value of the database is realized when it is used to point us back to the written responses to the questionnaire. It goes without saying that the database is by no means a replacement for these responses, nor can it fully represent the perspectives or opinions of the respondents surveyed. As Geoffrey Barraclough notes, quantitative history should not be used as a replacement for qualitative analysis, and neither does it seek to be [2]. However, by providing an efficient way to begin establishing trends and correlations within the information collected, databases can indicate what areas and what specific responses of the questionnaire to focus on when analyzing such patterns. In short, as Dr. Ruth Marshall is explicit in pointing out in her book, one must be careful not to become reductionist in analysis when dealing with a topic as complex and abstract as religion and its interaction with other social domains [1]. However, it seems to me that approaching this area of study using a mixed research methodology allows for the benefits that both quantitative and qualitative methods have to offer.

Conclusion

Acknowledgements I would like to offer my sincere thanks to the University of Guelph for funding my work with Dr. Femi Kolapo of the History Department as an undergraduate research assistant. This research assistantship is funded as a component of the University of Guelph Pamela Wallin Chancellor’s Scholarship. I would also like to thank those who reviewed earlier drafts of this article and Dr. Kolapo who approved this final version. My experience as a research assistant over the spring and summer months of 2011 has been very rewarding, and working with Dr. Kolapo has been both an honour and an absolute pleasure. I can say with certainty that I have learned a great deal about research methodology and the research process through both our discussions and the numerous learning opportunities Dr. Kolapo has provided me with. Working on this study has led me to respect much more the research process and the amount of time and effort that is invested into a research project. Not only have I gained a new interest in the impact of religion and faith on other societal domains, particularly at the state level, I am confident that the skills I have learned and the opportunities I have had during this assistantship have prepared me for future involvement in similar studies.

References 1. Marshall, Ruth. Political Spiritualities: The Pentecostal Revolution in Nigeria. Chicago: University of Chicago Press, 2009. 2. Barraclough, Geoffrey, and Michael Burns. Main Trends in History. New York: Holes & Meier, 1991 [1979].

Endnotes a.

From the final note of the questionnaire administered by Dr. Femi Kolapo for his research on the impact of the acclaimed Pentecostal revival in Nigeria on the political realm therein.

b.

Ibid.

.

Volume 6 • Issue 1 • Fall 2012

The questionnaire developed by Dr. Kolapo for this research project is in line with the goals of the study and will enable said goals to be accomplished. Even though there is always room for improvement, it seems to me that the overall organization of the questionnaire is excellent and the types of questions it includes foster relevant and specific discussion on both the personal and societal impact of the Pentecostal movement. This is particularly important when viewed in the context of the philosophy underlying Dr. Ruth Marshall’s analysis of this movement and its impact. Dr. Marshall emphasizes the Pentecostal movement’s focus on personal, internal revival, and self-mastery, all of which are equally represented in the first two sections of this questionnaire, which provide the respondent opportunity to detail the impact of Pentecostal Christianity on their life. Just as Dr. Marshall investigates the implication of this personal revival on the political realm in Nigeria, this questionnaire fluidly transitions into questions related to the respondent’s perspective of, and the movement’s role in this domain. Moreover, creating an electronic database to organize the responses to this questionnaire furthers our ability to more thoroughly investigate the perspectives of Pentecostal Nigerians, and to engage directly with the ideas and conclusions set forth by Dr. Marshall. Having spent many hours reading the responses to the questionnaires and running queries through our

database, I have already begun to see patterns that both support and conflict with Dr. Marshall’s findings. I eagerly anticipate the opportunity to expand the scope of this study to include more respondents, thereby providing more opportunity for analysis.

Studies by Undergraduate Researchers at Guelph (SURG) 20


Supplementary Information

Analyzing a Pentecostal “revolution” (Burton)

Appendix 1 The questions and note listed below comprised the questionnaire used in Dr. Kolapo’s study. Current Church Affiliation ____________________________ [No need for church address, pls.] Your native town/village/city & the state __________________________________ Your profession: Pastor, yes[ ] no [ ] Other _______________________________ Your age bracket [ ]15 -20 [ ]21-30 [ ]31-49 [ ]50-70 [ ]71-90 THE STRUCTURE OF NIGERIAN PENTECOSTAL SPIRITUALITY To remain anonymous, you do not have to write down your name or the address of your church. 1.

IDENTITY a. Who is a Pentecostal Christian [?] b. How are Pentecostal Christians different from non-Pentecostal Christians [?] c. Is it what happens to them or how they handle life that makes them different [?]

2.

PROCESS a. Why did you become a Pentecostal Christian [?] b. How did you become a Pentecostal Christian [?] {Your experience of becoming} c. Was your process of conversion instantaneous or gradual over time [?]

3. RUPTURE? How did your conversion to Christianity affect your relationships with a. Family (parents, brothers and sisters & extended family) [?] b. Immediate family (spouse and children) [?] c. Friends you had before you were born again [?] d. Work / or study [?] e. Your life style [?] f. Participation in traditions and customs of your town or village [?] g. Development organizations that include Christians and non-Christians [?] h. Town meetings [?] i. Do witches and wizards still have an impact on the life of a Pentecostal Christian [?] j. What part of or type of past life can still affect the Pentecostal Christian [?] PERSONAL DEVELOPMENT a. Can/ Should / Did being a Pentecostal Christian affect your progress (career, professional, business, income, etc.?) b. What professions or careers do you consider to be out of bounds for Pentecostal Christians [?] [Tick the appropriate ones.] [ ] comedian, [ ] musician, [ ] modeling, [ ] politician, [ ] criminal lawyers, [ ] soldiering, [ ] police, [ ] transport [ ] business, [ ] oil business, etc.

5.

NATIONAL DEVELOPMENT a. Why do you think Nigeria is in the backward and underdeveloped state it is [?] b. In what area of political life does the Pentecostal movement in Nigeria already have an impact [?] c. How can the Pentecostal movement affect national development and good governance, especially given that more people in Nigeria are non-Pentecostal Christians, Muslims, and traditional worshippers [?]

d. e. f. g. h. i. j.

Politics Should Pentecostal Christians go into politics, joining any of the political parties to contest for posts [?] [ ]Yes [ ]No Can pastors comment publicly about bad government or bad government policies [?] [ ]Yes [ ]No What should the proper response of Pentecostal Christians be toward bad government [?] What should guide who Pentecostal Christians should vote for in Nigerian elections (e.g. records of performance; being B.A. or Christian; or being popular) [?] What specific requests do you make to God for better government and development in Nigeria [?] Which, do you prefer, [ ] democracy, [ ] military rule, or [ ] rule by a powerful king [?] What is the responsibility of a Pentecostal Christian when a bad government gets worse and rigs election to stay on [sic] in power [?]

Volume 6 • Issue 1 • Fall 2012

4.

Studies by Undergraduate Researchers at Guelph (SURG) 21


Analyzing a Pentecostal “revolution” (Burton) k. l.

What are the responsibilities of the Pentecostal Christian in the face of Christian-Muslim crises in the northern part of the country [?] What are your three list [sic] of how to solve the problem of poor or bad governance in Nigeria [?]

6.

DOCTRINE: What is the biblical position that you accept as guide for your life in the following areas: a. Sin [?] b. Poverty [?] c. Prosperity [?] d. Government [?] e. Ill-health [?] f. Turning the other cheek [?] g. Prayer versus action [?] h. Managing versus changing reality [?] i. Family planning [?] j. What must we allow God to do and what does God allow Pentecostal Christians to do to change bad government to good [?] [ ] protest, [ ] party politics, [ ] revolutions, [ ] voting, [ ] other _________________________________________ k. Are Pastors called to prepare church members for heaven or to make them better people to enjoy and recreate the earth [?]

7.

How many times have you changed church a. In the past 3 years [?] _______________________________________________ b. In the past 5 years [?] _______________________________________________ c. In the past 10 years [?] ______________________________________________

8.

How many times did you attend a different Sunday or weekday service in other churches in the last 12 months [?] [ ] once [ ] twice [ ] less than 5 times [ ] more than 5 times [ ] up to 10 times

9.

Are there Christian churches you cannot go [?] [ ] Yes [ ] No Why?

10. If you come from a village or a town (not a city); how many Pentecostal churches do you have there [?] _______________________________________________________________________________________ Note: My major goal in this research is to try to map out, however subtle it may be, how or whether Pentecostal Christianity or what some people call the Born Again movement has impacted the Pentecostal Christian and the larger Nigerian society, especially, its politics and governance at federal, state and local levels, and whether the Pentecostal Christians see themselves as a force with concrete ability to effect change by impacting policies and governance.  You are at liberty to withdraw your agreement to participate in this research during the study and before any publication on it goes to press and have your data withdrawn.  Thanks for answering as many of the questions above as you can. As a guarantee that I will not disclose your name and your responses to these questions, in any type of publication, I ask that don’t put down your name, except if you want to.  The findings from this research will be available to all only after publications as journal articles or as a book or part of a book.  In addition to your exercising of the right to not write your name on the questionnaire, envelopes are provided for you to put the completed questionnaire, so that when you drop it off, nobody would be able to identify its content with you.

Volume 6 • Issue 1 • Fall 2012

Thank you very much for taking your time to assist me in answering these questions.

Studies by Undergraduate Researchers at Guelph (SURG) 22


Research Article

National parks and protected areas in African countries: A free market environmentalism approach for social and environmental sustainability Zuzanna Drewniak†, Kaitlyn Finnegan†, Charlotte Miles†, and Meredith Miller† Dept. of Food, Agriculture, and Resource Economics, Ontario Agricultural College, University of Guelph, Guelph, ON Canada. Faculty supervisor: Glenn Fox. For correspondence, please email: mmille05@uoguelph.ca. † denotes equal contributions.

Abstract There are two primary options for the successful preservation of national resources in African national parks: centralized government management and decentralized privatized management. In this article we argue that a free market environmentalism approach to the management of national parks in Africa is preferable to centralized government management. We begin by discussing social, economic, and biophysical trends related to the operations of national parks in Africa. Next, we describe the institutional and political structures of management options, including the conventional centralized planning model, and present alternatives such as co-management and privatization. We then identify current conflicts and controversies regarding national park management in the African context, which include land tenure and expropriation, poverty, and the protection of large mammalian endangered species. Finally, we apply the free market environmentalism approach to African national park management and make a case for why this approach would allow for better protection of endangered large mammal species, benefit African citizens in the surrounding communities, eliminate the free-rider incentive which can lead to acts such as poaching, and create incentives that are necessary for the preservation of African national resources. We conclude that this market-based system is effective in protecting natural resources in areas of Africa where the private owners are willing to pay for the preservation of the environment, and on privatized and which can be successfully profitable through the aid of competition in the market. Keywords: national parks; Africa; free market environmentalism; sustainability (environmental, social)

Introduction resources and amenities, therefore government controlled land containing natural resources such as national parks and protected areas are required [1]. It has also been argued, however, that market forces, through a decentralized market structure, are a better option, especially for developing countries, which have consistently undervalued natural resources, pricing them below their actual market values [4,5]. This market-based system would create economic incentives for managers and would result in improved environmental quality and sustainability in areas (such as national parks) within developing countries in Africa that are currently under government ownership [4]. To answer the question of what is the most viable management option for African National Parks, it is

Volume 6 • Issue 1 • Fall 2012

Government ownership of natural areas in the form of natural parks is a relatively new phenomenon, developing globally in the late 19th to early 20th century [1,2]. The phenomenon developed in light of the realization that there could be a potential scarcity of natural resources in the future from unsustainable use by private landowners – examples include over-grazing, slash-and-burn brush clearing techniques, and exploitation of the natural resources present on the land [1]. The first purpose of government owned (or managed) national parks is to protect the natural, historic and cultural heritage values on an area of land [3]. The second purpose is to provide opportunities for recreation and tourism [3]. There is a general consensus globally among resource managers, (including government agencies and populations), that private ownership is unable to preserve environmental

Studies by Undergraduate Researchers at Guelph (SURG) 23


National parks and protected areas in African countries (Drewniak et al.)

important to observe the current trends, institutional and political structure, and the problems and controversies related to national parks and protected areas in African countries. These components will be discussed in the following sections of this article in order to move towards identifying a solution for African National Park management. Here, we argue that a free market environmentalism approach to the management of national parks in Africa is preferable to centralized government management because it would better manage the protection of endangered large mammal species, as well as benefit African citizens in the surrounding communities. The free market environmentalism ideology is based on a decentralized market structure, and argues that this system (which is based on private property and tort law) is the best way to preserve the environment and its resources [6]. A free market approach and the commercial use of natural resources will create incentives that are necessary for the preservation of Africa’s natural resources. The way in which African National Parks are managed influences the conservation and preservation of natural resources generally in Africa. Lessons can be learned from the African National Parks that can inform National Park management elsewhere. Distinguishing between management approaches such as free market environmentalism and centralized government management techniques illustrates the pros and cons of each management system. A successful management system can be defined as one which protects valuable natural resources in the short and in the long term. This perspective is important because, without proper protection, over-exploitation of natural resources occurs, and this inevitably leads to a lack of resources required to sustain the human population. Our conclusion is that instituting a free market environmentalism approach for African National Park Management would provide the necessary protection of natural resources as well as the rights of local indigenous peoples. The encouragement and involvement of local indigenous peoples surrounding the protected parks allows for both scientific and local knowledge to be incorporated in the protection of these natural areas through the use of private property.

Based on the IUCN category system, Africa has a total of 1166 protected areas, which cover a total of 2,808,250 square kilometers (23.3%) of eastern, southern, western and central Africa, and 1,272,840 square kilometers (9.7%) of northern Africa and the Middle East [10,11]. The largest percentage of protected areas in the African continent consists of both habitat and species management areas and national parks [10]. The continent of Africa has a total of 338 national parks, with the majority residing within the borders of South Africa, Kenya, and Madagascar [10]. The majority of Africa’s 50 largest national parks (established between 1970 and 1980), are within the borders of Botswana, Malawi, Namibia, Tanzania, Uganda, Zambia and Zimbabwe (Table 1) [10]. In Kenya and South Africa, only one national park in each country is considered to be one of the 50 largest because although there are numerous national parks in these countries, each of the parks’ individual areas are relatively small [10,12]. Revenue to finance operations of African national parks and game reserves is obtained from tourism revenues, sport hunting revenues, government funding and global donation/support programs [13]. The tourism revenue that a park receives depends on park entry fees and the number of tourists visiting the park [13]. National park entrance fees for tourists vary widely between individual African countries [14]. In most countries, less than the equilibrium price is charged for park entry [1]. In 2002, national park visitor entry fees varied from free entry in Kiang West National Park in Gambia, to upwards of USD 100 in Tanzanian National Parks [14]. Africa has experienced major growth in the tourism industry in the last decade, indicating increased visitation and tourism revenues [15]. This growth indicates an increase in tourism in all parts of Africa between 2000 and 2008 [13]. Table 1. National parks in African countries (2010). Note: African countries omitted from this Table may have established protected areas, but do not have established national parks under the IUCN classification system. Reproduced with permission from [11].

Volume 6 • Issue 1 • Fall 2012

Social, Economic, and Biophysical Trends of African Protected Areas

Distribution of protected areas and national parks

A protected area is defined as a recognized geographical space that is dedicated and managed by the government or privately to achieve the long-term conservation of natural resources, ecosystem services and cultural values [7]. The International Union for Conservation of Nature (IUCN) divides protected areas into categories according to management objectives; these include areas such as nature reserves, natural monuments and national parks [8]. The percent of the earth’s surface covered in national parks and protected areas has increased from 3% in 1962 to 11.5% in 2005 [9].

Studies by Undergraduate Researchers at Guelph (SURG) 24


National parks and protected areas in African countries (Drewniak et al.) Funding for African protected areas and national parks Although tourist entrance fees can contribute to the funds used for park support in protected areas like game reserves, the majority of the funds used for this purpose are obtained from sport and trophy hunting revenues [13]. In the Selous Game Reserve in Tanzania, both tourism and hunting increased between 1991 and 2001, but hunting revenues consistently amounted to approximately 80-90% of the total park income compared to approximately 10% from tourism alone throughout the decade [13]. Although global spending on protected areas is USD 6.5 million, the required amount to fully support conservation programs and parks has been estimated to be between USD 12 and 45 billion [5]. National parks and protected areas in Africa receive less than 30% of the funding that is necessary for adequate park management. In the past several decades, there has been a decline in the amount of funding available to African National Parks. Estimates have stated that already inadequate park budgets have been cut in half in the last ten years [16]. Budget shortfalls in selected African countries ranged from approximately USD 220,000 to over 7,000,000 per year [16]. One consequence of these budget shortfalls is the existence of “paper parks”, parks that only exist on paper due to the lack of financial support for staff, vehicles and ongoing conservation programs [5].

Institutional and Political Context of National Parks in Africa

The conventional national park model Within the conventional national park model, land is publically owned, reserved for conservation, and has been withdrawn from the market and from the effects of market

Co-managed protected areas Co-managed protected areas have been designated by national, sub-national or local governments but are managed by all stakeholders that depend on the area culturally and/or for their livelihoods. The parties involved in a co-managed protected area negotiate a management plan, which is generally part of a broader agreement that includes complementary initiatives, by-laws, incentives and compensation [2]. The main difference between co-management and the conventional management model is that indigenous and local communities who occupy or use the lands and resources within the protected area can claim customary and/or legal rights to the land and resources when the areas are comanaged [2]. Other stakeholders, however, have different incentives and entitlements within the protected area and comanagement agreements do not give them equal weight in consultation and decision-making as those with customary or legal rights. Privatization of protected areas An alternative to publically owned natural resources and national parks is privatization, which is the conversion of publically owned land to private ownership [7]. The idea that privatization would be a better alternative to government management of parks and protected areas than government

Volume 6 • Issue 1 • Fall 2012

The conservation practices in place in African countries prior to colonization likely had highly adapted patterns of natural resource use [17,18]. The development of official government protected areas worldwide began with the model established in Yellowstone National Park in 1872 in the United States [2]. Developing countries in Africa have followed this historical model of conservation after colonization. The model’s purpose is to “defend the established boundaries (of the park) against the enemy” [16]. In this context, the enemy refers to the indigenous people in the surrounding geographical area who desire to use the natural resources inside the park boundaries. However, in recent years, there has been a push for more community involvement in the management of African national parks and protected areas – in other words, the privatization of publicly owned land [2,5].

forces [1]. National parks and protected areas in Africa are mainly governed by decision-making bodies created within governmental or semi-governmental institutions in accordance with national and/or regional legislation and policies [1]. Within Africa, like other countries around the world, the widely accepted reasoning behind publically owned national parks is to increase biodiversity and to create goods and services such as tourism for public use [9]. The conventional national park model in Africa focuses on policies and practices that discourage local participation, and emphasize centralized government decision-making [19]. Historically, in many African countries, parks and protected areas have operated in isolation from the surrounding indigenous societies, and do not include these societies in developmental activities and management decisions [19]. The ideology behind this method of management used globally can be seen in an example from the Imperial Institute of Forestry in Oxford, U.K. [20]. Their philosophy regarding natural resources states that “the public good was best served through the protection of forests and water resources, even if this meant the displacement of local communities” [21]. Decisions in African countries on what areas of land or water should be incorporated in the national parks are made by the state, with little or no public involvement in the design, establishment and execution of the park management plans [19].

Studies by Undergraduate Researchers at Guelph (SURG) 25


National parks and protected areas in African countries (Drewniak et al.) decision makers stems from the belief that private owners have stronger incentives to conserve natural resources, especially large mammal populations such as those found in many African National Parks and nature reserves. [22]. Within Africa, many private protected areas already exist, including a multitude of game reserves such as the Selous Game Reserve in Tanzania [13].

Challenges and Controversies with National Parks and Protected Areas Land use conflicts are typically defined as disagreements surrounding the current management of national parks and protected areas, the surrounding indigenous communities, and other stakeholders who have an investment in the specific area of land [23]. Challenges and controversies related to natural parks and protected areas in Africa include: expropriation and land tenure conflicts, poverty in communities surrounding protected areas, and the survival of populations of large mammalian endangered species. Expropriation and land tenure conflicts

Poverty There is controversy about whether the establishment of a national park exacerbates poverty in developing countries due to expropriation of land by the government from citizens without just compensation during park establishment [20]. In developing countries outside of the African continent, a relationship can be seen between park establishment using expropriation and poverty [25]. It is acknowledged that in the protected areas of the rainforest in central Africa, customary land rights are not acknowledged in conservation projects despite a World Bank policy that stresses that citizens displaced from protected areas should be compensated for their losses [25]. Instead, displaced citizens must find alternative sources of revenue to compensate for the revenue lost from their displaced land; this replacement income is typically much less and therefore exacerbates poverty conditions. Individuals in local communities can also potentially lose their sources of income with the creation of national parks [25]. As the definition of a national park does

Volume 6 • Issue 1 • Fall 2012

Land tenure can be defined as “the various relationships that people establish in order to access and manage land” [24]. In developing countries, land tenure conflicts usually arise as a result of the high rates of population growth and urbanization [24]. In African countries, there are several types of land tenure conflict: Unequal geographical distribution of natural resources, land utilization, legal recognition of ownership over parcels of land (e.g. customary rights) and, lastly, the conflict over the control of the land and its resources [24]. Major land conflicts come from public policies regarding the management of renewable resources, which maintain or reinforce existing logics of exclusion [24]. This leads to a “military”-based model, where “commanders” are appointed to “defend” the established boundaries against the “enemy”, usually the local people who require use of the resources within the boundary [16]. Although these policies partly develop from the intent to protect natural resources, they also allow centralized managers to monopolize the income they generate. Certain governments have encouraged the expropriation of land without the establishment of regulatory mechanisms to buffer the local communities from loss of these lands [24]. An example of a land tenure conflict related to expropriation is the 1969 expansion of Kruger National Park in South Africa [23]. Expropriation is a typical component in the establishment and functions of a protected area [5]. In 1969, the expansion of Kruger National Park in South Africa was completed, and the Makuleke indigenous community in the surrounding area was forced to move elsewhere without

compensation [23]. In 1995, the Makuleke community argued their land claim in the form of the Restitution of Land Rights Act and after two years of negotiations and conflict, settled on a co-management plan with the South African National Parks (SANParks) to ensure that conservation of the land was not at the expense of the needs of the indigenous community [23]. Other examples of land tenure conflict include the Mkomazi Game Reserve in Tanzania, where in 1988 more than 8,000 Maasai and Parakuyo cattle farmers and 100,000 cattle were displaced after conservationists became concerned over endangering wildlife species, habitats and tourism in the area [9]. Similarly, when the Lake Mburo National Park was created in 1982, 4,500 families were evicted from the area without compensation. This caused ongoing conflicts between the farmers and park authorities over the grazing of cattle within the park [9]. Finally, without notification in 2002, 30,000 people were evicted in Uganda when the game corridor linking Queen Elizabeth National Park to Kibale Forest Reserve was established [9]. According to the IUCN, national parks are “natural areas of land and/or sea, designed to protect the ecological integrity of one of more ecosystems for this and future generations, exclude exploitation or occupation inimical to the purposes of designation of that area, and to provide a foundation for spiritual, scientific, educational, recreational and visitor opportunities, all of which must be environmentally and culturally compatible” [9]. Considering this definition, it is easy to see how national parks can be areas of conflict, especially because many countries have communities of people living within their boundaries who are at risk of being evicted [9]. For example, in the central African Region, about 70% of national parks are inhabited and/or used by local people who are at risk from expropriation [9].

Studies by Undergraduate Researchers at Guelph (SURG) 26


National parks and protected areas in African countries (Drewniak et al.)

not allow for the use of the resources within the park (unless regulated heavily by the governing body [9]) local communities who would traditionally use the area for hunting and other related activities are left without a source of income [25]. During both the 1993 and 2003 World Park Congress, it was recognized that while national parks have global benefits, it is the poorer communities that have to carry the majority of the burden of conservation [25]. In 2008, a study consisting of 12 case studies in six countries used research regarding displacement and poverty collected over the last three decades to study the results using the Impoverishment Risks and Reconstruction Model for involuntary resettlement. [25]. It was found that by displacing inhabitants, individuals lost the income of the land within the National Park. An example of this phenomenon was seen when the scarcely populated area of Dzanga-Ndoki National Park in the Central African Republic lost access to 1,000 km2. Each individual lost access to forested land with a value of 55,000 dollars per person [25]. The study also found that the majority (67%) of an average central African’s total income came from hunting and gathering, while only 33% came from agriculture, labour and formal employment. Therefore, income generated in central African communities was found to be largely dependent on access to forests [25]. As the per capita per year income is one dollar per day in the Central African Republic, many individuals must rely on areas found in national parks to survive [25]. It is also of note that in African countries, there is evidence of large numbers of land sales and land abandonment after national park establishment, which may be caused by people moving to cities to find alternate employment and income [20]. Areas close to parks are typically much poorer than communities farther away as they do not have as much privately owned land [20]. Protection of large mammalian endangered species

The Free Market Environmentalism Perspective Free market environmentalism is based on a privatized market-based economy and emphasizes incentives associated with prices, profits and entrepreneurship. At the heart of free market environmentalism is a system of welldefined property rights which are specified and enforced by the government [6]. Research has documented the ability of this approach to preserve natural resources [6]. An example of the use of this approach can be seen in the operation of the Nature Conservatory in the United States. The Nature Conservancy is an environmental group that operates extensive land holdings. [6]. It has been able to successfully protect over 100 million acres of land worldwide through private ownership, without the aid of government funds or management [7]. There has been some push for a free market environmental approach to be applied to the management of national parks in Africa. For example, it has been well documented that wild animal populations have been declining within African national parks for many years. Reasons for this may include poaching, agriculture, mining, intrusive development and land clearing, among others [4]. Dissatisfaction with the public sector’s record of protecting endangered species has caused a push for the use of marketbased, commercialized approach [4]. A change to a de-centralized private market-based management approach includes the purchasing of land through private groups of likeminded individuals (such as The Nature Conservatory) or fully private individual firms [7]. The trading of live animals, breeding programs, hunting, and non-consumptive tourism are some examples of private entrepreneurship within free market environmentalism that may be successful in addressing the controversies associated with the current management of African national parks.

Volume 6 • Issue 1 • Fall 2012

Eastern and southern African countries are home to seven of the world’s biodiversity ‘hotspots’, which means that many of the species in these areas cannot be found anywhere else in the world. This has contributed to demands for the establishment of natural parks, game reserves, and other protected areas [9]. Controversies often arise due to conflicting opinions on how species habitat and how endangered species themselves should be managed. The main controversy related to large African wildlife species is that although a large number of protected areas and other conservation methods such as trade bans have been established under the assumption that centralized planning is the best option, they have done little to preserve the dwindling numbers of threatened, endangered or vulnerable species [4,22,26]. An example of how national parks and trade bans can actually threaten species can be seen in rhino populations. In 1975, rhinos were placed under the CITES

Appendix I policy, which meant that the hunting and trading of rhinos and rhino horn were completely banned [22]. This caused the price of rhino horn to rise dramatically in consumer markets and only fuelled poaching [22]. Another shortfall of national parks with regard to large mammalian species is that many of the parks are too small to contain these populations [16]. In many cases, mammalian populations migrate outside of the unfenced boundaries at certain times of the year [16]. Due to the ineffectiveness of endangered species protection thus far, the privatization of protected areas which makes use of the market structure of supply and demand, would create incentives that would aid in the conservation of large mammals and other natural resources [4,22]. Generally, the idea of the privatization of national parks and protected areas is not widely accepted. Many resource managers are reluctant even to consider it as an option [11].

Studies by Undergraduate Researchers at Guelph (SURG) 27


National parks and protected areas in African countries (Drewniak et al.) Evaluation of the free market environmentalism approach to national park management

Volume 6 • Issue 1 • Fall 2012

Free market environmentalism has the capability to address many of the current problems associated with government-run centralized national park management in Africa, if carefully implemented [22,26]. A market-based approach will add value to many of the resources within African national parks that are currently illegal to use, especially large mammalian endangered species. This allows the surrounding community to receive the benefits of these resources, which currently are very limited [26]. The market is simply a medium through which people voluntarily exchange items and property in a manner that conveys information about their values [26]. If individuals are willing to pay for what they value, their demands will be met by other individuals who are willing to supply the requested services. For example, animal welfare activists could pay land owners to preserve animals such as elephants from uses such as hunting [26]. A more likely scenario, however, is that the price for large mammals would be paid and preserved as hunters and consumers of large mammal products (such as elephant ivory and rhino horn) would be willing to pay the highest prices, protecting them from such purposes [26]. This would add a positive value to these large mammals and aid in their long-term survival, while eliminating the negative effects that currently encompass them such as poaching [26]. In Zimbabwe and South Africa, where it is legal to trade rhino horn, endangered rhino populations have increased, and so has the economic wealth of the surrounding community [22]. South Africa also introduced white and black rhino populations on private land, which caused the mammal to become a “draw card” species, meaning that it could be legally used for both trophy hunting and nonconsumptive tourism [22]. The increased demand and price for live rhinos has ensured that private land owners have a strong incentive to conserve and breed rhino populations. Profits from the selling of live rhinos have also allowed the National Park Board in South Africa to reinvest the proceeds from these sales in conservation initiatives. The privatization of African National Parks will also free up funds within African governments as they will no longer be responsible for upkeep and other park services [1]. While governments have little incentive to reduce spending, competitive firms are driven to minimize costs and run more efficiently in order to be successful in the marketplace [1]. National parks will benefit from a wider range of potential suppliers and therefore push towards achieving a better balance between cost of maintenance and profit. Thus, the government will benefit from the cost savings associated with the removal of government funding from national parks and will be able to invest the monetary reserves in other social or economic problems [1]. Privatization of African national parks would also allow for fair entry into the market and allow for the

surrounding community to take advantage of the economic benefits available [7]. Benefits include permitting traditional activities within the park, or compensation paid directly to the community [18]. The associated benefit activities on land within national parks to local communities can range from free entrance into the park for traditional cultural activities, to the permission for harvesting of fuel wood, medicinal plants, and approved sources of sustenance. Other benefits given to surrounding communities include lease payments paid by parks into tribal authority trust funds, preferential hiring of community members. and direct allocation of a percentage of the park fees and profits gained from tourism [18]. Programs have also been created to compensate for communities that have been negatively affected by the establishment of national parks and protected areas. These programs include the African Wildlife Foundation’s Good Neighbourliness program, which is aimed towards distributing benefits to the surrounding communities of protected areas in Tanzania, Kenya and Uganda where infrastructure must be built to accommodate the communities (i.e. schools, water supply outlets). Another compensation program is the Communal Area Management Program for Indigenous Resources (CAMPFIRE), in which community leaders offer wildlife revenues to locals on a per capita basis, as well as compensate farmers who experience damages due to wildlife encroachment on their private lands [18]. Although the intention of compensation programs is to directly benefit local communities, compensation and benefits can be mismanaged through actions of unequal distribution or through the withholding of subsidies. In turn, this can cause tension between the communities and the representatives involved in park privatization, potentially causing upheaval and unrest in community settlements [18]. These recurrent negatives associated with compensation from national parks are being dealt with through the institutionalization of commodity-based incentives programs and better management of potential benefits [18]. Any individual outside of the local communities that surround a national park and engage in traditional activities must be willing to pay to receive the benefits of the park. This aspect of the free market solves the free-rider problem (i.e. the use of resources without paying for use – for example poaching), as well as reduces the negative externalities that affect the surrounding community including damages caused by large endangered mammals on private farmland and personal property [7]. If land is privately owned, it is in the landowner’s best interest to keep problems to a minimum, as they are personally liable for damage to their own property. Publicly owned national parks do not have any incentive to stop damage to surrounding areas by wildlife, and so damage may be more frequent [1]. Despite all of its benefits, there are also several limitations to the free market environmentalism method that must be considered. Only areas of land capable of making a profit will be protected in a free market system [7]. This will

Studies by Undergraduate Researchers at Guelph (SURG) 28


National parks and protected areas in African countries (Drewniak et al.)

ignore factors such as the value of the ecological integrity ofan area and there is no guarantee against future development or alternative uses as individual values may change [7]. Moreover, markets are a medium through which people voluntarily exchange items of property and thereby express their values. If an individual’s values change, then there is no promise that national parks will continue to be protected [26]. A lack of land tenure security is a problem in Africa [26]. African countries need to ensure that there is a medium in which these rights are enforced, which will require government action [7]. As some African countries do not have governments fully committed to this, or these governments are compromised, it may be difficult to properly enforce the free market environmentalism method.

Conclusions

We wish to thank Professor Glen Fox, Department of Food, Agricultural and Resource Economics, in the Ontario Agricultural College at University of Guelph. The formulation of this paper was completed under his guidance and his expertise and feedback was greatly appreciated in the creation of the final report.

References 1.

Beckwith, J. (1981). Parks, property rights, and the possibilities of private law. Cato Journal, 1(2), 473499.

2.

Borrini-Feyerabend, G., & Kothari, G.O. Indigenous and local communities and protected areas towards equity and enhanced conversation. Cambridge: The World Conservation Union.

3.

Cessford, G.H., & Muhar, A. (2003). Monitoring options for visitor numbers in national parks and natural areas. Journal of Nature Conservation, 11(4). 240-250.

4.

Damania, R., & Hatch, R. (2005). Protecting Eden: markets or government? Ecological Economics, 53(3) 339-351.

5.

Saporiti, N. (2006). Managing national parks: how public-private partnerships can aid conservation. Washington: The World Bank.

6.

Anderson, T.L., & Leal, D.R. (2001). Free Market Environmentalism. New York, NY: Palgrave MacMillan.

7.

More, T.A. (2005). From public to private: five concepts of park management and their consequences. The George Wright Forum, 22(5), 12-20.

8.

ICUN. (2012). IUCN protected area management categories. Retrieved February 14, 2012: http://www.unep-wcmc.org/iucn-protected-areamanagement-categories_591.html

9.

Mombeshora, S., & Le Bel, S. (2009). Park people conflicts: the case of Gonarezhou national park and the Chitsa community in south-east Zimbabwe. Biodiversity Conservation, 18, 2601-2623.

Volume 6 • Issue 1 • Fall 2012

There are two primary options for the preservation of natural resources in African National Parks: Government management or privatized management. A free market environmentalism approach to the management of national parks in Africa is preferable to centralized government management as it would better manage the protection of endangered large mammal species as well as benefit African citizens in the surrounding communities. A free market approach, including the commercial use of natural resources, will create incentives that are necessary for the preservation of Africa’s national resources. It will also eliminate the freerider problem associated with those who want to preserve natural resources in national parks without directly paying for these benefits. In recent years, there has been recognition of the benefits of a market based system, as seen in the shift from a conventional national park model to co-managed parks. While full privatization of national parks is not generally an accepted view as of yet, there have been considerations of this course of action in some areas. It is important to consider both the benefits and the limitations of the free market environmental approach when considering the application of this approach to African parks. The free market environmental approach for resource protection will not be successful in all circumstances [7] [26]. It is best to take a more pragmatic approach to the implementation of free market environmentalism to National Park management in Africa; in other words, privatization should be decided on a case-by-case basis. In addition, national governments in Africa will need to strengthen legal protections of property rights if a free market environmental approach is to be successful. This way, national parks in Africa will be managed as efficiently as possible in every situation to preserve natural resources for future generations.

Acknowledgements

Studies by Undergraduate Researchers at Guelph (SURG) 29


National parks and protected areas in African countries (Drewniak et al.)

10. Protected Planet. (2012). Protected areas Africa. Retrieved February 12, 2012: http://www.protectedplanet.net/search

19. Pimbert, M.P., & Pretty, J.N. (1995). Parks, people, and professionals. Geneva: United Nations ResearchInstitute for Social Development.

11. United Nations Environment Program. (2003). United Nations list of protected areas. Cambridge: The World Conservation Union.

20. Naughton-Treves, L., Alix-Garcia, J., & Chapman, C.A. (2011). Biodiversity conservation and poverty traps special feature: lessons about parks and poverty from a decade of forest loss and economic growth around Kibale National Park, Uganda. Proceedings of the National Academy of Sciences, 108(34), 13919-13924.

12. Africa Almanac. (2012). Africa’s 50 largest national parks. Retrieved February 1, 2012: http://www.safarisafricana.com/50-largest-nationalparks/ 13. Baldus, R., & Kibonde, S. (2003). Seeking conservation partnerships in the Selous game reserve, Tanzania. Parks, 13(1): 50-60. 14. Pagiola, S., Martin-Hurtado, R., Shyamsundar, P., Mani, M., & Silva, P. (2002). Generating private sector resources to finance sustainable development: revenue and incentive effects. Washington: The World Bank. 15. Kaori, I. (2009). Overview of tourism in Africa. New York: Regional Bureau for Africa, United Nations Development Program. 16. Hanks, J., & Attwell, C.A.M. (2002). Funding and financial arrangements for TBPAs. In Peterman, T. & Braack, L.E.O. (eds.) Protected Areas: Guidelines for Good Practice and Implementation. Zschortau: Internationale Weiterbildung und Entwicklung. 17. Borrini-Feyerabend, G., & Sandwith, T. (2004). From ‘guns and fences’ to paternalism to partnerships: the slow disentangling of Africa’s protected areas. Parks: 13(1), 1-5.

21. McCracken, J. (1987). Conversation priorities and local communities. In Anderson, D. & Grove, R. (eds.) Conservation in Africa: People, Policies and Practice. Cambridge: Cambridge University Press. 22. Sas-Rolfes, M. (1998). Does CITES work? Four case studies. London: Institute of Economic Affairs. 23. Bosch, D. (2003). Land conflict management in South Africa: lessons learned from a land rights approach. Rome: Food and Agriculture Organization of the United Nations. 24. Delville, P.L., & Durand-Lasserve, A. (2009). Land governance and security of tenure in developing countries. Paris: French Ministry of Foreign and European Affairs, Land Tenure and Development Committee. 25. Cernea, M.M., & Schmidt-Soltau, K. (2006). Poverty risks and national parks: policy issues in conservation and resettlement. World Development, 34(10), 1808-1830. 26. Sugg, I.C., & Kreuter, U.P. (1994). Elephants and ivory: lessons from the trade ban. London: The Institute of Economic Affairs.

18. Schroeder, R.A. (1999). Geographies of environmental intervention in Africa. Progress in Human Geography, 23(3), 359-370.

Volume 6 • Issue 1 • Fall 2012

Studies by Undergraduate Researchers at Guelph (SURG) 30


Research Article

Tourism in Kenya’s national parks: A cost-benefit analysis Hubert Cheung Department of Integrative Biology, College of Biological Science, University of Guelph, Guelph, ON Canada. Faculty supervisor: John Fryxell. For correspondence, please email: hcheung7@hotmail.com.

Abstract East Africa is home to some of the most stunning wildlife in the world. With tourism in the region’s wildlife parks growing in popularity, it is imperative to evaluate the socioeconomic and environmental costs and benefits of this expanding industry. This study conducted a cost-benefit analysis of the various impacts that tourism has brought to Kenya’s national parks by monetarily valuating each impact. While the results of this cost-benefit analysis suggest that the benefits far outweigh the costs, even when non-measurable costs are considered, a number of fundamental issues must be addressed in order to improve the cost-benefit balance. The results are likely to be representative of the overall state of tourism in Kenya’s national parks and expose key areas where improvements can be made. Improvements to tourism in Kenya’s national parks can have positive implications for local people, the environment, wildlife species, tourists, and biodiversity conservation. Keywords: tourism; national parks; Kenya; cost-benefit analysis

Introduction The purpose of this study is to holistically investigate the costs and benefits of tourism in Kenya's national parks using a cost-benefit analysis. Each socioeconomic and environmental cost and benefit is investigated and given both a conservative and a liberal monetary valuation. By including both liberal and conservative estimations, this study is able to produce a range of possible scenarios in assessing the costs and benefits involved. This provides a broader representation of the net contribution of tourism in Kenya's national parks. It is hypothesized that while the socioeconomic and environmental benefits of tourism in Kenya's national parks are significant, the costs of associated problems created are also considerable and may outweigh the benefits when examined comprehensively in a cost-benefit analysis. While a cost-benefit analysis is admittedly less than complete, it is capable of providing valuable information to policy-making. As a holistic investigation of the socioeconomic and environmental impacts, it brings together a number of indicators into one study. This cost-benefit analysis can be used as a broad measure of the net contribution of tourism to Kenya's national parks [14,15]. Finally, given that improvements to the industry may benefit local people, the environment, tourist experience and biodiversity conservation, recommendations are made as to how benefits can be maximized and costs minimized.

Volume 6 • Issue 1 • Fall 2012

Ecotourism is the fastest-growing sector in tourism, growing at 10-15% per annum [1,2]. Ecotourism is defined as responsible travel to natural areas that conserves the environment and improves the wellbeing of local people [3]. Indeed, the iconic megafauna of East Africa has attracted tourists to Kenya's 65 national parks and reserves, which are operated by the Kenya Wildlife Service (KWS), and have stimulated growth in the country's tourism industry [4,5]. These parks represent around 10% of Kenya's land mass, created as conservation areas to protect wildlife [6]. Tourism is important to the Kenyan economy, providing a large number of jobs and accounting for 15% of GDP in 2009 with tourist revenue totalling KES 73.68 billion [7]. Supporters believe that the socioeconomic and environmental contributions brought by tourism in the national parks seem to overwhelmingly support the tourism industry and the direction it is headed. However, further investigation into tourism in these areas reveals that the industry has brought with it both costs and benefits. Detractors have found that an established neo-colonial structure surrounding the industry has greatly disempowered local communities [8,9]. Culture and language have been marginalized with the growth of tourism, which has sparked social problems such as drug use [10,11]. The physical environment has been degraded by tourism, and species populations continue to drop with biodiversity loss [11-13].

Studies by Undergraduate Researchers at Guelph (SURG) 31


Tourism in Kenya’s national parks (Cheung) Table 1. Summary of monetary valuations for costs and benefits associated with tourism in Kenya’s national parks. Impact Tourism revenue attributable to national parks

Conservative Valuation KES 73,680,000,000 x 53%

Liberal Valuation KES 73,680,000,000 x 75% KES 39,050,400,000

Operational cost of national parks

No calculations

KES 55,260,000,000 No calculations

-KES 1,540,000,000 Revenue sharing

KES 2,706,255,000 x 25% x 40%

-KES 1,540,000,000 KES 2,706,255,000 x 25% x 20%

KES 135,312,750 Employment opportunities

490,000 x 53% x KES 31,353

KES 270,625,500 490,000 x 75% x KES 31,353

KES 8,142,374,100 Opportunity cost

No calculations

KES 11,522,227,500 No calculations

-KES 23,370,000,000

-KES 23,370,000,000

Biodiversity loss and population declines

58%

Roads and off-road driving

8,800 km x 0.003 km x 100 (conversion to ha) x KES 19,024/ha -KES 50,223,360

8,800 km x 0.006 km x 100 (conversion to ha) x KES 19,024/ha -KES 100,466,720

Vehicle emissions

850 vehicles x 3.19 tCO2/vehicle x KES 410/tCO2 -KES 2,846,488

850 vehicles x 3.19 tCO2/vehicle x KES 6,260/tCO2 -KES 61,635,617

Improper discharge of waste by lodges

110 lodges x 0.5 ha/lodge x KES 19,024/lodge -KES 1,046,320

156 lodges x 0.5 ha/lodge x KES 19,024/lodge -KES 1,483,872

Electrical generation by lodges

110 lodges x 687 kWh/lodge/day x 365 days x 0.0002517 tCO2/kWh x KES 410/tCO2 -KES 313,203

156 lodges x 687 kWh/lodge/day x 365 days x 0.0002517 tCO2/kWh x KES 410/tCO2 -KES 10,022,501

KES 47,328,086,850 KES 3,864,954,704

KES 67,052,853,000 KES 4,624,588,710

30 years x KES 381,300,000 -KES 72,925,333

OVERALL BENEFITS OVERALL COSTS

58%

30 years x KES 36,900,000,000 -KES 713,400,000

Methodology and United States Dollars (USD). In order to work with this data, all data was converted to KES using the exchange rate USD 1 to KES 82, a rate which was chosen as a reflection of daily exchange rates observed over the month of January, 2012. This conversion rate was applied universally and did not correct for values given for past years. While this assumption may have affected the accuracy of the monetary valuations, it is unlikely that the result of the cost-benefit analysis would have been significantly skewed. Assumptions pertaining to each individual valuation are discussed in the following section, which explores each cost and benefit separately. A number of impacts are omitted from this cost-benefit analysis. Some impacts inherently cannot be assigned a monetary valuation. In other cases, the magnitude of the assumptions that had to be taken in order to derive a monetary valuation removed any meaningfulness this valuation would have offered. Specific omissions will be detailed in the following section.

Volume 6 • Issue 1 • Fall 2012

Although there is no standardized procedure for conducting a cost-benefit analysis, the process typically follows several general stages: project definition, identifying project impacts, determining which impacts are economically relevant, quantification of impacts and monetary valuation of effects [16]. For this project, socioeconomic and environmental impacts were examined with regards to tourism in Kenya's terrestrial national parks and reserves. Data was obtained from a variety of available literature and publications. An in-depth review of the material found was conducted to establish which impacts were economically significant and to quantify each impact. Once relevant parameters were established for each cost and benefit, the appropriate figures were combined to produce two annual monetary valuations, a conservative and a liberal estimate. Given the indirect nature of the valuation procedure used, assumptions had to be made in order to produce results. Due to the large variety of sources from which data was obtained, inconsistencies were not surprising. Data involving monetary values was found in both Kenyan Shillings (KES)

Studies by Undergraduate Researchers at Guelph (SURG) 32


Tourism in Kenya’s national parks (Cheung) Monetary valuation This section details the conservative and liberal monetary valuation of each cost and benefit. See Table 1 for a summary of calculations detailed in this section. Socioeconomic impacts

Volume 6 • Issue 1 • Fall 2012

Tourism in Kenya's national parks has produced a number of economic benefits, the greatest of which is the contribution to Kenya's overall tourism revenue. This benefit was conservatively valued at KES 39,050,400,000 and liberally valued at KES 55,260,000,000 per year. These estimates were made by multiplying the country's tourism revenue of KES 73,680,000,000 reported for 2010 by the Republic of Kenya Ministry of Tourism and Wildlife (KMoT) with the percentage of tourists citing game park visitation as their primary motivation for visiting Kenya [7]. This percentage was reported conservatively as 53% by Akama and Keiti and liberally as 70-80% (averaged to 75%) by Ikiara and Okech [5,17]. It is assumed that tourists' primary motivation for visiting Kenya is indicative of the proportion of overall tourism revenue attributable to the national parks. However, just because tourists declare game park visitation as their primary motivation does not mean they only stay in the parks; they spend money elsewhere, meaning benefits and costs span a much larger area than the parks alone [17]. The operational cost of Kenya's national parks, reported to be KES 1,540,000,000 for 2010, must also be subtracted from this figure [18]. Revenue sharing, whereby 25% of park fees are ceded to a local authority, is a system aimed at providing local communities with a direct share of tourism benefits [19, 20]. These payments are given to compensate for the communities' attenuated land use rights [21]. In 2010, income generated from park fees was reported to be KES 2,706,255,000 [18]. This figure was multiplied by the 25% rate to yield a supposed compensation of KES 676,563,750. However, the compensation process has been found to be slow and inadequate, with local communities receiving very little of the money as most of the money is sifted away by county councils [19,20]. An interviewee in a study by Bruyere et al. [20] described these benefits as "very minimal". To account for this lack of efficiency in the process, conservative and liberal discount rates of 20% and 40% were applied to the compensation. These rates were adopted based on an analysis of the qualitative accounts that unanimously describe the system's gross inadequacy, as no credible rates could be found. Once these rates were applied to the initial valuation, revenue sharing was deemed to provide benefits worth KES 135,312,750 in the conservative valuation and KES 270,625,500 in the liberal valuation. The KMoT reported in 2006 that the country's tourism industry generates a total of 490,000 jobs, of which 130,000 are formal positions and 360,000 are informal [22]. In order

to determine how many of these employment opportunities are attributable to tourism in the national parks, this figure was multiplied by the conservative and liberal percentages of tourists citing game park visitation as their primary motivation for visiting Kenya, reported as 53% and 75% respectively [5,17]. This assumes that tourists' primary motivation for visiting Kenya is indicative of the proportion of created jobs attributable to the national parks. This means that tourism has created 259,700 jobs in the conservative valuation and 367,500 jobs in the liberal valuation. Multiplied by the nation's annual average income of KES 31,353, the conservative valuation of employment opportunities is KES 8,142,374,100 and the liberal valuation is KES 11,522,227,500 [23]. These valuations do not take into account that most of the jobs created by tourism pay low wages and may not be stable [11]. An investigation of the true costs of tourism must consider the opportunity cost, which is the forgone value of the next best alternative to tourism [24]. In the case of Kenya's national parks, some 10% of total land mass was set aside as protected areas for biodiversity conservation, in which a booming wildlife tourism industry thrives [6,24]. Norton-Griffiths and Southey calculated the forgone gross revenues that could have been generated from agricultural and livestock production instead of setting aside land for parks and reserves to be KES 23,370,000,000 [24]. This figure was not modified further to give a conservative and liberal estimate as the methodology of the source was deemed to be a fair estimate of opportunity cost. Considering the limited ways in which the land could be used, it is safe to assume that agricultural and livestock production continues to be representative of how the land in national parks would have been developed. One of the socioeconomic benefits that tourism has brought is improvements to infrastructure and social welfare. This includes improvements to schools and education, access to medical care and road access [20]. The most important aspect of these benefits to the local communities is tourism's impact on assets. Tourism has enhanced natural capital, such as the establishment of emergency grazing areas, raised human capital, funding for education, and increased access to physical capital, such as emergency access [25]. While these benefits have positive effects on quality of life, other impacts are more harmful. In some regions, booms in tourism have sparked social problems, such as an increase in drug use and prostitution [11]. In some areas, tourism has indirectly encouraged children to drop out of school, as it has become in their economic interests to take up work in tourism [11]. Furthermore, conflicting cultural norms are often experienced when tourists travel and interact with local people, which inevitably create negative feelings towards tourism [11]. While it is understood that culture is dynamic and ever-changing, tourism has likely influenced and diluted the cultures and languages of local people [10,11]. As these dynamics are far too complex and intertwined to characterize, no meaningful monetary valuation could be drawn for these impacts.

Studies by Undergraduate Researchers at Guelph (SURG) 33


Tourism in Kenya’s national parks (Cheung) Another negative aspect of tourism in Kenya's national parks is the neo-colonial structure that has been established, whereby a small number of foreign investors control tourism resources [26]. Even community-based enterprises, aimed at benefiting the wider community and establishing developmental routes in the long run, reflect such a structure [9,27]. This contributes to the economic, psychological, social and political disempowerment of local communities [8]. Furthermore, there is an over-reliance on the tourism in some regions, such that the local economy is completely dependent on the continued success and growth of the tourism industry [11]. While there are significant downsides to these issues, to generate monetary valuations for these impacts is beyond the scope of this study. Environmental impacts

Volume 6 • Issue 1 • Fall 2012

The establishment of conservation and management programs and raised awareness brought by tourism in Kenya's national parks has been beneficial, but there has still been a steady decline of wildlife populations and a net loss of biodiversity. Between 1977 and 1997, the total of all nonmigratory species in the Masai Mara dropped by 58%; of these, the populations of warthog and buffalo have both declined by over 80%, while those of the giraffe, eland, waterbuck and topi have all declined by over 70% [12]. Wildebeest populations have also been found to have declined by 75% from 1980 to 2000 [13]. Ogutu et al. monitored the populations of seven ungulate species over 15 years and found that land-use change, including increased human settlement, is correlated with declines in species abundance [28]. While population densities of large predators, including lions and cheetahs, are relatively higher than other regions of Africa, declines are still experienced, largely due to increased human-predator conflict [29-32]. To monetarily valuate these losses, the 58% average loss documented by Ottichilo et al. was determined to be most representative of overall losses due to the study's comprehensiveness and its lengthy 30-year study period [12]. Assuming a linear population decline, the annual population decline rate was determined as 1.93%. Moran calculated total tourist surplus of non-consumptive use of Kenya's national parks to conservatively range from KES 3,813,000,000 to KES 36,900,000,000 per annum using willingness-to-pay estimates [21]. While tourism is not the single reason behind biodiversity loss, no figure could be found illustrating how much of these declines could be attributed to the tourism industry alone [28]. Thus, assuming tourism to be the sole factor leading to population declines, biodiversity loss can be estimated to conservatively cost KES 72,925,333 and liberally cost KES 713,400,000 per annum. Additionally, the increased presence of tourists, vehicles and lodges has impacted the behaviors of wildlife. Traditional wildebeest breeding grounds have been affected so heavily that the schedule of the annual Great Migration has become unpredictable [5]. Leopards, which normally hunt in the day time, have been forced to hunt at night due to

the noise pollution created by tourists and tour vehicles [5]. Frequently, packs of vehicles crowd around individual or groups of animals due to lack of regulation and enforcement, heavily affecting natural behavior [5]. Altered behaviors and migratory patterns may have consequences for the ecosystem. For instance, predators feeding on migrating wildebeest and zebra, such as lions and leopards, may experience a shortage of prey due to delays in the start of the Great Migration. Furthermore, the presence of tour vehicles may affect the hunting success of large predators. These impacts were determined to be contributing factors to overall population declines and biodiversity loss. Off-road driving by tour vehicles to seek out rare animals scars the terrain, while paved roadways fragment habitats and disrupt migratory routes [4,21,33]. The Kenya Roads Board declared that 6,000 km of unclassified roads and 2,800 km of unclassified roads contracted by local authorities existed in all national parks and reserves in 1995, the latest year for which statistics were available [34]. Assuming a width of 3 m for single-lane roads in the conservative valuation and 6 m for double-lane roads for the liberal valuation, the total land area converted to roadways is 2,640 ha and 5,280 ha respectively. These areas can be translated to a monetary cost through a valuation of losses in ecosystem services from watershed and erosion protection, pharmaceuticals and carbon sequestration [24]. Costanza et al. reported in 1997 that the value of ecosystem services provided by grasslands and rangelands - the dominant biome type throughout Kenya's national parks and reserves was approximately KES 19,024 per hectare per annum [35]. Combining these figures produces an annual conservative cost of KES 50,223,360 and an annual liberal cost of KES 100,446,720. These estimations were also made with the assumption that the conversion into roads has stripped the land of all ecosystem services. Tour vehicles, like all other vehicles, emit greenhouse gases [36]. Because there are limited statistics on either the total number of tour vehicles in operation in Kenya's national parks or the average distance driven by tour cars, these had to be deduced from other data. The latest data available regarding the number of tour vehicles in operation is from 1992, during which 33,110 vehicles entered the Maasai Mara National Park at a rate of 90.71 cars per day [4]. This figure can be conservatively assumed to be representative of all tour vehicles in operation in the Maasai Mara National Park, as the entry statistic included repeat entries by the same vehicle over 1992. From 1995 to 2009, the Maasai Mara National Park has consistently accounted for an average of 10.67% of all visitor arrivals to Kenya's national parks [37]. Combining these figures, it can be deduced that roughly 850 tour vehicles operated in all of Kenya's national parks. The average vehicle in Nairobi, Kenya, travels 17,284.5 km per year [38]. Given the UNEP's declaration that the average vehicle in Kenya emits 184.65 gCO2/km, this means the average vehicle in Nairobi emits 3,191,582.925 gCO2 per year [36]. Combining these numbers with the estimated range of carbon dioxide pricing from KES 410 to KES 6,260

Studies by Undergraduate Researchers at Guelph (SURG) 34


Tourism in Kenya’s national parks (Cheung) improper waste disposal by lodges is KES 1,046,320 while the liberal valuation is KES 1,483,872. It should be noted that, due to a lack of data, the valuations for this particular impact is likely to be an underestimation of the actual cost of improper waste disposal by lodges in Kenya's national parks. The remote locations of most safari lodges means that most are unable to access urban electrical infrastructure, and electricity is generally produced locally using technologically inferior and highly polluting diesel power generators [5]. A hypothetical case study of a hotel's electrical usage by the United States Agency for International Development in 2009 showed that a small lodge with six rooms requires 50.12 kWh per day to operate [42]. Assuming that this figure represents general usage patterns and can be proportionally scaled up for a hotel with 82.28 beds, the average lodge in Kenya's national parks requires about 687 kWh of energy per day. Using the estimation of 110 operational lodges in the conservative valuation and 156 lodges in the liberal valuation, it can be estimated that 75,570 kWh and 107,172 kWh per day is used respectively. Then, assuming that all of this electricity is being generated from diesel fuel, which has a conversion factor to carbon dioxide of 0.2517 kgCO2e/kWh, these figures are translated to an estimated 6942.65 tCO2e and 9845.95 tCO2e emitted per year by lodges in the parks [43]. To get an estimate of carbon cost, the price of emitting carbon dioxide needs to be factored in. Using the prices of KES 410 per tCO2 for the conservative scenario and KES 6,260 per tCO2 for the liberal scenario, a conservative valuation of KES 2,846,488 and a liberal valuation of KES 61,635,617 are derived for emissions from electrical generation by lodges [37]. Cost-benefit analysis The benefits associated with tourism in Kenya's national parks were conservatively valuated at KES 47,328,086,850 and liberally valuated to be worth KES 67,052,853,000 (Table 1). On the other hand, the costs were valuated conservatively at KES 3,864,954,704 and liberally at KES 4,624,588,710 (Table 1).Combining these valuations in a simple cost-benefit analysis demonstrates in all conservative and liberal pairings that the benefits brought by tourism in Kenya's national parks outweigh the costs (Figure 1). In a scenario where conservative benefits are coupled with liberal costs, the net benefit is lowered to KES 21,531,098,140 per annum. In contrast, combining liberal benefits with conservative costs yields the greatest net benefit of KES 42,015,498,2956 per annum.

Volume 6 • Issue 1 • Fall 2012

per tCO2, this estimates a total cost of tour vehicle carbon emissions to be KES 1,112,267 in the conservative valuation and KES 169,824,121 in the liberal valuation [39]. Despite the fact that commercial vehicles generally cover more distance than the average vehicle, these figures provide a conservative estimation of the cost of tour vehicle emissions. Emissions from civil aviation were also considered for inclusion. Globally, aviation emissions account for 4.9% of anthropogenic radiative forcing of climate change [40]. While data on the origins of international tourists to Kenya and general aviation emissions data are available, flight data could not be deduced meaningfully, deeming aviation emissions beyond the scope of this study. Monetary valuations would have required large assumptions to be made, so aviation emissions were not included as part of a larger estimate for transportation emissions. Lodges are often built near watering holes or breeding grounds, driving animals away from their traditional territories [5]. Lack of infrastructure, access and regulation has led to the discharge of waste and sewage from lodges into surrounding areas, including waterways from which wildlife and local residents draw drinking water [5,33]. Furthermore, not a single hotel has a proper disposal system for waste oil from generators [5]. Due to a lack of available statistics pertaining to the total number of lodges in operation in Kenya's national parks, an estimation of this figure was deduced from other data. In 2009, a total of 6,242,800 hotel bed-nights were occupied, meaning 17,103.56 beds are occupied per night [41]. In order to determine how many of these occupied bed-nights were attributable to tourism in the national parks, this figure was multiplied by the percentage of tourists citing game park visitation as their primary motivation for visiting Kenya, reported as 53% and 75% [5,17]. This assumes that tourists' primary motivation for visiting Kenya is indicative of the proportion of bed-night occupied in the national parks. Combining these figures, the number of beds occupied per night in the national parks is estimated to be 9,064.89 in the conservative valuation and 12,827.67 in the liberal valuation. In 1999, the Maasai Mara National Park's 25 hotels had a bed capacity of 2,057, averaging 82.28 beds per lodge [4]. Assuming full occupancy, this gives a conservative estimate of total lodges in operation in Kenya's national parks to be 110 in the conservative valuation and 156 in the liberal valuation. Given that no data is available for the area affected by waste disposal by lodges, a very conservative estimate of 0.5 ha per lodge was hypothesized. Combined with the KES 19,024 per hectare per annum value of ecosystem services lost by grasslands and rangelands, the conservative valuation of

Studies by Undergraduate Researchers at Guelph (SURG) 35


Tourism in Kenya’s national parks (Cheung)

Value of Costs and Benefits in Billions

KES 50 KES 42.015

KES 41.256

Liberal Benefits Conservative Costs

Liberal Benefits Liberal Costs

KES 40

KES 30 KES 22.291 KES 21.531 KES 20

KES 10

KES 0 Conservative Benefits Conservative Costs

Conservative Benefits Liberal Costs

Scenarios Pairing Conservative and Liberal Valuations

Figure 1. Comparison of four scenarios pairing conservative and liberal valuations for costs and benefits.

Discussion

Volume 6 • Issue 1 • Fall 2012

The very nature of cost-benefit analyses makes them highly controversial when used in making regulatory policy decisions. Heinzerling and Ackerman argue that the method itself lacks transparency and is weak in its attempts to put a dollar value on life, health and the natural world [15]. Ethical issues are raised at this practice [44]. It has been argued that cost-benefit analyses offer biased and misleading results, leading to inferior environmental protection and overall social welfare in the end [15]. In addition, while the discounting system used makes financial sense, it reduces the importance of environmental regulation and trivializes longterm risks [15]. Also ignored are the social and economic characteristics of stakeholders, with the result of ultimately reinforcing existing social and economic inequalities [15]. Moreover, many impacts have not been quantified and are inherently difficult to quantify, whereby major assumptions must be taken in order to produce numerical values [15,44]. However, despite these shortcomings, the results of a costbenefit analysis can contribute greatly to policy-making. Wells argues that even an incomplete or a partial cost-benefit analysis can be beneficial, yielding important and useful information valuable to policy-making [45]. Cost-benefit analysis offers a comparison of favorable and unfavorable effects of different policies [14]. Arrow et al. argue that, although cost-benefit analysis may not be sufficient in designing sensible policy due to the uncertainties and assumptions involved, it can help organize disparate information to the overall benefit of policy-making [14]. As outlined in the methodology, a number of costs and benefits are not accounted for in this study. Furthermore, this study could not address the possibility of overlapping

estimates. For instance, the costs of wildlife declines may have been double counted by other costs such as roads or habitat loss due to waste disposal. While the study accounts for many of the larger issues surrounding tourism in Kenya's national parks, some costs and benefits may have been overlooked and not factored into the cost-benefit analysis. Therefore, the specific results of the cost-benefit analysis conducted in this study should not be taken as definite or final, especially given the assumptions and omissions made. Rather, the overall trend that the socioeconomic and environmental benefits seem to outweigh the costs should be observed as a broad representation of the impacts of tourism in Kenya's national parks. Given the results of this cost-benefit analysis, it would be easy to assume that the benefits brought by tourism in Kenya's national parks far outweigh the costs, even when factoring in the several omissions made. However, these figures may not be accurate portrayals of the overall performance of wildlife tourism in Kenya. A number of fundamental issues must be addressed in order to maximize benefits and minimize costs. Despite the large revenues experienced by the tourism industry, the way it is conducted largely dilutes socioeconomic benefits for the majority of locals. The neocolonial structure of the industry makes it difficult for local citizens and communities to be involved in decision-making processes [9,26,46]. Outsiders generally focus on maximizing benefits for themselves, neglecting the need for costs to be minimized, since many of the socioeconomic and environmental costs do not directly affect them [47]. Although some opportunities exist for local entrepreneurs and Kenya's development agenda has emphasized encouragement of micro and small businesses over the past two decades, a divide of power continues to disengage local communities from tourism [48]. With over 50% of Kenyans living below the poverty line, and particularly with the incidence of poverty highest where tourist activity is highest, poverty is the most serious problem faced by local people [9,17]. It is crucial to many of these people's livelihoods and sustainable local community development that the neocolonial structure of tourism be addressed [17]. While tourism has the potential to aid in economic development and reducing poverty, a revamp of the current model of community-based enterprises, which emphasizes partnerships between foreign tour operators and local communities, is needed in order for this potential to be realized [17]. Poverty must be placed at the forefront of tourism development policies, and the involvement of local communities in tourism enterprises can create linkages to the local and national economies that are bypassed by many foreign operators [17,47]. Backed by a multitude of support organizations, the focal point of external intervention is the enhancement of conservation initiatives; a shift in focus towards the needs and priorities of local communities is imperative for community-based enterprises to achieve longterm sustainability [9]. A higher level of empowerment for local communities can be achieved by focusing on greater

Studies by Undergraduate Researchers at Guelph (SURG) 36


Tourism in Kenya’s national parks (Cheung) the benefits are felt throughout the country by different interest groups [20]. Social issues exacerbated by tourism, such as drug use, prostitution and the encouragement of children to work instead of complete school, must be addressed through policy and enforcement [11]. Although it is difficult to prevent some level of cultural dilution and marginalization of language as people continue to adopt and embrace Western values and customs, traditional values can be instilled through education [10,11]. Government policies must reflect the value of and the need to preserve culture and language, and must encourage tourism to be culturally sensitive and respectful [10]. It is absolutely crucial that species populations be continuously monitored and studied in an organized and structured manner by a government agency, such as the KWS. The collection of relevant data, such as species populations, species territory and migration patterns and behavioral changes, is critical for the impacts of tourism to be properly assessed and subsequent regulatory policies to be appropriately developed to help mitigate impacts and prevent further damage. Conservation plans can only be effective if detailed knowledge of species is incorporated [52]. While there are a number of studies documenting species populations, a centralized monitoring scheme must be established to yield consistent data. More research should also be conducted regarding the specific impacts of tourism on wildlife, in particular research that could quantify the causality between tourism and species population declines and biodiversity loss. Such detailed research will also allow for the evaluation of implemented solutions, with reduced ecological impact being the ultimate goal. Realistic, shortterm targets must be set for the recovery of wildlife populations and lost habitat. It can safely be assumed that tourists who cite game park visitation as their prime motivation for visiting Kenya are most attracted to the wildlife. However, despite there being a total of 65 national parks and reserves in the country, the majority of tourists choose to congregate in a few specific parks, with the top five parks accounting for 54% of all visitor arrivals in 2009 and the top ten accounting for 80% [37]. In fact, only half of Kenya’s parks had any visitation or visitor services [5]. Thus, a key strategy would be to rebrand and develop the rest of the national reserves, promoting and, in some cases, enabling tourists to visit other parks. This approach may minimize overconcentration of visitors in a select few parks, which will lessen the environmental impacts felt in these parks and allow for easier ecological recovery from anthropogenic damage. However, this will only be effective if a visitor cap or limit is set up for each park. The Galapagos Islands, Ecuador, serves as a case study where the number of visitors outmatched the ecosystem's carrying capacity for tourism. In 1974, a visitor limit of 12,000 was presented in the Galapagos National Park Management Plan. However, the visitor cap was surpassed every single year, leading to the abandonment of tourist limits altogether and a subsequent surge in visitor numbers. Not surprisingly, the rapid increase in visitors led to

Volume 6 • Issue 1 • Fall 2012

leadership, independence and local community priorities, and actively discouraging foreign elitism [9,49]. It is urgent for the government to find a different approach to tourism development policies to ensure equitable distribution of tourism opportunities and revenue among stakeholders [50]. Facilitating greater involvement of local stakeholders and emphasizing the priorities of local communities in the decision-making process may also address the lack of local representation on workforces. While the KMoT may boast that 490,000 jobs have been created by the tourism industry, these employment opportunities are inequitably distributed, with local people often losing out on these jobs to an outside workforce as evidenced by a lack of local representation in the staff of protected areas and lodges [20,22]. 64% of local community members find tourism employment opportunities to be insufficient, with not a single community member identifying local employment opportunities to be a positive benefit of tourism [20]. Contrastingly, 33% of protected area staff felt local job opportunities are a positive benefit of tourism [20]. These findings reveal symptoms of the industry's neo-colonial structure, further highlighting the need for local community priorities to be given greater antecedence [9]. Greater community involvement in the decision-making process can address the need to guarantee employment opportunities to locals. Although 25% of park fees are supposedly ceded to a local authority for the benefit of local communities, the shortcomings of the established revenue sharing system, characterized as inefficient and inadequate, have been documented for over a decade [19, 20]. With the majority of money sifted away in the administrative process, the description of revenue sharing benefits by local stakeholders as being "very minimal" reflects the abysmal degree of the system's weaknesses [20]. It is imperative for such compensation programs to be properly conducted. Drawing from studies in damage compensation programs for livestock depredation losses to protected predators, failure to compensate every time losses are suffered in a timely fashion is highly detrimental, exacerbating feelings of resentment against predators and conservation efforts [32,51]. In essence, if revenue sharing is not done properly, it may be better if it were not done at all. The current approach to revenue sharing should be reviewed and rejuvenated to become more organized, transparent and efficient. However, this may be far easier said than done. Addressing policy issues is difficult enough even in the most developed countries, let alone a government whose policies have been generally inconsistent, controversial, lacking in resources for implementation and enforcement and further hindered by corruption [20]. Although this study was unable to produce an estimate of the impact of infrastructure and social welfare improvements sparked by tourism in Kenya's national parks, these impacts are generally positive for local communities. Continued investment in these important and crucial changes is necessary to facilitate improvements in quality of life. These improvements should be equitably made to ensure that

Studies by Undergraduate Researchers at Guelph (SURG) 37


Tourism in Kenya’s national parks (Cheung) behavioral aspects are considered, especially where they apply to the local people's lives and environment - must be taken by such a process, with traditional and "extended expertise" given fair consideration for how a tourism operation is run [56]. Ongoing monitoring would ensure that, were socioeconomic and environmental factors to deteriorate beyond acceptable standards, tourism operations could be halted or held liable for costs. This would require the incorporation of aspects of the precautionary principle in the assessment structure, whereby a tourism operation could be halted and placed under review [57]. Additionally, were the neo-colonial structure of the tourism industry to persist, public participation mandated by an environmental impact assessment process would be able to draw some level of public participation [58]. It is recommended that legislation and standards are established to implement a tourism-specific environmental impact assessment process.

Conclusion While the results of this cost-benefit analysis indicate that the benefits associated with tourism in Kenya's national parks outweigh the costs, the specific figures should not be taken as definite given the multitude of assumptions and uncertainties involved in the monetary valuation process. However, as a holistic investigation of the socioeconomic and environmental impacts involved, the results should be interpreted as representative of the overall state of tourism in Kenya's national parks. A number of critical changes must be made to the existing tourism framwork to ensure benefits are maximized and costs are minimized. The neo-colonial structure of the tourism industry must be addressed to spur greater empowerment of local communities. Prioritization of the needs of local communities is imperative. Revenue sharing schemes must be reorganized to iron out inefficiencies and ensure that local communities actually benefit from these payments. Tourism operations must be culturally sensitive and respectful, and should promote the preservation of local culture and language. Systematic monitoring of species populations must be conducted for conservation to be successful, with realistic recovery targets needed to achieve long-term goals. Environmentally-damaging actions must be curbed and regulated, with supporting infrastructure established where necessary to reduce environmental degradation. Finally, the establishment of an environmental impact assessment process that is tourism-specific will be able to evaluate tourism operations to ensure the industry's sustainability. Further research regarding the holistic impacts of tourism should be conducted. Ultimately, improvements to the way tourism in Kenya's national parks is conducted will benefit local people, the environment, wildlife species, tourist experience and the effectiveness of biodiversity conservation.

Volume 6 • Issue 1 • Fall 2012

environmental erosion and disturbance to plants and animals, not to mention a general decline in tourism experience [53]. If visitor limits are to be implemented, these limits must be strictly adhered to. While negative impacts of tourism on habitat and wildlife and are inevitable, the actions that cause damage can be reduced with proper training and enforcement of regulation. Tour drivers are currently hired by independent tour operators and are mostly given insufficient training. Standardized training by a government authority like the KWS and the implementation of a licensing scheme may reduce practices that excessively interfere with wildlife or damage the environment, such as driving too close to the animals, crowding them and driving over delicate habitat. Enforcement, whether it is by a government agency or by tour drivers themselves in the form of community policing, must be conducted to effectively implement regulations and restrictions [5]. Tour vehicles and power generation at lodges release greenhouse gases, contributing to global climate change. In recent years, the Kenyan government has acknowledged the fact that the current vehicle inspection system is inefficient, with 70% of the country's vehicles failing to meet the country's emissions criteria, and is working towards establishing an institutional and regulatory framework for vehicle inspections [54]. Foreign tourism operators should be encouraged to perform periodic inspections to ensure tour vehicles are in good shape. A strategic plan to support such a rapidly growing industry must be developed to establish sufficient infrastructure, both internally within parks themselves and externally to supplement the parks [50]. Increasing the energy efficiency of lodges can reduce the amount of electricity used, reducing the need for on-site power generation, as lodges are often built far from access to main electrical grids [5]. For some lodges, depending on location, it may be economically worthwhile to establish connections to the national electrical grid [55]. A system of waste disposal must also be established to prevent the discharge of garbage, sewage and waste oil directly into the environment [33]. The least environmentally-damaging solution is to collect waste and refuse from lodges and transport it to waste management facilities. To ensure the long-term sustainability of tourism in Kenya's national parks, a structure for systematic review and examination of tourism operations and programs is needed to evaluate the costs and benefits. A tourism-specific environmental impact assessment process is proposed, which can systematically review the appropriateness of the tourism operation and evaluate construction impacts and operational impacts. With a tourism-specific system, the assessment process for new proposals can be accelerated if monitoring in and around the national parks is kept ongoing and dialogue with local communities - who are the most directly affected stakeholders in wildlife tourism - is continuously maintained. A holistic approach - in which moral, ethical, cultural and

Studies by Undergraduate Researchers at Guelph (SURG) 38


Tourism in Kenya’s national parks (Cheung)

Acknowledgements I would like to thank Dr. John Fryxell for his guidance and feedback on this research project. I am very grateful for the opportunity to have learned more about how human activity is affecting the people and ecosystems of East Africa. I would also like to thank Dr. Teresa Crease for providing me with advice and direction. In support of my project, I would also like to thank Dr. Richard Kuhn. I would also like to thank Michelle Lok, Christian Carlucci and everyone else who helped improve the paper.

References 1. Kamauro, O. (1996). Ecotourism: suicide or development? Voices from Africa Number 6: Sustainable Development Part 2. United Nations Non-Governmental Liaison Service. 2. Ramser, T. (2007). Evaluating Ecotourism in Laikipia, Kenya: Assessing the Socio-economic Impact and Conservation Attitudes. Thesis, University of Berne. Berne, Switzerland. 3. Sekercioğlu, C.H. (2002). Impacts of birdwatching on human and avian communities. Environmental Conservation, 29(3), 282-289.

10. Eastman, C.M. (1995). Tourism in Kenya and the marginalization of Swahili. Annals of Tourism Research, 22(1), 172-185. 11. Van Marwijk, R., & Joosten, M. (2003). A Small contribution? Small Tourism Entrepreneurs and Sustainable Development in Malindi. Thesis, Wageningen University. Wageningen, Netherlands. 12. Ottichilo, W.K., De Leeuw, J., Skidmore, A.K., Prins, H.H.T., & Said, M.Y. (2000). Population trends of large non-migratory wild herbivores and livestock in the Masai Mara ecosystem, Kenya, between 1977 and 1997. African Journal of Ecology, 38, 202-216. 13. Homewood, K., Lambin, E.F., Coast, E., Kariuki, A., Kikula, I., Kivelia, J., Said, M., Serneels, S., & Thompson, M. (2001). Long-term changes in SarengetiMara wildebeest and land cover: pastoralism, population, or policies? Proceedings of the National Academy of Sciences of the United States of America, 98(22), 1254412549. 14. Arrow, K.J., Cropper, M.L., Eads, G.C., Hahn, R.W., Lave, L.B., Noll, R.G., Portney, P.R., Russell, M., Schmalensee, R., Kerry-Smith, V., & Stavins, R.N. (1996). Is there a role for benefit-cost analysis in environmental, health, and safety regulation? Science, 272, 221-222.

4. Bhandari, M. (1999). Tourism raised problems in Masai Mara National Park in Narok, Kenya: Asia-Pacific Economic Conference.

15. Heinzerling, L., & Ackerman, F. (2002). Pricing the priceless: cost-benefit analysis of environmental protection. Washington: Georgetown Environmental Law and Policy Institute.

5. Ikiara, M., & Okech, C. (2002). Impact of tourism on environment in Kenya: status and policy. KIPPRA Discussion Paper Number 19. Kenya Institute for Public Policy Research and Analysis.

16. Hanley, N., & Spash, C.L. (1993). Cost-Benefit Analysis and the Environment. Cheltenham, UK: Edward Elgar Publishing.

6. Weaver, D.B. (1999). Magnitude of ecotourism in Costa Rica and Kenya. Annals of Tourism Research, 26(4), 792-816.

17. Akama, J.S., & Keiti, D. (2007). Tourism and socioeconomic development in developing countries: a case study of Mombasa Resort in Kenya. Journal of Sustainable Tourism, 15(6), 735-748. 18. Kenya Wildlife Service. (2010). Annual Report.

8. Scheyvens, R. (1999). Ecotourism and the empowerment of local communities. Tourism Management, 20, 245249.

19. Sindiga, I. (1995). Wildlife-based tourism in Kenya: land use conflicts and government compensation policies over protected lands. Journal of Tourism Studies, 6(2), 45-55.

9. Manyara, G., & Jones, E. (2007). Community-based tourism enterprises development in Kenya: an exploration of their potential as avenues of poverty reduction. Journal of Sustainable Tourism, 15(6), 628-644.

20. Bruyere, B.L., Beh, A.W., & Lelengula, G. (2009). Differences in perceptions of communication, tourism benefits, and management issues in a protected area of rural Kenya. Environmental Management, 43, 49-59.

Volume 6 • Issue 1 • Fall 2012

7. Republic of Kenya Ministry of Tourism and Wildlife. (2011). Tourism performance overview 2010. Retrieved on January 12, 2012 from http://www.tourism.go.ke/ministry.nsf/pages/facts_figures

Studies by Undergraduate Researchers at Guelph (SURG) 39


Tourism in Kenya’s national parks (Cheung) 21. Moran, D. (1994). Contingent valuation and biodiversity: measuring the user surplus of Kenyan protected areas. Biodiversity and Conservation, 3, 663-684. 22. Republic of Kenya Ministry of Tourism and Wildlife. (2006). Statistical analysis of tourism trends (globally and locally). 23. Kareithi, S. (2003). Coping with declining tourism, examples from communities in Kenya. London. 24. Norton-Griffiths, M., & Southey, C. (1995). The opportunity costs of biodiversity conservation in Kenya. Ecological Economics, 12, 125-139. 25. Ashley, C. (2000). Applying livelihood approaches to natural resource management initiatives: experiences in Namibia and Kenya. London: Overseas Development Institute. 26. Akama, J.S. (2004). Neo-colonialism, Dependency and External Control of Africa's Tourism Industry. London, UK: Taylor & Francis. 27. Rutten, M. (2002). Park beyond parks: genuine community-based wildlife eco-tourism or just another loss of land for Masaai pastoralists in Kenya. London: International Institute for Environment and Development. 28. Ogutu, J.O., Piepho, H.-P., Dublin, H.T., Bhola, N., & Reid, R.S. (2009). Dynamics of Mara-Serengeti ungulates in relation to land use changes. Journal of Zoology, 278(1), 1-14. 29. Gros, P.M. (2002). The status and conservation of the cheetah Acinonyx jubatus in Tanzania. Biological Conservation, 103(2), 177-185. 30. Bauer, H., & De Iongh, H.H. (2005). Lion (Panthera leo) home ranges and livestock conflicts in Waza National Park, Cameroon. African Journal of Ecology, 43, 208-214.

32. Hazzah, L., Mulder, M.B., & Frank, L. (2009). Lions and warriors: social factors underlying declining African lion populations and the effect of incentive-based management in Kenya. Biological Conservation, 142, 2428-2437.

34. Kenya Roads Board. (1995). Kenya road network. Retrieved on February 3, 2012 from: http://www.krb.go.ke/road-network/road-network.html 35. Costanza, R., D'Arge, R., De Groot, R., Farber, S., Grasso, M., Hannon, B., Limburg, K., Naeem, S., O'Neill, R.V., Paruelo, J., Raskin, R.G., Sutton, P., & Van den Belt, M. (1997). The value of the world's ecosystem services and natural capital. Nature, 387, 253-260. 36. United Nations Environment Programme. (2011). Developing a national fuel economy database and baseline: Kenya case study. Nairobi: United Nations Environment Programme. 37. Republic of Kenya Ministry of Tourism and Wildlife. (2010). Visitor arrivals, park stats. Retrieved on January 18, 2012 from: http://www.tourism.go.ke/ministry.nsf/doc/Visitor%20Ar rivals,Parks%20Stats.xls/$file/Visitor%20Arrivals,Parks %20Stats.xls 38. International Sustainable Systems Research Center. (2002). Nairobi, Kenya vehicle activity study. La Habra: International Sustainable Systems Research Center. 39. Duong, M.H. (2009). What is the price of carbon? Five definitions. Survey and Perspectives Integrating Environment & Society, 2(1). 40. Lee, D.S., Pitari, G., Grewe, V., Gierens, K., Penner, J.E., Petzold, A., Prather, M.J., Schumann, U., Bais, A., Berntsen, T., Iachetti, D., Lim, L.L., & Sausen, R. (2010). Transport impacts on atmosphere and climate: aviation. Atmosperic Environment, 44, 4678-4734. 41. Republic of Kenya Ministry of Tourism and Wildlife. (2010). Hotel bed occupancy by zone. Retrieved on January 18, 2012 from: http://www.tourism.go.ke/ministry.nsf/doc/Hotel%20Bed %20Occupancy%20by%20zone.xls/$file/Hotel%20Bed% 20Occupancy%20by%20zone.xls 42. United (2009). options States

States Agency for International Development. Powering tourism: electrification and efficiency for rural tourism facilities. Washington: United Agency for International Development,

43. Carbon Trust. (2011). Conversion factors: energy and carbon conversions - 2011 update. London: Carbon Trust.

Volume 6 • Issue 1 • Fall 2012

31. Frank, L., Hemson, G., Kushnir, H., & Packer, C. (2006). Lions, conflict and conservation in Eastern and Southern Africa. Background Paper for the Eastern and Southern African Lion Conservation Workshop, Johannesburg, South Africa, 11-13 January 2006.

33. Okello, M.M., & Kiringe, K.W. (2004). Threats to biodiversity and their implications in protected and adjacent dispersal areas of Kenya. Journal of Sustainable Tourism, 12(1), 55-69.

Studies by Undergraduate Researchers at Guelph (SURG) 40


Tourism in Kenya’s national parks (Cheung) 44. Pearce, D. (1998). Cost-benefit analysis and environmental policy. Oxford Review of Economic Policy, 14(4), 84-100. 45. Wells, M.P. (1997). Economic perspectives on nature tourism, conservation and development. Washington: The World Bank Environment Department. 46. Coast, E. (2002). Maasai socioeconomic conditions: a cross-border comparison. Human Ecology, 30(1): 79-105. 47. Ashley, C., Boyd, C., & Goodwin, H. (2000). Pro-poor tourism: putting poverty at the heart of the tourism agenda. London: Overseas Development Institute Natural Resource Perspectives. 48. Mshenga, P.M., & Owuor, G. (2009). Opportunities for micro and small scale businesses in the tourism sector: the case of the Kenya coast. KCA Journal of Business Management, 2(2), 52-68. 49. Thompson, M., & Homewood, K. (2002). Entrepreneurs, elites, and exclusion in Maasailand: trends in wildlife conservation and pastoralist development. Human Ecology, 30(1), 107-138. 50. Akama, J.S. (2002_. The role of government in the development of tourism in Kenya. International Journal of Tourism Research, 4, 1-13. 51. Gusset, M., Swarner, M.J., Mponwane, L., Keletile, K., & McNutt, J.W. (2009). Human-wildlife conflict in northern Botswana: livestock predation by endangered African wild dog Lycaon pictus and other carnivores. Oryx, 43(1), 67-72.

52. Thirgood, S., Mosser, A., Tham, S., Hopcraft, G., Mwangomo, E., Mlengeya, T., Kilewo, M., Fryxell, J., Sinclair, A.R.E., & Borner, M. (2004). Can parks protect migratory ungulates? The case of the Serengeti wildebeest. Animal Conservation, 7, 113-120. 53. Brandon, K. (1996). Ecotourism and conservation: a review of key issues. Washington: World Bank Environment. 54. Republic of Kenya Ministry of Environment and Mineral Resources. (2011). Motor Vehicle Emissions Control Stakeholders Workshop. 55. Parshall, L., Pillai, D., Mohan, S., Sanoh, A., & Modi, V. (2009). National electricity planning in settings with low pre-existing grid coverage: development of a spatial model and case study of Kenya. Energy Policy, 37, 2395-2410. 56. Eden, S. (1996). Public participation in environmental policy: considering scientific, counter-scientific and nonscientific contributions. Public Understanding of Science, 5, 183-204. 57. Gustavson, K.R. (2003). Applying the precautionary principle in environmental assessment: the case of reviews in British Columbia. Journal of Environmental Planning and Management, 46(3), 365-379. 58. O'Faircheallaigh, C. (2010). Public participation and environmental impact assessment: purposes, implications and lessons for public policy making. Environmental Impact Assessment Review, 30(1), 19-27.

Volume 6 • Issue 1 • Fall 2012

Studies by Undergraduate Researchers at Guelph (SURG) 41


Research Article

Educational assortative mating in the United States and the effect on income inequality by household from 1960 to 2005 Kathryn Swierzewski Department of Economics, College of Management and Economics, University of Guelph, Guelph, ON Canada. Faculty supervisor: Louise Grogan. For correspondence, please email: swierzkt@mcmaster.ca.

Abstract This study examines the effect assortative mating by education has on income inequality by household. In contrast to the majority of other literature in this field which focus on the United States (U.S.) as a whole, this study makes use of statelevel data to examine the marriage mating market with respect to education attainment. It also examines how homogamous partnerships increase income inequality across households by analyzing changes in the Gini coefficient over time. Panel data for this analysis is from the U.S. Bureau of Economic Analysis and the Integrated Public Use Microdata Series (IPUMS-International and IPUMS-USA) from the U.S. Census of the Population. Assortative mating by education is shown in this analysis to be a contributing factor to increasing inequality among homogamous heterosexual partnerships in the U.S. from 1960 to 2005. Keywords: assortative mating; education level; United States (state-level, from 1960-2005); income inequality (household); labour economics; welfare economics

Introduction state-level data as opposed to data pertaining to the country as a whole [5,11]. In similar studies by Kalmijn, Mare, and Mare and Schwartz, recent contracted partnerships and newlyweds were analyzed [1,2,6]. The present study steps away from this approach as it can cause bias; for example, individuals have the choice and opportunity to attend school after marriage. Looking solely at newlywed partnership patterns would not take into consideration the number of people who upgrade their degree later in life. Instead, the present study focuses on marriages and cohabitating partnerships, since it is believed that individuals go through the same sort of screening process in finding a long-term partner as they do in finding a spouse. Lu et al. noted in their analysis that there has been a trend in highly educated women marrying down (hypogamy) [5]. The present study does not examine partnership patterns of those who have attained a Bachelor’s degree level of education or higher. Rather, it groups higher levels of education attainment together under the university category. It investigates what is happening at all levels of education – from less than primary schooling completed to university completed – and looks at the effects assortative mating by education has on income inequality by household.

Volume 6 • Issue 1 • Fall 2012

Educational assortative mating has become a topic of increasing importance in welfare and labour economics over the past few decades. There has been increasing income inequality resulting from marriage market mating due to individuals’ abilities to choose if, when, and with whom they will enter into marriage. Mare, in two separate studies, and Schwartz noticed increases in educational homogamy over time generated by decreasing intermarriage among groups of relatively well-educated individuals [1,2]. Blau and Smits et al. observed increases in income inequality in the U.S. that may have reduced the likelihood of educational intermarriage by increasing economic and social distances between education groups [3,4]. These barriers are partially due to the fact that finding a prospective spouse is more likely at higher levels of schooling than at lower levels. The present study analyzes the effect of the relationship between educational assortative mating and income inequality by household from 1960 to 2005 in the U.S. This is done by examining the effects of homogamy within educational groups, and analyzing implications on income inequality by comparing Gini coefficients at the state level. This is similar to the methodology of Fortine and Schirle and Lu et al. and their use of Gini coefficients on Canadian and U.S. data, but the present study makes use of

Studies by Undergraduate Researchers at Guelph (SURG) 42


Educational assortative mating in the U.S. and the effect on income inequality by household (Swierzewski)

Data Collection and Analysis Sources of data The Integrated Public Use Microdata Series data sets (IPUMS-International and IPUMS-USA) from the U.S. Census of Population and Housing from the U.S. Census Bureau were used in this analysis from 1960 to 2005 in 10year intervals (except from 1990-2005, which was a 15-year interval) [12,13]. IPUMS supplies the micro data for 1%

harmonized samples of the population census, organized by household. The majority of the data used were from IPUMSInternational and harmonized across countries, which enables the results of this analysis to be compared to that of other counties in future studies. IPUMS data on the U.S. is not standardly weighted across years. For all 50 states, the data were collected on April 1st for 1960-1990, and on July 1st for 2005. The unemployment variable was generated using data from IPUMS-USA. Concerning theIMPUS data sets, the long form of the census was administered differently for all years of the data. In 1960, the long form census was given to every fourth enumeration unit; in 1970, the long form census was given to 1 out of every 5 households with 15% of the population receiving type one of the census and 5% receiving type two of the census. For the 1980 and 1990 data, one half of all housing units were required to fill out the long form census in areas with populations under 2,500, and one sixth of the population in all other areas. Finally, data from 2005 was received from the census administered to one out of every one hundred households. The definition of what constitutes a household varied from year to year, but does not affect the present analysis. Broadly, households were defined as dwelling places not including institutions and transient quarters. Households in the 1960, 1970, and 1980 censuses were defined as five or fewer people unrelated to the head of household; in 1990, they were defined as ten or fewer people unrelated to the head; and in 2005, there was not a restriction imposed on the number of unrelated persons. The gross Domestic Product (GDP) was acquired through the U.S. Bureau of Economic Analysis where data is collected through federally-administered surveys. Educational attainment and assortative mating When looking at married and cohabitating partnerships, the sample was restricted to households containing one heterosexual couple with both individuals between 18 to 65 years of age. More than one couple living in a household would cause bias and increase the level of inequality observed because income would be pooled and living costs such as property taxes, hydro, water, gas, and groceries would be split among residents of the household. Similarly, taking into consideration same-sex partnerships could also lead to bias due to a gender differential in the labour market in terms of potential earnings. Only a small portion of partnerships in the U.S. are reported to be same sex in the census data (less than 0.5%), so eliminating them from the analysis is unlikely to skew results. For the purpose of looking at education attainment of males and females in heterosexual relationships, the variable analyzed was the categorical variable “edattan”, which was used to group people into the following categories: less than primary education completed, primary education completed, secondary schooling completed, and university degree completed. This variable is a good indicator for the general level of education in a sample but is not detailed enough to

Volume 6 • Issue 1 • Fall 2012

In their study, Daly and Valetta noted increases in the proportion of heads of families who were unmarried women and concluded that this could be having an effect on increasing family earnings inequality [7]. Social norms have been altered through time; society is now more accepting of partnerships out of wedlock. Women are also increasingly likely to pursue more advanced levels of education and to partake in the labour force, which has resulted in shifting family dynamics [7]. For example, the socially accepted role of a female is now more likely to resemble that of a male head of household than in previous decades, and women are also increasingly self-reliant from an economic standpoint. Rose and Qian noticed a marriage decline among women over time; since women are no longer as reliant on a spouse for economic stability, they are more concerned with the quality of a potential spouse [8,9]. This trend could also, in part, be attributed to the findings of Juhn, who noticed that the number of less-educated men between the ages of 20 to 24 who are unemployed or not enrolled in school had been on the rise since 1970. Ultimately, this is a less attractive quality to females in the marriage market [10]. In a separate study, Rose noticed an increase in the number of divorces in recent years, resulting in a greater number of lone parents [8]. Consequently, the present study examines statistics of unmarried males and females (never married, divorced, or widowed) to see if there has been a change in single statistics over time that could be contributing to income inequality. Mare and Schwartz and Qian used crossing models, which determine educational differences between partners that are serious barriers to intermarriage, and used this model to analyze the difficulties that people face in finding a partner with a different education level than their own [2,9]. This can in part be attributed to the fact that people tend to find a spouse while in school, given that school can act as a mating market that brings compatible people together. The present study does not address issues of barriers in the marriage market, but rather assumes such barriers exist. Assortative mating by education is of importance because it is thought to cause higher levels of income inequality across a country. Stating that there is an increase in inequality is not sufficient – increasing inequality is a crucial topic that needs addressing through policy intervention. Policy suggestions that could potentially aid in correcting the income inequality problem in the U.S. are provided at the end of this study.

Studies by Undergraduate Researchers at Guelph (SURG) 43


Educational assortative mating in the U.S. and the effect on income inequality by household (Swierzewski) Table 1. Assortative mating of males and females in heterosexual relationships by education in a) 1960 and b) 2005.

a.

and thus a greater gap between households with higher and lower levels of education attainment. This relationship will be looked at in greater detail later in this study. Income differences between males and females After looking at marriage patterns of female and male partnerships with respect to education attainment, the percentage of women earning 1.5 times or greater than their partner’s salary, approximately equal to their partner’s salary (defined as being between 0.5 times and 1.5 times the partner’s salary), and 0.5 times or less than their partner’s salary were examined by cohort (Table 2). Table 2. Relative salaries of females as compared to their husbands’ salaries (%) by cohort.

Volume 6 • Issue 1 • Fall 2012

give insight into the dispersion of education at the top end of the education distribution. Tables 1a and 1b show the joint distribution of men’s and women’s education attainment, the former table for 1960 data, and the latter table for 2005 data. Similar data exists for 1970, 1980, and 1990, but is not shown here. Collectively, these data show that for each year, the majority of partnerships are with a partner of the same level of education; thus, there is strong evidence of homogamy by education for all years assessed in this study. While homogamy is present, there are still a large number of males partnering outside of their education level. In 1960, 56% of men with primary education completed partnered with females with less than primary education completed. This indicates heterogamy with respect to men’s education (men marrying down) at the bottom end of the education distribution. In the same year, 60% of men who completed secondary school partnered with women who obtained a university degree. This indicates heterogamy at the top end of the education distribution (men marrying up). Data from other years also depicts a similar relationship: men tend to marry down at the bottom end of the education distribution, but tend to marry up at the top. While there is still a considerable amount of marrying up present, data for 1980, 1990, and 2005 show that homogamy by education seems to be increasing over time. In 2005, homogamy at the top end of the education distribution is more noticeable than in other years. The percentage of males and females finding a partner outside of their education level is considerably lower than in 1960. Findings suggest that not only is homogamy more prevalent in more recent years, people are also choosing to obtain higher levels of education. Trostel notes a link between higher education and earning potential [14]. This link combined with assortative mating by education creates potential for higher household earnings,

b.

Studies by Undergraduate Researchers at Guelph (SURG) 44


Educational assortative mating in the U.S. and the effect on income inequality by household (Swierzewski) Table 3. Single and never-married, divorced, spouse-absent, and widowed males and females in the U.S. per cohort (%) between a) 18 to 65 years of age and b) 24 to 35 years of age.

a.

b.

Singles statistics Singles statistics are important to analyze because they help in understanding individuals’ partner preferences and how these may have changed over time. Examining whether there is an increased number of single individuals at an older age helps give insight into determining what factors are contributing to the changing marriage market. For the purpose of analyzing singles statistics of males and females, the sample was first restricted to males

and females not in a relationship between the ages of 18 to 65. Table 3a depicts summary statistics of single males and females from 18 to 65 years of age by cohort, significant within a 99% confidence interval. The table shows that the percentage of single, never-married males has remained relatively constant from 1960 to 2005, while the percentage of single, never-married females has increased from 20% in 1960 to 37% in 2005. The percentage of males who are divorced, separated, or have a spouse who is absent has increased by 15% from 1960 to 2005, whereas for females, the percent increase was 19%. Conversely, the number of widowed females has fallen drastically over the years – in 1960, 47% of females were widowed whereas in 2005, this percentage was only 11%. There has also been a considerable decrease in the number of widowed men – in 1960, 19% of men were widowed, whereas in 2005, only 4% were widowed. In a second analysis, the single men and women sample was restricted to individuals between 24 and 35 years of age, since the majority of people within this age group tend to be entering into partnerships with the opposite sex. Table 3b shows that the number of single, never-married males between 24 and 35 years of age increased from 73% in 1960 to 82% in 2005, while the number of single, nevermarried females between 24 and 35 years of age increased from 31% in 1960 to 72% in 2005. The percentage of those divorced, separated, or with an absentee spouse decreased by 6% for males and by 30% for females between 1960 and 2005. There was also a notable decrease in the percentage of the population between 24 and 35 years who were widowed

Volume 6 • Issue 1 • Fall 2012

Over time, women’s salaries appear to be approaching men’s salaries, though men continue to make considerably more, even in recent years. In 1960, 69% of women made 0.5 times their husband’s salary or less, compared to 50% in 2005. In 1960, only 7% of women earned 1.5 times their husband’s salary or greater but in 2005, this percentage increased to 18%. Bias exists in the statistics at both ends of the distribution (i.e. where women are earning 1.5 times their husband’s salary or greater and where women are earning 0.5 times their husband’s salary or less) because the statistics shown also take into account individuals who are not working. In earlier years, it was a social norm for women to be stay-at-home wives and for men to be the primary breadwinners of the household whereas in later years, stay-at-home fathers with working wives have become more prevalent. Given the data set, there is no way to tell which individuals are unemployed and searching for work, and which have made the choice to be stay-at-homeparents and to not partake in the workforce.

Studies by Undergraduate Researchers at Guelph (SURG) 45


Educational assortative mating in the U.S. and the effect on income inequality by household (Swierzewski) the number of widowed males dropped from 25% in 1960 to 0.32% in 2005, and the number of widowed females dropped from 12% in 1960 to 1% in 2005. Collectively, singles statistics show that women and men are staying single longer and those that do marry are more prone to divorce, separate, or have an absentee spouse than at any other time in history. In addition, there are a decreasing number of marriages lasting until ages where widowing would commonly occur. Increasing age of marriage can be attributed to several factors including increasing pressure to complete higher levels of education and later entry into the work force, as well as higher numbers of women pursuing an education and entering into the workforce. The latter factor results in lowered dependency of women on a spouse for financial support and raised expectations for what constitutes a good mate. Although there does not seem to be a shortage of single males statistically, it seems that women are increasingly searching for spouses with higher levels of educational attainment. This is supported by data which indicates that males with lower levels of education stay single for longer periods of time. . Independent variables

where and represent income earned by males and females respectively in heterosexual partnerships within a particular state during a particular year (1960, 1970, 1980, 1990, or 2005). This variable does not take into consideration a child or teen residing at their parents’ residence and their income contribution to the household. Mean household earnings in the U.S. have changed considerably over time; in 1960, the mean annual household income was USD 4,872.60, as compared to USD 9,440.90, USD 19,455, USD 36,572.40, and USD 68,508.30 in 1970, 1980, 1990, and 2005, respectively. Although these income levels are in nominal dollars, this does not affect analysis due to the presence of a time trend dummy variable. Gross Domestic Product “GDPstate” is a variable that represents the market value of all final goods and services at the state level. This is an important variable in measuring the overall well-being of a state because it acts as an indicator of standard or living. The unemployment rate variable is calculated by taking the number of unemployed people in a state and dividing this value by the total number of individuals capable of being in the labour force. This variable was generated from IPUMS-USA data and is a raw variable not weighted or corrected for specific factors in the economy that the Bureau of Labour Statistics generally accounts for. Unemployment rate captures the standard of well-being of a state. This variable is somewhat correlated with mean household earnings because the greater the number of people reported as unemployed, the lower household earnings tend to be. The percentage of the mean number of households in a particular state in a given year that own their own residence is a variable expressed as a decimal ranging in value from 0 to 1. More households own their own residence in recent years of analysis as compared to earlier years. In 1960, the percentage of the mean number of households in a particular state who owned a residence (expressed as a decimal) was 0.67, compared to 0.87 in 2005. There is some bias associated with this variable because some people with considerably high incomes may choose to rent a residence instead of own one; this is especially true for business professionals who travel. However, such individuals would only account for a small portion of the population. In general, the higher the poverty in a particular state, the less likely households are to own a residence; thus the percentage of the mean number of households that own their own residence is used as a proxy for poverty. There is a small degree of multicollinearity between ownership of residence and unemployment rate.

Volume 6 • Issue 1 • Fall 2012

The effect of educational assortative mating on income inequality by household was assessed in this study using fixed effects. There are six main variables of interest: mean women’s education attainment, mean household earnings, state-level Gross Domestic Product, state-level unemployment rate, state-level mean household ownership, and a time trend variable. The mean women’s educational attainment variable was created to determine the causal effect of assortative mating on income inequality at the state level. This variable is the percentage of the mean number of women in heterosexual partnerships in a particular state that have the same or greater level of education as their partner expressed as a decimal ranging in value from 0 to 1. The mean education level of women relative to that of their partners was fairly constant over time and by state – in 1960, the mean value of women having obtained an education equal to or higher than that of her spouse was 0.82. This only increased slightly to 0.85 by 2005. This small change in mean can be attributed to the fact that in earlier years, it was more common for both males and females to have attained lower levels of education, but over time, higher education has become increasingly prevalent. The high value of the mean women’s education attainment variable in both 1960 and 2005 indicates that there are not a lot of women in the samples who have lower education levels than their partners. This suggests that institutes of higher education act as filters in the marriage market and can help individuals with similar levels of education find compatible spouses. The mean household earnings variable is an unweighted variable that was created by considering households containing a single heterosexual couple. It is equal to the sum of the couple’s earnings and all other partnership earnings

including government transfers in the state divided by the number of households in the state sample:

Studies by Undergraduate Researchers at Guelph (SURG) 46


Educational assortative mating in the U.S. and the effect on income inequality by household (Swierzewski) Finally, a time trend variable is used for each year of the data and is used to capture changes to the level of inequality that can be explained by a change in time.

Table 4. Gini coefficients by U.S. state by cohort.

Dependent variable: the Gini coefficient The Gini coefficient dependent variable was created using the ineqdeco command by Stephen Jenkins, which is a Stata routine [15]. The Gini coefficient measures the level of inequality in a given sample. Slight gaps exist in this variable for 1970 data because information for seven of the states was unavailable. For the purposes of this study, the Gini coefficient was calculated at the state level by cohort (Table 4). In 1960, New Hampshire had the lowest Gini coefficient (0.249) and thus the highest level of equality, whereas Mississippi had the highest Gini coefficient (0.416) and thus the highest level of inequality. In 2005, New Hampshire had the lowest Gini coefficient (0.329) and thus the highest level of equality, whereas North Carolina had the highest Gini coefficient (0.421) and thus the highest level of inequality. The overall range of Gini coefficients has increased over the years because the level of income inequality by household has increased in most states. U.S. data from 1960 shows a Gini coefficient of 0.304 for 1960 and 0.392 for 2005.

Empirical Methodology Fixed effects regression

Volume 6 • Issue 1 • Fall 2012

Regression was used to attempt to explain the increase in inequality measured by the Gini coefficient due to assortative mating ceteris paribus. Increases in the Gini coefficient indicate increases in inequality. The empirical model examines the conditional association of educational assortative mating on income inequality by households at the state level using fixed effects regression analysis that accounts for state-fixed effects. Fixed effects regression is best suited for these data because the model imposes timeindependent effects for each entry that are possibly correlated with the regressors. This study controls for omitted variables in panel data since the omitted variables vary across states but do not change over time. The variable “women’s education attainment of equal or greater value than partner’s education attainment” shows that as partners are selectively matched based on educational attainment, income inequality results. The general framework used in this analysis is as follows: where

Studies by Undergraduate Researchers at Guelph (SURG) 47


Educational assortative mating in the U.S. and the effect on income inequality by household (Swierzewski) Fixed effects regression analysis assumes that is not independent of or . The variable is treated as the unknown intercept that is being estimated for each state in the United States; it is the entry-fixed effect. The variable is the time invariant regressor and is the unobserved variable that varies from state to state – for example, statespecific laws, policies, or beliefs. The variable is the vector of time dummy variables. Economic intuition was used to select appropriate independent variables. Women’s education attainment equal or greater than the partner’s education attainment “meanedsog” is the causal independent variable for assortative mating by education. The greater the percentage of the mean represented as a decimal, the higher a woman’s education is relative to her male partner. Higher educational attainment of women as compared to their male partners indicates that assortative mating is present, because individuals view education status as an important determinant in selecting a partner. Mean household earnings of a partnership “mnhhearn” was another potential independent variable. The greater the gap between mean household incomes in a particular state is, the greater the income inequality. This variable was initially used in the fixed effects regression but later dropped due to multicolinearity with mean women’s education attainment because higher education is associated with greater returns in terms of earned income in the labour force. Analysis attempts to control for omitted variable bias by including as many relevant variables as possible in the regression. Economic intuition fueled the following preliminary regression:

After detecting the presence of heteroscedasticity in the fixed effects model, the regression was corrected and errors were robust (Table 5). The within R-squared, which is the Rsquared from the mean deviation regression, is 0.842. The intraclass correlation indicates that 67% of the variance was due to differences across panels. The errors are correlated with the regressors by -0.128. The F-statistic of the fixed effects model is 7.91 with a p-value of 0, indicating that the coefficients differ from zero. Table 5. Regression results with robust standard errors.

The conditional association variable of interest, women’s education greater or equivalent to partner’s education, has a p-value of 0 and a t-statistic of 3.99. Therefore, this variable is statistically significant with a coefficient of 0.441, indicating that an increase of 0.1 in the mean number of years of education in a particular state would increase the Gini coefficient by 0.044 ceteris paribus with standard errors of 0.111. The large value of the coefficient indicates that assortative mating has a substantial and, due to the standard errors, significant role in explaining the Gini coefficient. An interpretation of these results is that as a woman’s education level rises to meet and potentially exceed her partner’s education level, she is more likely to participate in the workforce (i.e. put her education to use) and thus contribute to a higher household income. Mean women’s education equal or greater than that of partner’s education is a good variable for measuring assortative mating because it takes into account the fact that both men and women are increasingly pursuing higher levels of education over time by considering the relative difference between men and women’s educational attainment. The presence of educational assortative mating in the U.S. is a likely contributor to income inequality. Because men and women with similar education levels are likely to marry, households consisting of a highly educated couple will make, on average, substantially more than households consisting of a couple with a lower level of education. Thus, the income

Volume 6 • Issue 1 • Fall 2012

From this regression, the statistical significance of each of the variables and the overall fit of the model (the Rsquared and the F-statistic) were taken into consideration and the fixed effects regression was reformed. A multicolinearity test was conducted, and the regression was amended. After this point, the fixed effects regression was then robust to check for heteroscedasticity. Tests were then conducted to check for serial correlation, endogeneity, as well as the joint significance of the independent variables.

Results and Discussion

Studies by Undergraduate Researchers at Guelph (SURG) 48


Educational assortative mating in the U.S. and the effect on income inequality by household (Swierzewski) Assortative mating has increased income inequality the most in Connecticut, with an increase in the Gini coefficient of 0.062. Evidently male-female mating by education attainment has contributed to increasing income inequality by household in 94% of all states within the United States of America, a result that is consistent with the findings of Blau, Kalmijn, Mare, Mare and Schwartz, and Smits et al. [1-4, 6]. Table 6. Effect of assortative mating on change in Gini coefficient by state.

Volume 6 • Issue 1 • Fall 2012

differential between couples of high versus low education attainment is exemplified. The variable GDP per state is statistically significant with a t-statistic of 3.65 and a p-value of 0.001. The beta coefficient of this variable is 5.20e-08, which is very small and indicates that GDP per state does not play a major role in explaining the Gini coefficient’s potential value by state, even though it is significant in its determination. Unemployment rate is not statistically significant with a p-value of 0.797 and a t-statistic of -0.26. The coefficient is -0.036 which indicates that as unemployment increases in a state, equality will also increase ceteris paribus. This initially seems counterintuitive, but can be explained by the fact that the Gini coefficient measures the level of inequality in a state and does not explain how inequality is distributed. Greater unemployment means that the distribution of income in a particular state is more equal because fewer people are generating an income. Even though this variable is not statistically significant using a 95% confidence interval, it is intuitively important in explaining income inequality. The ownership of residence variable is statistically significant with a p-value of 0 and a t-statistic of -4.23 on a 95% confidence interval. The coefficient for home ownership is -0.232 indicating that an increase of 0.1 in the percentage of the mean number of home ownerships in a state expressed as a decimal would decrease the Gini coefficient by 0.023 ceteris paribus. The more people who own a home in a state there are, the more likely there will be income equality. Because residence ownership is used as a proxy for poverty in this study, more residence ownership implies that there are less people living in poverty. If households are unable to afford home ownership and are forced to rent, then they most likely do not have the income required to own a home. The time trend dummy variables for 1970, 1980, 1990, and 2005 are all statistically significant ranging in p-values from 0 to 0.003 and t-statistics from 3.15 to 9.16. These variables make it possible to capture state-specific trends since the time dummy variables capture the difference in inequality that is occurring with respect to time. These variables are extremely important with panel data. The time dummy variable for 1960 was dropped due to multicolinearity. The variable for 2005 can be interpreted as having a positive impact on inequality meaning that assuming the base year for comparison is 1960, time trend for 2005 accounts for a change in the Gini coefficient by 0.084 all else equal. Overall, assortative mating in the U.S. by education level appears to explain, by conditional association, income inequality by household. Table 6 shows the effect that women’s educational attainment that is the same or higher than that of their partner’s educational attainment has on the Gini coefficient in each state, from 1960 to 2005. In Arkansas, Mississippi, and North Carolina, assortative mating has decreased income inequality levels with changes in the Gini coefficient being -0.008, -0.023, and -0.001, respectively. All other states show a positive correlation between educational attainment and income inequality.

Studies by Undergraduate Researchers at Guelph (SURG) 49


Educational assortative mating in the U.S. and the effect on income inequality by household (Swierzewski)

Conclusions and Limitations

Acknowledgements I wish to thank Fraser Summerfield and Michael Batu for their help in the creation of this paper. I would also like to thank Professor Louise Grogan for her guidance and support throughout the process of constructing this paper.

References 1. Mare, R.D. (1991). Five decades of educational assortative mating. American Sociological Review, 56, 15-32. 2. Mare, R.D. & Schwartz, C.R. (2005). Trends in educational assortative marriage from 1940 to 2003. Demography, 42, 621-646. 3. Blau, P.M. (1977). Inequality and Heterogeneity. New York, NY: Free Press. 4. Smits, J., Ultee, W., & Lammers, J. (1998). Educational homogamy in 65 Counties: an explanation in differences in openness using country level explanatory variables. American Sociological Review, 63, 264-185.

Volume 6 • Issue 1 • Fall 2012

The present study shows a conditional association between assortative mating by education and state-level income inequality in the U.S. between 1960 and 2005. Variables of interest include women’s level of education attainment equal to or greater than that of their partner, mean household income by state, Gross Domestic Product of each state per year, unemployment rate of each state per year, residence ownership status of each household by state per year, and a time trend variable; however, more explanatory variables are required to fully capture a causal effect. Fixed effects regression analysis of panel data from the U.S. Bureau of Economic Analysis and the U.S. Census of the Population were used to determine conditional association. Statistical significance of the causal variable of interest, women’s level of education attainment equal to or greater than that of their partner, as well of other independent variables (except unemployment rate) was shown. The Fstatistic showed joint significance of regressors, indicating that the fixed effects model shows strong conditional association between assortative mating based on education level and income inequality, as measured by the Gini coefficient. Some limitations exist for the regression analysis conducted in this study. The education attainment variable for men and women only categorized education into four groups. Another variable measuring education was available, but was not consistent for all years of observations. Further research and analysis should be conducted to assess assortative mating patterns at the top end of the education distribution, for example those with Bachelor’s, Master’s, doctoral, or professional degrees. Another limitation of this study is that the Gini coefficient only addresses that fact that inequality exists, but says nothing with regards to whether or not there are a large percentage of individuals at the top or bottom ends of the income distribution (i.e. it does not explain how income inequality is distributed). Two states with similar Gini coefficients could have very different income distributions. Further research on this topic is required to analyze how inequality is distributed. In future studies, looking at comparative statistics on the Gini coefficient that compare percentiles of the samples by earnings would be helpful. For example, dividing the average income of the top 5 and 10% of the population by the average income of the bottom 5 and 10% would give insight into the distribution of income in each state over time. Further research should also assess how many households in the sample live below the poverty line because in this way, it would be possible to evaluate whether rising inequality is really a problem. If there is an increase in inequality over time but the same number or fewer households living below the poverty line, then is the rising inequality really a problem?

Increasing income inequality due to educational assortative mating is an issue that deserves government attention. Since 1960, income inequality has been on the rise in the U.S., and policy intervention is necessary to mitigate the problem. One policy suggestion is to strengthen the social safety net. This can be done by decreasing the cost of obtaining a post-secondary education, because removing financial barriers from education would increase accessibility and allow more people to pursue a higher education. Another policy option is to strengthen unions. A lot of unions tend to represent workers at the bottom end of the income and education distribution – for example, factory workers. Strengthening these unions can lead to increased wages and worker benefits. Implementing a Pigovian tax to generate negative externalities is another policy worth considering to aid in correcting income inequality. A negative externality is an action of a product on consumers (in this case, higher education) that imposes a negative side effect, or a social cost, on a third party (in this case, lowered income). A Pigovian tax would mean that those who make more money (and presumably, are more educated) would be taxed more than those who do not. However, a limitation of this system is that the tax money generated may not go directly to those citizens who most need it and may instead be used by the government for other purposes that appear to be more pressing. Thus, another option would be to give greater tax breaks to the wealthy for charity donations and positive financial community involvement, because this would have a direct positive effect on decreasing inequality with respect to income.

Studies by Undergraduate Researchers at Guelph (SURG) 50


Educational assortative mating in the U.S. and the effect on income inequality by household (Swierzewski) 5. Lu, Y., Morissette, R., & Schirle, T. (2011). The growth of family earnings inequality in Canada. Income and Wealth, 57, 23-29. 6. Kalmijn, M. (1991). Shifting boundaries: trends in religious and educational homogamy. American Sociological Review, 56, 786-800. 7. Daly, M.C. & Valetta, R.G. (2006). Inequality and poverty in the United States: the effects of rising dispersion of men’s earnings and changing family behavior. Economica, 73, 75-89. 8. Rose, E. (2001). Marriage and assortative mating: how have the patterns changed? Center for Statistics in the Social Sciences at the University of Washington. 9. Qian, Z. (1998). Changes in assortative mating: the impact of age and education, 1970-1990. Demography, 35, 279-292. 10. Juhn, C. (1992). The decline of male labour market participation: the role of declining market opportunities. Quarterly Journal of Economics, 107, 79-121.

11. Fortin, N. & Schirle, T. (2006). Gender dimensions of changes in earnings inequality in Canada. In Green, D.A. & Kesselman, J.R. (eds.) Dimensions of Inequality in Canada. Vancouver, BC: UBC Press. 12. Minnesota Population Center. (2011). Integrated Public Use Microdata Series, International: Version 6.1. Minneapolis: University of Minnesota. [Machinereadable database]. 13. Ruggles, S., Alexander, T., Genadek, K., Goeken, R., Schroeder, M.B., & Sobek, M. (2010). Integrated Public Use Microdata Series, Version 5.0. Minneapolis: University of Minnesota. [Machine-readable database]. 14. Trostel, P.A. (2010). The impact of new college graduates on instrastate labour markets. Journal of Education Finance, 36, 186-213. 15. Jenkins, S.P. (1999). "INEQDEC0: Stata module to calculate inequality indices with decomposition by subgroup," Statistical Software Components S366007, Boston College Department of Economics, revised 24 Feb 2010.

Volume 6 • Issue 1 • Fall 2012

Studies by Undergraduate Researchers at Guelph (SURG) 51


Research Article

Differences in composition of territories in relation to behaviour, stage, and depth of the three-spot damselfish, Stegastes planifrons, in Caribbean coral reefs Maria J. Arroyo Gerez Department of Integrative Biology, College of Biological Sciences, University of Guelph, Guelph, ON Canada. Faculty supervisor: Elizabeth G. Boulding. For correspondence, please email: mjarroyogerez@gmail.com.

Abstract Coral reef diversity is correlated with the depth at which the reefs are found, the energy available for biological processes, and the species’ roles and presence throughout the food chain. Can a specific species activity alter the whole ecosystem? Can a small-scale, short-term activity such as fish behaviour have a long-term effect on a larger scale, that of the reef? Can the life stage of a species mediate substrate competition? The three-spot damselfish (Stegastes planifrons) is hypothesized to regulate competition between substrate coverage by actively farming – protecting from herbivores and weeding – in order to regulate the algal species composition and percent coverage of the reef. This behaviour is observed in both juvenile and adult fish. Deeper patches are predicted to have less coral diversity and higher algal diversity; juvenile fish are predicted to have less diversity in their patches than adults. Coral and algal diversity are hypothesized to be negatively correlated. In this study, behaviour of the S. planifrons was classified into one of four categories (active patrolling, passing patrolling, farming, and hiding) and palatable algae surface area coverage was digitized from photographs; life stage was either juvenile or adult. Coral and algal genus diversity were measured along a depth gradient of 0-16 m where the diversity of the reef was thought to be the highest. A three-way ANCOVA was performed to test whether fish behaviour (a small-scale, short-term process), depth, or fish development stage (juvenile or adult) had a significant effect on coral or algal diversity (a large-scale, long-term process). Results showed a significant effect of fish behaviour category on algal genera diversity, and a significant effect of depth on both algal and coral genera diversity. Farming yielded significantly more algal coverage than hiding. This study shows that small-scale, short-term behaviours by S. planifrons can have an effect on algal genera diversity on coral reefs in Utila, Honduras. Keywords: Stegastes planifrons (three-spotted damselfish); behaviours (short-term, small-scale); stage of development (juvenile, adult); depth; coral reef and algal diversity (long-term, large-scale); damselfish territories

Introduction (Pomacentridae) are important species for marine benthic ecosystems [3,6-8]. Through herbivory on competitive dominants and controlled selective farming practices, they regulate the algal species composition and percent coverage, especially in shallow parts of the reef [3,7,9-12]. Herbivory consists of consuming algae while farming involves selective weeding of algae to increase the presence of another species [7,11]. Algae coverage is positively correlated with damselfish farming techniques since they crop the turf and palatable algae to maintain them at their optimum growth rate while removing unpalatable algal species [4,13,14]. Damselfish behaviour such as defence and farming can alter the percent algal coverage within their specific territory [7].

Volume 6 • Issue 1 • Fall 2012

Coral reefs are important for diversity and productivity in marine ecosystems [1,2]. Patterns of coral reef diversity are predictable with depth – an increase in coral diversity with depth for the first 20m, followed by a decrease as light energy decreases with depth [3]. The shallower parts of the reef (above 12m) are the most vulnerable sections for disturbance because a higher abundance of fish, living coral, and algal coverage are found in these areas [3,4]. However, natural and human induced intermediate disturbances have in more recent years led to a higher diversity of algae, coral and fish at these depths [1-3,5]. Ecosystem biodiversity can be altered by certain species that act as mediators within natural processes. Damselfish

Studies by Undergraduate Researchers at Guelph (SURG) 52


Differences in composition of territories in Caribbean coral reefs (Arroyo Gerez) behaviours, then territory composition is not expected to differ significantly within the shallow range down to 15m. Also, if farming increases the benthic surface area coverage and genera of algae present in territories, then patches with higher algal diversity should have lower coral diversity due to inter-functional group competition for primary and secondary space on the substrate.

Materials and Methods Data collection The study sites were located on the south side of Utila, Bay Islands, Honduras (16°6'N, 86°55'W), on the Mesoamerican Barrier Reef. Daily data collection occurred over a two-week period in July 2011 throughout the coral reef: back reef, reef flat and outer slope [4]. Ranges of depths of the observed territories were from 0.4 m-15.6 m, and were sorted by depth classes (0.0-1.0m, 1.1-5.0m, 5.1-7.0m, 7.1-9.0m, 9.1-11.0m, 11.1-16.0m), fish behaviour, fish stage and substrate composition. Depth classes were based on the equipment used to collect the data; the first class was collected using snorkelling gear, while the deeper ones were done with SCUBA equipment. Depth, fish stage (juvenile or adult), territory composition and diversity (algae, coral, invertebrate, and other fish genera), and fish behaviour (active patrolling: chasing of other fish out of their territories; passive patrolling: swimming around their territories; hiding; farming: weeding, cutting and spitting algae outside their territories) were noted at each territory. Behaviours were noted at three stages of the observation period: initial, when first spotted; intermediate, when other observations were taking place; and final, after all observations were done, and the fish were no longer being physically disturbed by any external activity done by the collector. The behaviour that was used was the intermediate one as it appeared to be the most stable one of the three. Territories were chosen where an S. planifrons was spotted and observed for approximately five minutes or until all data were collected [14]. The territories’ boundaries were recognized by the turf coverage and the defensive behaviour of the fish whenever their territory was invaded [14]. Digital photographs (Olympus Tough 8000 with Ikelite underwater housing, 1.6x1.2mp) were taken of each territory. Photographs of six territories from each depth class, and six replicates from each depth class (n=36 territories) were analyzed for palatable algal coverage using ImageJ software [27]. Photographs included either a central area of a territory, a larger area, or the full area of a territory depending on its location and the feasibility of obtaining the full territory. The algal coverage was determined as the area covered by turf and Dictyota spp. divided by the total area of the photographic frames. This proportion was later transformed to arcsine values for statistical analysis. Diversity was determined as the number of algal genera

Volume 6 • Issue 1 • Fall 2012

By selectively farming preferred palatable algae, damselfish affect interspecific competition between massive coral species (Montastrea spp.) and branching species (Acropora spp.), as well as coral and algal competition, for primary substrate since the main limiting resources that algae and coral compete for in coral reefs are light availability and space on the appropriate substrate [3,4,9,12,15-17]. Patches of algae on dead coral shade live coral heads, and prevent photosynthesis by the zooxanthellae that live inside hermatypic coral [18]. Algae also can colonize live coral tissue and induce abrasion by bacteria that settle next to the algae; this harms the coral and diverts energy resources into defence mechanisms rather than into growth [18]. Disturbances caused by damselfish have been described as small-scale (1m2), short duration (days or weeks) and frequent (lasting for months or years), thus contributing positively to a higher diversity in coral reefs through algal control (i.e. more productive algal patches, and more diverse sites) leading to more diverse fish populations and coral species presence [1,2,13]. Damselfish territories can be found covering up to 40-50% of substratum on shallow reefs [14]. Both coral and algal diversity are decreased with depth. The algae farmed the most by damselfish – Lobophora, Dictyota and Halimeda – are the most common algae found in the Caribbean [14,15]. Farming by certain damselfish species creates monocultures of these algae [19-22]. These patches of monoculture algae are created by either active exclusion of other herbivore species by direct attack, or by selective weeding-cutting of the preferred algae which is left the most abundant in their specific territory [3,13,14,22,23]. Farming practices may also result in species-rich patches that differ between damselfish species [7,24]. In Japan, Stegastes nigricans farms red algae monocultures, affecting the algal species’ ability to recover from disturbance [22,23,25]. In the Caribbean, other Pomacentridae species have been studied but no research has been conducted on the effects of their farming habits on algae composition, or on the effect that algae have on coral growth and recruitment [7,9,17,26]. Other studies have been performed in localized areas where high densities of aggressive and territorial Pomacentridae are found. Stegastes planifrons was observed in Utila due to their easily distinguishable features and their passive territoriality [personal observations]. The purpose of this study was to determine the effect that S. planifrons have on benthic composition inside their territories regardless of the depth at which they are found. Algal and coral diversity were chosen as variables due to their contribution to coral reef productivity. Damselfish control over algal diversity can indirectly lead to changes in coral diversity in that species of coral that colonize and grow faster will have better chances of survival within damselfish territories. Short observations of fish activities yielded an estimated regular activity that could be linked to the diversity of their territories. If S. planifrons controls algae and coral benthic coverage by its farming techniques and territorial

Studies by Undergraduate Researchers at Guelph (SURG) 53


Differences in composition of territories in Caribbean coral reefs (Arroyo Gerez) present in each territory. Data analysis A total of 87 territories (N =87) were observed. Fish behaviour, fish stage and depth class were analyzed as independent variables in a three-way ANCOVA using the General Linear Model module of SYSTATW5 (SYSTAT Software Inc.). The analysis was done separately for the two territory composition dependent variables: number of algal genera and number of coral genera. Depth and behaviour were the two variables then tested for effect on genera diversity. The independent variables were depth (continuous: 0-16m), fish stage (categorical: adult or juvenile) and fish behaviour (categorical: farming, passive patrolling, active patrolling and hiding). As fish stage did not have a significant effect on algae genera diversity, it was removed from both models. Behaviour was significant for the algal diversity analysis and therefore a Tukey post hoc test was used to determine which of the four behaviour categories was significantly different from the others with respect to its effect of the number of algal genera present.

Results and Discussion Depth and substrate composition There was a significant difference in the diversity of algae (M=2.4 genera) and coral genera (M=1.5) present (p<0.001) along the depth gradient (Figure 1). The general trend of diversity is an increase with depth, for both algae (r2=0.135, SE=1.024) and coral (r2=0.076, SE=0.983) (Figure 1). Algae surface coverage increased in shallow waters (until 11 m), then decreased after this depth (Table 1). These results support evidence for higher productivity in shallower reefs and in damselfish territories [13].

Table 1. Depth categories and their correspondent arcsine (ᵒ) value of algal benthic coverage of territories at different depths. Six territories per depth category were randomly chosen and analyzed for the algal coverage (n=36, SE=2.58). Depth category (m) Arcsine (ᵒ) algae coverage

0.0– 1.0

1.1 – 5.0

5.1 – 7.0

7.1 – 9.0

9.1 – 11.0

11.1 – 16.0

20.30

34.00

34.72

25.12

34.42

24.32

There was not a significant difference between the means of algal and coral genera; this provides indirect evidence that there is a low level of competition between these two functional groups for primary and secondary space. In the reefs surrounding Utila, there has been no evidence of either of the groups becoming the dominant competitor for space [personal observations]. However, diversity increased (Figure 1) with depth for both functional groups, and algae coverage decreased after the first 12 m within the depth gradient studied (Table 1) [3,4]. Algal diversity was more affected by depth (r2=0.135, p<0.001) than coral diversity (r2=0.076, p=0.10). The depth categories that had the lowest diversity were the first two (0.0-5.0m); this may be due to the physical location of the sites. The shallow reef flats might be more vulnerable to human-induced disturbances, since the sites studied are located closer to hotels and houses than the deeper sites located at the reef crest and along the wall. These results disagree with the Intermediate Disturbance Hypothesis where the most diverse sites should be the ones where the disturbances are promoting biodiversity instead of limiting it [5]. Human settlements can be an intermediate disturbance if controlled, or an excessive one that tends to limit biodiversity [28]. In Utila, human influence has a negative effect on the overall biodiversity and structure of marine and coastal ecosystems [28]. The diversity and depth relationship found in this study follow patterns previously described in coral reefs [3,4] where the coral and algae are found to be increasing within the depth studied (0-16m). Coral and algae diversity were significantly affected by depth rather than by fish behaviour. Behaviour and stage

Volume 6 • Issue 1 • Fall 2012

Figure 1. Coral and algae genera present along the depth gradient where S. planiforms territories were observed. Data represent the mean number of genera (SE shown as error bar) present in each category. Algae was significant more diverse than coral (p<0.001) and more affected by depth.

Fish were found either isolated inside their territories or sharing a territory with another damselfish. 80 sites of single individuals were observed. Adults were found to be the most active farmers (19 sites) compared to juveniles (9 sites), as well as hiding the most (9:1 for adults and juveniles, respectively). The two patrolling behaviours showed a smaller difference between fish stages. Adults and juveniles were found to be more frequently passively patrolling than actively (19 and 7 sites respectively for adults and 14 and 2 sites for juveniles) (Figure 2). Seven territories were found to have more than one individual; four behaviours were present. The ratio between adults and juveniles was the same

Studies by Undergraduate Researchers at Guelph (SURG) 54


Differences in composition of territories in Caribbean coral reefs (Arroyo Gerez) for passive patrolling and hiding (1 adult: 1 juvenile) but differed for farming (3 adults: 1 juvenile) and active patrolling (4 adults: 1 juvenile).

Figure 2. Frequency of fish behaviour categories observed throughout the territories where a single first was found (N=80). Adult (n=54) fish are more abundant in general than juveniles (n=26). Eight territories were found to have more than one fish and are not included in this graph. Fish stages were not significant for algal and coral diversity changes along the depth gradient studied.

Table 2. Two-way ANCOVA analyses with ocean depth (0-16m), and fish behaviour (farming, active patrolling, passive patrolling, and hiding) as the independent variables but with different dependent variables (A: number of genera of algae and B) number of genera of coral). Fish life stage did not have a significant effect on either algae or coral genera diversity; therefore it was removed from both models. * = significant at Îą=0.050. SS

df

MS

F

p

A: Algae no. genera Depth Behaviour* Error

10.420 9.669 79.469

1.000 3.000 82.000

10.420 3.223 0.969

10.752 3.426

0.002 0.024*

B: Coral no. genera Depth* Behaviour Error

7.464 1.860 80.328

1.000 3.000 82.000

7.464 0.620 0.980

7.620 0.633

0.007* 0.596

Behaviour, algal, and coral coverage The most abundant behaviour was passive patrolling (42.04%) followed by farming (32.95%), hiding (12.50%) and active patrolling (11.36%). Behaviour had no significant effect on coral genera diversity (p=0.596) (Figure 3, Table 2). Algae, however, was affected by fish behaviour (p=0.024, df=3). (Figure 4, Table 2). The most significant difference between behaviours was found to be between hiding and farming (p=0.024, df=82) (Figure 4, Table 3). Results show that farming yielded the most algal coverage followed by active patrolling, passive patrolling and hiding (Figure 4, Table 3).

Figure 4. Mean number of algae genera present in relation to the four behavioural categories of fish. Farming had the most genera present (M=3.10, SD=1.135) followed by active patrolling (M=2.80, SD = 1.135), passive patrolling (M=2.351, SD=0.977) and hiding (M=2.272, SD=0.646). Table 3. Probability values for post hoc Turkey test. Multiple comparisons test the effect of fish behaviour (farming, active patrolling, passive patrolling, and hiding) on the number of genera of algae present (see Table 2 for ANCOVA).

Figure 3. Mean number of coral genera present in relation to the four behavioural categories of fish. Hiding had the most genera present (M=2.27, SD=1.103) followed by active patrolling (M=2.20, SD = 1.229), farming (M=1.79, SD=1.205) and passive patrolling (M=1.78, SD=1.104).

Active 1.000 0.944 0.248 0.764

Farming

Hiding

Passive

1.000 0.024* 0.142

1.000 0.529

1.000

Overall, the results of this study are in agreement with literature pertaining to algal coverage increasing due to damselfish behaviour as farmers, excluders and grazercontrollers, and diversity trends [3,4,9,13,14,16,17,23]. The effect that algae cover has on coral appears to be insignificant, at least in the section of Mesoamerican reef that was studied, which supports the positive effect of damselfish on healthy coral reefs [17]. Algal coverage does not represent a threat to healthy living coral [23], as there is not significant

Volume 6 â&#x20AC;˘ Issue 1 â&#x20AC;˘ Fall 2012

Active Farming Hiding Passive

Studies by Undergraduate Researchers at Guelph (SURG) 55


Differences in composition of territories in Caribbean coral reefs (Arroyo Gerez) evidence from the results of studies of interspecific competition [16,17]. Algal diversity was not found to alter the coral diversity or adult coral mortality; instead, depth was the factor that had the most significant effect on coral diversity changes (p=0.007, df=1) (Table 2) [23]. The results found in this study are estimates of algal coverage within damselfish territories signalling greater productivity within these patches [13]. Further studies of algae chlorophyll-α concentrations will provide more evidence on the actual relevance of these territories, the effect that depth and fish behaviour have on them, and their overall role in marine ecosystems [29-31]. If chlorophyll-α concentrations were merged with the obtained results, then damselfish regulation of primary productivity and substrate coverage competition might be better understood [17]. Further work will yield better explanations on the function of the coral reef and the role of these fish.

Acknowledgements I would like to dedicate this paper to my father, whose life has inspired me and given me the strength I needed to continue, and thank my mother and sister for their support all along the way. I would like to thank Dr. Elizabeth Boulding (University of Guelph) for running the ANCOVA analyses and for her advice all along the preparation of this project, Dr. Alan Pinder (Dalhousie University) for his fieldwork supervision and advice, and Dr. Kevin McCann (University of Guelph) for being the second reader of this paper. I also thank Dr. Justin Hines (Operation Wallacea Canada), Dr. Alison Curtis and Thomas Fey (Operation Wallacea UK), Sarah Laverty and Maria Isabel Chaves (Coral View Dive Resort, Utila) for their diving accommodations and facilities. My thanks to my dive buddy, Chloe Ready for her constant support and the editors of my drafts, Jenna McDermott, Katlynd Treiber-Vajda, and Nathaniel McLaren as well as two anonymous reviewers for all the hard work they put into improving my manuscript.

References

2. Sebens, K.P. (1994). Biodiversity of coral reefs: what are we losing and why? American Zoologist, 34, 115-133. 3. Huston, M.A. (1985). Patterns of Species Diversity on Coral Reefs. Annual Review of Ecology, Evolution, and Systematics, 16, 149-177.

5. Connell, J.H. (1978). Diversity in Tropical Rain Forests and Coral Reefs. Science, 199, 1302-1310. 6. de Ruyter van Steveninck, E.D. (1984). The composition of algal vegetation in and outside Damselfish territories on a Florida reef. Aquatic Botany, 20, 11-19. 7. Ceccarelli, D.M. (2007). Modification of benthic communities by territorial damselfish: a multispecies comparison. Coral Reefs, 26, 853–866. 8. Allison, A.W., Cook, M.M., DiGirolamo, A.L., Eme, J., Grim, J.M., Hohmann, B.C., Conner, S.L., McGill, C.J., Pomory, C.M., & Bennet, W.A. (2008). A comparison of damselfish densities on live staghorn coral (Acropora cervicornis) and coral rubble in Dry Tortugas National Park. Southeastern Naturalist, 7, 483-492. 9. McCook, L.J., Jompa, J., & Diaz-Pulido, G. (2001). Competition between coral and algae on coral reefs: a review of evidence and mechanisms. Coral Reefs, 19, 400-417. 10. Ceccarelli, D.M., Jones, G.P., & McCook, L.J. (2001). Territorial damselfishes as determinants of the structure of benthic communities on coral reefs. Oceanography and Marine Biology: Annual Review, 39, 355–389. 11. Ceccarelli, D.M., Jones, G.P., & McCook, L.J. (2005). Foragers versus farmers: contrasting effects of two behavioural groups of herbivores on coral reefs. Oecologia, 145, 445–453. 12. Hinds, P.A., & Ballantine, D.L. (1987). Effects of the Caribbean threespot damselfish, Stegastes planifrons (Cuvier), on algal composition. Aquatic Botany, 27, 299-308. 13. Klumpp, D.W., McKinnon, D., & Daniel, P. (1987). Damselfish territories: zones of high productivity on coral reefs. Marine Ecology Progress Series, 40, 41-51. 14. Klumpp, D.W., & Polunin, N.V.C. (1989). Partitioning among grazers of food resources within damselfish territories on a coral reef. Journal of Experimental Marine Biology and Ecology, 125, 145-169.

Volume 6 • Issue 1 • Fall 2012

1. Jackson, J.B.C. (1991). Adaptation and Diversity of Reef Corals. Bioscience, 41, 475-482.

4. Chabanet, P., Ralambondrainy, H., Amanieu, M., Faure, G., & Galzin, R. (1997). Relationships between coral reef substrata and fish. Coral Reefs, 16, 93- 102.

Studies by Undergraduate Researchers at Guelph (SURG) 56


Differences in composition of territories in Caribbean coral reefs (Arroyo Gerez)

15. Scott, F.J., & Russ, G.R. (1987). Effects of grazing on species composition of the epilithic algal community on coral reefs of the central Great Barrier Reef. Marine Ecology Progress Series, 39, 293-304.

23. Sammarco, P.W. (1983). Effects of fish grazing and damselfish territoriality on coral reef algae. I. Algal community structure. Marine Ecology Progress Series, 13, 1- 14.

16. Letourneur, Y. (2000). Spatial and temporal variability in territoriality of a tropical benthic damselfish on a coral reef (Reunion Island). Environmental Biology of Fishes, 57, 377–391.

24. Hixon M.A., & Brostoff, W.N. (1983). Damselfish as keystone species in reverse: intermediate disturbance and diversity of reef algae. Science, 230, 511- 513.

17. Williams, A.H. (1981). An analysis of competitive interactions in a patch back-reef environment. Ecology, 62, 1107-1120. 18. Hughes, T.P., Baird, A.H., Bellwood, D.R., Card, M., Connolly, S.R., Folke, C., Grosberg, R., Hoegh-Guldberg, O., Jackson, J.B.C., Kleypas, J., Lough, J.M., Marshall, P., Nyström, M., Palumbi, S.R., Pandolfi, J.M., Rosen, B., & Roughgarden, J. (2003). Climate change, human impacts and the resilience of coral reefs. Science, 301, 929-933. 19. Vine, P.J. (1974). Effects of algal grazing and aggressive behaviour of the fishes Pomacentrus lividus and Acanthurus sohal on coral-reef ecology. Marine Biology, 24, 131-136. 20. Montgomery, W.L. (1980a). The impact of nonselective grazing by the giant blue damselfish, Microspathodon dorsalis, on algal communities in the Gulf of California, Mexico. Bulletin of Marine Science, 30, 290-303. 21. Montgomery, W.L. (1980b). Comparative feeding ecology of two herbivorous damselfishes (Pomacentridae: Teleostei) from the Gulf of California, Mexico. Journal of Experimental Marine Biology and Ecology, 47, 9-24.

26. Frédérich, B., Fabri, G., Lepoint, G., Vandewalle, P., & Parmentier, E. (2009). Trophic niches of thirteen damselfishes (Pomacentridae) at the Grand Récif of Toliara, Madagascar. Ichthyological Research, 56, 10-17. 27. Rasband, W.S. (2011). ImageJ. U.S. National Institutes of Health, Bethesda, Maryland, USA. http://imagej.nih.gov/ij/ 28. Canty, S.W.J. (2007). Positive and Negative Impacts of Dive Tourism: The Case Study of Utila, Honduras. Thesis, Lund University. Lund, Sweden. 29. Tyler, J.E. (1974). In situ detection and estimation of chlorophyll and other pigments-preliminary results. Botany, 51, 671-678. 30. Shulenberger, E., & Reid, J.L. (1981). The Pacific shallow oxygen maximum, deep chlorophyll maximum, and primary productivity, reconsidered. Deep Sea Research, 28, 901-919. 31. Marwood, C.A., Solomon, K.R., & Greenberg, B.M. (2001). Chlorophyll fluorescence as a bioindicator of effects on growth in aquatic macrophytes from mixtures of polycyclic aromatic hydrocarbons. Environmental Toxicology and Chemistry, 20, 890–898.

Volume 6 • Issue 1 • Fall 2012

22. Hata, H., & Kato, M. (2002). Weeding by the herbivorous damselfish Stegastes nigricans in nearly monocultural algae farms. Marine Ecology Progress Series, 237, 227–231.

25. Hata, H., Watanabe, K., & Kato, M. (2010). Geographic variation in the damselfish- red alga cultivation mutualism in the Indo-West Pacific. BMC Evolutionary Biology, 10, 185-195.

Studies by Undergraduate Researchers at Guelph (SURG) 57


Research Article

In silico modelling of Arp1 and Arp2 as targets of Photox ADP-ribosylation Lianne P. Davidge† and Kristen V.L. Daly† Department of Molecular and Cellular Biology, College of Biological Sciences, University of Guelph, Guelph, ON Canada. Faculty supervisor: John Dawson. For correspondence, please email: davidgelp@gmail.com. † denotes equal contributions.

Abstract Photox has been identified as an ADP-ribosylating toxin, produced by Photorhabdus luminescens, which targets actin. Recent unpublished studies have suggested that Photox may also target actin-related proteins (Arps). Herein we report possible interacting residues of actin, Arp1 and Arp2, with Photox. Both actin and Arp1 were predicted to interact with + Photox in six main groups across the NAD binding surface whereas four interaction groups were proposed between Arp2 and Photox. The possible interactions evaluated included hydrogen bonds, salt bridges, and electrostatic and hydrophobic interactions. Our results suggest that Arp1 is a very likely target of Photox, while Arp2 is also a possible candidate; however, the possible interactions between Photox and Arp2 appear to be weaker than the possible interactions between Photox and Arp1. Understanding the various interactions of Photox and its targets may help to provide insight into its biological action, and may potentially lead to a treatment plan for similar toxin-related diseases. Keywords: ADP-ribosylating toxins; Photox (Photorhabdus luminescens); actin-related proteins (Arp1, Arp2) (targeting of); in silico modeling of protein interactions

Introduction and subsequent transfer of the ADP-ribose moiety to the target proteins [5,6]. Most ARTs also have NAD+ glycohydrolase activity, which generates a free ADP-ribose from NAD+ [7]. There are four families of ARTs, which are defined based on their respective targets: type I target heteromeric GTP-binding proteins; type II modify elongation factor two; type III ADP-ribosylate small GTP-binding proteins; and type IV ADP-ribosylate actin [5]. In each case, the ADPribosylation inhibits function of the modified protein and cannot be removed by host enzymes [6]. For example, ADP ribosylation of actin inhibits actin polymerization, preventing synthesis of microfilaments. Thus, as the existing cytoskeleton is depolymerized, it is not replaced and eventually results in cell death. Photorhabdus luminescens are motile, bioluminescent, Gram-negative bacteria, which live symbiotically with nematodes, and infect insects [8]. The P. luminescens genome encodes a wide variety of toxins and hydrolytic enzymes, which when released, kill the insect host within 24 to 48 hours [8,9]. Among these toxins is Photox, a type IV ART actin targeting protein encoded by the plu0822 gene of P. luminescens [9]. Further analysis of the protein determined that it shares primary sequence identity with the

Volume 6 • Issue 1 • Fall 2012

Bacterial infections remain a major cause of human disease and illness in both developed and developing countries. Many bacterial diseases, such as Legionnaires’ disease, Lyme disease, and Escherichia coli 0157:H7 were once thought to have been eradicated; however, in recent years, many have been re-emerging, often with increased resistance [1]. The primary means of bacterial pathogenesis is bacterial multiplication within the host, which elicits an immune response [2]. In order to combat the host’s immune system, many bacteria have evolved additional pathogenic strategies; these include microbial adhesion, the injection of virulence factors to the host, and the expression or secretion of a variety of toxins [2]. Several of the toxins produced by pathogenic bacteria are ADP-ribosyltransferases (ARTs), toxins that act to destroy host cells in the same way that antibiotics act to kill bacteria. Examples of diseases caused by ARTs in humans include diphtheria, cholera, pertussis, botulism, and typhoid fever [3-5]. Given the increase in multidrug-resistant bacterial infections such as Salmonella, there is an increase in interest and urgency to study and understand these enzymes [6]. ARTs act by causing irreversible modifications to host proteins, carried out by cleavage of the N-glycosyl bond between the nicotinamide and the N-ribose of NAD+

Studies by Undergraduate Researchers at Guelph (SURG) 58


In silico modeling of Arp1 and Arp1 as targets of Photox (Daly and Davidge) of Photox on these proteins may help to further the discovery of a treatment for similar toxin-related diseases.

Methods Modelling Photox structure The Photox gene was determined to be plu0822; the amino acid sequence for this gene was then located using the NCBI protein database [9]. The EMBL-EBI ClustalW 2.0 server, used to align three or more sequences highlighting areas of conservation and similarity, was used to identify possible homologues of Photox by a multiple sequence alignment (MSA) with other actin-targeting ARTs [10,15]. A search using the basic local alignment search tool (BLAST), a sequence comparison algorithm used to search databases of sequences, was performed in order to identify any close homologs to the N-domain Photox with that of other ARTs [16]. SignalP 4.0, an internet tool capable of identifying signal peptide sequences in an amino acid sequence, was used to identify possible signal peptide sequences in Photox [17]. The amino acid sequence was used to create a model structure using the Swiss-Model protein structure homologymodeling server [18-20]. Swiss-Model identifies a structural template, aligns the target sequence and the template structure in order build a model, and evaluates the model built [18-20]. Based on the results of the MSA, SpvB was selected as a model structure (PDB accession code: 2GWL). The Photox model generated by Swiss-Model was submitted to the ConSurf 3.0 server, used to estimate and visualize the conservation of amino acid positions in proteins, [21-23] along with the MSA of Photox and the other actin binding proteins, using Photox as the query sequence. The position of NAD+ relative to Photox was determined by using the coordinates of NAD+ bound to SpvB. The hydrogen bonds of Photox and NAD+ were identified based on the results of Margarit et al. [6]. Modelling of Photox:actin As no structure of SpvB:actin was available, a model of Iota:actin was selected for the superposition of Photox in order to model the interactions between Photox and actin (PDB accession code: 3BUZ). Determination of Photox:actin interactions Electrostatic and hydrophobic surface models were used to assess possible interactions between Photox and actin. esidues rom hoto and actin ithin . o each other were considered for potential hydrogen bonds and salt bridges (Appendix 1). Interactions between Iota and actin were considered using a MSA previously generated [5].

Volume 6 â&#x20AC;˘ Issue 1 â&#x20AC;˘ Fall 2012

SpvB ART toxin produced by Salmonella enterica [a]. Photox shares key catalytic residues found in SpvB and other ARTs [9]. The addition of an ADP-ribose group to the R177 of actin completely inhibits actin polymerization [9]. Photox is considered highly toxic to eukaryotic cells as basal expression in yeast caused a severe growth-defective phenotype [9]. Based on primary sequence conservation, ARTs can be divided into two major categories: cholera toxin-like (CT) ARTs and diphtheria toxin-like (DT) ARTs [10]. ARTs have three primary regions of conservation. The CT toxins contain a conserved arginine residue in region one whereas the DT toxins contain a conserved histidine residue [10]. Region two, the region containing the NAD + binding site, contains an STS motif in the CT group and two tyrosine residues in the DT group [10]. Region three contains a conserved tertiary motif involved in ADP-ribosylation activity in both the CT and DT toxins [10]. CT ARTs can be further sub-divided into four main subgroups [10]. The ARTs which specifically target R177 of actin primarily fall into the C2-like subgroup. These include: Bacillus thuringiensis Vip2, Clostridium perfringens Iota Ia, C. difficile CdtA, C. spiroforme Sa, C. botulinum C2-I, and S. enterica SpvB [5,9]. The C2-like toxins have two common structural features: C domain and an N domain [5]. The C domain is essential for catalytic activity and consists of an EXE motif: the first glutamate residue catalyzes the ADPribosyltransfer and the second glutamate residue binds to the nicotinamide ribose [5]. The N domain is responsible for binding and translocation [5]. Although type IV ARTs, by definition, have primarily been shown to target actin, recent unpublished results suggest that actin-related proteins (Arps) in Saccharomyces cerevisiae yeast may also be targets of Photox [b]. Arps, as suggested by their name, are proteins similar, but not identical, to actin [11]. A total of ten Arps have been identified in S. cerevisiae and are grouped based on the similarity of their primary structure to that of actin; the most similar Arps are in the Arp1 family, and as the family number increases, the similarity to actin decreases [11,12]. Arps share between 17% and 60% conservation with actin, particularly in regions critical to the tertiary structure of actin [11,12]. Thus, Arps are suggested to display similar folding patterns to conventional actin. Arps are thought to have cytoskeletal roles, including dynein motor activity for Arp1 and actin polymerization for the Arp2/3 complex [13,14]. They are also suggested to have roles in nuclear activities such as chromatin remodeling and transcription regulation [13]. Arps have been found in widely divergent organisms from yeast to human, thus suggesting conserved and fundamental roles in cells [13]. In this study, we evaluate the possibility that Photox may target yeast Arps using in silico protein modeling and analysis. Both Arp1 and Arp2 are identified as probable targets of Photox, though the interactions between Photox and Arp1 were stronger. Understanding the biological action

Studies by Undergraduate Researchers at Guelph (SURG) 59


In silico modeling of Arp1 and Arp1 as targets of Photox (Daly and Davidge) Identification of yeast Arps as potential targets of Photox Ten Arps in yeast were identified, and their sequences compared to actin using an MSA and the ClustalW 2.0 server [15]. Arps that shared a conserved modifiable arginine residue with actin were identified; an additional MSA was generated using the Arps containing the conserved, modifiable arginine, and actin. A ConSurf model was generated based on the structure of actin. Modelling of Photox:Arp The MSA of Arps with a conserved modifiable arginine residue and actin was evaluated in order to determine which two Arps were the best candidates to model for Photox binding. The evaluation was completed by examining the conservation of residues previously predicted to be involved in Photox:actin binding. Arps having the greatest number of identical residues, or residues of similar hydrophobicity and/or charge compared to actin were selected. To create a model of Arp2, the protein sequence was submitted to Swiss-Model [18-20] without a model structure. Swiss-Model located candidate structure models and generated a theoretical structure for Arp2; this model was based on rabbit actin and was the model which depicted the entire Arp2 structure (PDB accession code: 2Y83 chain O). For Arp1, the PDB was searched using the primary sequence to identify candidate model structures with an E-value cut-off of 0.001; yeast actin selected as a model (PDB accession code: 1YAG). Swiss-Model was used to generate a theoretical structure [18-20]. In order to create a model of the interactions between Photox and the Arps, each Arp was modeled onto the Iota:actin structure (PDB accession code: 3BUZ), using DALILITE [24]. The resulting root-mean-square deviation (RMSD) alue aided in the determination o ho related the rps ere to actin a M abo e 2 meant the structures ere completely unrelated hile a M o . indicated that the structures were very closely related. The PDB text files were edited to create models of Photox:Arp2 and Photox:Arp1. The quality of these models was assessed using the QMEAN server [25].

All theoretical models were selected based on how much of the sequence was modeled on the template as well as identity to the template. For example, the template that had the highest identity with Arp2 did not model the entire Arp2 protein; therefore, the template with the second highest identity was selected as it modeled the entire protein. All protein models were visualized using PyMOL. The tools used were chosen as they are user-friendly, easily accessible and due to the high target-template sequence identities. In particular, Swiss-Model was chosen because the program contains a refinement process and it is capable of selecting the best template for modeling.

Results Photox and other ARTs that bind actin were compared using MSA with specific consideration of the active site (Figure S1). The NAD+ binding site and active site region showed the greatest conservation between the various ARTs. A search using BLAST did not identify any ART homologs to the N-domain of Photox; however, an unknown protein from P. luminescens had 80% homology to the N-domain of Photox [16,27]. SignalP 4.0 did not detect the presence of a signal peptide sequence in Photox [17]. Given the high sequence conservation between the actin binding site of Photox and SpvB seen in the MSA, a structural model of Photox was created based on the SpvB structure (PDB accession code: 2GWL) (Figure 1).

Both Photox:Arp models were analyzed for potential interactions using the same criteria as above: proximity less than 4.0 Å, as well as electrostatic and hydrophobic interactions. The PDBePISA server was used to predict hydrogen bonds and salt bridges [26]. The residues predicted to be involved in Photox:actin interaction were also considered in each Arp, using the MSA of the Arps and actin.

Figure 1. Photox structure. (A) Photox chainbow representation + with NAD in the binding cleft based on the SpvB structure (PDB accession code: 2GWL). (B) Topology diagram of Photox using the same colour scheme as the chainbow representation with the + NAD binding cleft indicated by the black arrow. (C) Close-up view of the catalytic site of Photox depicting important residues for + NAD binding and active site stabilization. Residues are coloured as per the chainbow representation.

Volume 6 • Issue 1 • Fall 2012

Identification of potential Photox:Arp interactions

Studies by Undergraduate Researchers at Guelph (SURG) 60


In silico modeling of Arp1 and Arp1 as targets of Photox (Daly and Davidge)

Figure 2. Identification of yeast Arps as potential targets of Photox. (A) Multiple sequence alignment of the region of actin containing R177, the modifiable arginine residue, and the Arps in yeast using the EMBL-EBI ClustalW 2.0 server [15]. The arrow identifies the conserved modifiable arginine residue. (B) Multiple sequence alignment of actin and the four Arps which contain the modifiable arginine residue, preformed by the EMBL-EBI ClustalW 2.0 server [15]. Residues highlighted in green are those which are proposed to be involved in Photox:actin binding. (C) The ConSurf surface rendering of conservation in the four Arps which contain the modifiable arginine residue, showing the portion of the molecule which interacts with Photox generated using the ConSurf 3.0 server [18-20]. Purple indicates regions of high conservation, whereas cyan represents regions of very low conservation.

Modelling of Photo:actin The superposition of the Photox structure onto an Iotoa:actin structure generated a theoretical model of the Photox:actin binding interface (PDB accession code: 3BUZ) using DALILITE to evaluate the structural similarities between Photox and Iota by maximizing the overlap of their distance matrices [24]. Determination of Photox:actin interactions The DALILITE superposition of Photox onto Iota generated a structural sequence alignment (Figure S3). The actin binding residues on Photox were grouped based on their overall positions on Photox, the region of actin with which they interact, as well as how they relate to the residues which interact in the Iota:actin model (Table 1) [5,9]. Electrostatic and hydrophobic surface models generated in order to identify potential interacting regions between Photox and actin (Figure S4).

Volume 6 â&#x20AC;˘ Issue 1 â&#x20AC;˘ Fall 2012

QMEAN analysis, a server capable of deriving global, entire structure, and local, per residue, error estimates on the basis of a single model, of the Photox model resulted in a QMEAN Z-score of 0.213 [25]. A QMEAN Z-score of one is indicative of a very good model and a strongly negative score is indicative of a model of very low quality [28]. QMEAN found 64.322% structural identity between the C-domains of Photox and SpvB, and the E-value, a measure of accuracy of an alignment with values approaching zero being of higher accuracy and significance, equaled 9.06e-71 [25]. Based on this Photox model, a topology map was generated, highlighting the NAD+ binding region (Figure 2). The active site of the Photox model was analyzed based on the active site of SpvB (Figure 2) [6,9]. The hydrogen bonds involved in NAD+ binding were identified: E355 interacted with the nicotinamide ribose, G289 interacted with nicotinamide, R288 and R231 interacted with the phosphates of NAD+, and K291 and K294 interacted with the adenosine ribose [9]. The catalytic residues were found to include an STS motif (residues 317-319) and E355, the second glutamine residue in the EXE motif. N227 and L290 were identified as active site stabilizing residues [9].

Studies by Undergraduate Researchers at Guelph (SURG) 61


In silico modeling of Arp1 and Arp1 as targets of Photox (Daly and Davidge) Table 1. Proposed interactions between Iota and actin, Photox and actin, Photox and Arp2 (Figure 4), as well as Photox and Arp1 (Figure 6). Residues were categorized in groups based on their location in t e ter ar structure as ell as side c ain orienta on. esidues ere ta en into considera on or ossible interac ons based on ro i it it in , electrostatic properties, as well as hydrophobic properties. Iota

Actin

Group 1

Photox

Actin

H238 E242

K68

Photox

Arp1

H238 E242 E254

K46 K73 R75

Y217 Q221 Q239 T243 S246

K79 V83 E84 D85 W86

Y223

D187

G249 R248

I75 I76 T77 W79

Y217 Q221 S246

T77 W79 D80

Group 3

Y251

R177 M176

Y223

R177 D179

Group 4

G374

L110

H349 F350

L110 P112

L212 H349 F350

M113 P115 K117

H349 F350

M118 N119 P120

Group 5

S347 R295 R352

S265 E270 S271 A272 E276

L290 K291 F292 D293 N237

G268 M269 E270 S271

K291 K294 R396

D271 V272 E273

K291

G278 L279 G280

Group 6

Y60 D61 Y62

Y279 N280 K284

A390 S391

E276 N280

D389 A390 S391 A392

E279 L280 N283 R327

A390 S391 A392

D286 Q290

superposition results or rp2 had a -score o . , a se uence identity o , and an M o 2. . Both Photox interaction models were evaluated using the QMEAN server [25]. The Photox:Arp1 model) had a QMEAN Z-score of -1.32 (Figure 3) [28]. QMEAN server was unable to evaluate the Photox:Arp2 model.

Modelling of Photox:Arp Models o both rp2 and rp ere generated, and subse uently superposed onto actin to allo modeling o their hoto interactions. he LILI superposition results or rp had a -score o . , a se uence identity o , and an M o . [24]. The DALILITE

Figure 3. Photox:Arp1 interactions. A cartoon overview of + + Photox:NAD :Arp1 coloured by local QMEAN scores, showing NAD and the modifiable arginine residue [25]. Colours represent local error in the model with blue indicating the least amount of error and red the most.

Volume 6 â&#x20AC;˘ Issue 1 â&#x20AC;˘ Fall 2012

The sequences of the ten yeast Arps were aligned with the sequence of actin (Figure 2A); Arp2, Arp1, Arp4, and Arp5 contain the modifiable arginine residue. Based on this, these Arps were aligned with actin to better investigate other conserved residues (Figure 2B). A ConSurf model, a colour coded structural representation, was generated based on the structure of actin (Figure 2C) (PDB accession code: 3BUZ) [18-20]. Dark purple areas on the ConSurf model indicate a high degree of conservation, whereas areas of light blue are indicative of the least degree of conservation. The modifiable arginine residue was found to be highly conserved. The actin residues predicted to interact with Photox were evaluated for conservation in the Arps; both identical residues and residues with similar properties were considered. Based on this evaluation, Arp2 and Arp1 were found to have the most conservation in these critical residues, and as a result were evaluated further.

K79 W81

Photox

Group 2

Identification of yeast Arps as potential targets of Photox

H238 E242 S246

Arp2

Studies by Undergraduate Researchers at Guelph (SURG) 62


In silico modeling of Arp1 and Arp1 as targets of Photox (Daly and Davidge)

+

Figure 4. Investigation of potential Photox:Arp2 interactions. (A) Representation of the Photox:NAD :Arp2:ATP binding interface depicting residues which are predicted to be involved in binding between Photox and Arp2. The circled sections of Arp2 indicate major structural differences between Arp2 and actin. (B) Proposed intermolecular interactions between Photox and Arp2. Hydrogen bonds are represented by red dashed lines and salt bridges are represented by green dashed lines.

Investigation of potential Photox:Arp interactions

Figure 5. Surface analysis of Photox:Arp2 interactions. (A) Electrostatic surfaces of the binding interface of Photox and + Arp2 including NAD . Circles represent areas which are proposed to interact with each other. (B) Hydrophobic surfaces of the binding + interface of Photox and Arp2 with NAD present.

Volume 6 â&#x20AC;˘ Issue 1 â&#x20AC;˘ Fall 2012

The Arp2 binding residues on Photox were categorized into four groups based on the same criteria used for Photox:actin and Iota:actin (Figure 4, Table 1). Several of these groups are very similar to those defined for the Photox:actin interactions; however, some groups are absent entirely. For consistency, the group names were kept the same as the assignments used for Photox:actin, hence the absence of groups one and three. Salt bridges are predicted to form between R396 of Photox and E173 of Arp2, K294 of Photox and D171 of Arp2, and E242 of Photox and K79 of Arp2. The hydrogen bonds predicted are very similar to those predicted for Photox:actin; however, the hydrogen bond between Y223 of Photox and the modifiable arginine residue of Arp2 is absent (Figure 4). The electrostatic interactions between Photox and Arp2 were modeled (Figure 5). Similar to the Photox:actin model, interactions were predicted between the electronegative region to the lower left of NAD+ on Photox and the electropositive region to the lower right of NAD + on Arp2. A second interacting region was predicted above and to the right of NAD+ on Photox and above and to the left of NAD + on Arp2. As with the Photox:actin interactions, the hydrophobic model generated for Photox:Arp2 suggests a lack of large interacting regions, but many small interactions were noted (Figure 5).

Studies by Undergraduate Researchers at Guelph (SURG) 63


In silico modeling of Arp1 and Arp1 as targets of Photox (Daly and Davidge)

+

Figure 6. Investigation of potential Photox:Arp1 interactions. (A) Representation of the Photox:NAD :Arp1:ATP binding interface depicting residues predicted to be involved in binding between Photox and Arp1. The circled region of Arp1 indicates the major structural difference between Arp1 and actin. (B) Proposed intermolecular interactions between Photox and Arp1. Hydrogen bonds are represented by red dashed lines and salt bridges are represented by green dashed lines.

Figure 7. Surface analysis of Photox:Arp1 interactions. (A) Electrostatic surfaces of the binding interface between Photox + and Arp1 including NAD . Circles represent areas which are predicted to interact with each other. (B) Hydrophobic surfaces of + the binding interface between Photox and Arp1 with NAD present.

Volume 6 • Issue 1 • Fall 2012

The overall structure of Arp2 was compared to that of actin (Figure 4). The area containing groups one and two is highly expanded on Arp2, containing two β-strands in place o the α-helices. Further, group two on actin is composed of an α-helix, but not in Arp2. The area between group two and four have similar secondary structural elements, but a different folding pattern. The interacting residues between Photox and Arp1 were categorized in the same manner as the Photox:Arp2 interactions (Figure 6). The interactions predicted were found to be very similar to those predicted between Photox and actin, though some differences were noted. Group one in Photox:Arp1 was predicted to involve more residues compared to Photox:actin, while group five was predicted to display the opposite trend. A salt bridge was predicted between Photox E242 and Arp1 R75, and a hydrogen bond was predicted between Y223 and D187 of Photox and Arp1 respectively (Figure 6). The electrostatic interactions modeled (Figure 7) suggested one interacting region. This is between the electronegative region to the upper right of NAD + on Photox and the electropositive region to the upper left of NAD + on Arp1.As with the Photox:actin and Photox:Arp2 models, the hydrophobic interactions are spread throughout the structure and are not clustered (Figure 7). When compared to actin, the theoretical structure of Arp1 only differed significantly in one area (Figure 6). This area is located near groups one and two, and was observed to be slightly less compact on Arp1; however, the secondary structure remains conserved.

Studies by Undergraduate Researchers at Guelph (SURG) 64


In silico modeling of Arp1 and Arp1 as targets of Photox (Daly and Davidge)

Discussion

Volume 6 • Issue 1 • Fall 2012

The structure of Photox was modeled based on SpvB, which shares very high sequence conservation with the C-domain of Photox. Compared to other ARTs, the SpvB active site has a 30-residue insertion, which is shared by Photox [6]. The QMEAN Z-score was 0.231, indicating a high quality model [28]. Further support for the quality of the model was a sequence identity of 64.322% between the Cdomains of SpvB and Photox. Also, the E value reported was 9.06e-71, indicating a very high probability that Photox and SpvB are homologues. Therefore, due to the high sequence identiy, the high QMEAN Z-score and the low E value, SpvB was determined to be the structural model for Photox. Photox residues 209 to 488 were modeled based on SpvB, which represents the actin-binding domain. Residues one to 208 were not modeled as they do not share a conserved sequence with SpvB. The N-domain of Photox did not exhibit homology with any ART; however, it showed 80% homology to an unknown protein from P. luminescens [16,27]. The N-domain of most actin targeting ARTs is responsible for binding and translocation and would therefore contain signal peptide sequences [5]. As the N-domain of Photox does not contain a signal peptide sequence and as it is predicted to be disordered by GeneSilico, it is unlikely that is behaves in the same manner as other actin targeting ARTs [23]. Given that this study only aimed to evaluate the binding interaction between Photox and Arps, having a more accurate model of the binding domain was determined to be of greater importance than a less accurate model of the entire protein. Therefore, the N-domain of Photox was not modeled. A model of SpvB:actin was not available; therefore, a model of Iota:actin was used to model Photox:actin. Iota, another actin targeting ART, does not contain the same insertion consisting of an extension of one α-helix, the additional α-helix that appears in Photox and SpvB. As such, the model of Photox:actin may not be as accurate as the Cdomains of Iota and Photox had a sequence identity of only 23.7% . Although few residues were found to be completely conserved, the properties of residues were conserved in many positions (Figure S1). The proposed binding residues between Photox and actin were categorized into six groups according to their positions on the two proteins [6,9]. Although residues from different groups may be in close proximity along the protein backbone, the orientation of the side chains was the determining factor for the grouping. As a result, residues that appear to be close together along the backbone may not be included in the same group. For example, Y223 and Q221 of Photox are sequentially very close; however, the Y223 side chain is angled towards the NAD+ binding site, while Q221 points in the opposite direction.

There are ten various Arps in yeast; however, only four were found to have the conserved, modifiable arginine residue: Arp2, Arp1, Arp4, and Arp5. These four Arps were evaluated for similarities with actin. This comparison was completed based on primary structure; only residues proposed to be involved in Photox:actin interactions were analyzed. Residues were considered to be conserved if they were identical to the actin residue, or shared similar properties and charges. Arp2 was found to share the greatest number of similar residues with actin, and Arp1 shared the second highest number of similar residues. Both Photox:Arp models were generated by superposing the Arp structural models onto the actin portion of the Iota:actin model used previously. The analysis of Photox and actin interactions was completed using an experimentally determined structure of actin, as opposed to a modeled structure. Thus, the Photox:actin interactions are likely more accurate than both sets of Photox:Arp interactions, which were completed using modeled structures for both Photox and the Arps. The DALILITE superposition of Arp2 onto actin had a Z-score of 33.7, suggesting a good quality model. The 50% sequence identity between Arp2 and actin, as well as the RMSD of 2.4 urther support the uality o this model. However, the QMEAN server was unable to assess the quality of the Photox:Arp2 model. This suggests significant errors present in this theoretical structure. These errors could be attributed to an incorrect model of Arp2 or could suggest that Photox does not in fact interact with Arp2 in this way. The model of Arp1 was found to have a Z-score of 53.9, a sequence identity of 47%, and an RMSD of 1.0 by DALILITE, all of which indicate a model of high quality. The QMEAN evaluation of the Photox:Arp1 model had a QMEAN Z-score of -1.32. This very negative value calls into question the quality of the Photox:Arp1 interaction, perhaps indicating that this is not a true interaction. The proposed interactions between Photox and actin, Photox and Arp2, and Photox and Arp1 are very similar. Though differences exist in the tertiary structures of actin and the Arps, the Photox-binding regions have similar structures. However, the ConSurf data for the Arps and actin, the Photox-binding surface did not indicate many conserved regions aside from the modifiable arginine residue. As a result, it is unlikely that all Arps would share a comparable Photox-binding surface. Based on the results presented, it is likely that Arp1 is a target of Photox, though the exact interactions determined using these models may not be entirely accurate. Arp2 is also a likely target of Photox; however, to determine the precise interactions would likely require more accurate structures. The models generated in this study suggest fewer interactions between Photox and Arp2, compared to Photox and Arp1; suggesting that the Photox:Arp2 interaction may be more transient. Arp4 and Arp5 also share the conserved arginine residue, and thus are possible targets of Photox, though their possible interactions were not evaluated in this study. Given the lack of conservation in the Photox-binding

Studies by Undergraduate Researchers at Guelph (SURG) 65


In silico modeling of Arp1 and Arp1 as targets of Photox (Daly and Davidge) surface of the Arps, it is possible these Arps have very different tertiary structures; thus Photox interactions are likely to be very different, if present at all. Given the proximity of Photox Y223 and actin R177, a hydrogen bond was predicted; however, this interaction was not predicted in either of the Photox:Arp models. This suggests that the modifiable arginine residue in Arps is not positioned in close proximity to Photox, and thus may not be modified by Photox. It would be beneficial to evaluate the enzymatic activity of Photox modification of Arps, similar to the study completed by Visschedyk et al. for Photox modification of actin [9]. An additional tool for determining whether interactions are present would be the use of green fluorescent protein (GFP)-fragment reassembly [29-31]. This technique involves creating fusion proteins of the two potential interacting proteins; one is fused to the N terminus of a GFP-fragment, and the other is fused to the C terminus of the GFP-fragment. If the proteins do in fact interact, the GFP fragments are brought in close enough proximity to one another to allow reassembly. Once reassembled, GFP provides a distinct fluorescent marker, confirming the interaction [29-31]. This would be a useful experiment to evaluate the presence of Photox:Arp interactions. Experimental evidence for the interactions proposed above could be attained through site-directed mutagenesis of Photox, actin, Arp2, and Arp1. By mutating the individual residues suspected to be involved in the interaction, and observing the effects on binding, the role that the mutated residues play in the interaction could be determined. All of the models used to evaluate potential Photox targets and interactions are theoretical, and as described above, some are of better quality than others. Particularly in the case of Photox and the Arps, having an accurate structure would be beneficial in more accurately assessing potential interactions.

Conclusion

Acknowledgements We thank Dr. John Dawson for all of his assitance during the writing and editing of this paper. We are grateful for the unpublished data from Dr. Rod Merrill. We are additionally thankful for the guidance and suggestions of Ms. Vicki Nowell.

References 1.

Cohen, M.L. (1998). Candidate bacterial conditions. Bulletin of the World Health Organization, 2, 61-63.

2.

Lee, V.T., & Schneewind, O. (2001). Protein secretion and the pathogenesis of bacterial infections. Genes and Development, 15, 1725-1752.

3.

Pallen, M.J., Lam, A.C., Loman, N.J., & McBride, A. (2001). An abundance of bacterial ADPribosyltransferases: implications for the origin of exotoxins and their human homologues. Trends in Microbiology, 9, 302-307.

4.

Corda, D., & Di Girolamo, M. (2003). Functional aspects of protein mono-ADP-ribosylation. The EMBO Journal, 22, 1953-1958.

5.

Tsuge, H., Nagahama, M., Oda, M., Iwamoto, S., Utsunomiya, H., Marquez, V.E., Katunuma, N., Nishizawa, M., & Sukurai, J. (2008). Structural basis of actin recognition and arginine ADP-ribosylation by Clostridium perfringens ι-toxin. Proceedings of the National Academy of Sciences USA, 105, 7399-7404.

6.

Margarit, S.M., Davidson, W., Frego, L., & Stebbins, C.E. (2006). A steric antagonism of actin polymerization by a Salmonella virulence protein. Structure, 14, 1219-1229.

7.

Paone, G., Stevens, L.A., Levine, R.L., Bourgeois, C., Steagall, W.K., Gochuico, B.R., & Moss. J. (2006). Arginine-specific mono ADP-ribosylation in vitro of

Volume 6 • Issue 1 • Fall 2012

Based on the models presented in this study Arp1 is a probable target of Photox. Arp2 is also a likely target of Photox, though the interactions are predicted to be weaker than those between Photox and Arp1. Improved structural models and activity studies would help to confirm these observations. Arp4 and Arp5 are also possible targets of Photox, as they possess the modifiable arginine residue, though their interactions have yet to be modeled. The Arps lacking the modifiable arginine residue are unlikely to be targets of Photox. Based on these results, the targets of bacterial toxins in the cell may be more extensive than previously determined. Elucidation of all cellular targets of these toxins will be required to fully understand and treat bacterial infections. Future in vitro studies should further investigate the binding of Photox to Arp1 and Arp2. Enzyme kinetics experiments would be beneficial in order to determine if

Photox truly ADP-ribosylates Arps as well as actin. These studies would assist in the determination of the mechanism of toxicity of Photox. Understanding the biological action of Photox on various proteins may help to further the discovery of a treatment for similar toxin-related diseases. These models may aid in understanding how ARTs destroy host cells and may lead to in vitro experiments which examine various ART targets. This understanding may lead to alternate methods of treatment for diseases caused by ARTs, such as diphtheria, cholera, pertussis, botulism, and typhoid [3-5].

Studies by Undergraduate Researchers at Guelph (SURG) 66


In silico modeling of Arp1 and Arp1 as targets of Photox (Daly and Davidge) antimicrobial peptides by ADP-ribsoylating toxins. Journal of Biological Chemistry, 281, 17054-17060. 8.

9.

Bowen, D., Rocheleau, T.A., Blackburn, M., Andreev, O., Golubeva, E., Bhartia, R., & ffrench-Constant, R.H. (1998). Insecticidal toxins from the bacterium Photorhabdus luminescens. Science, 280, 2129-2132.

19. Kiefer, F., Arnold, K., Künzli, M., Bordoli, L., & Schwede, T. (2009). The SWISS-MODEL repository and associated resources. Nucleic Acids Research, 37, D387-D392. 20. Peitsch, M.C. 1995. Protein modeling by e-mail. Biotechnology, 13, 658-660.

Visschedyk, D.D., Perieteanu, A.A., Turgeon, Z.J., Fieldhouse, R.J., Dawson, J.F., & Merrill, A.R. (2010). Photox, a novel actin-targetting mono-ADPribosyltransferase from Photorhabdus luminescens. Journal of Biological Chemistry, 285, 13525-13534.

21. Ashkenazy, H., Erez, E., Martz, E., Pupko, T., & Ben-Tal, N. (2010). ConSurf 2010: calculating evolutionary conversation in sequence and structure of proteins and nucleic acids. Nucleic Acids Research, 38, W529-W533.

10. Fieldhouse, R.J., & Merrill, A.R. (2008). Needle in the haystack: structure-based toxin discovery. Trends in Biochemical Sciences, 33, 546-556.

22. Landau, M., Mayrose, I., Rosenberg, Y., Glaser, F., Martz, E., Pupko, T., & Ben-Tal, N. (2005). ConSurf 2005: the projection of evolutionary conservation scores of residues on protein structures. Nucleic Acids Research, 33, W299-W30.

11. Schafer, D.A., & Schroer, T.A. (2000). Actin-related proteins. Annual Review of Cell and Development Biology, 15, 341-363. 12. Poch, O., & Winsor, B. (1997). Who’s ho among the Saccharomyces cerevisiae actin-related proteins – a classification and nomenclature proposal for a large family. Yeast, 13, 1053-1058. 13. Goodson, H.V., & Hawse, W.F. (2002). Molecular evolution of the actin family. Journal of Cell Science, 115, 2619-2622. 14. Harata, M., Oma, Y., Tabuchi, T., Zhang, Y., Stillman, D.J., & Mizuno, S. (2000). Multiple actin-related proteins of Saccharomyces cerevisiae are present in the nucleus. The Journal of Biochemistry, 128, 665-671 15. Larkin, M.A., Blackshields, G., Brown, N.P., Chenna, R., McGettigan, P.A., McWilliam, H., Valentin, F., Wallace, I.M., Wilm, A., Lopez, R., Thompson, J.D., Gibson, T.J., & Higgins, D.G. (2007). Clustal W and Clustal X version 2.0. Bioinformatics, 23, 2947-2948.

17. Petersen, T.N., Brunak, S., von Heijne, G., & Nielsen, H. (2011). SignalP 4.0: discriminating signal peptides from transmembrane regions. Nature Methods, 8, 785-756. 18. Arnold, K., Bordoli, L., Kopp, J., & Schwede, T. (2006). The SWISS-MODEL workspace: a web-based environment for protein structure homology modeling. Bioinformatics, 22, 195-201.

24. Holm, L., & Park, J. (2000). DaliLite workbench for protein structure comparison. Bioinformatics, 16, 566-567. 25. Benkert, P., Künzli, M., & Schwede, T. (2009). QMEAN server for protein model quality estimation. Nucleic Acids Research, 37, W510-W514. 26. Krissinel, E., & Henrick, K. (2007). Interference of macromolecular assemblies from crystalline state. Journal of Molecular Biology, 372, 774-797. 27. Waterfield, N.R., Bowen, D.J., Fetherston, J.D., Perry, R.D., & ffrench-Constant, R.H. (2001). The tc genes of Photorhabus: a growing family. Trends in Microbiology, 9(4), 185-191. 28. Benkert, P., Biasini, M., & Schwede, T. (2011). Towards the estimation of the absolute quality of individual protein structure models. Bioinformatics, 27, 343-350. 29. Wilson, C.G.M., Magliery, T.J., & Regan, L. (2004). Detecting protein-protein interactions with GFPfragment assembly. Nature Methods, 1, 255-262. 30. Sung, M.K., & Huh, W.K. (2010). In vivo quantification of protein-protein interactions in Saccharomyces cerevisiae using bimolecular fluorescence complementation assay. Journal of Microbiological Methods, 83(2), 194-201.

Volume 6 • Issue 1 • Fall 2012

16. Altschul, S.F., Gish, W., Miller, W., Meyers, E.W., & Lipman, D.J. (1990). Basic local alignment search tool. Journal of Molecular Biology, 215, 403-410.

23. Glaser, F., Pupko, T., Paz, I., Bell, R.E., Bechor, D., Martz, E., & Ben-Tal, N. (2003). ConSurf: identification of functional regions in proteins by surface-mapping of phylogenetic information. Bioinformatics, 19, 163-164.

Studies by Undergraduate Researchers at Guelph (SURG) 67


In silico modeling of Arp1 and Arp1 as targets of Photox (Daly and Davidge)

31. Sung, M.K., & Huh, W.K. (2007). Bimo77lecular fluorescence complementation analysis system for in vivo detection of protein-protein interaction in Saccharomyces cerevisiae. Yeast, 24(9), 767-775.

Endnotes a.

In this article, reference to sequence identity refers to amino acid sequence identity.

b.

Dr.Rod Merrill, Dept. of Molecular & Cellular Biology (Guelph, ON Canada), personal communication.

Volume 6 â&#x20AC;˘ Issue 1 â&#x20AC;˘ Fall 2012

Studies by Undergraduate Researchers at Guelph (SURG) 68


Supplementary Information

In silico modeling of Arp1 and Arp1 as targets of Photox (Daly and Davidge)

Methods Supplement Modelling of Photox:actin In order to model the interactions of Photox and actin, DALILITE was used to superpose the Photox model onto the structure of a homologous protein bound with actin [24]. A structural sequence alignment was also obtained from DALILITE [24]. As no structure of SpvB:actin was available, a model of Iota:actin was selected (PDB accession code: 3BUZ). A theoretical model of the Photox:actin interaction was created by editing the PDB text files of Iota:actin and the output file of Photox from DALILITE [24]. The quality of this model was assessed using the QMEAN server [25]. Determination of Photox:actin structure Various models of the Photox:actin theoretical structure were created and used to assess possible interactions, including electrostatic surface models and hydrophobic surface models. The proximity of amino acids was evaluated to identify possible hydrogen bonds and salt bridges; residues within 4.0 Å of each other were considered. The PDBePISA server was used to assess possible protein interfaces, and the results were evaluated based on the above criteria [26]. The interactions described for Iota:actin were also considered using the MSA previously generated [5].

Results Supplement Modelling Photox structure Photox and other ARTs that bind actin were compared using a MSA, with specific consideration of the active site (Figure S1). The ConSurf server was also used to generate a visual representation of conserved residues; purple residues represent the most conserved regions, and cyan residues represent the least (Figure S1) [21,22,23]. The NAD+ binding site and active site region showed the greatest conservation. Photox residues identified for NAD + binding and catalytic activity included N-XXX-R, RGLK-XX-K-XX-L, and STS-X[35]-E. Other conserved residues were identified; however, they were not located on the protein surface.

Volume 6 • Issue 1 • Fall 2012

Figure S1. Photox homology model. (A) Multiple sequence alignment of Photox and other ARTs in the C2-like class, which inhibit actin polymerization. Black arrows indicate residues which participate in NAD+ binding and red arrows indicate residues which participate in the activity of the toxins. This alignment was produced with the EMBL-EBI ClustalW 2.0 server using the region of Photox that was aligned to SpvB (PDB accession code: 2GWL) [15]. (B) Visualization of the conservation between Photox and the toxins in the C2-like family produced with the ConSurf version 3.0 server [25-27]. Regions of purple represent high conservation, whereas cyan regions represent low conservation. (C) Sideby-side representations of Photox and SpvB.

Studies by Undergraduate Researchers at Guelph (SURG) 69


In silico modeling of Arp1 and Arp1 as targets of Photox (Daly and Davidge) Modelling Photox:actin A theoretical model of the Photox:actin binding interface was generated by superposing the Photox structure onto an Iota:actin structure (PDB accession code: 3BUZ) using DALILITE [24]. The resulting DALILITE structure had a Z score of 12.3, sequence identity o and an o . [24]. This model was then assessed using the QMEAN server; the model was coloured based on the QMEAN results, with colour representing the potential error in the structure for each residue (Figure S2) [25]. The spectrum ranges from blue to red, where blue is the least amount of error, and red is the most. The areas with the least amount of potential error include t e β-s eet oti in P oto and t e β-sandwich in subdomain three of actin. The QMEAN Z-score was -0.85 [28].

Figure S2. Overview of Photox:actin interactions. A cartoon overview of the Photox:NAD+:actin structure coloured with local QMEAN scores showing NAD + and the modifiable arginine residue (R177) on actin [24]. Colours represent local error in the model with blue indicating the least amount of error and red the most. Determination of Photox:actin interactions The DALILITE superposition of Photox onto Iota generated a structural sequence alignment (Figure S3) [24]. Few residues were found to be completely conserved; however, residues with similar properties were conserved in some positions. The actin binding residues on Photox were grouped based on their overall positions on the Photox protein and the region of actin with which they interact (Figure S3) [6,9]. The overall tertiary structure of Photox was compared to that of the Iota actin-binding region. Photox was found to have an insertion consisting o an e tension to one αeli as ell as an additional α-helix. The potential interactions between Photox and actin were predicted based on proximity, electrostatics, and hydrophobicity (Table 1). Salt bridges were predicted between K291 of Photox and E270 of actin, as well as E242 of Photox and K68 of actin. Multiple hydrogen bonds were also predicted: Photox Y223 and actin R177, Photox S391 and actin E276, Photox Y217 and actin T75, and Photox S246 and actin D80 (Figure S3) [5,9]. The electrostatic surfaces of Photox and actin were modeled (Figure S4) and two potential interacting regions were identified. The negative region in the bottom left of Photox (relative to NAD+) may interact with the electropositive region on the bottom right of actin (relative to NAD+). Also, the electronegative pocket above and to the right of + NAD on Photox may interact with the electropositive protrusion to the upper left of NAD + on actin. The hydrophobic surfaces of Photox and actin were also modeled (Figure S4), with orange representing hydrophobic regions, and green representing hydrophilic regions. The hydrophobic interactions were not found to be clustered in regions on the proteins; however, many individual interactions are likely to play a role in the overall interaction between Photox and actin.

Volume 6 • Issue 1 • Fall 2012

Figure S3. Determination of Photox:actin interactions. (A) Structural sequence alignment of Photox and Iota toxins produced by DALILITE [24]. Purple arrows indicate residues on Iota which are required for actin binding whereas the orange arrows indicate residues on Photox which are predicted to be required for actin binding. (B) Representation of the Photox:NAD+:actin:ATP binding interface depicting predicted residues involved in binding between Photox and actin. The circled section of Photox indicates the major st ructural difference between Photox and Iota. (C) Proposed intermolecular interactions between Photox and actin. Hydrogen bonds are represented by red dashed lines and salt bridges are represented by green dashed lines.

Studies by Undergraduate Researchers at Guelph (SURG) 70


Supplementary Information

In silico modeling of Arp1 and Arp1 as targets of Photox (Daly and Davidge)

+

Figure S4. Surface analysis of Photox:actin interactions. (A) Electrostatic surfaces of the binding interface between Photox and actin including NAD . Circles represent + areas which are proposed to interact with each other. (B) Hydrophobic surfaces of the binding interface between Photox and actin with NAD present.

Discussion Supplement The sequence of Photox was aligned with ART proteins. ConSurf modeling indicated high conservation within the active site region [21-23]. Other conserved residues were not located on the protein surface, suggesting a possible role in tertiary structure conservation. Since no structures of SpvB:actin were available, a structure of Iota:actin was selected to model the Photox:actin interactions. Although Photox shares more conserved residues with SpvB than with Iota, the actin-binding region is well conserved, which allowed modeling by superposition. The QMEAN Z-score for this Photox:actin model was -0.85, indicating that there are likely errors within the model [28]. The DALILITE Z-score for superposition was 12.3, suggesting a model of moderate quality. DALILITE found 23.7% sequence identity between Photox and Iota, which as expected, is less than the sequence conservation between Photox and SpvB. This indicatesthat the model was based on conserved areas between these proteins, and thus is reasonably accurate. The RMSD value was 2. , suggesting that the Photox model and Iota have closely related structures. The structural differences between the actin binding domains of Iota and Photox were evaluated. An insertion of an α- eli as ell as t e e tension o an e isting α-helix in the group one region were observed on Photox. As a result of this addition, six groups of actin binding residues were observe d for Photox, compared to the five groups previously determined for Iota [5]. Further differences between the Iota:actin binding groups and Photox:actin binding groups were observed using the structural sequence alignment from DALILITE [24]. Although few residues were found to be completely conserved in the primary sequences, it was noted that the overall tertiary structures of Photox and Iota were well conserved. For example, group six of Iota is structurally comparable to group six of Photox; however, these residues are located in very different positions of the primary structure. Given the modeled structural similarities, as well as the sequence conservation, it is likely that Photox and Iota are homologous. However, given the differences in folding patterns, it is likely that these proteins have evolved significantly since divergence. Overall, both proteins share enough structural similarity to perform similar functions, as both Photox and Iota modify R177 of actin [5,9].

Volume 6 • Issue 1 • Fall 2012

Studies by Undergraduate Researchers at Guelph (SURG) 71


Review Article

Iron in the brain (with apologies to Jean-Paul Sartre): Heavy metal mismanagement Erin L. Stephenson Collaborative Program in Neuroscience, Department of Molecular and Cellular Biology, College of Biological Sciences, University of Guelph, Guelph, ON Canada. Faculty supervisor: George Harauz. For correspondence, please email: elstephenson@sympatico.ca.

Abstract Iron’s activity in the body can be two-faced. On the one hand it is integral to many enzymatic reactions; on the other hand it is toxic, with a great capacity for cellular damage. This review examines iron in the brain through the lens of multiple sclerosis (MS), reviewing the functions of intracellular and extracellular iron and their impact on the disease, as well as highlighting the focus of new research and controversial therapies. The primary cause of MS has remained enigmatic, ever since its first clinical documentation. Several studies have suggested a link between MS and iron. Abnormal iron accumulation has been found as deposits in MS lesions around cerebral veins, in the macrophages surrounding MS lesions, and also in deep brain structures. There are features of MS, such as the inflammatory environment and altered vasculature, which are important in highlighting mechanisms of how iron can accumulate, and also how iron dysregulation can create a positive feedback cycle that further promotes neurodegeneration and increased iron accumulation. This potential link between iron and MS has gained widespread attention, in part due to the controversial cerebrospinal venous insufficiency (CCSVI) hypothesis and “Liberation Therapy” first introduced by Dr. Zamboni in 2008. Determining the role of iron in MS will help provide a better insight to the different factions of scientists who disagree on whether MS is primarily an autoimmune disorder, or whether a neurodegenerative mechanism is the instigator. Keywords: multiple sclerosis (MS) and therapies; iron (regulation in the brain, link to disease); CCSVI hypothesis; Liberation Therapy (Zamboni, 2008); review

Introduction they have inoperable cancer. Hope is an emotion in its own right, and the physician may be wrong” [3]. While there have been improvements in the treatments of MS, there is still no cure or efficient way to halt the progression of the disease, and the causes and etiology of MS have remained elusive. 250 200 150 100 50 0

Prevalence of MS (per 100,000 individuals)

Figure 1. Prevalence of MS in the world [1].

Volume 6 • Issue 1 • Fall 2012

Canada has one of the highest incidences of multiple sclerosis (MS) in the world (Figure 1) [1]. The Multiple Sclerosis Society of Canada states that three people are diagnosed with this chronic disease every day. MS tends to be clustered geographically, with higher incidences in temperate regions amongst Caucasians, and in women [2]. The symptoms of MS are complex and unpredictable, but all individuals suffer from progressive muscle weakness and fatigue [2]. Ultimately balance, memory, vision, and hearing may be impacted, depending on what region(s) of the brain are affected by the disease [2]. There has historically been a general reluctance to diagnose patients with MS. Even as recently as the mid1900s, neurologists viewed the diagnosis as a drawn-out death sentence. Foster Kennedy exemplified this sentiment when he said in 1950 that “one should no more tell our patients they have multiple sclerosis than we should tell them

Studies by Undergraduate Researchers at Guelph (SURG) 72


Iron in the brain: Heavy metal mismanagement (Stephenson)

CCSVI and the Internet Due to its uncertain nature, there are many debates surrounding MS. While scientists still disagree about the cause of MS and whether it is autoimmune or neurodegenerative in origin, this is not the issue that has gained widespread public interest. The sudden increase in interest of MS began in 2008, with a hypothesis called Chronic Cerebrospinal Venous Insufficiency (CCSVI) [4]. Differing views on support for the hypothesis resulted in public outcries from MS patients and their families against the government and the scientific community.

aspect that might have carried the greatest weight was that sufferers of what had been hitherto regarded as an untreatable disease were now promised a “cure”. The remedy that Zamboni proposed was simple: his method was to open congested veins with an inflatable balloon, a procedure known as a balloon-angioplasty. Zamboni’s surgical angioplasty intervention then gained a very provocatively chosen name: ‘Liberation Therapy’. Due to the lack of alternative treatments that promised such an easy cure, MS sufferers looked for anything to free them from this debilitating disease. The divide

Chronic cerebrospinal venous insufficiency (CCSVI) The public awareness and interest in medical research at this level has created an interesting dichotomy between the rigors of the scientific process and the demands of MS sufferers. While the medical community and government are following well-established protocols for determining the viability of this procedure and the validity of Zamboni’s results, MS sufferers have been outraged at being denied access to this treatment by the government. In response, thousands of patients have paid large sums of money and travelled to countries that conduct the procedure for the promise of this miracle cure [9]. The media has taken advantage of the general public interest and has responded with numerous reports. Since CTV carried out an exclusive interview with Dr. Zamboni in 2009, there have been 22 follow-up CTV news stories on Liberation Therapy in the past two years alone [10]. In February 2012, the Canadian Broadcasting Corporation (CBC) aired MS Wars: Hope, Science, and the Internet on “The Nature of Things” television program [9]. The impact of social media has been astounding. There are thousands of blogs, websites, and YouTube videos promoting a “miracle cure” for MS (Figure 2). There is even a directory encompassing all of the social network sites for CCSVI [11]. These have allowed MS patients to focus on the reportedly positive results from this treatment. As the positive affirmations spread with the Internet as their primary medium, they provided what every MS sufferer and their families so desperately needed: hope for a cure.

Figure 2. A selection of CCSVI logos found readily on the Internet.

Volume 6 • Issue 1 • Fall 2012

The CCSVI hypothesis, first published online in late 2008 by Dr. Zamboni, was based on an observation of iron deposits surrounding veins in MS. In this hypothesis, iron took a lead role; it purported MS to be a neurodegenerative disease occurring as a result of venous pathology. Zamboni based his theory on his opinion that iron deposits in the MS brain had some similarities to the iron pathology around veins in legs in a chronic venous disorder [4]. Therefore, he hypothesized that MS was due to poor drainage of venous blood from the brain. The congestion of blood would cause increased pressure across the vascular wall and promote the release of hemoglobin to the area surrounding the vein. As will be reviewed in more detail later, hemoglobin and its breakdown products are able to damage endothelial cells and contribute to oxidative stress. Macrophages that break down hemoglobin also release iron [5]. Zamboni’s publications, detailing the findings of vascular anomalies in MS patients and purporting CCSVI as a cause of MS, however alluring, were treated with skepticism in the scientific community due to the lack of a double-blind randomized control normally required for clinical trials. In surveys, such as the one conducted by Zamboni, the Hill Criteria can be used to infer causality [6]. Criteria on the Hill checklist include temporality and consistency. As Awad et al. deftly review, Zamboni’s hypothesis as CCSVI being a cause of MS fails a number of the criteria [6]. In the case of temporality, it requires that if CCSVI is indeed a cause of MS, then it should always be present before the development of the disease [6]. It was observed that CCSVI is rare at clinical onset of MS, but its occurrence increases as the disease progresses [5,7]. This fact already suggests that CCSVI is, at best, a secondary phenomenon. There has also been a problem with consistency. Studies conducted afterwards failed to reproduce the results found by Zamboni [6]. A potential explanation of this lack of consistency could be due to inadequate methods. As noted by Chambers et al., neither Zamboni nor others who attempted to replicate his results state how their control subjects were selected [8]. However, the hypothesis of CCSVI conjured up a whirlwind of interest in the public MS community. One

Studies by Undergraduate Researchers at Guelph (SURG) 73


Iron in the brain: Heavy metal mismanagement (Stephenson) Recently, a case-control study published in 2012 demonstrated that CCSVI does not have a causal role in MS, as it was not present in those with early stages of MS disability [8]. Another 2012 study indicated that there was no relationship between cerebral vein abnormalities and disease progression in patients [12]. As CCSVI is likely soon to be retired as another false theory in the history of MS, balloonangioplasty will become just another example of the miracle of faith-healing, also known as the placebo effect. The attention that the Zamboni article gained from the general public brings up significant questions about the traditional practices of how scientific research is communicated and refuted. It also raises the question of whether patient pressure and social media should affect the timing of clinical studies. It has also shown the divide that exists between scientists and the public. While releasing the results of scientific research widely on the Internet will potentially help advance scientific discovery at a faster pace, there is still a disconnect between the discoveries of science and public knowledge. When patients and the public are using social media to advocate and mobilize, scientists who want their findings to reach the public and make an impact must do the same. A long term goal of science should be to provide more resources to educate and engage with the public, politicians, and the media. Improved scientific literacy would help ensure that resources are focused on effective scientific research, and not on ineffective therapies. As the theory of CCSVI is increasingly being refuted by objective clinical data suggesting that, at most, it is a secondary phenomenon, the greatest question on the lips of both the scientific and public community remains: what is causing MS? To investigate this question further, aspects of potential links between MS pathology and iron will be examined. This review will address how iron may play a role in this disease, either initially or during its progression, and aims to emphasize the need for more research in this field.

Myelin and Multiple Sclerosis Myelin in the central nervous system

Multiple sclerosis MS is a now accepted as being a multifactorial disease; it is grouped into a variety of different forms on the basis of the pattern of its progression. While it is known that MS destroys the protective myelin sheath of neurons, it is not known what initiates this destruction. Relapsing-remitting is the most common form of MS in which periods of clinical episodes are followed by a partial recovery in symptoms [2]. In approximately 15% to 20% of cases, the pathology gradually worsens over time with no improvement, and is referred to as primary progressive [2]. The disease can take this course from the onset, but relapsing-remitting cases can also eventually convert to this progressive course [2]. This latter form is called secondary progressive. In the progressive form of the disease, primary or secondary lesions do not heal; the degeneration is continuous and irreversible [15]. There are currently no effective treatments available to halt the progressive course of MS.

Iron in the Cells and CNS Despite the scientific evidence accumulating against the CCSVI hypothesis, there are other mechanisms that could lead to iron injury playing a role in MS. Iron’s activity in the body can be two-faced; on one hand it is integral as a cofactor in many enzymatic reactions but on the other, when it is unbalanced or unregulated, it has the potential to be toxic and cause extensive cellular damage. The physiology of intracellular iron Iron is integral to nervous system function. It plays a role in neurotransmitter synthesis, DNA synthesis, and energy production via the electron transport chain [16]. It also plays a role in biogenesis which occurs when nutrients or oxygen are in short supply, and involves an increase in the number of mitochondria to meet energy demands [15]. Iron is also essential in myelin formation, as well as the maturation of the cells that synthesize myelin, known as oligodendrocytes, which are the main iron-containing cells in the adult CNS [16]. Iron is especially prevalent in the loops of myelin due to the high metabolic demands of myelin synthesis [16]. Toxic iron Free or unbound iron is available to interact with reactive oxygen species, thereby producing free radicals. This process can be especially damaging if it occurs in the

Volume 6 • Issue 1 • Fall 2012

Throughout the brain and spinal cord, together referred to as the central nervous system (CNS), nerves function like electrical wires, transmitting information in the form of electrochemical impulses. An insulating membrane wraps itself around these nerves, and is responsible for protecting the nerves and promoting the transmission of electrical information. Known as myelin, this lipid-rich sheath has been termed ‘white matter’, referring to its white glossy appearance due to its lipid composition [13]. In MS there is a profound and preferential destruction of the myelin sheath. In the process of demyelination, the myelin sheath is stripped off neurons in a pattern that follows the venous system [13]. The loss of myelin insulation and plaque formation not only interferes

with the transmission of the nerve impulse, but also means loss of support for the neuron, potentially leading to cell death [14]. Eventually the neurodegeneration spreads, encompassing the entire CNS [15].

Studies by Undergraduate Researchers at Guelph (SURG) 74


Iron in the brain: Heavy metal mismanagement (Stephenson) brain because neuronal membranes are high in polyunsaturated fatty acids, making them extremely vulnerable to free radicals [17]. Lipid peroxidation products from this damage can be a source of pathogenesis by their ability to modify proteins, as well as become a target for an immune attack [18]. One mechanism of free radical production is through reaction of redox-active iron and the reactive oxygen species, hydrogen peroxide, which is a natural by-product of aerobic metabolism and is relatively non-toxic. However, hydrogen peroxide and free ferrous iron (Fe 2+) can react via the Fenton reaction, producing the extremely reactive hydroxyl radical [17]. Whereas the hydrogen peroxide molecule has limited reactivity with biological molecules, the hydroxyl radical’s reactivity is only limited by its rate of diffusion [19]. The hydroxyl radical potently reacts with membrane lipids via a free radical mechanism, and thus propagates membrane damage [20]. The outcome of the Fenton reaction is the conversion of ferrous iron to ferric iron (Fe3+). The superoxide anion is another reactive oxygen species produced as a by-product from normal aerobic metabolism [21]. Superoxide can react with ferric iron to reform ferrous iron (Figure 3) [22]. This allows the newly formed ferrous iron to produce hydroxyl radicals, and thereby facilitate inflammation and damage [21]. Another way that iron can cycle back to its redox-active state is through the high levels of ascorbate in the brain [23]. In the absence of transition metals, ascorbate has important antioxidant properties. However, in the presence of iron or other transition-metals, such as copper, ascorbate creates free radicals by reducing ferric iron and recycling iron back to its redox-active state [17,23]. This ability of iron to recycle back to its redox-active state is one way that damage by iron can become self-sustainable.

Iron also has a close link to the immune system. It acts as a chemoattractant to attract immune cells [17]. It can stimulate changes in vasculature, promoting the expression of adhesion molecules [17]. Iron can promote the release of certain proteinases that degrade the cells surrounding the vascular system, thereby promoting adhesion and migration of autoimmune cells into the surrounding tissue, which become activated and can then release pro-inflammatory molecules [17]. Activated immune cells, such as microglia, can undergo oxidative bursts which involve the release of reactive oxygen species; activated microglia have been

An extracellular source of iron Iron’s properties make it a useful cofactor in important enzymes. It exists as part of a heme component in enzymes such as catalase and peroxidases, which are important in protection against oxidative damage. Heme consists of rings of carbon and nitrogen with iron at the core. By far the most abundant source of heme is hemoglobin, and there is no doubt that hemoglobin plays a vital role in sustaining mammalian life. Despite its usefulness, hemoglobin is neurotoxic and has an immense capacity to cause damage to cells [17]. The endothelium is protected from this damage by erythrocytes which transport hemoglobin in an environment containing a lot of antioxidant enzymes [24]. However, vasculature problems promote erythrocyte rupture, and hemoglobin can be released either intravascularly or extravascularly. When released from erythrocytes, hemoglobin is very reactive. Once hemoglobin is oxidized into its methemoglobin form, it can release its heme moieties readily [25]. Due to its hydrophobic nature, heme can cross lipid membranes. This property is an integral aspect of its ability to cause cellular damage [25]. Once inside the plasma membrane, heme can react directly with peroxides, promoting oxidative stress. This reaction breaks down heme, and releases free iron, which is then able to participate in other reactions [26]. In this manner, heme acts as a Trojan horse to transport hydrophobic iron into the interior of cells. Iron itself is then able to cause damage. Damage due to hemoglobin release can be measured by the presence of lipid peroxide breakdown products [27]. The lipid peroxide products can activate the immune system as well as react with hemoglobin, promoting the conversion of the protein to its methemoglobin form, and thereby facilitating heme release [25]. Iron regulatory proteins Due to iron’s integral but damaging nature, iron regulatory proteins are key to managing the delicate balance between the body’s need for iron, as well as limiting iron’s

Volume 6 • Issue 1 • Fall 2012

Figure 3. Cyclic of iron between ferrous and ferric states can promote the continual production of hydroxyl radical [22].

identified in MS as mediators of demyelination [17]. An oxidative burst not only increases the concentration of free radicals, but can also liberate iron from ferritin [15]. The free iron, converted into its reactive ferrous form, is then able to react and produce hydroxyl radicals. Thus, this mechanism results in the amplification of oxidative damage by promoting iron’s ability to create toxic hydroxyl radicals [7]. Iron promotes oxidative damage when it facilitates mitochondrial stress [17]. The demyelination that occurs in MS creates a greater energy demand for the cell’s mitochondria. Moreover, iron accumulation promotes inflammation, which further disturbs the already stressed mitochondria. Iron can also inhibit base repair of DNA, contributing in yet another way to oxidative stress [17].

Studies by Undergraduate Researchers at Guelph (SURG) 75


Iron in the brain: Heavy metal mismanagement (Stephenson) damage. One protective mechanism that the body has developed is to sequester iron in a non-reactive form in the cell. The protein ferritin is responsible for sequestering iron, and is conserved amongst diverse life forms, from bacteria to humans [23]. Due to iron’s importance in cell function, neurons usually contain an endogenous supply of iron bound to ferritin [16]. There is also an exogenous delivery supply of iron – transferrin is the major iron delivery protein for neurons [28]. As mentioned earlier, as a source of iron, heme can promote oxidative damage by producing radicals, but also by releasing free iron to react further. A protective mechanism is in place to prevent this damaging degradation of heme. This enzyme complex consists of the enzyme heme oxygenase (HO), as well a source of electrons for the reaction, derived from NADPH cytochrome P450 reductase (referred to as NADPH henceforth) [29]. The mechanism consumes oxygen and seven electrons supplied by NADPH to cleave the heme ring, releasing carbon monoxide, biliverdin, and free ferrous iron [29]. HO protects against damage from heme by safely breaking down heme without production of free radicals [23]. Biliverdin, converted to bilirubin, also has antioxidant effects [23]. For the protective mechanism of HO to be effective, the redox-active iron must be then bound to ferritin [23,28-30]. Dysfunction of iron regulatory proteins

Linking Iron to Multiple Sclerosis Neurodegeneration and iron Iron dysregulation is not unique to MS. Iron’s mismanagement occurs in numerous disease states and can occur through more than one mechanism. However, there are features of MS, such as the inflammatory environment and altered vasculature, which are important in highlighting mechanisms of how iron can accumulate, and also how iron dysregulation can create a positive feedback cycle that further promotes neurodegeneration. Studies of iron in multiple sclerosis There are a number of studies linking iron to MS, ranging from large-scale analyses studying the brain as a whole, to small-scale analyses involving the measurement of molecular interactions. Whole brain scans, such as magnetic resonance imaging (MRI), analyze iron content in areas of the brain. Another technique, called susceptibility-weighted neuroimaging (SWI), measures iron concentration by assessing magnetic field susceptibility differences in different tissues [31]. The disadvantage of these studies is that they cannot differentiate whether iron deposits in MS are a benign result of the disease progression, or a key promoting factor of neurodegeneration. What these studies emphasize is that the importance lies not in total brain iron levels, but in the areas of the brain with abnormally accumulated iron and iron in a redox-active form [17]. It is important to note that iron accumulation occurs as a natural process of aging. Under pathological conditions, iron levels may accumulate to the point where they overcome the capacity of the normal mechanisms in place to properly manage iron [32]. Abnormal iron deposition has been observed in a number of neurodegenerative diseases, and may contribute to neuronal damage and progress degeneration [32]. This situation could occur in MS, since there is evidence of abnormal iron accumulation [16]. In white matter, there is accumulation in oligodendrocytes, macrophages, and microglia around sites of inflammation associated with veins [16,33,34]. Higher than normal levels of iron are also found in deep gray matter structures in the brain. The significance of this finding is that some of these structures are involved in message relay. For example, the thalamus, one structure that is affected, is a diffuse relay center important for the control of sensory and motor functions [35]. If iron accumulation is responsible for damage to these structures, it could have farreaching effects on the many brain regions that are connected by these relay stations [35,36]. In general, the extent of iron accumulation in gray matter structures and lesions has been

Volume 6 • Issue 1 • Fall 2012

Iron regulatory protein levels are controlled by their responsiveness to iron levels, and dysfunction in these proteins can be a source of iron mismanagement. A dysfunction in transferrin, leading to an alteration in iron transport, can lead to a multitude of neurological impairments, and is one of the symptoms in MS patients; it also occurs in iron deficiency, inflammation, cancer, Alzheimer’s disease, Parkinson’s disease, and Restless Legs Syndrome [28]. HO acts as an antioxidant during short-term instances of heme overload. However, chronic activation of this system can have deleterious effects. One of the reasons that chronic activation of HO can have damaging effects is the requirement of NADPH for the breakdown of heme. NADPH is also required for the production of glutathione, an important antioxidant for the cell [30]. Therefore, chronic depletion of NADPH due to HO hyperactivity can slow the recovery of glutathione. Depletion of glutathione and NADPH increases the availability of hydrogen peroxide to react and produce radicals and cell death [30]. Another cause of damage by overexpression of HO lies in the free iron produced from heme breakdown. Normally, this iron is quickly sequestered with ferritin. In times of iron overload, however, ferritin binding sites can become saturated, leading to increased levels of free ferrous iron to react. In summary, inadequate storage of iron allows the redox active iron to produce reactive hydroxyl radicals that can lead to oxidation of lipid membranes, denatured proteins, and damaged DNA [3]. Iron can cause oxidative damage and

promote inflammation, thereby promoting further release of iron. This process can then become self-sustaining.

Studies by Undergraduate Researchers at Guelph (SURG) 76


Iron in the brain: Heavy metal mismanagement (Stephenson) shown to be a good predictor of disability progression in MS, as well as the extent of lesion accumulation and level of cell death [16,33,37]. Histological studies have analyzed the lesions in the brains of MS patients, as well as accumulation of molecular players in lesion sites. They have shown that lesions are surrounded by iron-loaded macrophages. A likely cause of iron accumulation in these cells may be macrophage uptake of cell debris from degeneration, such as the remnants of oligodendrocytes or blood extravasation. Measuring levels of iron-regulatory proteins in the blood or cerebrospinal fluid yields a general idea of brain iron levels. Evidence of abnormal iron homeostasis can come from analysis of iron and iron-related proteins in cerebrospinal fluid, blood, and post-mortem analysis [36]. Levels of ferritin have been found to be elevated in patients with some forms of MS, suggesting local iron overload [4]. In a situation like this, there is potential for medical research; however, what is currently missing is the ability to follow the disease progression before the clinical onset. The problem of causality Studies documenting the occurrence of iron accumulation and MS show that there is a connection between the occurrence of MS and iron mismanagement. What they do not show is whether iron is a cause of the degeneration in MS, or whether iron accumulation in plaques is a protective mechanism, a degenerative mechanism, or merely a by-product from the degeneration. To determine causality, there must be a way to identify the beginning of the degeneration. This point is not easily identifiable, as it occurs when there are no outward signs. Clinical trials are further complicated by the multifactorial nature of MS, as well as uncertainty of where the disease is in its progression when a diagnosis is made [3].

Work to be Done

Vasculature anomalies in multiple sclerosis Even as Zamboni’s CCSVI theory is turning out to be false, the evidence of a vascular pathology in MS on which the theory was founded remains valid. Studies have found dilated cerebral veins in the brain of individuals with MS as well as fibrin cuffs found around the veins [4].

Conclusion The question still remains regarding iron’s role in the pathogenesis of MS. Is iron an instigator of the disease, a by-product of an inherent neurodegenerative process, or merely an epiphenomenon and not a real player in this illness? What is certain, however, is that iron dysregulation is a potent source of damage, and should be focused on as a future source of research and therapies. Determining iron’s role in MS will hopefully move the scientific world one step closer to elucidating a mechanism for this disease. Another significant note is the divide that still exists between scientific research and the public. While patients are using social media to mobilize, scientists in this aspect are falling short. A statement of this is in the impact that CCSVI has made in the public sector. A quick Internet search reveals a plethora of CCSVI organizations (some examples are shown in Figure 2). To search for the articles that disprove CCSVI as a cause of MS, one must search harder. Scientists need to become more effective in public communication, not only on this issue, but in general. Science will never be independent of society. The discoveries of science will impact the lives of the public, and funding from the public and government impact the type and amount of research that can be conducted. Therefore, improving scientific literacy will create a better appreciation for the current scientific discoveries that are occurring, and

Volume 6 • Issue 1 • Fall 2012

There still remain dichotomous views of whether MS is a neurodegenerative disorder or primarily an autoimmune disorder. However, evidence is accumulating suggesting that MS might have a neurodegenerative rather than autoimmune origin. Demyelination and death of myelinproducing oligodendrocytes can occur before inflammation or the activation of immune cells [38,39]. Inflammation due to the presence of immune cells in the brain could occur as a secondary response to a neurodegenerative process [39].

It is important to investigate further the venous abnormalities in MS patients in order to understand the physiological mechanisms. Vasculature problems can cause nutrient deficiency and hypoxia, which is a lower oxygen concentration than normal [22]. The brain is especially susceptible to hypoxic damage due to its high metabolic rate and therefore steep oxygen requirements [22]. The hypoxia not only causes the death of neurons and glia directly affected, but also increases the levels of reactive oxygen species and raises the level of glutamate, a prevalent excitatory neurotransmitter, to damaging levels [22]. This process can become cyclic, as the vasculature damage due to hemoglobin release can promote further damage to the vasculature and surrounding tissue. However, what remains to be determined is whether the greatest potentiator of vasculature damage is from hemoglobin, heme, iron, or another unknown player. The oxidative stress due to venous haemorrhage and iron deposits can have real implications on the process in demyelination. Aside from MS, iron deposition has been observed in other demyelinating diseases, such as in Hurst acute haemorrhagic leukoencephalitis [14]. Future research into the vascular pathology in MS will be of benefit. This avenue of inquiry will hopefully lend new insight into the role of iron deposits in MS, the potential source of hemoglobin as a source of iron overload, and the role of hemoglobin and its breakdown products in venous pathology.

Studies by Undergraduate Researchers at Guelph (SURG) 77


Iron in the brain: Heavy metal mismanagement (Stephenson) hopefully bridge the divide that has emerged between science and society.

Acknowledgements This work was performed under the auspices of a Summer Research Assistantship awarded by the Multiple Sclerosis Society of Canada, and the Ontario-Manitoba endMS Regional Research and Training Network. The laboratory is supported by the Natural Sciences and Engineering Research Council of Canada, and by the Canadian Institutes of Health Research. The author is grateful for helpful discussions with and comments by Ms. Agata Zienowicz, Dr. Vladimir Bamm, and Professor George Harauz.

9.

Canadian Broadcasting Corporation (CBC) – The Nature of Things. (2012). MS wars: hope, science, and the internet. Retrieved on June 18, 2012 from: http://www.cbc.ca/natureofthings/episode/ms-warshope-science-and-the-internet.html

10. CTV – CTV News. (2009). The liberation treatment: a whole new approach to MS. Retrieved on June 18, 2012 from:http://www.ctv.ca/CTVNews/WFive/20091120/W 5_liberation_091121/ 11. CCSVI Locator. Retrieved on June 18, 2012 from: http://ccsvi-ms.ning.com

References

12. Garaci, F.G., Marziali, S., Meschini, A., et al. (2012). Brain hemodynamic changes associated with chronic cerebrospinal venous insufficiency are not specific to multiple sclerosis and do not increase its severity. Radiology, [Epub, ahead of print].

1.

MS Atlas. (2012). Retrieved on June 18, 2012 from: http://www.atlasofms.org/index.aspx

13. Lumsden, C.E. (1971). The immunogenesis of the multiple sclerosis plaque. Brain Research, 28, 365-390.

2.

Multiple Sclerosis Society of Canada. (2)12. Retrieved on June 18, 2012 from: http://mssociety.ca/en/information/default.htm

3.

Minagar, A. (2008). Neurobiology of multiple sclerosis. International Review of Neurobiology, 79, 1-732.

4.

Zamboni, P. (2006). The big idea: iron-dependent inflammation in venous disease and proposed parallel in multiple sclerosis. Journal of the Royal Society of Medicine, 99, 589-593.

5.

6.

Zivadinov, R, Weinstock-Guttman, B., & Pirko, I. (2011). Iron deposition and inflammation in multiple sclerosis. Which one comes first? BMC Neuroscience, 12, 12-60.

7.

Yamout, B., Herlopian, A., Issa, Z., Habib, R.H., Fawaz, A., Salame, J., et al. (2010). Extracranial venous stenosis is an unlikely cause of multiple sclerosis. Multiple Sclerosis Journal, 16, 1341-1348.

8.

Chambers, B., Chambers, J., Cameron, H., & Macdonell, R. (2012). Chronic cerebrospinal venous insufficiency is not more prevalent in patients with mild multiple sclerosis: a sonographer-blinded, case-control, ultrasound study. Multiple Sclerosis Journal, [Epub ahead of print].

15. Lassman, H., & van Horssen, J. (2011). The molecular basis of neurodegeneration in multiple sclerosis. FEBS Letters, 585, 3715-3723. 16. Gonzalo, H., Brieva, L., Tatzber, F., et al. (2012). Lipidome analysis in multiple sclerosis reveals protein lipoxidative damage as a potential pathogenic mechanism. Journal of Neurochemistry, [Epub, ahead of print]. 17. Williams, R., Buchheit, C.L., Berman, N.E.J., & Levine, S.M. (2012). Pathogenic implications of iron accumulation in multiple sclerosis. Journal of Neurochemistry, 120, 7-25. 18. Gadoth, N., & Hilmar, H. (2011). Oxidative Stress and Free Radical Damage in Neurology. Human Press. 19. Packer, L., Hiramatsu, M., & Yoshikawa, T. (1996). Free Radicals in Brain Physiology and Disorders. San Diego, CA: Academic Press. 20. Ferretti, G., & Bacchetti, T. (2011). Peroxidation of lipoproteins in multiple sclerosis. Journal of Neurological Science, 311, 92-97. 21. Yoshida, T., & Migita, C.T. (2000). Mechanism of heme degradation by heme oxygenase. Journal of Inorganic Biochemistry, 82, 33-41.

Volume 6 • Issue 1 • Fall 2012

Awad, A., Marder, E., Mio, R., & Stuve, O. (2011). Multiple sclerosis and chronic cerebrospinal vascular insufficiency: a critical review. Therapeutic Advances in Neurological Advances, 4, 231-235.

14. Waschbisch, A., Manzel, A., Linker, R., & Lee, D. (2011). Vascular pathology in multiple sclerosis: mind boosting or myth busting? Experimental & Translational Stroke Medicine, 3.

Studies by Undergraduate Researchers at Guelph (SURG) 78


Iron in the brain: Heavy metal mismanagement (Stephenson) 22. Kalinowski, D.S., & Richardson, D.R. (2005). The evolution of iron chelators for the treatment of iron overload disease and cancer. Pharmacological Reviews, 57, 547-583.

31. Haacke, E.M., Cheng, N.Y., House, M.J., Liu, Q., Neelavalli, J., Ogg, R.J., et al. (2005). Imaging iron stores in the brain using magnetic resonance imaging. Magnetic Resonance Imaging, 23, 1-25.

23. Halliwell, B., & Gutteridge, J.M.C. (1985). Free Radicals in Biology and Medicine. New York, NY: Oxford University Press.

32. Riederer, P., & Youdim, M.B.H. (1993). Iron in Central Nervous System Disorders. Wien, NY: Springer-Verlag.

24. Tsemakhovich, V.A., Bamm, V.V., Shaklai, M., & Shaklai, N. (2005). Vascular damage by unstable hemoglobins: the role of heme-depleted globin. Archives of Biochemistry and Biophysics, 436, 307-315. 25. Balla, J. (2003). Haem, heme oxygenase and ferritin in vascular endothelial cell injury. Nephrology Dialysis Transplantation, 18 [suppl 5], v8-v12. 26. Grinshtein, N., Bamm, V.V., Tsemakhovich, V.A., & Shaklai, N. Mechanism of low-density lipoprotein oxidation by hemoglobin-derived iron. Biochemistry, 42, 6977-6985. 27. Belcher, J.D., Beckman, J.D., Balla, G., Balla, J, & Vercellotti, G. (2010). Heme degradation and vascular injury. Antioxidants & Redox Signalling, 12, 233-248. 28. Leitner, D.F., & Connor, J.F. (2011). Functional roles of transferring in the brain. Biochimica et Biophysica Acta, 1820, 393-402. 29. Robinson, S.R., Dang, T.N., Dringen, R., & Bishop, G.M. (2009). Hemin toxicity: a preventable source of brain damage following hemorrhagic stroke. Redox Report, 14, 228-235. 30. Schipper, H.M., Song, W., Zukor, J., Hascalovici, J.R., & Zeligman, D. (2009). Heme oxygenase-1 and neurodegeneration: expanding frontiers of engagement. Journal of Neurochemistry, 110, 469-485.

33. Habib, C.A., Liu, M., Bawany, N., Garbern, J., Krumbein, I., I., Mentzel, H.J. et al. (2012). Assessing abnormal iron content in the deep gray matter of patients with multiple sclerosis versus healthy controls. American Journal of Neuroradiology, 33, 252-258. 34. Bagnato, F., Hametner, S., Yao, B., Van Gelderen, P., Merkle, H., Cantor, F.K., et al. (2011). Tracking iron in multiple sclerosis: a combined imaging and histopathological study at 7 tesla. Brain, 134, 36023615. 35. Lebel, R.M., Eissa, Al., Seres, P., Blevins, G., & Wilman, A.H. (2011). Quantitative high-field imaging of sub-cortical gray matter in multiple sclerosis. Multiple Sclerosis Journal, 18, 433-441. 36. Khalil, M., Teunissen, C., & Langkammer, C. (2011). Iron and neurodegeneration in multiple sclerosis. Multiple Sclerosis International, [Epub, ahead of print]. 37. Neema, M., Arora, A., Healy, B.C., Guss, Z.D., Brass, S.D., Duan, Y., et al. (2009). Deep gray matter involvement on brain MRI scans is associated with clinical progression in multiple sclerorsis. Jounral of Neuroimaging, 19, 3-8. 38. Dutta, R., & Trapp, B.D. (2011). Mechanisms of neuronal dysfunction and degeneration in multiple sclerosis. Progress in Neurobiology, 93, 1-12. 39. Barnett, M.H., Henderson, A.P.D., & Prineas, J.W. (2006). The macrophage in MS: just a scavenger after all? Pathology and pathogenesis of the acute MS lesion. Multiple Sclerosis Journal, 12, 121-132.

Volume 6 â&#x20AC;˘ Issue 1 â&#x20AC;˘ Fall 2012

Studies by Undergraduate Researchers at Guelph (SURG) 79


SURG Fall 2012 (Volume 6, Issue 1)  

A peer-reviewed undergraduate publication of the University of Guelph.

Read more
Read more
Similar to
Popular now
Just for you