Page 1

Trends in Cowpea Research

Proceedings of the Cowpea Research Seminar Harare, Zimbabwe 25-26 September 1991

SADC/EEC/IlTA Cowpea Research Project

Contents Opening Address .._.................................................................._....... _......_..........


R.I. Fenner, Director, Deparhnent of Research (Specialist Services) Harare, Zimbllbwe Cowpea Improvement Program in Botswana .............. _.............. _...... _........


E.S. Mosilrwe Past, Present, and Future Status of Cowpea Research in Zimbabwe............ R. Dubt and B.!. Nyob


Development of the Cowpea Improvement Program in ambia ...._.............


Review of Cowpea Research in Lesotho ............................ ...............................


1M. Muli/a路MiIti S.S. Moima

Evaluation of Contrasting Cowpea Genotypes at Different Dates of ........... Planting under Rainfed Condi tions


~B. Kwapala

Intercropping Cowpea with Maize In Tanzania ....................._......_............... F.A. M yaka and J. K. Mligo


Intercropping Cowpea in Maize and Sorghum under....... _...... _......._..... _.... Zimbabwean Condi tions


T. Nleya On路Fann Research Experience With Improved Cultivars .............................. G. A. S. Milli and J. M. Mulila Mitti


Effect of Plant Density and Time of Planting on the ....................................... Grain Yield of Cowpea J.C Musanya


Cowpeas: A Basic Component of a Multiple Crop System............................. In the District of Marracuene, Maputo, Mozambique


I. AriaTFandino Cotton and Cowpea Relay Intercropping: Preliminary Results on its.......... Economics and Effect on Some Agronomic Characters of Both Crops F.A. Myab and I.CB. KAbissa


Recent Research on the Cropping System with ............................................... Cowpeas In Zambia I.e. Musanya


Selection and Evaluation of Cowpea Crosses between TV x 3236 and......... A/ectra vogelii resistant landraces from Botswana M.e. Phillips



An Evaluation of Maize.Cowpea Intercropping ............................................. Systems in Swaziland lohn Pali-Shikhulu and ZcdWll MIlmba


Cowpea Pathology in Botswana .... .....................................................................


Baikilbile MIlti/o Evaluation of Cowpea [Vigna unguiculata (L.) Walp.) Cultivars Against Root-Knot Nematode Meloidogynt jaVilnica......................................................... V. W. Sakil Adverse Effect of a YeUow Mosaic Disorder on Growth Nodulation and Seed Yield of Two Cowpea Varieties at Umbeluzi in Southern Mozambique

97 100

A.Z. MIltangue, A.L Doto, and D.M. Naik Importance and Extent of Yield Losses Caused by Major Insect Pests on.... Cowpea in Zambia


P.H. Sohati and S. Sithanantham Field Losses Caused by Cowpea Pests at Umbeluzi, Souihem Mozambique


Richard A. Amable Exploratory Field Testing of Plant Extracts Against Some Pests on................ Cowpea in Zambia


P.H. Sohati Il1Id S. Saithananthan Prevalence of Cowpea Diseases and Extent of Yield Losses Caused by Scab in Zambia


H.C. Haciwa and J. Kannaiyan Recent Research on Cowpea Aphid-bome Mosaic Virus in Zambia ............


J. Kannaiyan, H.C. Haciwa, S. Sithananthan, .H. Sohati, and JM. Mu/i/a Ust of Participants ............................ ....................................................................



Opening Address R.

J. Fenner

During August there was a report on BBC radio about an appropriate and environmentally sound way (vis a vis smallholders) to control blister beetles (Milabris sp.) in cowpea crops. The red and black striped variety of these beetles is also known as the CMR (cape mounted rifles) beetle! The suggested method of control was to place blue, plastic dishes full of water among cowpea plants in the fi eld . Apparently, the beetles are attracted by the blue color and drawn to their deaths by drowning! The recommendation was proposed by a private entrepreneur in Botswana. The relevance of this story will become clearer as we progress. The broad objective of the Southern Africa Development Community (SADC) Cowpea Research Project is to generate technologies which could enhance the production of cowpeas in the SAOC countries commensurate with national demand. In order to adrieve this goal, the project envisages strengthening national research capability to conduct research on cowpeas through the development of improved varieties and appro priate production practices in collaboration with national scientists. To facilitate the realization of this collaboration, project activities include exchange of germplasm, information, and training. We who are gathered here are aware that research on cow peas has been on going among the various national programs for five to 20 years. The question that I w ould like you to ask yourselves is: what precise benefit have cowpea farmers received from this research effort, which has undoubtedly consumed much time, effort and resources? The next and related question to bear in mind is: in what ways can a regional project facilitate our individual abilities to contribute to the production efforts of o ur farmers? The catchword in agricultural develo pment project circles these days is "lust~jn~bi1ity". The aim, weare told, should be to generate production technologies which not only improve productivity but also preserve the resource base. The most important resource base our smallholders have is the soil. Is there a role for cow peas and other legumes known to these farmers in helping to ameliorate and preserve the soil? Some of these farmers possess draught animals to till theirsoil. Again, can legume crops such as cowpeas be instrumental in improving and sustaining the performance of the draught animals? On these smallholdings, human labor occupies a paramount place in the production system. We know that in these communities, cowpeas are consumed in various forms : leaves, fresh seeds and pods, and dry g rain. Could improved varieties contribute to the balanced nutrition of our target households and thereby contribute to the health of individuals and sus tain their production efforts? I began with a story about beetles, blue di shes, and farmers. Are we, who fo rm a communityof research scientists reaching the farmer in a similar, direct way? I believe that this is the most important challenge we face, and with it, I am pleased to declare this Cowpea Research Seminar open . v

Cowpea Improvement Program in Botswana 拢.5. Mosarwe

Advanced variety adaptation trial Multilocational experiments were conducted during 1990/91 at Sebele, Mahalapye, Goodhope, Pandamatenga, Matsaudi, Boro, Etsha 6, and Etsha 8. The purpose of the study was to evaluate further the performance of promising cowpea varieties, both local and from lITA, Nigeria, in different agroecological zones th ro ughout the country.

Results revealed Significant differences among the varieties in five of the eight locations (Mahalapye, Goodhope, Matsaudi, Etsha 6, and ElSha 8). All varieties performed best at Etsha 8 and worst at Pandamatenga. The overa ll mea n y ields of the varieties ranged from 0 to 1039 kg/ha. Varietal seed yield means fluctuated from one environment to another. This indicates that environmental di fferences could be the main factor that determines most of the varietal yield differences. Variety 8342 recorded the highest overall mean y ield of 1039 kg/ha, followed b y B462 (894 kg / hal, IT835-742-2 (890 kg/hal, TVx 3236 (825 kg / h al, IT82E-16 (771 kg/ hal, and 1T85F-2020 (706 kg/hal while 8345 recorded zero y ield. The results show that medium-maturing varieties such as 8342 are env ironment路specific although not all varieties specifically adapted to these environments were medium-maturing. Conversely, the late-maturin g varieties tended to be better adap ted to mo re adverse environments characteri zed by prolonged seasonal dry spells than earl y- and medium-matu ring varieties . Generally, yields were low compared to those ob tained during past cropping seasons. One of the major constrain ts which contributed to the low prod uction was high infestation by insect pests, most especia lly pod s ucking bugs, which caused seriou s damage to the crop. Effects of the majo r cowpea diseases (cowpea a phid-bome mosaic virus, bac terial blight, and ashy stem blight) were mod era tely severe. Although, varieties were planted very late due to late rains, there was adequate moistu re for crop survival in all experimental sites. It is concluded tha t despite these constraints, these varieties have potential since they have shown stability across most, if not all of the environmen tal conditions. Some of them are still being tested in on-farm research trials and w ill be released to farmers in the near future.

Cowpea Genetic Resources Some work in cowpea genetic resources collection and conservation was started in the mid seventies. Initially, collections were carried out in a rather ra ndom manner until 1982 when the Cowpea Improvement Program was transfered to the Collaborative 1

Research Support Programme (CRSP)-a USAID-funded project. As a result of the transfer, more cowpea germ plasm was collected and documented. To date, a total of 852 cowpea collections have been assembled and are now being handled by the Plant Genetic Resources Management (PCRM) Unit. The number of collections has remained unchanged for some time now because there is no storage space for new collections. A large number of these collections have been systematically characterized and evaluated for 35 morphological characters and this information is contained in the catalogs together with passport data . Some of the collections have been evaluated further for resistance to three major diseases: cowpea aphid-borne mosaic virus, Microphomina phaseo/i, and Xnnthomonas Cllmpestris. Research is also on-going to identity resistance to Aphid craccivora and Alectra vogelii.


Past, Present, and Future Status of Cowpea Research in Zimbabwe R. Dube and B.I. Nyoka

Cowpea production in Zimbabwe has for long been confined to the smallholder farmer. Research on cowpea improvement and utilization was minimal as the crop was regarded as a rural crop with no industrial use. The lack of high-yielding pestand disease resistant varieties of suitable maturity coupled with unavailability of a guaranteed market to diSpose excess grains have slowly relegated this nutritious pulse, which was once ranked first, down to nearobscuri ty. Elite lines from lITA tested so far have shown a nearly threefold increase in yield over the local varieties and mature in half the time it takes the local varieties to mature. Cultivation of cowpea [Vigna unguiculata (L.) Walp.], locally known as nyemba in Shona and indumba in Ndebele, in Zimbabwe dates back to before 1890 (Reid 1977). The crop is mainly grown for its leaves, and green and dry pods which can be consumed in various preparations. The green leaves can be cooked as vegetable (munyemba), fresh green leaves can be boiled, salted, dried, and served as relish after recooking (mufushwa), fresh green pods can be boiled and eaten as a snack (mulcove) while dried grain can be cooked, mixed with cereals, (mutakura) or roasted, pounded, and cooked to form a paste (rupiza) normally eaten with thick porridge (sadza).

Cowpea production in Zimbabwe In Zimbabwe, cowpea is grown in diverse ecological areas which vary in rainfall, soil type, and temperature. Rainfall distribution in Zimbabwe is very uneven. It ranges from less than 162 mm per year in natural region IV and V to more than 1800 mm per year in natural region I. The altitude ranges from 400 m in region IV and V with an average temperature of JOoC to over 2000 mm in natural region I with an average temperature of 18째C. The bulk of the cowpea is produced by communal farmers who live mainly in the drier parts of the country. Varieties grown by the farmers are late maturing (J50days) and do not fit in with the short growing seasons in natural regions IV & V. Indigenous cultivars are indeterminate and give good leaf yields but low grain yields. Low grain yields are also attributed to the planting patterns of the communal farmers. Cowpea is a minor component of the cropping systems in the communal areas. It is rarely grown as a sole crop and is usually mixed with maize, pearl millet, finger millet, sorghum, and occasionally with groundnuts in no particular planting pattem. This reduces the cowpea plant population 10 suboptimal levels, resulting in low yields. Even when grown in pure stands, cowpea is oflen widely spaced, resulting in low yields.


Productions constraints The major constraints to cowpea production identified in Zimbabwe include low yields, due to lack of improved cultivars, pests and diseases, poor agronomic practices, and lack of specific markets. Agronomic practices Cowpea is grown as a low input crop in Zimbabwe. Seeds are home-grown and of low quality. There is no certified seed for cowpea. The seeds are stored under traditiona: conditions where they often get infested with weevils resulting in poor emergence. Plant density is often low whether under intercroppingor in pure stands. No fertilizer or manure is applied ; no pest control is effected. Productivity of cowpea is further red uced by the plucking of tender leaves for use as relish throughout the crop's growth. In addition, the foliar productivity of the indigenous varieties has not been evaluated and varieties grown are unimproved and low yielding. Diseases and pests The major cowpea diseases identified in Zimbabwe include leaf spot (Aschochyta phase%rum), scab (ÂŁ/sinoe phaseo/I), Fusarium wilt (Fusarium oxsporum), bacteria wilt (Xallthomonas uignico/a), and some unidentified viral diseases. Among the pests identified , pnd borers (Clavigralla species), the green stink bug (Nezara viriduia), the foliage beeUe (Ootheca mutabilis), and the cowpea weevil (Cal/osobruchus rhodesianus), have been found to be quite devastating. Infestation begins in the fi eld and continues during storage, making control a problem. Cowpea re.earch in Zimbabwe Research on cowpea in Zimbabwe began in 1921-22 with the introduction of the upright elite lines from USA. One of the elite lines, Turianai, was subsequently crossed with astrainofthe upright New Era from Potchefstroom(S.Africa)in 1938-39 (Arnold 1947). Prolific strains, S.E.S.D3andS.E.5.G4, were isolated but were lostover the years when research on them ceased Gohnson 1970). Cowpea research resumed in 1980 with the establishment of a cowpea breeding program which aims to develop high-yielding. stable cowpea genotypes with multiple resistance to pests and diseases. The program will also develop varieties that are able to retain leaves late in the season as the consumers utilize both the leaf and the grain in their diet. This will be done by: (i) Testing and evaluating introduced lines for yield and adaptation (ii) Collecting and characterizing local germplasm (iii) Identifying the major constraints limiting cowpea production in Zimbabwe. So far, a number of improved lines from Botswana, Brazil, the International Institute of Tropical Agriculture (liT A), the Semi-Arid Fond Grains Research and Development (SAFGRAD) Project, and local germplasm materials have been tested. Most of the lines from Botswana, SAFGRAD, and Brazil performed badly. In the 1981-82 season, a numberoflIT A lines were tested at two sites. Most of the introduced lines outyielded the local variety, Dr Saunders' Upright (Table 1). The highest grain yield was obtained with TVx 2724-01F at both sites. 4

Table 1.

Yield of liTA line. and one local variety (Cowpea Intemdional trial No. 2) Zimbabwe, Summer, 1981~2 Site

Variety Vita-4 Vita-7 lie-Brown TVx-2724-01F TVx-2724-017 TVx-3808-02/ TVX-3871-02F TVx-4262-014D TVx-4272-05D Dr. Saunders' Upright Mean



1267 1314 886 1897 1466 306 978 1383 978 395 1087

2235 1436 1887 2270 2196 1434 2005 2235 1236 1708 1874

Source: Agronomy Institute Annual Report, 1982.

In the 1989-90 season, 40 local germpla sm materials were planted at the Harare Research Station for characterization and yield evaluation. Three prominent accessions were isolated: 475 /89, 243/89, and 1293/89. The three varieties flowered and matured well ahead of Vita 4, a standard check. Two local materials, 475/89 and 243/ 89, were included in the multilocational trials in the 1990/91 season. A crossing program was initiated in the 1990/91 season following the single seed descent (SSD) breeding method. Crosses were made between the local materials and the introduced elite lines. A number ofIITA lines in the advanced variety trial (A VI) under two groups, medium maturity and early maturity, were tested at five sites in the 1990/91 season. Included in the medium maturity group were local accessions 475/ 89 and 243/89 . . Most of the lines did not perform well at Chiredzi due to lack of rainfall (256: 3 rom). The local varieties, 475/89 and 243/89, yielded least at Chiredzi but attained high yields at Gwebi and Kadoma . Vita 4 and IT82D-875 were stable across four sites (Table 2). Table 3 shows the yield performance of lITA early maturing lines in five sites. Lowest grain yields were obtained at Chinedzi and also at Mlezu Agricultural Institute. Mlezu received about 546 rom of rainfall which was not evenly distributed. A dry spell was experienced at flowering and pod initiation which led to a drastic reduction in yield. IT 82[)-889 yielded highest (an average about 1395.7 kg/ha across the sites).


Table 2- Yield (kglha) of the medium maturity cowpea line. and two local cultivaB, Summer, 1990-91 Site Matopos





Vita-4 IT82D-699 1T83S-990 1T82D-8n 1T82D-875 IT82D-709 1T83D-219 1T82D-952 IT82D-513-3 1T83S-87 1T82D-927 1T83S-841 1T81D-975 IT83S-852 IAR-48 1VX-3236 475/89 243/89

1058 0943 0935 1114 1175 1090 0510 0775 0931 0842 0481 1274 1028 0670

0368 0291 0297 0340 0414 0278 0259 0220 0421 0225 0354 0184 0259 0148

1716 1350 0931 1668 1129 1282 1018 1034 1482 1501 1155 1234 1264

1398 1371 1281 1615 1896 1433 1643 1366 1582

1157 1157


1306 1284 1217

1607 1259 0611 0983 1368 0893 1541 0397 0748 0611 0652 0793 0303 0934 0274 0517 1940 1974

Mean CVO/O ISD(5%)



0857 0920

0347 0102 0092


0940 33 0429

0273 44 0168

1256 20.2 0351

1115 33 0510

0864 1275 1429 0744 1406 1142 1161 1792 1749 1650 145 0331

Source: Crop Breeding Institute Annual Report, 1991.

Varieties in trials designated as "bruchid resistant", "aphid resistant", "vegetable type", "early and medium maturity" and "dual purpose" were evaluated in preliminary variety trials (PVT) at two sites, Gwebi and Kadoma, during 1991. Most of the lines tested were quite promising and will be included in the intermediate variety trials (IVT) next season. Cowpea international nurseries of bruchid resistant, medium-and-extra early maturity, aphid resistant, vegetable and dual purpose cowpea lines are also being tested under local conditions.


â&#x20AC;˘â&#x20AC;˘ ble 3. Yield (kglhd of early-m.turity UTA lines, Summer, 1990-91 Site Variety 1'8SF-867-S 1T830-328-S [1'820-789 [1'820-889 IT82E-32 IT84E-124 [T830-442 1T835-850 IT820-812 IT83S-844 IT83S-%2 ITB4S-2221-2 IT83S-818 IT85F-1517 K59

Matopos 0807 1287 1078 0938 1045 0944 0850


1T82D-889 1T82D-M6

1093 1099 0599 0838 101 9 0791 1012 1028 0954 1051

Mean CV% LSD (5%)

1126 5.5 0339


Chiredzi 0030 0351 0539 0273 0387 0266 0347 0351 0469 0287 0209 0188 0075 0185 0111 0122

Mlezu 0542



0789 0629 0620 0629 0568 0641 0383 0671 0679 0586 0419 0569 0578 0693 0650 0516



30.9 0120

27.9 0226



1168 1309 1334 1184 1216 1252 1230 1239 1302 1420 1274 1117 0927 0983 1157 1495 1378 1134 1438 16.9 0289

Gwebi 1643 1457 1143 1759 1079 1188 1553 1504

0665 1267 0579 1203 1024 0829 1133

0464 1855 1600 1426 24.5 0414

Soum.: Crop Breeding Institute Annual Report, 1990/91 Future reseuch prospects In the past, research was centered on testing in troduced lines for yield and adaptation, collecting and characterizing local germ plasm, and identifying the major constraints to cowpea production in Zimbabwe. The breeding program will continue to eval uate introduced lines . The program will collaborate with the University of Zimbabwe to screen promising cowpea materials fo r resistance to the cowpea weevil and with the Plant Protection Research Instit ute (PPRI) to screen for resistance to vira l diseases as well as identify the viruses affecting cowpea. Sui table genotypes identified forspe<:ific characters such a s yield, seed size, disease and pest resistance, and drought tolerance will be used in the hybridization program. The program wilt a lso test some elite lines under on-farm conditions and will evaluate them for acceptability and adaptability. Thevariabilityof weather and seasons between regions in Zimbabwe necessitates the development of stable genotypes w hich can give reasonable yields in different environments. Cowpea materials will be evaluated under dry and low fertility conditions.


Conclusion Although cowpea is a cheap source of protein affordable by all, it is underproduced in Zimbabwe. Cowpea breeding research is still in its infancy but specific objectives forfuture research have been identified . These include evaluatio n ofintroduced lines for yield and adaptation, screening for viral disease and weevil damage, selection for dual purpose types, and initiation of a hybridization program.

References Agronomy Institute. 1982 . Annual Repo rt. Department of Research and Specialist Services, Ha .. re. Zimbabwe. Arnold. H.C. 1947.Annual Report 1945-46. Salisbury Agricultural Experim entStation . Rhodesia Agricultu ral Journal 44: 306-326. JCihnson. D.T. 1970. Thecowpea in the African areas of Rhodesia . Rhodesia Agricultural)oumal 67: 61-64. Reid. M.G. 1977. Theearly ag riculture of !Illata beland and Mashonaland. Rhodesia Agricultural Joumal74 (4): 97-104. Crop Breeding Institute Annual Report. 1991. Annual Report . Department of Research and Specialist Services. Ha rare, Zimbabwe.

Discussion Q. H.ciw.: What are the major d iseases ilnd pests?

A. Dub.: Bacterial blight, scab. FlIsnrillln wilt, and viral diseases (wh ich have not yet been identified). Major pests are pod borers, green stink bug, bruchids, and aphids. Q . Data: How serious is marketing in limiting production and what is being done to add ress this problem? A. Dube: Marketing is limiting cowpea production since the farmer is not sure whether the surplus will have a market. TIle Department of Economics is looking into marketing of all the different crops but we do not yet have a solution. Q. P.Ii路Shikhulu: How does the cowpea resea rch program intend to reconciie the need for fresh green leaf harvests and the need for grain at the end of the season?

A. Dub.: This will done with the initiation of a hybridization program, combining the locil1 germplasm and introduced elite lines with upright plant type. The local germplasm has good leaf yieid but poor grain yield, whereas the introduced elite lines have opposite characteristics. Q. Doto: Have these crosses been made already?

A. Dub.: Yes . Crosses were made during the summer in the 1990-91 season. Fl seeds were planted in the greenhouse in the winter and F2 seeds will be planted in the summer of 1991-92. Q. Phillips: Shouid new varieties be tested under intercropping with cereais, since

this seems to be how most farmers grow cowpea in the region? A. Dub.: Yes. New varieties shouid be tested first under intercropping before release farmer since the crop is still a minor component of the cropping system. With

10 the


the limited land available to [communal land) farmers , they will be unwilling to allocate land specifically to cowpea, ie. monocrop cowpea. Comment. Sak.: Intercrnpping of cowpea with maize is dictated by the size of landholding, which currently is about 1 ha per family [in Malawi). Therefore, technologies have to address this issue but at the same time take into consideration the need s of the estate sector. Comment. Haciwa: Although farmers grow cowpea in intercrops, it might be necessary to carry out some research work using monocrops. This is because disease and pest pressure is higher in monocrops SO screening for disease and pest resistance wou Id be easier under these conditions. Identified resistant genotypes may then be tried in intercrops. Q. Uriyo: We have seen in the presentations made from Malawi and Zimbabwe that farmers would use fertilizers on their maize crop but not use any purchased inputs like insecticides, which would increase yields per hectare, on cowpea. Research has shown that with the right cowpea varieties and a minimum insecticide spray regime, yields of up to 2 t/ha can be obtained. CouId Dr. Amable, who has recently participated in a cowpea market study in the SADCC region comment on whether the pricing and marketing structure of cowpea discourages farmers from using purchased inputs in cowpea production?

Comment. Amable: This study is still in progress and no definitive conclusions can be drawn yet. However, of the five countries visited, only Botswana had an official pricing and marketing system. In all the other countries, cowpea marketing was through priva te traders who bought surplus production from farmers at prices determined by market factors. The bulk o f production appeared to be retained on farm, with the period of bruchid路freestorage greatly influencing the amount retained . Cowpea is definitely a rural smallholder crop. 'Consumptionand production in many countries appeared to be declining, sometimes even in rural areas, largely because of a perceived association with low status (eg. bulk feeding in schools, and miners' and caneculters' hostels). With this background , there wasa dear disincentive to produce large surpluses of grain and thus use purchased inputs.


Development of the Cowpea Improvement Program in Zambia I.M. Mulilil-Mitti

The pape< discusses cowpea research work prior to the establishmenl of the Groin Legumes Research Team in 19821nd oullines the objectives of the cowpea Improv&menl program in Zambia. II further disc.- the cowpea variety trials, _lilts obtained. and the newly initiated cowpea hybridization program.

Cowpea [Vigna unguiCII/afil (L.) Walp.) Is one of the importantfood legumes grown in almost all parts of Zambia but mainly in the southern, western, and eastern provinces. 1he crop is particularly important in smallholder rainfed agriculture. Late maturing, prostrate or spreading local landraces are traditionally intercropped with maize, cassava, sorghum, or bulrush millet. 1lte crop is grown for grain, fresh or dry leaves for relish, and occassionally for fresh pods. Cowpea research work in Zambia was started in 1966 when introduced and local cultivars were evaluated at the Misamfu Regional Research Station in the northern province with the objective of selecting cultivars resistant to ascochyta blight (Phoma exigua vardiversispora Bub. (Boerema) and scab (Spmac/cmra sp.= Elsinoephaseoli]enkins), the two diseases found to be of economic importance for the crop. 1he program was, however, abandoned as none of the lines introduced was resistant. 1lte program was started again in 1970 in thewestern and southern provinces with the objective of producing drought-tolerant cultivars for these drought-prone areas. In the 1972-73 season, 17 Jines were screened for yield, reaction to diseases and insect pests, drought tolerance, and performance under sole and intercrop systems (prior 1979). Of these, only5 survived the severe drought with yields ranging between 92 and 306 kg/ha (Mbewe et at 1985). The variety New Era Grey was released after a series of variety trials for local and introduced materials in different sites in the southemand western provinces. From the mid-seventies to early eighties cowpea research in the western province has resulted in thereleaseof two lITA varieties, TVx 3O-1G(Muliana) and 1Vx 456-01G (Shipepo) (Kannaiyan et at 1986). Multidisciplinary research in cowpea was initiated in 1982 with the establishment of the Grain Legumes Research Team (now known as the Food Legumes Research Team) under the Eastern Province Agricultural Development (EPAD) project.

Cowpea variety improvement program The major objective of the cowpea variety improvement program is to increase the productivity and production of cowpea in order to achieve self-sufficiency. To attain this goal, the team has been conducting research aimed at developing high-yielding, insect pest and diseases resistant cowpea varieties adapted to the major agroecologicaJ


To achieve the stated goal, the team will: • collect, evaluate, and maintain local and exotic germplasm • introduce high-yielding cowpea varieties/advanced breeding lines • screen for resistance/tolerance to major biotic stress factors limiting cowpea production • incorporate resistance/tolerance (through hybridization) to the major biotic stress factors limiting cowpea production_ Germplasm evaluation Since the establishment of the Food Legumes Research Team, about 1,000 cowpea accessions have been collected from indigenous and exotic Sources. A large number of locallandraces have been collected through various missions in the major growing areas of the country. Majority of the accessions were received from the International Institute of Tropical Agriculture (TITA), Tbadan, Nigeria. Germplasmevaluation isconducted under Msekera conditions (except for 1985/ 1986 when the germplasm was evaluated at Msekera and Masumba). The major traits considered for evaluation include growth habit, flower color, vigor, reaction to major diseases and insect pests as well as yield potential. The available genetic resources have been successfully exploited as entries to multilocational variety trials or as parents in the hybridization program. Germplasm of recent additions to the advanced variety trial (ZVu 588, ZVu 98, ZVu 238, and ZVu 268) were evaluated in the 1988/1989 season. They were tested further in a preliminary trial in 1989/1990 before being included in the advanced national variety trial. From these studies conducted during 1988/1989, two lines possessing good resistance to major insect pests (ZVu 151 and ZVu 950) and two lines with good resistance to diseases (ZVu 258 and ZVu 590) have been identified. These lines are being used in the hybridization program. Variety trials Promising cowpea genotypes, selected from local and exotic gennplasm, progress through preliminary trials (plots of 2 rows of at least 2 replicates) at Msekera to advanced national variety trials (plots of4 rOWS with 4 replicates) at 4-5 locations. In addition to Msekera and Masumba (Eastern province), the other locations are: Kaoma (Western Province), Mochipapa and Siatwinda (Southern Province), and, most recently, Kabwe (Central Province). Other entries in the preliminary trial are from thediseaseand insect pest nurseries and advanced breeding lines from IITA. Advanced national variety trials From 1983 to 1987, IlTA Cowpea International Trials, CIT 1 (extra early varieties), CIT 2 (medium maturity varieties), CIT 3 (bruchid resistant varieties), CIT 4 (vegetable cowpea varieties), CIT 5 (dual purpose varieties), and CIT 6 (aphid/virus resistant varieties), were conducted to identify varieties with good performance and suited to Zambian conditions. In addition to the cowpea international trials, a Zambian national variety trial was also conducted between 1984 and 1986. Entries in this trial were 8 liTA lines, a local landrace, and Muliana as control. Two outstanding lines in this trial were TVx 465411

44E and 1T820-889 (early maturing and resistant to cowpea aphid-borne mosaic visus [CAMV))_ In a second national variety trial tested at 3 sites during the 1986/1987 season with 7 previously tested lITAselections,2Iocallandraces,and Muliana as control, four lines (ZVu 83, 1Vx 4654-44E, K 80, and 11'820-875) were found prOmising. However, only IT820-875 was resistant to CAMV (Kannaiyan et al. 1989). All promising lines from all of the previous years cowpea international trials and Zambia national variety trials were reconstituted into three groups: medium maturity grain purpose, medium maturity dual purpose, and early maturity. Table 1_

National cowpea variety trial: medium maturity, groin purpose (grain yield in kglha) at M.ekero, 1988/1989 -199011991

Year Variety

1988/ 1989

1989/ 1990

1990/ 1991

ZVu83 ZVu 237 White Wonder lTD82D-875 K80 TVx 3381-02F Muliana (C) IT820-812

991 937 848 1270 983 956 653 864

810 991 709 471 794

1097 837 1165

Mean C.V.(%) SEÂą LSO (0.05)

938 21.0 100 287

966 696 554 894 830

460 966 726


719 26.9

29.0 176 359

90 258


% Control

961 922 908 902 824 773 757 718

127 122 120 119 109 102 100 95

845 23.9 101 217

Medium maturity grain purpose Eight promising lines from lITA, two local landraces (ZVu 83 and ZVu 237), with Muliana and Shipepo as local controls, have been tested for their performance, with emphasis on grain yield. Results from these trials have been inconsistent with entry performance, being variable across years and locations due to insect pest damage and outbreaks of diseases such as scab and ascochyta blight. Insect pest pressure has been exceptionally high at Masumba and Siatwinda while severe scab and ascochyta blight outbreaks were recorded at Kaoma and Mochipapa in the 1989/1990 and 1990/1991 seasons. Generally, yields have been good at Msekera where both insect pest and disease incidence have been less severe. Yields have been moderate in Kaoma and rather poor in Mochipapa and Siatwinda. Table 1 shows grain yield results for the Msekera region for three seasons. The results indicate that the two local land races, ZVu 83 and ZVu 12

237, are the most stable varieties. In the 1989/1990 season, ZVu 237 had the highest overall grain yield at Msekera and Kaoma. 1he landracesare,however, susceptible to CAMV and have been included in the hybidization program for incorporation of resis tance to CAMV. Medium maturity dual purpose Since in most parts of Zambia cowpea is grown for its leaves, varieties are also evaluated for leaf production. The cowpea dual purpose national variety trial meant for this purpose has been tested in the same location as the grain purpose trial. The grain yield results for this trial have also been inconsistent, but to a lesser extent than the grain purpose trial. Grain yields obtained in this trial have been generally low, partly due to leaf picking. Agronomic trials on the effect of leaf picking on grain yield have shown that varieties respond differently. Results from these trials have indicated that TVx 3381-Q2F is the most stable in grain yield when subjected to leaf picking while Muliana's grain yield declines drastically as a result of leaf picking (Musanya 1991). Leaf yields have tended to be quite high at Msekera with very small differences among the entries. Preliminary visual quality ratings indicate that TVx 3381-Q2F and ZVu 83 a re preferred for their leaves. Tables 2 and 3 show results of combined analysis for grain yield and leaf yield respectively for Msekera and Kaorna . ZVu 83 has the highest overall mean grain yield but there were no significant differences between varieties. Table 2. National cowpea variety trial: medium maturity, dual purpose (grain yield in kglha) at Msekera and Kaoma for 198911990 and 199011991 Kaorna

Msekera Variety\ year 1990

1989/ 1991

ZVu83 ZVu 237 IT82D-875 TVx 4654-44E K80 TVx 3381-02F M66 Muliana (C) White Wonder IT83S-872 Shipepo (C)

454 838 693 579 627 781 483 533 575 202

Mean CV(%) SEÂąI04 LSD (0.05)

559 37.1 127 298


1990/ 1990 776 404 585 689 601 375 631 538 480 671 169


1990/ 1991

298 266 193

116 124 160 130 118 126 150 109 150


278 213 214 280 191

524 343

94 259 13

Variety Mean





490 36.1 129 269


33.5 107 263

411 408 408

408 406 374 370 365 349 327 257 332 20.9 214



113 112 112 112 111 104 101 100 % 90




734 749 646












TVx 4654-44E

Muliana (C)

White wonder





Shipepo (C)








LSD (0.05)









C.V. (%)















608 673














129 263







808 456

94.7 810










111 .0

% Control










Variety Mean



904 1109






1990/ 1991



397 434



684 428


1989/ 1990



1988/ 1989








1990/ 1991







ZVu 237

1989/ 1990

1988/ 1989

Variety /Year


Table 3. National cowpea variety trials: medium maturity, dual purpose, leaf yield (kglha> atMsekera and Kaoma for 1988189 to 199011991

Leaf yield was not significant between varieties across the locations over three years. The local land race, ZVu 237, had the highest overall mean for leaf yield. While Muliana had a low grain yield, its leaf yield ranked third in the combined analysis.

Early maturity This trial comprising 10 lITA early maturing varieties (1T82D-889 as control), was begun in 1988/1989. Results obtained have been more consistent than those obtained for the medium maturity trials. Outstanding entries in this trial, IT 82E-16 and IT 82E-32, have performed better than the control for three seasons (1988/1989 to 199(11991) at Msekera. Table 4 shows the results of combined analysis for early maturity trials conducted in 1988/1989 and 1989/1990 seasons. Three en tries: IT 82E-16, IT 82E-32, and IT 830442 have significantly out-yielded the control across the locations over the two years. IT 82E-16 and IT 82E-32 have also been found promising in on-farm trials in the Western province. The two varieties will be tested extensively on-farm in other provinces next season before being considered for pre-release.

Hybridization program The cowpea hybridization program, initiated in 1988, has made a total of 42 crosses. These crosses have been designed to generate greater variability to facilitate selection of desirable genotypes. Efforts have mostly been directed at combining high yield potential with insect pests and disease resistance, and developing dual purpose materials with desirable leaf qualities. The major diseases being considered are scab and cowpea aphid-borne mosaic visus. A single seed descent technique based on the modified pedigree method has been used (or advancing breeding populations from F2 generation onwards. Effective selection o( desirable genotypes is undertaken in the FS generation population bulks, when a reasonable degree of homozygosity is expected to be achieved.

Conclusion While progress has been made in identifying cowpea lines with high yield potential, .t is evident that stable yields from such lines wiU only be realized if crop damage by insect pests and diseases is minimized. Good sources of tolerance to major diseases and insect pests have been identified from germplasm, and disease and insect pest nurseries. Only a few of these have so (ar been utilized in the hybridization program. There is still scope for further exploitation of the available genetic variability in developing disease and insect pest tolerant varieties. To complement host plant resistance work, there is need for intensified agronomic research on planting dates and intercropping to ensure stable cowpea yields on farrners' fields . Our experience at Msekera has shown that when planted early cowpea tends to suffer less damage from insects pests and diseases. In addition; early maturing materials yield better than medium maturing ones and suffer lessdarnage from insect 15



56 163

80 252




LSD (0.05)




C.V. (%)











69 201



214 32.0


244 119



108 313


















159 143

756 681 602

% Control

Variety Mean










1162 819

1989/ 1990















269 250



246 403




394 192





1988/ 1989 500

1989/ 1990


367 344

361 302

1988/ 1989





11'85F-867 -5




IT82D-889 (C)




11'830 -442


742 729

1406 1394

IT82E-16 11'82E-32


1989/ 1990

1988/ 1989



Table 4. cowpea variety trials: early maturity, grain yield (kglha) at Msekera, Kaoma and Mochipapa for 1988/89 and 1989/1990

pests. There is a need for systematic experiments to verify the apparent beneficial effect of early planting in all major cowpea growing areas.

References Kannaiyan,j., M.N. Mbewe, D.C. Greenberg, H .C. Haciwa, N.S.irving, and P.H. Sohati. 1986. Cowpea production and research in Zambia. World Cowpea Research Conference (Country papers), !>-9 November 1984, lIT A. ibadan, Nigeria. Tropical Grain Legume Bulletin (Special issue) 32: 144-148. Kannaiyan').,).M. Mulila-Mitti, D.C. Greenberg, and M.N . Mbewe. 1989. Varietal improvement of bean and cowpea in Zambia.1n Proceedings of the National Workshop on Food Legumes and Improvement in Zambia, 9-11 Mar 1988, Mfuwe, Zambia . Mbewe, M.N., D.C. Greenberg.j. Kannaiyan, H.c. Haciwa, N.5. irving, a nd P.H . Sohati . 1985. Cowpeas in Zambia. Paper presented a t the EEC ISAFGRA D join t Planning and Evalua tion Workshop (Maize and Cowpea), Contonou, Peoples Republic of Benin, 16-20 Sep 1985. MU!:anya,j.c. 1991. Recent research on the cropping systems with cowpeas in Zambia. Paper presented at the SADCC/nTA Cowpea Seminar in Harare, Zimbabwe, 2!>-26 Sep 1991 . Prior, A.). 1979. Cowpeas in Zambia, a review paper, Mt. Makulu Research Station. Department of Agriculture, Lusaka, Zambia. (Cyclostyled) .


Review of Cowpea Research in Lesotho 5.5. Maima

During the 1987-1991 cropping seasons, cowpea variety trials were conducted. to identify varieties adapted to the variousCTopping seasons in Lesotho using a complete randomized block design with four replicates. Results from these trials suggest that some cowpea varieties are mo re s uperior than others and that climatic variation has an effect on cowpea yield. Yields obtained in 1990/ 91 were highest due to lavorable clim:>tic conditions. IT82E·32, 1TB3S-990, and IT83D-442 w .... ;dentified as rugh yielding and stable across seasons, compared with other varieties.

In Lesotho, cowpea is produced only along the lower southern and foot hills zones. In these areas, the area planted to the crop is insignificant when compared with areas planted to major food crops (maize, sorghum, wheat, beans, and peas). Cowpea [Vigna unguiculata {L.)Walp.] can improve the nutritional status of people in the country as it is an important source of cheap protein and other essential nutrients. As food, the leaves and grain are given equal importance. Sometimes however, the crop is used as fodderfor animals. On account of its ability to fix nitrogen (up to 140 kg N/ha) efficiently, it provides a high proportion of its own nitrogen requirements, besides leaving a fixed N deposit of up to ~70 kg/ha in the soil for the succeeding cereal crop. In Botswana, the crop is tolerant of drought and heat hence, it may be important in the southern region of the country where severe drought has been experienced during the last decade. 1herefore selection of well adapted, high-yielding, drought and disease tolerant cowpea varieties can boost production. This study aims to: • Test and evaluate yield and adaptability of newly introduced cowpea cultivars, obtained from genetic selections of other cowpea international research centers, under environmental conditions at Lesotho. • Select promising high-yielding cowpea materials for further testing. • Provide research workers with opportunities to compare local and introduced materials.

Materials and methods Cowpea variety yield trials were conducted during the summers of the 1987-1991 croppingseasonsat the farm of the Maseru Research Station. Seed was obtained from cowpea/soybean improvement programs in Zimbabwe and Botswana. A complete randomized block design with four replicates was used for these studies. Data on disease incidence, plant maturity, and grain yield were collected and analyzed.


Results and discussion The results show that yields fluctuate every year due to seasonal climatic variation. The yields for 1987/88 were the lowest. (Table 1, Table 2, and Table 3). Disease incidence was regarded as very minor as cowpea mosaic virus (CMV), though an outstanding disease, had no effect on cowpea yield (Table 1). 1Vx4262, BOO9-B, and B242A were identified as higher yielding during 1987/88. IT82E-32, IT83S-990, and IT83D442 were also identified as high yielding and stable across seasons (Tables 2 and 3). These results indicate that yield potential is higher with improved cowpea. Table 1. Performance of 8 best cowpea varieties obtained from Botswana during 1987/88 aopping season


Yield (kg/ha)

Maturity days

1Vx4262 BOO9-B B242A 1Vu 1185 EL 7 B044 B097 IT82-889

700 700 736 735 696 617 546 520

130 130 130 130 130 130 130 100

Mean SEÂąCV(%)

669 166.5 37.0



Reactionto rust



5 R R R


-R= No Symptoms, S- Symptoms", CMV â&#x20AC;˘ Cowpea mosaic virus

Table 2. Yield of cowpea varieties, 1989190


Yield (kg/ha)

IT82E-32 IT83S-990 IT85F-867-5 IT830-442 17820-889 IT835-852 IT82D-875

1283 1088 1007 813 574

IT835-8n IT84E-124

458 395 258 117


6flJ.7 19

Table 3. Yield. of 9 local cowpea varieties, 1990/91 season Variety

Yield' kg/ha

IT82E·32 IT82D-a89 IT83S-990 IT83D-442 IT82D-875 Black~ed (California) IT83D·328·5 IT85F-867·5 IT83S-852

2889 a 1884 ab 1686 ab 1199 bcb 1180 bcb 1036 d 1003 cd 671.8d 433.3d

Mean SE± CV(%)

1196.7 110.2 195

Means within columns followed by the same letter are not significantly different at a.OSlevel as determined by Dwu:an's multiple range test.

Conclusion Of 20 newly introduced varieties, six (IT82E·32, IT83S-990, IT83D-442, JT82D-889, TVx4262, and 8009·8) have been selected for further evaluation.

Discussion Q. Mligo: Some of the lines showing superiority in Lesotho, have also performed well

elsewhere, for example the red seeded IT 82D·889 which has been released in Tanzania. Is there any preference for color in your country? A. Moim..: There is no preference for color. Q. Amable: Could you describe the agroecological conditions in the main cowpea growing area of Lesotho, ie. the lowland?

A. Moim.: The lowlands are between l,5(X)to 1,600 mabovesea level. Temperatures range from 2~oC during the summer and between - 2-10"C in the winter. The summer growing season starts with the rain in the middle of October and continues till mid·May. However, frosts may occur from around the middle of April, and the cold dry winter continues till the end of September. Q. Mosarwe: Most of the material from lITA, Nigeria tends to be early to medium

maturing. However, in Lesotho the material has been late maturing, taking up to 130 days. It appears this could be due to genotype x environment interaction. A. Moim... This would appear to be the case with lower temperatures playing the major role. 20

Q. Mamba: How importantisCDwpeacompared to bean (Phaseolusvulgaris) in Lesotho?

A. Maima: At present, bean is the more important crop. However, cowpea was an important crop before the introduction of bean. Comment. Nley.: The cowpea varieties which the presenter indicated were obta ined from Zimbabwe are not actually Zimbabwean varieties but are lITA lines which were supplied by Dr. D.M. Naik [then Regional Coordinator, UTA-CLIP, Southern Africa] who was based in Zimbabwe.


Evaluation of Contrasting Cowpea Genotypes at Different Dates of Planting Under Rainfed Conditions Moses B. Kwapata The yield components of ~ contrasting cowpea [Vig ... ""gui",,.,. (Ll WoIp.1 gent>types were evalualed ., two diIlerent dates at planting; nUd-/...uary and nUdFebruary, 1991; under ,aWed conditions. The yields 01 dry grain were significanUy differenl among the genotypes """ ranged from 106-160 kg/halo< the early planted vorieties to ~162S kg/ha to. the Iale pIanled varieties. Although varieties "",fat noed differendy ., the two planting dales, UCR路12S and UCR-231 consislenUy peifOiDoed betl... than the others, wher... UCR418 and UCR~I performed poorly. Crain yield correlated with the apparenl ha<vest index (0.16-0.51) and number at seeds per pod (7-13). In enntrost, PUlIlber of pods pet plant (2~), seed size (7-21 g per 100 seeds), and stem dry matter w,,", poorly correlated with yield. 11 is hypothesized that the high yields of the few cowpea genotypes _ dependent on the efficient partitioning of photoassimulales 10 the grain. In contras~ high vegetable biomass and heavy pod set poooibly created intense intr...planl competition for photoassimilales al the expense of good grain fill fo< many of the loW-yielding genotypes. In addition. pooc adaplabilily to the prevailing climatic conditions might account for the eXln!mely poor pod set of some of the low-yielding cowpea genotypes. It isconcJuded that planting ofcowpca at either mid-January or mid-february hasonly sligN effect on the yield at many of the genolypes.

Cowpea [Vigna unguiculata (L.) Walp.) is mostly grown in the warm and dry areas of Malawi. A number of local varieties are grown by smallholder farmers. Most of these varieties are landraces of the prostrate types, have low yield potential, and are often intercropped with maize, cassava, and sorghum. 'The low yielding potential of the varieties used, and poor agronomic practices contribute to the poor yields of cowpea often obtained by the smallholder farmers . Cowpea yields depend on the variety grown, the prevailing climatic conditions, and the agronomic practices used (Nangju 1979, Singh and Rachie 1985, Kwapata and Hall 1990 a&b). Hence, it is necessary to evaluate the performance of many cowpea genotypes with different plant morphological characteristics under local climatic coo<Utions and agronomic practices, in order to select varieties that are highly productive and well adapted to the local environmental conditions. 'The study attempted to evaluate the yield potential of different cowpea genotypes with upright and prostrate growth habits at different dates of planting, and to determine yield components associated with high-yielding genotypes.

Materials and methods 'The studies were conducted at the Bunda College of Agriculture, Ulongwe, Malawi, at two different dates of planting: mid-January (early) and mid-february (late) during the 1991 rainy season.


The first experiment was planted early and consisted of 34 cowpea genotypes 176,399,231, 182,74,239,391,232,290,235,230, 107,218,434, 108, 234,431,233,236, 425,406,384, 180,426,219,436, 430,432, 418 and 420). Thesecend experiment was planted late and consisted of 19 genotypes (UCR: 390, 125, 231,239, CBSD, 457, 395, CSSE, 472, Bambey 21 , ~, 469, 398, 391, 421, 245, 176,431, and 418). Atboth planting dates, eight genotypes (UCR: 125,176,231, 239,391, 390, 418,and 431) were included as checks. In both experiments, a completely randomized block design was used with genotypes as treatments and four replicates. The treatment plots consisted of three rows, 6 m long and 0.6 m apart. Cowpea seeds were sown 0.2 m apart within the row and thinned to two plants per hill . The experimental plots were weeded using hoes or by hand. No fertilizer or pesticide was applied. Irrigation was not necessary as the crop was grown during the rainy season (December to April). Weather data werecoUected(Table 1). All data were collected from the net mid-row of each treatment plot. (tK:~179, 145,125, 73,

Table 1. Average daily weather conditions during the 1990-1991 the growing season

Temp. (DC)


Rainfall (cm)



January February March April

295 265 265 22.0 26.2

18.2 18.0 17.4 17.5 14.5

8.5 16.8 8.1 6.7 1.9






Results and discussion Yield The grain yields of cowpea were significantly different among the genotypes at both planting dates. The yields ranged from 104 to 1609 kg/ ha for the early planting date to 80 to 1625 kg/ha for the late planting date. Nine out of 35 genotypes planted early and eight out of nineteen genotypes planted late produced yields of toOO kg/ha or above. The yields of eight genotypes which were planted at both dates were comparable to the yields of other genotypes. Two genotypes, UCR 125 and 231, produced high yields, while two other genotypes, UCR 418 and 431, produced low yields consistently at both the early and late planting dates. In general, the yields of prostrate genotypes were lower than those of upright genotypes. Similar observations have been made by Kwapata and Hall (1990 a&b). The yields of most of the upright 23

genotypes were comparable at both dates of planting, although that of many prostrate genotypes tended to be lower when planted late. Yield components lhere was variability in apparent harvest indices (AHI) among the genotypes but not between the planting dates (Tables 2 and 3). In general ,high AHI were associated with many high-yielding genotypes. Genotypes differed in the number of seeds per pod and seed size. Different dates of planting had little effect on these yield components. Number of pods per plant was greater with low plant stand, but this (plant stand) did not appear to have adversely affected the other yield components which were genetically specific to individual genotypes. The yield of different cowpea genotypes correlated to AHI (r '" 0.4), pods per unit area (r '" 0.8), and seeds per pod (r '" 0.4). On the other hand there were poor correlations between stem and leaf biomass, and pods per plant and seed size. These relationships suggest that there are complex competing processes in plants. In a genotype with high stem and leaf biomass, grain yield may be constrained by translocation of more photoassimulates to supporting vegetative tissues. Furthermore, high foliage volume could impair light distribution within the plant canopy and this in tum would affect the quality of light needed for initiation and development of floral organs (Kwapata et al. 1990. Kasperbauer and Karlen 1986; Adams 1982). 1he apparent increase in pods per plant associated with low plant density may be a reflection of the compensatory growth characteristics of cowpea, which is considerably very plastic. In addition, fewer plants per unit area allow more open canopies to develop and this pennits better light distribution within the canopies. This promotes initiation and development of more floral organs. but does not completely compensate for the reduction in the number of pods per unit area which contributes to low yield overall. Conclusion The results of these studies indicate that a number of cowpea genotypes have high yield potential. These genotypes indude: UCR 125,231390.179, 145, andCB5D. Many of these genotypes had high apparent harvest indices, suggesting that they partition considerable amounts of their dry matter to pods. On the other hand. low yielding genotypes tended to have lower apparent harvest indices, and majority had prostrate plant morphology. The planting of cowpea at either mid-January or mid-February had only a slight effect on the yield components of many of the genotypes. These genotypes, however, need further evaluation for several seasons and in different agroecological zones before they can be selected as suitable varieties for smallholder farmers.


Table 2. Yield and yield components of cowpea planted In mid-January, 1991 Genotype ranlcing

Genotype code

Yield (kg/ha)

1 2 3' 4 5' 6 7' 8 9 10' 11' 12

179 145 125 73 176 399 231 182 74 239 391 232 390 235

1609 1557 1523 1510 ]418 ]293 ]237 1159 1084 978 976


14 15 16 17 18 19 20

21' 22 23 24 25 26 27 28 2~

30 31 32 33' 34

Mean LSD(0.5)


107 218 434

lOS 234 431 233 236 425 406 384 180 426 219 436 430 432 418 420

Harvest Seed size


832 818 779 767 729 709 705 648 628 622 620 601 487 444 40S 338 211 162 124 116 116 104

Seeds/ pod

Pods/ plant

10 13

13 14



0.13 0.17 0.30 0.22 0.22 0.16

9.2 7.1 8.3 7.8 11.9 12.3 10.1 8.0 9.7 8.2 10.9 15.5 9.1 155 12.1 8.3 20.9 95 10.2 17.1 14.4 10.8 9.6 16.4 9.2 17.1 9.7 11.6 15.5 135 11.4 8.8 145 11.2

0.28 0.16

11.8 3.2



0.31 0.37 0.38 0.32 0.25 0.39 0.26 0.27 0.23 0.32 0.36 0.32 0.33 0.31 0.21 0.32 0.36 0.36 0.32 0.20 0.25 0.16 051 0.32 0.25 0.19 0.16 0.29

768.8 479

'Genotypes which were included in the mid路January planting.


12 12 12 12 11 10

11 9 9

13 8 11 9

13 10

13 11 11

10 2 16 10 9

6 6 15




29 6 3 5

11 8 9 10 7



9 19




9 9 10 9 8 10

5 7 10 12 4 17 13 9 3

10 2.0

10.6 14

10 10

Table 3. Yield and yield components of cowpea planted in mid-February, 1991 Genotype ranking l' 2' 3'

4' 5 6 7 8 9 10 11 12 13 14" 15 16 17" 18' 19"

Yield (kg/ha)

Harvest index

Seed size

Seed/ pod



390 125 231 239

1625 1525 1389 1100 1069 1045 1003 1003 939

883 862 772 631 597 524 430 405 376 80

0.18 0.31 0.31 12.0 0.26 0.28 0.19 0.23 0.31 0.18 0.30 0.27 0.33 0.37 0.25 0.29 0.31 0 .25 0.20

11.2 10.1 11.9 11 17.9 115 16.9 16.7 10.0 14.8 15.9 15.4 14.4 14.4 15.9 17.2 12.7 24.3 18.3

15 13 14 17 9 13 10 10 13 9 lO 12 12 11 12 11 12 10 8

20 11 16 17 10 15 13 11 11 9 10 42 42 15 18 18 21 18

869 341

0.27 0.09


11 .3 2.0

CBSD 459 395

CBSE 472 Bambey 21 8006 469 398 391 421 245 176 431 418




Genotype code


Genotypes which were included in the mid¡February planting .

Table 4. Cowpea yield response to date of planting (kg/ha) Planting dates Genotype code 125 176 231 239 390 391 418 431 Mean



1523 1418 1247 978 832 976 116 628

1525 405 1389 1100 1625 597 80 376





30 18.4 10

References Adams, M.W. 1982. Plant architecture and yield breeding. Iowa Journal for Rest'arch 56: 22>254. Kasperbauer, M.f., and Karlen, D.L. 1986. light mediated bioregula tion of tillering and photosynlhate partitioning in wheat. Physiology of Plants 66: 159--163. )(wapata, M.B. and A.E. Hall. l'19Oa. Response of contrasting vegetabl~owp.a cultivars to plantdensity and harvesting of young green pods. 1:Pod production. Field C rops Research 24: 1- 10. )(wapata, M.B. and A.E. Hall 1990b. Detemin.nts of cowpea (V igna unguiculata) seed yield at extremely high plant density. Field Crops Research 24: 23-32. Kwapata, M.B., A.E. Hall, and M.A. Madore. 1990. Response of contrasting vegetable-cowpea cultivars to plant density and harvesting of young green pods. II. Dry-matter production photosynthesis. Field Crops Research 24 : 11-21. Nangju, D. 1979. Effects of density, plant type and season on growth and yield of cowpea . Journal of the American Society of Horticultural Science 104: 466-470. Singh, S.R. and K.o.Rachie. 1985.Cowpea research, production and utilization. WileyPublishing Co, New York.

Discussion Q_ Mligo: Yield coula be easily presented as a function of pods/ plant, seeds/pod, and l00-seed weight. in your paper you indicated that as seed size increases, yield of acultivardecreases. I would have thought tha tChanges in any of the yield components would result in adjustments of the other components in tenns of the sink. Thus we could have fewer pods/plant o rfewer seeds/pod in response to larger seed size rather than yield being so severely affected as to result in a negative co rrelation. Could you enlighten me on this please?

A. Kw.p.ta: True, yield is a function of total aggregates of yield components. However, the rela tive levd of compensation associated with changes in some yield ,omponents, eg., increased seed size, is not large enough to overcome the decreased number of pods/plant ar unit area . The possible explanation is that large seeds create large sinks that demand more translocation of assimilates at the expense of pod set or seeds / poC:. Furthermo re, the c"anges in yield components are not equal in magnitude across many cowpea varie ties. This is why the increase in seed size would not result in increased yield per sewhen considering a large pool of varieties grown under the same soil and climatic conditicns_ Q. Amable: What are the regreSSion parameters for the seed size versus grain yield

per plot and was the lineal function the best fit? A_ Kwapata? The r value was around 0.5 and simple linear regression gave th:? best fitting line. Q_ Phillips: How much genotype x date of planting interaction d o you think there is and are you confident that if planting is deliberately delayed the rainfdll pattern is reliable enough in that part of Malawi [to get a good crop]?

A_ Kwapata: There is considecable genotype x date of planting interaction, suggestin E that recommendations fordateof planting should be specific for those g~n otypes that


do well at particular dates of planting. Blanket recommendations for planting date should be avoided if maximum cowpea yield is to be obtained. Regarding reliability, climatic data has been collected at the experimental site for about 20 years. From these records, the rainfall pattern is very reliable and has a confidence level of about 80 percent. Q. Haciwa: Were there any other factors associated with the different planting dates, for example disease and pest pressure'

A. Kwapata: The late planted crop was sprayed three times against aphids but, by and large, there were very few problems with diseases. The results reflect the influence of different dates of planting as they relate to temperature, moisture, and, possibly, photoperiod. Q. Mosarw.: What could be the optimal plant population denSity?

A. Kwapata: The optimal density varies with the growth habit of the variety. Our experience is that for spreading habit, the optimal population is around 200,OOO/ha and up to 400,OOO/ha for upright and compact types.


Intercropping Cowpea with Maize in Tanzania F.A. MyalaJ and J.K. Mligo

ExpertmentswerHonducted during 1989 and 1990 to study theeffectofcowpea (Vig"" ""guidl••• (Ll Walp.l variety. lime of planting. and planting patterns on the overall productivity of a maize-cowpea mlercroppiJ'lg system. Treatments were arranged in a 3 x 2 x 2 factorial with four replications. Results revealed that cowpea yields were affected by planting pattern as well .. time of planting cowpea. By planting a single row of cowpea between p<ilied rows of maize, cowpea yields were increased by 24%

compared to planting between single rows of maize. Late planting of cowpea reduced. cowpea yield by 67% when compor«l to early planting. Maize yield was not affected by any of the f.ctors studied. Yield advantage showed a falling trend willi delayed sowing of rowpea. It is concluded. that in low· to mid-altitude areas with monomodal to weakly bimodal rainfall patterns, an early determinate or late, semi-determinate cowpea cultivar could be interaoppcd with late maturing maiZe by simultaneous planting or relay intercropping cowpea between paired rows of maize not more than hoVo weeks after planting maize. 11 most parts oleastern Africa, intercropping often involves acereal and a legume with a cereal being considered as the main crop (Willey 1979). Maize/cowpea intercrop is one of the common combinations in this system with cowpea being planted after maize. This leads to a reduction in cowpea yield due to, among other things, the shading effect of maize. Cowpea is usually planted a few weeks afler maize. Although the effect of shading alone on cowpea has not been reported, several researchers (May and Misangu 1982 and Karel etal. 1982) have reported a yield reduction of more than 60% as a result of intercropping with maize. 'This could be attribu ted to several factors suchasshading by maize, reduced plant population of the cowpea, time of planting of the cowpea, the cowpea cultivar used, and the planting pattern of the intercrop. Several methods can be employed to reduce the effect of shading: using paired maize rows, choosing an appropriate time to plant cowpea, and using an appropriate cowpea variety in terms of plant type and maturity. This pa per discusses res ults offield tria Is cond ucted in 1989 and 1990 on maize and cowpea intercrops to assess the combined effect of row arrangement (planting pattern), time of planting cowpea and cowpea variety used on yield and productivity of both crops.

Materials and methods The field experiment was conducted during 1989 and 1990 on a sandy-clay loam soil at the experimental farm of the lIonga Agricultural Research Institute (IARl) (6 46' South, 37 02' East, and 506 m above sea level) Kilosa, Tanzania. Cowpea cultivars Tumaini and Vuli·l were either planted simultaneously with maize (Zea mays cv. Staha) or relay interplanted two and four weeks after maize in either single or double rows 29

alternating with single rows of maize. Maize was planted on 6 March 1989 while the first cowpea was planted on 24 February 1990. A randomized complete block desib'Tl (RCSO) with 4 replications was used with treatments arranged in a 3 x 2 x 2 factorial. Sole plots for both crops were also included in each replication at random with other treatments. Sole maize plots included both planting patterns while sole cowpea plots for both cultivars were planted at similar times as those in the intercrops. In alternate single rows intercrop rows were spaced 0.9 m for both crops. In alternate double rows, maize rows were paired with a space of 0.3 m between rows in a pair and 1.5 m Ix>tween pairs while cowpea rows were paired at 0.75 m and 050 m between rows in a pair for cultivars Tutll..ini and Vuli-I respectively and 1.05 m and 1.30 m between !,airs forTurnainiand Vuli-I respectively. Row arrangement in so le maize plots were similar to those in respective intercrops. In sole cowpea plots, rows were spaced at 0.75 m and 0.50 m for cultivars TUlnaini and Vuli-I respectively. Plant population of maize sole cropped, and intercropped were within the recommended density (Anon. 1987, Table I). For cowpea, recommended population density was utilized in sole plots only (Table 1). For pest control, Thiodan 35 EC (Endosulfan) was applied on cowpea during flower bud formation and then 10 days later at a rate of Ikg a.i/ha (Price et a!. 1982). Other experimental and management detail s are indicated in Table I. At crop maturity, plots were harvested and cowpea and maize yield recorded . For each plot, land equivalent ratio (lER) wascaleulated asdescribed by Willey and Osiru (1972) and d ata were analyzed statistically using analysis of variance. Table 1. Experimental and management details for the trials Alternate single row intercrt>E: M.,ize Cowpen Target populiltion (planl> / ha)


Alternate double row intercroE Ma ize

37,000 44,000 (CII . Tumilini)

BY,OOO (cv . VulJ -l

Intl'rplOlnt di"tmKl' (;'\)

N l' t an.-'il (m2 )

F...·rtilil l·r



37,000 {C\! . Tumaini} 139,000 (ev . Vuli-l )

44,000 66,000 (c v . Tumaini) 20,000 (ev. Vuli-I)



0.1(, (ell . Vlili· l )

0.16 (ell. Vu li -l)

27.0 27.0 60kg N / ha in th ... fo rm

60 kg / ha in thl' form o feAN



0.2 (. uti-I) 18.0



Ma ize

«(Y. Tumil ini) {(v. Tuma ini)

()f CAN NUI11~ r






60 kg / h., in the form of eAN

of pl,mhi per hill = Two pl.1nts for ma ize, Vuli·1, and monocropped T umai ni, ,md one

pl<1n t for Tumaini in th e intercrop.


Results Climatic data Rainfall started earlier in 1990 than it did in 1989 (Table 2), hence the first planting was done earlier in 1990 than in 1989. However, rainfall stopped early in 1990 with most of it falling during mid-March to early April. It was therefore better distributed thro ugho ut the growing period of 1989 than it was in 1990 (Table 2). Although temperature was slightly higher from the last week of March in 1990 than it wasd uring the same period in 1989 (Table 2), it fell as bo th seasons progressed. o

Table 2. Rainfall (mml in ten-day totals and daily mean temperatures ( Cl for the 1989 and 1990 growing seasons at lIonga Daily mean Temperature

Rainfall Months







1-10 11-10 21-28

0.0 22.0 8.1

11 .0 76.1 60.8

26.4 26.1 26.1

28.1 26.8 25.7


1-10 11-20 21-31

98.2 13.9 16.9

53.7 130.7 92.2

26.8 265 26.7

25.1 26.1 29.1


1-10 11-20 21-30 1-10 11-30 21-31

35.9 8.3 37.8 31.0 37.9 22.6

111.3 24.9 49.6 16.1 3.6 6.0

25.5 24.9 24 .9 24.4 23.6 22.9

26.2 25.5 26.0 25.5 24.5 23.6


1-10 11-20 21-31

34.8 15.6 6.5

0.0 0.0 0.0

22.9 22.1 21.0

23.3 23.3 22.4


1-10 11-20 21-31


0.0 0.0 1.7

21.4 21.9 23.2

21.5 21.8 21 .8




Cowpea grain yield All main interaction effects were non-significant (P > 0.05), except time of planting cowpea (P < 0.01), planting pattern (P < 0.(01), and yea r x variety interaction (P < 0.05) (Table 3). Higher cowpea yields were obtained for cv.Tumaini in 1989 than in 1990 (Table 4). Planting cowpea simultaneously with maize (lr two weeks after maize 31

resulted in similar cowpea yields 61% and 67% respectively higher than obtained when cowpea was planted four weeks after maize (Table 4). However intercropping maize and cowpea in alternate double rows increased cowpea yields by 24% as compared to relay intercropping in alternate single rows (Table 4). When sole cropped. the peforrnance of cowpea varieties was similar at all planting dates.

Table 3. Analysis of variance of maize and cowpea yields for two years Source of

Cowpea grain yield (kg/ha)


Year (Y) Replic.1tion in yrs Time of planting

cowpea (T) Cowpea varieties (V) Planting pattem (P) YxT YxV TxV YxP TxP YxP VxP YxTxV YxTxP TxVxP )'xTxVxP Error


Maize grain yield (kg / ha)

Land equivalent ratio

Mean square

f . value

Mean square

f. value

Mean square

f. value

df. 1 6

87882 62911

3.00 NS 2.15 NS

171358565 2588894

214.47'·· 3.37 NS

0.042 0.\04

0.77 NS InNS





0.15 NS



2.56 NS


1.12 NS


0.32 NS

1530118 427050 318225 120882 98937 821847 50017 562453 701780 16256 2080976 3629701 798963

1.92 NS O.53NS 0.39 NS 0.15 NS 0,\2 NS 1.02 NS 0.06 NS 0.70 NS 0.88 Ns 0.Q2 NS 2.60 NS 4.54'

0.187 0.016 0.273 0.030 0.187 0.116 0.017 0.069

3.46 NS 0.29 NS 5.04' 0.56 NS 3.46 NS 2.14 NS 0.32 NS 1.27 NS 2.06 NS 2.38NS 0.77 NS S.73-

74948 1 2 1 2 1 2 1 2 2 1 2 2 66 95

473859 9582 13383 71648

19589 91146

922 88697 60509 83216 26830 22528 29229

... =Significontol 0.001 level. significonl (P < 0.05)

16.21· ..•

0.32 NS 4.56' 2.45 NS 0.67 NS 0.34 NS 0.Q3 NS 0.03 NS 2.07 NS 2.85 NS 0.92 NS 0.77NS


0.129 0.042 0.310 0.054

=Significanl al 0.01 level•• =Significonl al 0.05 level. NS = Non

Maize grain yield All main and interaction effects were non-significant (P > 0.05) except year effect (P < 0.(01) and highest level interaction (P < 0.05) (Table 5). Maize yields in intercrops as well as monocrops were higher in 1989 than in 1990. Generally. planting pattern did not have any effect on maize grain yield under both cropping patterns. However. in 1989 maize grain yields were reduced when maize was planted simultaneously with cultivar Tumaini in alternate single rows and in 1990 when Vuli-l was reJay planted two weeks after maize in alternate double rows (P < 0.05. Table 5 ).


Table 4. Effect of time of planting cowpea, planting pattern, and year x cowpea variety interaction on cowpea grain yield in a maize/cowpearelayintercrop Cowpea Varieties Year 1989 1990 LSD (0.05) 99




279 298


Time of pbnting cowpea Simultaneous planting Two weeks after maize Four weeks after maize

Cowpea grain yield (kg/ha) 360 a 372 a

223 b

PI.nting pattern Alternative single rows Alternative double rows difference Sole crop Simultaneous planting Two weeks after maize Four weeks after maize LSD (0.05) = NS


389 141 ......

Vuli-I 741 920 923



1139 1021

Means fo llowed by the sa me letter do n ot differ significantly (p = 0.05) mean separa tion by LSD (for time of plnnting cowpea mean!' only) .

••• = Highly significant (P < 0 .001), NS = Non·significa nt.

Land equivalent ratio All main and interaction effects were non·significant (P < 0.05) except time of planting cowpea (P < 0.01 , Table 6), highest level interaction (P < 0.01), and year x variety interaction (P < 0.05, Table 6). At both planting dates, higher yield advantages were achieved with simultaneous planting or planting cowpea two weeks after maize (Table 6). It was only when cowpea was planted in alternate double rows four weeks after maize that a higher yield advantage was achieved . The high yield advantage was obtained with cv . Tumaini in 1989 and cv. VuH-} in 1990 (Table 6).

Discussion From the weather data, it is clear that intercropping wa s under a better moisture regime in 1989 than in 1990. This is furtherreflected by d ata from maize yields (Tables 2 and 5, P < 0.01); maize yields were higher in 1989 than in 1990. On the other hand, cowpea yield S from intercropping trials were not affected by seasonal rainfall variation and maize yield response under different intercropping patterns was similar across seasons.


Table 5. Effect of planting pattern, cowpea varieties, and time of planting cowpea on maize grain yield (kglha) in the maize/cowpea relay intercropping VulH

Tumaini A It("rflo'lll!'




Single rows

Double rows

Single rows

Doubl(> rows

1989 Simultaneous pbnting





Cowpea planted two weeks after maize





Cowpea planted four weeks after maize





1990 Simult;lOeous planting





CowpE'<l planted two weeks after mil iz拢>





Cowpea planted fo ur weeks after maize





LSD (0.05)


Sole Crop Single rows


Paired ro ws


4721 2704

19A9 1990








Ye;u mean difference;: 2017"路 Plnnting pilttem mean difference = 75 NS .. = Significant (P < 0.01 ), NS = Non !>ignificant

For all intercrop treatments, plant population of maize was kept at a recommended density (Table 1), only planting geometry (row arrangement and interplant distance) varied. This variation did not have any effect on the maize yield (Table5).ln most areas of eastern Africa, farmers usually consider a cereal as their main crop, it is therefore important that alteration of the planting pattern does not affect the maize yield negatively. By pairing the maize rows, cowpea yields were increased by 57% (Table 4). Cowpea plant populations in both planting patterns were similar, hence it can be concluded that the variation in cowpea yields between the two patterns was not due to the differences in plant population but probably due to reduced shading from the maize when maize was planted in pai red rows. Component crops population ratio (maize: cowpea) in relation to respective monocrops was 100 : 56. With this ratio it might be possible to further increase cowpea yield by increasing the cowpea popula34

tion interplanted between the paired rows of maize especially with early determinate cultivars such as Vuli-1. Table 6_ Ef(ectof planting paUern, cowpea varieties, and time of planting cowpea on LER in the maize/cowpea relay intercropping Vuli-l

Tumaini Alternate




Single rows

Double ro ws

Single rows

O(luble rows

Simultaneous planting





Cowpea planted two weeks after maize





Cowpea planted four weeks after maize





Simultaneous planting





Cowpea planted two weeks after maize





Cowpea planted four weeks after maize







LSD (0.05) 0.3 AI though pairing the maize rows increased cowpea yields in the intercrop, results have shown this to depend on time of planting cowpea in relation 10 time of planting maize. Planting cowpea simultaneously or two weeks after maize resulted in similar cowpea yield . Huwever, when cowpea was planted four weeks after maize, cowpea yields were significantly reduced (P < 0.05, Table 4). Ntare (1990) working on cowpea and pearl millet intercrops in the Sahel obtained results of similar trend. He recommended (Ntare 1990) early interplanling of cowpea to reduce competition from a cereal crop. Furthermore, in the present study, the cowpea monocrop yields for both varieties were similar across aU planting dates, indicating that reduced cowpea yields, when cowpea was planted four weeks after maize in the intercrop, was due to reasons associated with intercropping situation x dates of planting cowpea in the interaop interactio ns rather than late planting. With LER, the highest level interaction was significant (P < 0.01), however a declining trend for both cowpea cultivMs was observed under both planting patterns. This suggests that in the present intercropping system land could be utilized better. Yield advantages similar to those obtained with earlier planted intercrops were observed for Tumaini in 1989 and Vuli-l in 1990 when they were relay planted (our weeks after maize (Table 6). It might be possible that with a season of limited rainfall as was the case in 1990, V uli-l could even be relay interplanted four weeks after maize to obtain yields similar to those of earlier intercrops. 35

Q. Hadao: What is your comment about the likelihood of adoption of row planting

under cowpea-maize intercropping by farmers in Tanzania? A. Myalca: Because farmers in Tanzania already plant their maize in rows, I would think it would be easy for them to adopt a paired row system. However, the technology should be tested on-farm before final say on the possibility of its adoption. Q. Kwapata: Is there any explanation for the large differences in sole crop yield values not being [statistically) significant? A. Myaka: It might be possible that beca use the sole crops were analyzed separately, the number of observations from which the means were calculated were lower in sole crops than in the intercrops, resulting in lower precision and non-significance. CommenL Dolo: It would be more informative if some indicators, for example CV values were also included. I would make a similar remark regarding the other presentations. A. Myaka: The point is taken. However, experience has shown CV from intercropping experiments to be very high thus when presented in an intercropping situation it might, in most cases, not be of much value.


Intercropping Cowpea in Maize and Sorghum under Zimbabwean Conditions T. Nleya

Cowpea IVig1Ul wrgwcuIlftll (l.) Walp.] is an important component in the diet of the people in the communal areas of Zimbabwe. It is grown throughout the country primarily by communal farmers for both its leaves and grain. While systematic intercropping has been discouraged in communal agriculture over years in Zimbabwe. most communal farrm'1'S still randomly interplant cowpeas within ccreal-based cropping systems. To determine whether the yield of cowpea can be maximized without significantly reducing the yield of cereals, an exotic cowpea variety, IT82D-889, was intercropped with maize or sorghum \lnder two cropping systems: (a) a single row of cereal to a single row of cowpea (1: 1) and (b) two rows of cereal to a single row of cowpea (2 , 1).

lnterplanting cowpea between rows of cereals reduced cereal yields when compared to sole crops at all sites. 1ne degree of reduction in cereal yields was

gen<'raU y Ngher allow rainfall sites than at highrainfallsiles. Intercropping efficiency, recorded by land equivalent ratio (LER), was higher in 1 ; 1 cropping systems than in 2 : 1 cropping systems. Implications of the findings on future research strategies are discussed .

Cowpea [Vigna unguiculata (L.) Walp.] is an important component in the diet of the people in the communal areas of Zimbabwe. Cowpea, also know as nyemba (Shona) and induml>a (Ndebele), is grown throughout Zimbabwe primarily by communal farmers for its leaves, immature pods, and dry grain which are consumed in various preparations such as munyemba (fresh green leaves) mufushwa, (dried and stored leaves), mukove (green peas), mutakura/inkobe (dry grain boiled with maize), and rupiza/bhizha (cowpea paste) (Mariga, Giga, and Maramba 1985). Cowpea serves as an important source of protein to the communal population who cannot afford expensive animal protein as the seed contains 23-30% protein and the cooked leaves contain two-thirds of this (Bressani 1985). While a lot of effort ha s gone into the production of commercial cash crops in Zimbabwe, subsistence crops, such as cowpea and bambara groundnuts, have received little attention (Johnson 1970). Steps should be taken to improve yields of these crops so that farmers' dietary requirements can be produced ona smaller hectarage in order to make more land available for cash cropping. A cowpea research program was initiated by The Agronomy Institute, Causeway, Harare, in 1982. One of the major objectives of the program was to test cowpea lines developed at the International Institute of Tropical Agriculture (UTA) in Nigeria for grain yield under Zimbabwean conditions. Initial trials indicated that the short-duration, semi-determinate cowpea lines outyield the local, indeterminate varieties. However, before such exotic varieties are introduced to the farmer, the program has to determine whether these varieties can fit into the prod uction systems practiced by the farmers . Hence, a survey was conducted by the Department of Research and Specialist Services, in collaboration with lITA, to


determine the uses of cowpea, and production practices in Zimbabwe. The survey showed that most communal farmers intercrop cowp"a with maize and other cereals such as sorghum, finger millet, and pearl millet. Th" form 01 inte rcropping practiced by the larmers is additive since the populaton 01 th" c"real (the main crop) is maintained at optimum level and the cowpea is planted randomly within the cereal to geta ''bonus'' yi"ld olcowpea . However, the cowpea yield lrom such intercropping systems is very low . The need to improve the system was lelt by the Agronomy Institute. HPflce in 1987/88 an intercroppingtrial was designed with the objective 01 maximizing the yield 01 cowpea intercropped with either maize or sorghum without signilicantly reducing yields of the cereals.

Materials and methods The trial was conducted at four si tes, twoon-station(Panmureand Makoho li) and two on-farm (Rushinga and Marange). Panmure and Rushinga are sites with a high potential lor crop production in terms of rainlall and soil type while Makoholi and Marange are low potential sitt'S (rabies 1 and la). Table 1. Experimental details and site description (1987/88) O n-Farm

On-Station Site





Natural Region Soil texture Rainfall during crop growth (mm) Planting date

lib MgSCl


lib MgSCI


812 30-11-87

538 5-12-87

591 3-12-87






T.D Maize Sorghum

450 'AN' 350 'AN'

450 'AN' 450 'AN'

350 'AN' 350 'AN'

350'AN' 350'AN'

Design RCOB

4 replicates

Plot size

Gross-7.6 m 2 Nett-26.2 m 2

Fertilizer Init;al Maize Sorghum Cowpea


Seven treatments were tested: sole crops of maize, sorghum, and cowpea, and four intercrops; I maize row to I cowpea row (1M : IC), 2 maize rows to I cowpea row (2M : lC), I sorghum row to I cowpea row (IS: IC), and 2 sorghum rows to 1 cowpea row (2S: Ie). R20l, a short d",ation variety, was used as the maize component crop and SV-2 as the sorghum component crop. 1T820-889, a short duration, bush type developed at liT A, was used as the cowpea component crop. 39

Table la. Experimental details and site description (1988/89) On-Station Site Natural Region Soil texture Rainfall during Crop growth (mm) Planting date Fertilizer Inilial maize sorghum cowpea









lib MgSCI



636.5 29-J1.s8

257.7 2-12-88

873.7 3-12-88

524.8 6-12-88

300 'I.'


300 'L'

300 'L'


T.D Maize Sorghum

200 'AN'

200 'AN'

200 'AN'

200 'AN'

200 'AN'

200 'AN'

200 'AN




4 replicates

Plot size

Gross--57.6 m2 Nett-26.2 m2

Maize was planted at intervals of 90 em between rows and 45 cm within rows (25 000 plants/ha) in both sole and intercropped plots while sorghum was spaced 90 em between rows and 15 cm within rows (74 oooplants/ha) in bothsoleand intereropped plots. Cowpea was planted at intervals of 45 cm between rows and 20 cm within rows (111 000 plants/ha) in sole plots. The within row spacing for cowpea was kept at 20 cm in intercropped plots, thus 1:1 intercrops had a cowpea population of 55 000 plants/ha and 2 : I intercrops a cowpea population of 27 000 plants/ha. Data on yield as well as rainfall during crop growth al each site and each season were recorded.

Results In the sole croppings, Panmure recorded highest maize yields (6.2 t/ha and 4.1 t/ha) while Makoholi recorded lowest yield (3.1 t/ha and O.91/ha) for both seasons of the study. Highest yields of sorghum were recorded at Rushinga and lowest yields at Makoholi for both seasons (Tables 2 and 3). Rushinga recorded highest sole cowpea yields (875 kg/ha) in 1987188 while in 1988 / 89, the highest cowpea yields were recorded at Panmure. Interplanting cowpea between rows of the two cereals reduced cereal yield as compared to sole crop yields at most sites (Figs. 1 and 2). TIle degree of cereal yield red uction was significantly higher in drier environments with lighter textured soils (Marange and Makoholi) compared with reduction in wetter environments with heavier textured soil (Panmure and Rushinga). The higher cowpea population in the 40

Table 2. Maize, sorghum, and cowpea yields (kglha) in sole and intercrop plots (1987188)



M.iu SorgflW1\

Sole maize Sole 30rghum Solerowpea







JUI 2497



R_ ....


Mar짜lge North


4958 2021








tOOl 578



IM : IC 2M : Ie IS : Ie 2S , IC



211.00 164.23

Sign (S"') L.5.0 . CY. %




2296 2383 NS 6.'

18'l3 2636

ISO 22 34.61 298.36




15<7 15<1





178.OC 28.6

602.93 7.'

4532 4789



N5 117


106.66 1032.08 23.6 60.5


258.57 NS 10.9

352 197 311 146

2274 3740


184 311 208

2<89 2952



190.65 32.8

993.2 16.8

112.24 N5





Table 3. Maize, sorghum, and cowpea yields (kglha) in 80le and intercrop plolli


(1988/89) TRlltmml


Sole maize

Sorghwn COIoO'pe..

41 27




, 66 2419 235<

""50 SS

128.40 NS



277 323 286 89.9

538 61' 136 2IE IIXU


" I. 32




42 .3



So.... um



2560 2179

114 IE 134 88




29, 9 3278


NS 611 .9 ' .2


4023 2389

4092 3370

38 23.6




N5 277.0 37.7

...... um



3027 3206




Sole row pea

S.E S;gn (5%) LSD. C.V. %



Sole sorghu m

intercrops IM : IC 2M : 1C 15 : Ie 25 : 1C




96 7.



11 66




NS 11 2.4




78.2 285

121 .1

1 : 1 intercrops increased the competition pressure on cereal yields (Figs. 1 and 2) moreso in sites with low potential for crop production. As a way of determining the efficiency Q( the different cropping systems, land equivalent ratios (LER) were calculated for each cereal crop and each cropping system (Tables 4 and 5). The effect of cereal crop on LERs was not significant at all sites, meaning that there is no need to separate the crops when looking at the effect of cropping system on LERs. The 1 : 1 intercrops had relatively !tigher LERs than the 2 : 1 intercrops at all sites and for both seasons except at Makoholi in 1988/89. The same cropping system (1 : 1) had significantly higher LERs than sole crops at Panmure, Rushinga, and Makoholi during the 1987/88 season . 41

Discussion Maize and sorghum yields were reduced by intercropping. However, the degree of reduction in yield of these cereals was strongly dependent on the environment. Reduction was highly significant in drier environments with lighter textured soils than in wetter environments with heavier textured soils. By reducing the cowpea population (2 : 1 intercrop system), less competition was posed o n the cereals, and where moisture and nutrients were no t limiting (Panmure and Rushinga), the reduction in cereal yield was insignificant. The reduction in cereal yield was probably due to competition for light, soil nutrients, and moisture by the component crops (Ofori and Stem 1986; Ughtfoot 1'.l80). Lightfoot (1980), working with a similar intercropping system of sorghum and cowpea in Botswana, reported no reduction in sorghum yields. Thedifferenceof his study from the present results could bedue to population T~ble

4. Land equivalent ratios ~s offected by maize


Sorp.m Sip (5%)

,_.. ,_..

























Table 5. und equivalent r~tios as




Rushing. '!9IWM







by cropping system

M. koholi

Panmure Cereal



Panmure l












i .~



13 21






















Signs (5%)











03 93


differences. Lightfoot (1980) worked with very low populations of sorghum and cowpea in his intercrops. Intercropped cowpea always produced some grain. Mostly yields were low even in sole cowpea crops (Tables 2 and 3). The 1 : 1 intercrops attained reasonable yields of cowpea while the 2 : 1 intercrops plants/ha attained a maximum of 286 kg/ha. However, as indicated earlier, the higher cowpea population adversely affected cereal yields. TIle LERs obtained showed reasonable benefits from intercropping, (LERs were mostly above one). TIle benefits from intercropping were generally greater in high 路 potential wetter environments than in low potential environments. Lightfoot (1980) indicated that intercropping has greater advantages in wetter climates with long continuous growth periods and under high levels of inputs. This agrees with the results reported in this paper. Ironically, farmers who practice intercropping are located in low potential dry environments where the benefits from intercropping are very low.

Conclusion The study shows that cowpea is very competitive and can adversely affect maize and sorghum yields, particularly in drier environments. Although some yields can be obtained from cowpea, the significant reduction in cereal yields may not be adequately compensated for by the cowpea yield obtained . The objective of maintaining main crop yields was not met especially in those sites representative of environments where intercropping is the main practice. In light of these results, intercropping studies are being shifted towards screening cowpea to identify genotypes which are less competitive in cereal-based intercropping systems.

References Bressani,R . 1985. Nutritivevalueof cowpea . Pages3~jnCowpearesearch,productionand

utilisation, edited by S.R Singh and K.O. Rachie. John Wiley and Sons, USA. Johnson, D.T. 1970. The cowpea in the African areasof Rhodesia. Rhodesia AgriculturalJoumal 67(3): 1-4. Lightfoot, C. W.F. 1980. An initial report on the evaluation of intercropping. Evaluation of Farming Systems and Agricultural Implements Project Departmentof Ag ncultural Research, Ministry of Agriculture,Gaborone, Botswana (EFSAlP) . Ministryof Agriculture,Gaborone, Botswana.

Mariga, IX., D. Gigo, and P. Maramba . 1985. Cowpea production constra ints and resean:h in Zimbabwe. Grain Legume Bulletin 30: 9-14. Ofori, R , and W.R Stem . 1986. Maize; cowpea inten:rop system: effect of nitrogen fertilizer"" productivity and effeciency . Field Crops Research 14: 247-261.

Discussion Q. Phillips: Could you cJarifyhow the LER can be greater than 1.0 when the yield of cowpea in intercropping is less than the cereal lost as a result of intercropping.


A. Nleya: !he LER values were obtained by converting the yields of cowpea and cereal intercrops into fractions of their respective sole crop yields and then combining the two fractions. Comment. Dashiell: Some of the LER values reported in the paper were quite high (grearer than 1.5) because the yield of monocrop cowpea was low. Thus when the cowpea portion of the LER is calculated a high value will be obtained (ie. intercrop cowpea yield expressed as a proportion of monocrop cowpea yield is very close to 1.0) even before addition of the partial LER for the cereal component. Q. Pall-Shik\tuIu: Could you comment on the management (inputs) used in the two sets of sites, ie. on-station and on-farm.

A. Nleya: Management of the two sets of trials was the same. Both on-farm and onstation trials were researcher-managed and inputs were the same.

Q. Amable: Was there an economic analysis of the gains and losses under the vario us cropping systems?

A. Nleya: No, the economic analysis was not done.


On-Farm Research Experience With Improved Cowpea CuItivars G. A. S. Milti and J. M. Mulila Milti

In Zombia, particularly in the EaslemProvince, c~is cullivated under dlv.... soil .00 cllmalic conditions and is troditionalJy grownllCalteled in. malncropofmaizeor 5Crglnun. The local c~ is • vlgorou., prootnote to oemidimbing plant that yields little grain. Thestudy wasconducted to obtain fonne,,;' I ! i ... ~ofleaf.oo grainqualities of 10 ~ varieties and also to detennine how tolerant the farmers' selected varieties were to disease M'ld insect pressure. 1he varieties were planted in • randomized. complete block design wiihsik!SM replicals. Number oivarieties tested. diffen)d each season. Results revealed that fanner parlipalion played • very significant role in helping scientists to quickly understand what qualili.. the fanners requUed in • c~ variety. However. it was not possible to find a single variety with all characteristics preferred by the farmers. ZVu 83, however. met mast farmers' criteria.

Cowpea is one of the most co mmonly grown food legumes in Zambia particularly in the eastem Province. Although production figures are not available, it is estimated that all rural households grow a few plants mixed with local maize or sorghum each season. In the Eastem Province, cowpea can be grown on the plateau and in the Luangwa valley (one-third of the Province) where other proteinaceous foods such as beans are environmentally out of place, groundnuts are difficult to grow, and cattle are absent. Cowpea plays a major role in household diet In the province, it is utilized at various stages of growth (Mitti 1983/84): fresh leaves are eaten as spinach (often prepared with pounded groundnuts) throughout the season; some leaves are parboiled, sundried, and stored away for use in the dry season, green pods are also boiled and eaten as snack (along with pumpkin or green maize), starting aroWld January1 February when household food reserves are low or exhausted, and finally dry grain is a cherished relish (cooked like beans) in the dry season. While food legumes, because of their high protein content in general, constitute the natural protein supplement to the predominantly maize staple diet, in the Eastern Province, cowpea represents the legume of choice for the majority of farm families. With 24 percent protein, 62 percent soluble carbohydrates, and a relatively high lysine content (Singh and Rachie 1985), cowpea can help reduce pro tein deficiency malnutrition. A good harvest would provide women a ready relish for their families, particularly in the dry season when relish material is scarce. However, local cowpea produces very little grain; the plants are susceptible to-disease and insect pests, and often only a few survive to maturity. In the 1987/1988 season, on-farm work was initiated to test some of the varieties being produced by the cowpea improvement program. The main objectives of the onfarm work was to obtain fanners' criteria for evaluating cowpea materials and to


enable scientists assess acceptable tolerance by the variety to disease and insect pest pressure.

Materials and methods Ten improved cowea varieties including standard controls, ZVu83 and a local site variety which differed from site to site, were tested over four seasons: 1987/88, 1988/ 89, 1989/90,and 1990/1991 . 1henumberofvarieties tested varied each season as those which exhibited severe susceptibility to diseases/insect pests were dropped and replaced by new varieties. 1he varieties tested during each of the four seasons are indicated in Table 1. Farmer selection Most of the farmers were volunteers among the numerous cowpea growers in the target area where the Adaptive Research Planning Team (ARPT) normally conducts its on-farm research Program. While many farmers were willing to test the new cowpea varieties in their own fields only five farmers could be accommodated each season. Thus, there were five test sites in each of the four seasons, whose location varied seasonally accordingly to the five farmers selectp'~ to participate in the trial for the particular season. Table 1. Te.t materials and sea.ons Crop season Cultivars ZVu83 ZVu 237 Local TVx 465444E K80 "

IT82D路875 1T83S-872 Muliana Whi te wonder TV x 3381-02F



1989 / 90



x x x x x x x

x x x x x x

x x x x

x x



x x x

x =Sown; - = Not sown

Trials design and records Th trial was superimposed on local maize planted with two seeds per station at 75 cm x 9Ocm. Two seeds of each cowpea variety were sown 25 cm apart between the maize. Time of planting and other management aspects were controlled by the farmer. In some seasons, when an improved maize variety was used, some fertilizer was applied to the maize. The cowpea entries were planted in a completely randomized design with sites as replicates.


Trials were evaluated by the farmers' preference ranking of cowpea varieties, grain yield as a percentage of ZVu 83, and the number of plants surviving to fruition. Leaf harvest was also taken into account Similar management practices and evaluation procedures were followed at all sites. Grain yield data were subjected to analysis of variance.

Results 1987/88 season A zero grain yield of the local variety was recorded during the 1987/88 season despite it recording the highest leaf yield, and near zero yields for Muliana (Table 2). The zero yield of the local variety could be because it was just starting to bloom at the time the season came to an end. Also, Muliana was rendered nearly fruitless by the cowpea aphid-borne mosaic virus. Farmers preferred varieties KSOand ZVu 83 (control) for their high grain and leaf yield. However, some preferred ZVu '137 despite its low grain yield (24% of control). Table 2- Leaf and grain yield (kglha), 1987188 season Variety TVx4654 ZVu83 Local

K80 ZUv'137 Muliana

Leaf kg/ha

Grain kg/ha


586.4 875.0 0.0 890.0 208.3

1543 1994 1543


% of control

67 100 0 102

24 1

Fanners' ranking

2 1,2 3 1,2 1

1988/89 season Despite similarities between the three sites, the local cowpea yielded at only one site (Table3). The other test varieties however, competed stiffly yielding the same or even better grain than the control. Farmers preferred K80 most, followed closely by ZVu '137 and TVx 4654. In observation plots, however, IT 82D-875 and IT 835-872 gave very highgrain yield s and were consequently promoted to trial plots. 1herewas no sensible leaf yield data this season due to frequent errors by farmers such as mixing-up of the varieties. 1989/90 season Although the site local on average gave no grain yield, its leaf yield was far superior (over 10 times) to that of all other varieties put together (Table 4). However, the leaf yield data was not statistically analyzed. ZVu 83 significantly outyielded other varieties by over 200 kg/ha (Table 4). Farmers ranked it most preferred followed closely by ZVu '137, TVx 4654-44, and K80 much like in the previous season. However, due to its susceptibility to CAMV, K80 was dropped and replaced by a less-prone but more leafy variety, TVx 338HJ2F, the following season. White wonder was also included.


Table 3. Grain yield, 1988189 season (over 3 sites) Grain Variety


1<801127 TVx4654 ZVu83 ZVu237

125 1003 900.2 771 .6

Others 1 Local

1.2 111 100 85.7


3 1 3

(27) (94)

(241.8) 2 (848.8) 3 (1m4.0)

IT 820-875

Farmers' rankings

% of contral


1. 2.

Exduded in ANOVA All sites except one (725.3 kg/ ha) gave zero yield

3. 4.

One site only

Two sites only

Table 4. Leaf and grain yield. 1989190 season (over 5 sites)

Variety ZVu83 ZVu237 IT 820-875 TV x4654-44 E 1<80 ISO (0.05) eV(%) Local"

Grain" kg / ha


719.6 a 486.8 b 455.0b 370.4 b 338.6b

100 51 47

2. 4 3




169 129

104 175 232.5

loGlJ variety haVlng given zero yield, not included in ANOY A


data from one site only


% of central

68 63

1 4. 2 (2)

36.58 1534



Farmers' ranking

Leaf kg / ha

same letter not significantly different

1990/91 season

Site locals were clearly superior in leaf yield, but still gave no grain yield (see Table 5). The introduced varieties yielded similarly (around 500 kg/hal with no significant differences between them. Farmers again favored ZVu 237. followed by ZVu 83. and TVx3381-02F. The local variety was liked only for its leaf yield . 48

Table 5_ Leaf and grain yield, 1990191 (over 3 replicated .ites)

Variety ZVu237 ZVu83 TVx338HI2F

Leaf kg/ha

Grain kg/ha

267 492 192

536.9 506.0 459.8 303.4 42.8

ISD (0.05)

CV (%) Observations" Local 11'820.875 White wonder

17.44 411 169


%of control

106 100

rankings 1 2





407.5 367.0


6 5



â&#x20AC;˘ NOI. lru:luded in ANQVA

Overall performance Leaf and grain weights were taken as straight averages for each variety over the seasons (i.e. no ANOVA was d one). Farmers' comments for or against each variety were similarly averaged over the seasons . Overall, farmers' ranked ZVu 83 highest, followed by ZVu 237 and IT82o.875 (Table 6). This ranking was expected as the same varieties more or less ranked highest at most sites every season .

Discussion Cowpea varietal screening started with on-station tests many seasons back. Typical of on-station work, the materials were evaluated mainly on the basis of grain yield. As such, the initial set of varieties used for on-farm testing were selected because they were high-yielding (grainwise) and more tolerant to disease (on-station). But during the years of on-farm testing. when farmers themselves grew the varieties and evaluated them in their own way. a number of lessons were learnt about farmers' preferences: (a) farmers ' choice was based on both grain yield and leaf yield; (b) farmers ' preferred smooth (not hairy) and tender leaves; (c) preferred grain type was large. tasty/delicious, and quick to cook; (d) farmers preferred early-maturing varieties (though this quality sometimes leads to grain rot when harvesting is delayed). Farmers were very concerned about varieties that: (a) lodge and climbed onto maize heavily; (b) were susceptible to storage pests; (c) had a dull color; (d) took long to cook (fuel or time shortage); (e) were small-seeded (though most are).


T.ble 6. Average yields .nd formers' preference nnking


Le.f kg/h.

Grain kg/ha

%of control









More leaf, more pods, brown seed coat, big grain,tasty

Hard to cook






Large seed, early maturity, leafy, delicious.

Hard to cook (small seed?)






Delicious bigger seed .

Hard to cook, susceptible to sto rage pests, color not liked






Lots of leaves tasty (good at the table)

Few leaves, lodges Small seed, poor seed colur, hard to cook, poor storage.




More leafy more grain


Lodges, climbs maize,

small seed, sto rability questionable. White wonder










Local 1764







Leafy, long life Leafy

Late maturity, Iittle/ no seed Seed spoiled by rain


Data from one season or site

Leaf harvest or farmer ranking / comments not recorded .


TheJ'e were other lessons with implications on the methodologies used by scientists to evaluate the varieties: (a) although statistical procedures were used to detect mean differences, fanners' choice was not necessarily based on yield level, but also on quality of the grainorleaf, a criterion which cannot be easily analyzed statistically, (b) the leaf yield data provided by the farmers were highly subjective, as the amount picked by the farmers' depended on a number of factors including: • size of the family to be fed • availability of other preferred relish, e.g., meat • leaf quality rather than harvestable quantity of the variety • interest in preservation or capacity to do so for dry season use Hence leaf harvest data does not lend itself to rigorous statistical judgement. For these reasons farmers' choice was sometimes different from thatof scientists, which made it rather difficult for scientists to decide which variety to release to the farmers. However, looking at the overall performance and ranking, ZV u 83 was most preferred by farmers and will be proposed for released. In the meantime, ZVu 83, K80, and others ranked highly by farmers will be tested further, along with new materials, until suitable varieties are found. Farmers' participation will be increased and efforts will be made to come up wi th some standard evaluation criteria to use in grading the varieties. Farmers' practices such as planting dates, methods of planting, and time of harvests, will also be investigated.

Conclusion Farmers' participation played a very Significant role in helping scientists to quickly understand what qualities/characteristics the farmers wanted in a cowpea material. Most varieties were similar in grain and leaf yield as well as tolerance to some pests, hence, only the farmers' evaluation criteria that could separate them. Due to a shortage of improved or tes ted materials at this stage in the cowpea improvement program, it was not possible to find a variety that possessed all characteristics with farmers preferred. However, scientists are now in a better position to seek/develop suitable cowpea types. Meanwhile farmers will have to make do with ZVu 83 which some said took too long to cook. References Mitti, G. 1983 / 84. Informal survey of the cropping systems 01 Chip.til South and Katete districts. Adaptive Research Planning Team-E.1Stem Province(ARPT -EP) Annual Report, 1983/84. Singh, S.R., and K.O.Rachie. 1985. Cowpea Research Production and Utilization. UTA/World Cowpea Research Conference.

Discussion Q. Mligo: You ranked one of yo ur test materials high partly because ithad large seed size. However, you record small seed size as one of the complaints against it. Is the material small- or large-seeded?


A. Mulil.-Mitti: I am sorry for the confusion. The material is large seeded. Q. Dolo: Inany given year, testing was limited to about five participating farmers. Is

it possible to enlarge the sample size by involving more people-possibly neighborsin the evaluation process? A. Mulila-Mitti: Thisisa good suggestion. In fact in future trials, we intend to involve more farrners so as to get a larger sample size. It's also a good idea to invite the neighboring farmers to participate in the assessment of improved cultivars. Q. Matilo: Do the farmers consume the green pods? The lack of grain yield could have been due to pods having been eaten green.

A. Mulila-Mitti: Farmers do consume green pods, however, the lack of grain yield in the local variety is due to its late maturity. At the end of the season when the introduced varieties were being harvested, the local variety was only just starting to flower. Q. Nleya: 11 seems as if the major complaint from the farmers was the that the varieties are hard to cook, and all varieties seemed to get this complaint. Which variety were the farmers comparing these varieties to?

A. Mulila-Mitti: lam notquite sure. I suspect that the farmers compared the cooking time with that of their local varieties. Q_ Dashiell: Were any chemicals used for insect control?

A. Mulila-Mitti: No chemicals were used to control insect pests. Actually, no inputs were used in these trials.


Effect of Plant Density and TIme of Planting on the Grain Yield of Cowpea J.e. Musanya

A two-year study was carried out at the Mseke:ra Research Station" Zambia. to determine the moot productive spaciI1g and the best time for planting cowpea variety 1VX-4654-44E. A spUt路plot experimmW desigt1. with two factors replicated three times, was used. Better yields wereobtainedal highdensity early planting Plots with 90,000111,000planls/ha gnve16%higher grain yield than those with 55,000 planls/ ha. Delay in planting from 30 December 10 13 January resulted in a IS% decreaso> in grain yield and a furthetdelay in the plantingda,.,o 28 January resulted in a 92% de<:reaSe in grain yield. The large reduction in yield due to late planting was main1y because of severe incidence 01 cowpea aphid.borne mosaic virus (CAMV) and adult inse<lleaf eaters (Alc:idaUs and Systoles). CAMVand insect scor.. were significanUy higher.' the last date of planting compared to the earlier dates.

Cowpea is the third mostimportantfood legume crop in Zambia after groundnutsand beans (Kannaiyan et al. 1989). The crop can be grown throughout the whole country but the greatest potential exists in the Westem, Southern, Lusaka, Central, and in parts of Eastem provinces. The grains and leaves of cowpea are an important source of cheap dietary protein to many resource-poor farmers who cannot afford the expensive animal proteins. However, up till n ow, cowpea has been largely a subsistence crop and is predominantly grown as an intercrop (Tamala Tonga 1989). Cowpea types mostly grown in these intercrop systems are the local trailing/ prostrate land races which have a poor yield potential but are good leaf producers (Kannaiyan et al. 1985; 1989, Reddy et al. 1989) . Present research efforts are aimed at providing cultivars that can produce leaves and at the same time give reasonable grain yields. However, a lot of work still has to bedone to obtain information onoptimumspacingand the most ideal timeofplanting. lbis study evaluates the effect of plant density and time of planting on TVx-4654-44E, a promising cowpea variety introduced from IITA.

Materials and methods Experimentd site on cowpea was conducted at the Msekera Regional Research A field eXR"riment 0 0 Station (13 39'5,32 34'E) from 30 December 1985 to 8 April 1986 and from 30 December 1986 to 20 April 1987. Msekera is located 12 km west of Chipata town. Soil type at the experimental site is sandy路day loam of moderate fertility with a pH range of5.1 to 5.8 (CaCl2). CI in'late bfMsekera is tropical continental with an average annual rainfall of 1092 mm (Commissaris 1974). Total rainfall throughout the experimental period was 1129.5 mm during the first season and 923.6 mm the following season.


Treatments and experimental design Cowpea variety TVX 465444E was used in the experiment. A split-plot design wi Itt two factors replicated three times was used. Tim! of planting (TOP) was used as a factor in the main plots and spacing was used in the subplots. There were three TOPs at fortnightly intervals. Three inter-row spacings of 45, 60, and 75 cm together with three intra-row spacings of 15, 20, and 25 em were used (nine spacing combinations in all). Crop management Disc ploughing and harrowing were performed on the experimental site and D' compound fertilizer (l()% N, 20% P205, 10% K20, l~o 5) at the rate of 200 kg/ha was broadcast by hand before making ridges and planting. Planting was carried out on 30 December, and 13 and 28 January. After planting,only shallow cultivation was done between ridges using hand hoes. All weeds, especially those growing at the base of the plants, were pulled out by hand. 1here was no supplementary irrigation as the experiment depended on rainfall. Harvesting took place when pods changed from green to brown and this differed depending on TOP: 25 March for 1st TOP, 13 April for the 2nd TOP, and 28 April for the 3rd TOP. Ouring the following season, harvesting took place on Z7 March for the lst TOP, 9 April for the 2nd TOP, and 20 April for the 3rd TOP. Indices recorded in the experiment were stands at harvest, pod and grain yields, number of pods per plant, 100 seed weight, incidences of insects (leaf hole score [1- 9 scale) and diseases (CAMV score [1-9 scale)).

Results and discussion During the first season (1985/86), the number of plant stands recorded for the three TOPs was close to target levels. Stand densities obtained from various spacing combinations were different from what was expected, and these differences were not specific to any particular TOP. Grain yield was verydoselycorrelated to the numbers ofstands obtained in the 1st and 2nd TOP. Grain yield of the 3rd TOP was extremely low when compared to the 1st and 2nd TOPs (Table 1). In the 2nd season (1986/87), the number of stands recorded were dose to target levels but this time only in the 1st TOP and not in the 2nd and 3rd. Grain yields obtained were again high in the 1st and 2nd TOP but not the3rd (Table 2). Although the yields were high in the first two TOPs they did not follow the trends of the stands as in the first season. A positive correlation was found between stands and grain yield in the lsITOP of the first season only. In the 1st TOP of the first season, high stands were associated with high grain yields (r = 0.83). The relationship between the stands and yield was low (r = 0.5--0.6) for the other TOPs. Although there was no relatiooship between nurroerof the stands and yield obtained, it was obvious that stands /rom densities of 90,000 to 100,000 plants/ha, which were from spacings of (45 cm x 15 cm, 45 cm x 20 em, and 60 em x 15 em), gave yields of over 900 kg/ha especially in the first season, with an exception of the 3rd TOP (Table 1). 1he yields obtained in the second season especially those coming from the 1st and 2nd TOPs were all high irrespective of the number of stands 54

except again in the 3rd TOP (Table 2). This rendered data interpretation difficult. However, thecornblnations which gave high yields during the lstseason were among those that gave high yields. Due to the results obtained during the second season, the experiment did not succeed in bringing out clearly the advantages of planting the crop at any specific spacing for high yields. However, results of the first season and the performance of the crop during the second season suggest that the use of populations in the range of 90,000 to 100,000 plants/ha would assure high yields. The experiment did, however, bring out the advantages of planting cowpea early in the season. This was shown by the consistent low yields obtained in the 3rd TOP of both seasons. Table 1. Effed of time and spacing on the yi eld of TVx-4654-44E during the 19851 86 season Sporing (em)

TorgetPop. ('OOO/ha)

Achieved Pop. (,OOO/ha)

1 4Sx 15 4Sx20 4Sx25 60x 15 6Ox20 6Ox25 75x 15 75x20 75x25


67 89 67 53

113 83 69 84 65 55





89 111 83

C.Y ('Yo) S.E. (±) Spacing S.E. (±)TOP Sig (P5O/.) Spacing Sig (P5%) TOP Sig (P5%) Spacing X TOP

148 112




142 141 109 141 91 88 109 106 83 B:2 64 69 89 83 67 67 54 53 90 89 2.8 0.8 1.0

Crain Yield (kg/ha)

Insect Score (1-9 Scale)

(1-9 Scale)











1027 967 789 903 756

950 674 736 670 759

65 82

3 3 4 3 3 4 3 3 4

3 4 4 4 4 4 4 4 4

5 5 5 4 5 6 6 5 6

2 2 2

3 3 3

5 7 6 7 6 6 6 7 7

4 5 15.0 0.2 0.3 ns


821 642 840 558

850 714 590 724

68 66 76 38 58 63 46

838 714 63 242 43 57










3 2 3 4 3

2 1 2 2

3 6 23.0 0.3 0 .3


ns ns

uck of response to spacing cannot be explained with the present data due to very high fluctuations in the achieved plant stands. 1he low yields of the3rd TOP in all seasons is explained by the presence of CAMV and Insects. CAMV affects the plants by inducing leaf distortion and stunted growth which severely affects yields (Thottappilly and RosseI1985). The incidenceofCAMV has been observed to be on the increase in cowpea planted late (from February to end of March) at Msekera (Food Legume Annual Report 1990). This increase is suspected to be due to the presence of large populations of aphids which are vectors of CAMV (Thottappillyand RosseI1985). The use of insecticides to control aphids significantly reduced the incidence of aphid-spread CAMV at Msekera (Food Legume Annual Report 1990). Insects on the other hand (Alcidodes and Systates) fed on the foliage thereby reducing the leaf area necessary for optimum photosynthesis. The scores


obtained lor CAMV and insects were higher in the last TOP 01 all seasons than for the

other times. Table 2. Effect of time and .pacing on the yield of TVX-46S..44E during the 19861 87aeaaoft Spacing em

Target Pop. ('OOO/ha)

45x 15 45x20 45x25 60x 15 6Ox20 6Ox25 75x 15 75x20 75x25

148 111

89 111 &3 67 89 67 53

Achieved Pop. ('OOO/ha) 1 2 3 142 112 89 100 91

66 86 65 54

91 Mean 90 28.6 C.V (%) 65 78.7 S.E. (±) Spacing 1.8 2.4 107.0 S.E. {±)TOP ns Sig (Ps%) Spacing-' Sig (P5O/.) TOP Sig (Ps%) Spacing" TOP ..

123 97

88 93 72 62

80 63 52

135 102 87 91 75 63 78 62 52

81 &3 33.6 0.4 0.4

Crain Yield (kg/ha) 2 3 1 1215 1466 67 1087 1542 60 1322 1700 84 988 1092 75 1257 1155 140 1161 1321 48 75 965 1109 1242 909 &3 1116 955 99 81



CAMVScore (1-9 Scale) 3

1 2 1 2 2 1 1 1 1

2 4 3 4 3 4 3 3 3 3





6 7

S 6 6 5 6 6 6

.. .. .. ns ns


Conclusion The results suggest that cowpea should n ot be grown beyond the second week 01 January in Msekera . Should it be imperative to h ave them grown after that period, then some lonn of control lor aphids should be considered. This consideration is on the basis that aphids playa role in the transmission of CAMV. In case of spacing, it appears that anyone 01 the following spacings w ould give reasonable yields: 45 em x 15 em, 45 cm x 20 em, and 60 cm x 15 em.

References Commissaris, A.T.L,M. 1974. Detailed soil survey, Mseke£a Research Station. Soil Survey No 19, Soil SurveyUnit, Land Use Services Division, Ministry of Rural Development, Lusaka. Food Legume Annual Report 1989-1990. Ministry of Agriculture, Department of Agriculture, Research Branch. Kannaiyan, J., MN. Mbewe, D,C. Creenberg, H.e. Haciwa, N.S. Irvin, and P.H. Sohati . 1985. Cowpea production and research in Zambia , Tropical Grain Legume Bulletin 32: 144-148. Kanna iyan,)"J,M. Mulila-Mitti, D.e.Creenberg, .nd MN. Mbewe. 1989. Varietal improvement of bean and cowpea in Zambia . Pages 86-I03 in Proceedingsof the National Workshop on


Food Legumes Research and Improvement in Zombia, !l-l1 Mar 1988. Mfuwe, Eastern Province, Zombia. Ministry of Agriculture, Ch ipata, Zambia. Reddy, M.S., K. Kanenga, I.e. Musanya, and J. Kannaiyan. 1989. Agronomic and cropping systems research involving grain legumes. Pages 104-116 in Proceedings of the National Workshop on Food Legume Research and Improvement in Zambia, Mfuwe, Eastern Province, Zambia. Ministry of Agriculture, Chipata, Zambia. Tamala Tonga. 1989. Production of certified grain legumes in Zambia : problems and prospects. Pages 188-191 in Proceedings of the National Workshop on Food Legume Research and Improvement in Zambia, Mfuwe, Eastern Province, Zambia. Ministry of Agriculture, Chipata, Zambia . Thottappilly, G., and H.W. Rossel. 1985. Worldwide occurrence and distribution of virus diseases. Pages 155-171 in Cowpea research, production and utilization, edited by S.R. Singh and K.O. Rachie. John Wiley and Sons Ltd. USA. Summerfield, R.J., J.s. Pate, E.H . Roberts, and H.e. Wien. 1985. The physiology of cowpeas. Pages 65-101 in Cowpea research, production and utilization, edited by S.R. Singh and K.O. Rachie. John Wiley and Sons Ltd. USA .

Discussion Q_ My.k.: If you had waterlogging problems during the first year, I would expect higher yields for later-planted cowpea. Why was this not so?

A. Mus.nya: Rainfall at Msekera does not follow a normal pattern_ Sometimes there is more rain at a certain period of the year as was the case during this triaL Q. Amable: As a follow-up to Myaka's question, what about the effects of rainfall distribution on yields obtained during the first season?

A_ Musanya: It is just unfortunate that rainfall data was not provided. Otherwise our experience has it that rainfall at Msekera starts sometime in Novembercontinues until around Christmas when a dry spell sets in until the end of January. Rain returns from around mid-February and continues until March, SO wedo not have well-distributed rainfall. Q, Dashiell: Did you suggest that varieties resistant to aphids should bea benefit as

this might reduce the incidence of CAMV? A_ Musanya: No _ I said if planting has to take place after the second week of January, then some form of aphid control, for example pesticide application, would reduce the spread of CAMV via the aphid vector. CAMV is also seedborne, however, in the absence of aphids the disease would be confined to a few plants.


Cowpeas: A Basic Component of a Multiple Crop System In the District of Marracuene, Maputo, Mozambique J. Arias-Fandino The study describes the most common cropping syslem in the Distrid of Marracuene. This cropping system is representative of cowpea production in Mozambique. 1be most common cwtivars of cowpea found here are those that are prostrate, photoperiod.路scns.itiIlC, leafy, and produce big grains of good quality. It is suggested that thesevarietiesshouJd bemaintained, purified. identified. multiplied. evalUOlteci. and promoted.

Cowpea is the second most important legume in Mozambique. It is cultivated on over 100,000 ha/year and is surpassed only by groundnuts which is cultivated on lSO,1XXl ha/year. Marracuene is located 25 km north of Maputo. Most of its soils are highly sandy and it has a strong peasant farmer population. For these and several other reasons, it could be considered as an average representative sample of what cowpea production in Mozambique is like.

Cropping system In Marracuene, intercropping is very common. Plots planted to a mixture of maize,

groundnut, and cowpeas are seen. But far more common are mixtures comprising water melon, cabbage, pigeon pea, and jacinth beans. No fertilizers and/or complemetary inputs are used, and plots are cleared manually and are irrigated with rain water.

Cowpea cultivars It is thought that most of the varieties grown are local. However, there is a feeling that some cultivars could have been introduced from abroad . One definitivecharacteristic of all these cultivars is that they are a complex mixture of genotypes . The local cowpeas cultivars are mixtures of at least five large-seeded genotypes, and are cream, violet, black, and speckled gray. All are postrate and photoperiod sensitive. They produce a lot of leaves and also high grain yields. They also protect the soil against eolian erosion which is very heavy in the area.

Growing cycle Planting date Seeds are sown when the first rains of the year appear (early October). The cultivars used and the planting dates are the sound core of the system, because they allow a long-lasting leaf production season until the short day and very dry season arrive allowing for optimal grain development. 58

When early-and medium-maturing varieties are planted in the early months, their grain production is ruined by the normal rains of December, January, and February.

Vegetative cycle Growth period lasts for about three months (November-January). During this period, other crops such as maize and groundn u ts complete their life cycle, producing rather poor yields. However: as early as middle November, and throughout February, these local varieties start and keep producing a fairly large number of leaves. TIlese leaves are of excellent quality, by local standards, for cooking. Flowering Because of the sensitivity of these genotypes to changes in day-length, they start flowering by February when day length becomes shorter. Harvesting By March, the grain is ready for harvesting. Because the weather is very dry at this time, the quality of the grain is excellent. Valuable contribution of the leaves of local cowpea Under this Marracuene system of cowpea production, which local farmers, technicians, and agronomist say is representative of cowpea production in Mozambique, and other than the production of grain, there is an enomous production of leaves, because this food is produced: (a) as the first food after the long dry season (March through September) and (b) when there is no field crop in production. This means that cowpea leaf production in Mozambique is not only worthy, in terms of local human food needs, but also serves as food during the hungry period when almost no food is available.

Conclusion It is concluded that (i) Cowpea is a highly decisive component with local peasants in Mozambique. (ii) favorite local cuItivars are those of the postrate and photoperiod-sensitive genotypes. (iii) Cowpea leaf production. Production of cowpea leaf is important because the leaves are available as food during the season and yields are high. futhermore, it has a wide market.


Cotton and Cowpea Relay Intercropping: Preliminary Results on its Economics and Effect on Some Agronomic Characters of Both Crops F.A. Myaka and rCB. Kabissa

ThepoientloJ clreloyinten:ropping cotton with cowpea was investigated at the nongo Agricultural Reoearch _lute during the 1990 aopping.....,... The y;eld perforII\ar\a! of cowpea and _on .... eJWnined unde< two planting pottenw and two opnymg ...gimes (sprayed and noosprayed) in â&#x20AC;˘ foctoriol arrangement using a randomized complete block deoign. Und.,. inteio<>pping, spraymg incr.....-d seed cotton yield. cowpea grain y;eld. and intercropping yield aclvontage by 63%. 130%. and 101% respectively while under sole aopping. spraying increased seed cotton yield and cowpea grain yield by 76'4 and 94'4 respectively. Marginal analysis showed that fanners con expect to gain inretum for their lnve5bnent when they decide lochange from growing cotton as III sole crop to relay inf~pping cotton with cowpea ,

In the eastern cotton growing areas (ECGA) of Tanzania, up to six insecticide applications are needed foroplimal yieldof cotton (Gossypium hirsutum) (Kabissa 1989). Cowpea (Vigna unguiculata) is also an important crop in these areas and farmers are advised to spray this crop at least twice starting from flower bud initiation (priceet aI. 1982). However due to rising insecticide and application costs in recent times, profit margins realized from these crops have become lower. One way of optimizing profit margins would be to intercrop cotton and cowpea thereby increasing returns while maintaining production costs. Unfortunately, there have been relatively few studies on intercropping of cotton with cowpea. Beltrao et al. (1984, 1986a, and 1986b) evaluated several cowpea cultivars for suitability for intercropping with cotton. Their studies revealed that cowpea cultivars differ in their competitiveness wi th cotton. In one of their other studies,one of the cowpea cultivars (Tv 1836-(135) usedgavea 33.4% higher net income than cotton alone, and had a Ianduse efficiency ratio of 1 ~56 (Beltrao etal. 1986). Another cowpea cultivar (Pitiuba) was the most competitive and reduced cotton yield by 33%. In their other studies (Beltrao et aI.1984; 19863) intercropping significantly reduced seed cotton yield. Seshadri and Natarajan (1989) also studied the effect on yield of intercropping cowpea with cotton. In their study cowpea significantly increased returns compared with cotton in pure

stand. However there have been no studies on cowpea/cotton intercropping with reference to the effect on cowpea of pesticide use on cotton. By intercropping cotton with cowpea in such a way that the latter benefits from sprays applied on cotton, returns may be optimized by increased yield from cowpea. The present study was therefore undertaken to assess the effect of intercroppingcotton with cowpea on yield and returns from both crops.


Materials and methods The experiment was conducted at the nonga Research Station during the 1990 cropping 5eiOOI\. This station is situated 16° 46' south 01 the equator and 37" 02' east 01 the Greenwich Meridian at an altitude of 506 m above sea level. The site on which the experiment was laid out is characterized by sandy-clay loam soils. The experiment, comprising two planting patterns and two spray regimes, was laid out as a 2 x 2 factorial with four replications in a randomized complete block design. Cotton cultivar I174 was sown on 22 Februrary 1990 using a spacing of 0.9 m between rows and 0.3 m between plants within a row. Cowpea cultivar Vuli·1 was sown on 8 March 1990, two weeks after sowing cotton, using spacing as indicated in Table 1. Vuli·l is an early maturingcultivar with a deterrninategrowthhabit Cowpea was sown either iii single or double rows alternating with single rows of cotton. Treatments comprising cowpea and cotton as sole crops were also included in the experiment. Other experimental and management details are indicated in Table 1. Table 1. Experimental and management detail. for the trial

Alternate single row

Single row cotton alternating with

Sole crop

double row cowpea

Target population (plants/ha) Plot size (m) Interow distance (m)

Interplant distance (m~ Net area (m )


Cowpea Cotton











' .5x6

4.5x 6


4.5 x6















(between Pairs) 0.30 between rows within pair) 0.30





• Number of plant/hill - One and two plants for cotton and cowpea respectively Approximately 56 days after germination of cotton, cypermetrin 9'Yo ED was applied by electrodyne at 31.6 ha- 1 using 2.7 m track spacing to sole cotton and to cotton intercropped with cowpea in designated plots. Similarly some plots comprising sole cotton and cowpea as well as intercropped cotton were nonsprayed. Insecticide was applied six times at lO-day intervals. In view of the relative planting dates of the two crops, intercropped cowpea was expected to benefit only from two of the sprays applied on cotton. At crop maturity, plots were harvested and cowpea grain yield and seed cotton yield recorded. Land equivalent ratio (LER) was also calculated as described by Willey 61

and Osiru (1972). Data were then subjected to analysis of variance. Economic analysis was also done following the procedure described by ClMMYT (1988).

Results Considerable rainfall was received between late February and April. However starting from May therewas little or no rainfall. Temperature feU slightly as the season progressed (Table 2). Planting pattern had n o effect on seed cotton and cowpea grain yield as well as intercrop yield advantage (P > 0.05). Also interaction effects on all the three variables were not significant (P > 0.05). For both crops, higher yields and yield advantages were obtained where cotton was sprayed. Spraying increased seed cotton yield by 63% and 76% under intercropping and monoculture conditions respectively. Similarly, spraying increased cowpea yield by 130% and 94% underintercropping and monoculture conditions respectively. Spraying cotton also increased the yield advantage of the intercrop by l00'Yo (Table 3).

Table 2. Rainfall (mm) in ten-day total. and daily mean temperatures (0C) for the 1990 growing season at Jlonga

Dailymean Month



Temperature (째C)


1- 10 11-20 21-28

11.0 76.1 60.8

28.1 26.8 25:7


1-10 11-20 21-30

53.; 130.7 92.2

25.1 26.1 29.2


1-10 11-20 21-30

111.3 24.9 49.6

26.2 25.5 26.0


1-10 11- 20 21-31

16.1 3.6 6.0

25.5 24.5 23.6


0.0 0.0 0.0

23.3 23.3 23.3

21.5 0.0 1.7

21 .8 21.8


11- 20 21-30

July 1-10

0.0 11-20 21- 31


Relay intercropping cowpea with sprayed cotton gave higher returns than sprayed sole cowpea or cotton (Table 4). Single rows of cotton relay intercropped with double rows of cowpea gave 10% more returns than cotton relay intercropped in alternate single rows of cowpea (Table 4). Similarly relay intercropping single rows of cotton with double rows of cowpea under sprayed conditions gave 41% and 64% more returns than sole cotton and cowpea respectively. On the other hand, relay intercropping cotton and cowpea in alternate single rows under sprayed conditions produced 28% and 49% more returns than monOcropped cotton and cowpea respectively (Table 4). Marginal analysis showed a marginal rate of returns (MRR) of 5560/. by moving from monocrop cotton to cotton/cowpea relay intercropping under spraying conditions (Table 5). Table 3- Effect of apraying regime. and planting pattern on seed-c:otton yield, cowpea grain yield, and land equivalent ratio in the cotton/cowpea relay inlercropping Seed Colton Yield (kg/ha) P1antlng Pattern

Cowpea Crain Yield (kg/ha)

Sprayed Unsprayed Mean

Sprayed Unsprayed M ....

Land Equivalent Ratio Sprayed U""""yed Mean

Pattern 1




















1690 a 1752.


754. 17111.

327b 882b

1.41 a


Pattern meandif. Mean Sole Crop

242 ISffia


150 NS

0.22 NS

Patlem J â&#x20AC;˘ Alterna .. single rows, Patlem 2 K Single row cotton alternating with double rows of cowpea Paired .......... for each fa<ta< in the same row followed by diffeftnt letten ...... snificanUy diIferent at 5% level.

Discussion In the present study cotton was considered as the main crop in view of its importance .5 a traditional cash crop in the ECGA, therefore its population density was maintained at a recommended density of 37,000 plants ha-1 across all planting patterns (Table 1). However total returns from yields of intercrops under sprayed conditions were higher than yield of respective sole crops (Table 4). This is in agreement with results reported by Beltrao et a!. (19800) and Seshadri and Natarajan (1989) that cowpea significantly increased retumscompared with cotton in pure stand. Although Insectpestdata were not recorded in the present study it might be possible that, when not sprayed, the ability of cotton to recurperate after Insect damage is reduced under intercropping than when grown as a sole crop.


Table.. Monetary retums from different trutment combinations in the cottonl cowpea relay intercropping Returns

Monetary (Tsh/ha)

Treatment Intercrop (single row cowpea) with spraying on cotton Intercrop (single row cowpea) with no spraying Intercrop (double row cowpea) with spray on cotton Intercrop (double row cowpea) with no spray Monocrop cowpea sprayed Monocrop cowpea unsprayed Monocrop cotton sprayed Monocrop cotton unsprayed LSD (0.01)

158817 7fin7

175425 112311

106750 55094 124286

107033 6690

Cotton: Cowpea price ratio -1 : 0.88 228.8 Tsh = 1 US 5 (Mid-August 1991)

On the other hand, yieldsof cowpea intercropped with sprayed cotton were lower than yields from sprayed sole cowpea. This was probably due to reduced plant population of cowpea under intercropping (Tables 1 and 3) as well as competition for resources with cotton. Percentage component crop population ratios (cotton: cowpea) in relation to their respective sole crops were)oo: 78 and )00: 56 in intercrops with double row cowpea and single row cowpea respectively. This offers a possibility of increasing thecowpea yield in this intercropping system by manipulating the planting pattern and increasing the cowpea population density in the intercrop. Therefore more work regarding the effects of various planting patterns and cowpea population on yield is need. Betrao et al. (1984 and 1986a) reported that intercropping cowpea with cotton significantly reduced seed cotton yield. Their observation is in sharp contrast to results presented hen. where seed cotton yields under sprayed conditions were not significantly affected by intercropping. Such conflicting results seem to be due to differences in types of varieties used during the studies. In our study, a determinate and early-maturing cowpea cultivar was used. This type of cultivar could be less competitive to component crops in an intercropping situation than an indeterminate and late-maturing cultivar. Unfortunately Betrao et al. (1984 and 1986a) did not describe the growth habits of the cowpea cultivar used. Results of thepresentstudy show thatintercroppingcotton with cowpea increases returns as a result of additional yield from cowpea and that farmers in the ECGA can expect to gain in return for their investment if they relay intercrop cowpea with cotton (Table 5). It is worth noting that whereas sole cotton gave comparable seed cotton yields under sprayed and nonsprayed conditions, cowpea gave significantly less yield under nons prayed conditions whether or not it was intercropped with cotton. This may in part be due to compensatory growth of cotton following insect attack. This ability is of limited value in early-maturingcultivars with determinate growth such as Vuli-l. 64

Table 5. Partial budget andmarginalanalysesforthe cotton/cowpea relayintercrop and sole crop cotton Intercrop in Sole crop cotton

alternate single rows


(sprayed cotton)

Average seed cotton yield (kg/hal Average cowpea yield (kg/hal


Adjusted seed cotton yield (kg/hal Adjusted cowpea yield (kg/ha l



1473 614

Gross field benefit for cotton (I'sh/ha) Gross field benefit for cowpea (Tsh/ha)



Total gross field benefit (Tsh/ha)



Cost of cowpea seed (I'sh/ha) Cost of labor for planting cowpea (Tsh/ha)

o o

700 3750

Total cost that vary (Tsh/ha)









Net benefit MRR(%)

1637 682


10% Yield adjustment 228.8 Tshs = 1 US S (Mid-August 1991)

The present study showed that some level of pest control is essential for optimal yields of both cotton and cowpea. However, the study did not examine how target arthropod pests respond to control operations and how this could be enhanced given a multiplicity of pesticides, pesticide formulations, and applicators. These and other aspects of cotton and cowpea intercropping need to be investigated, bearing in mind that cowpea is grown mainly for food.

References De. Beltrao, N.EM., O.M.P. Azevedo, L.B. Oa Nobrega, D.C. Viera, I.R. Crisostomo, C.T. Bandeira, and A. De Viera. 1984. New methods of intercropping fo r north east Brazil using herbaceous cotton and cowpea . Communicado Tecnico Centro National de Resquisa do

Algodao No. 25. 10 pp.

De. Beltrao. N.E. 19860. Evaluation for oowpea cultivars under intercropping system with annual cotton . Resquisa Agropecuaria Brasileira 21(11): 1147-1153.


De. Beltrao, N.E. M., LB. Da Nobrega, DM.P Azeredo, OJ. De Viera, and S.R. Crisostomo. 1986b. Inlftcropping of upland coHon with cowpea cultivars. Resquisa Agropecuaria Brasileira 21(3): 271-284. CIMMYr. 1988. From agronomic data to farmer recommendation. An economic training manual. Page 79 in Agronomic data to farmer recommendation. Completely revised edition. Mexico, D.F. Kabissa, f.C.B. 1989. Evaluation of damage thresholds for insecticidal control of He1iroverpa .rmiger. (Hubner) (Lepdoptera: Noctuidae) on cotton in eastern Tanzania. Bulletin of Entomological Research 79: 95-$. Price, M, F. Machange, and lA. Assenga. 1982. Growing cowpea in the Morogoro Region. Tanzania Ministry of Agriculture. 46 pp. Seshadri, V., and IC.Natarajan. 1989. Effectofdifferentintercropson growth and yield of upland cotton (Gossypium hirsutum) and their economics, Indian Joumal of Agricultural Sciences 54(4): 227- 230. Willey, R. W., and D.s .O.Osiru, 1972. Studies on mixturesofmaize and beans (Phas<olus vulgaris) with particular reference to plant population, Joumal of Agricultural Science 79: 19-529.

Discussion Q. Comments. Dolo: In your partial budget and marginal analyses, you may wish to consider including an additional element, namely the added costs associated with harvesting and processing the cowpea in the intercrop. One may also wish to look at labor costs associated with weeding should the cotton-cowpea intercrop require more labor relative to the cotton monocrop. A. Myaka: Harvesting and processing costs are taken careof in calculation of the field price for cowpea . The field price is the market price minus the cost of all operations beginning with harvesting and up to the time the crop reaches the marketplace. These would include harvesting. threshing. transportation, etc. The field price is then used to calculate the gross field benefit for cowpea. On the second comment, the point is taken. However, determining costs of weeding an intercrop as opposed to a monocrop is complicated as far as labor is concerned because one may claIm also, that the weeding frequency is less in the intercrop thus resulting in lower labor costs. Q. Uriyo: You stated that future research in this study would include safety. Could you elaborate on this?

A. Myaka: Because the rates of pesticides used in this trial were recommendations for cotton, it is important to consider the residues which mightaccumulateinco,w pea seed and the implications for human safety.

Follow-up. Uriyo: In cotton, you are interested in both the lint and cottonseed. Edible oil and cake are obtained from the latter. Do you feel that til!! dosages of pesticides sprayed on cotton are likely to result in abnormally higher rates of residue accumulation in cowpea and become harmful to the people who would eat this cowpea? A. MyAko: It is possible that the pesticides sprayed on cotton might result in abnormally higher accumulation in edible cowpea seed thus posing a danger to humans. This is why we propose to study the safety aspects.


Comment. Amable: The pesticide used for cotton in your trial, cypermethrin, is the same as rerommended for cowpea. Safety ronsiderations would probably involve timing. for instance where cowpea leaves are consumed and sprays are made at times when leaves could be picked. Rates of application could also be a factor where these differ for cowpea and cotton, tho this would be unlikely as the range of pests is similar in both crops. A. Myaka: It might sometimes be misleading to generalize that pests of these crops are similar. Some important pests of cowpea are not important on cotton and vice versa; also some species are different, for example the aphid species found in cotton is different from that found in cowpea. Q. Kwapata: Did you observe any differences in growth pattern or plant and fruit morphology associated with dosage and frequent spraying of cotton chemicals?

A. Myaka: Unfortunately such observations were not made. However, cowpea planted two weeks after cotton benefitted from the first two cotton sprays and was harvested before the third spray. This follows the recommended spray regime for cowpea in Tanzania, therefore, one would probably not expect s uch an effect to occur.


Recent Research on the Cropping System with Cowpeas in Zambia J.e. Musanya Thepaperpr-..ts. briefhistowyofcowpea .....,ch, reviews studieo "" the cropping systems with cowpea in Zombi.. and concludeo that inteKropping cowpea with cereals appears more promising than relay croppir1g.

Cowpea is an important food crop which is grown and utilized all over Zambia. Both its leaves and grains provide good dietary protein to many rural families of the country (Kannaiyan et al. 1989; Reddy et al. 1989). While cowpea has much tooUer in terms of cheap vegetable proteins, it is largely a subsistence crop and is predominantly grown as an intercrop \I'amala Tonga 1989). In the local rommunities where cowpea is grown, people prefer types that are profuse leaf producers to those that produce high grain yields (Kannaiyan et al. 1985; Reddy et a!. 1989). Outlined in this paper are some of the agronomic investigations carried out to augment the breeder's efforts in the provision of suitable varieties.

Brief history Sustained research on cowpea in Zambia dates back to 1960 when screening of some cowpea lines for scab and ascochyta was doneat Misarnfu innorthem Zambia. No line was found resistant to any of the diseases. Research then shifted from Misarnfu, a high rainfall site, to the low rainfall sites in the western and southern provinces. 'The program at these sites concentrated on screening for drought tolerance. It was from this work that, around 1970, the first cowpea variety, New Era Gray, was released. Further work up to early 1980 resulted in the release of two lITA-<leveloped lines, lVX 3(}.IG and lVX 456-01G, renamed Muliana and Shipepo respectively(Kannaiyan et al. 1985). However, according to Tamaia Tonga (1989), Shipepo appears to be the only variety with adequate seed for commercial production. During the period following the development and release of the three cowpea varieties, cowpea research was handled under legumes research. This meant that scientists worked on the crop as a matter of interest and not asa policy. It was not until 1982, with the establishment of the World Bank-funded Eastern Province Agricultural Development Project, that a specific research team was formed which included cowpea in its terms of reference. This team known as the Grain Legumes Research Team had a mandate to conduct research on beans, cowpeas, bambara nuts, pigeon peas, and minor legumes. Work on breeding and plant protection was well under way in 1982 while agronomic research did not begin until 1985 (Reddy et al. 1989). 'The agronomic investigations that were initialed dealt with the development of appropriate recommendations for all promising new varieties from the breeders.


Materials and methods These trials were initiated in order 10 assess the performance of cowpea under different cropping situation as cowpea is intercropped most of the time (Kannaiyan et a!. 1985, Reddy et al. 1989). Cowpea was intercropped with maize and sorghum in areas representing plateau and valley conditions in Zambia . Maize is a national staple and does well on the plateau while sorghum is a major staple in the valley areas and performs well undet v,!lIey conditions. For maize, a 1 : 1, 2 : 1, and 3 : 1 row arrangement with plant population fluctuating between 50 and 100% for maize and between 25 and 67% for cowpea was used (Table 1). For sorghum, a constant row arrangement of 1 : 1 with a sorghum population of between 50 and 1000/. and cowpea population of between 50 and 75% was used (Table 2).

Results In the maize/cowpea intercrops, the achieved plant population of both component crops came close 10 target levels. However, with the increase innumberof maize rows, cowpea yields in thedifferent intercroppingcombinations were reduced Significantly. In the 1 : 1 combination with 50% maize and 50% cowpea population, 439 kg/ha of cowpea was obtained compared to 197kg/ha in the 3 : 1 combination with 100010 maize population and 60% cowpea population. Maize yields increased with increase in the number of rows for maize, i.e, from 3,368 kg/ha in the 3 : 1 comb ina tion wi th 50% maize and 50% cowpea to 6,598 kg/ha in the 3 : 1 combination with 1000/. maize population and 66% cowpea population (Table 1). In the sorghum/cowpea intercrop, the performance of cowpea was quite good with achieved plant population being very close to target levels. Sorghum performed badly mostly due to poor germination and shootfly problems. Cowpea achieved the highest yield of 1,676 kg/ha when sole cropped, which reduced to 864 kg/ha in the 1 : 1 row arrangement with 100% sorghum population and 50% cowpea. Sorghum yields, although highest under sole croppings (483 kg/ ha), were generally poor due to the low plant population (Table 2). With relay cropping the yields achieved were very low especially for cowpea. In the maize/cowpea relay where cowpea was planted 3 weeks after maize with the 2 : 1 row arrangement (1000/. maize population and 66% cowpea population), only 45 kg l ha of cowpea was obtained (Table 1). This low yield was as a result of cowpea aphidborne mosaic virus (CAMV) which severely affected the relay-planted cowpea. Results of cowpea relay-planted with sorghum showed cowpea to be equally affected by CAMV. Cowpea was again planted three weeks after sorghum. Low yields of cowpea were recorded from relay-planted plots compared to plots where cowpea was planted at the same time as sorghum. The cowpea that was planted at the same time as sorghum in the intercrop gave 52 to 66% of sole cowpea yields whereas cowpea relay planted three weeks after sorghum produced 32 to 38% of sole cowpea yields. With sorghum, more than 60% of sole sorghum grain yields were achieved in all the intercrop combinations (Table 2). The yields of sorghum were however generally poor because of low plant population and a long dry spell towards the end of the crop season.


Tablet. Effects of intereropping and relay planting of cowpea in maize


Achieved cowpea stand



Cowpea yield (kg/ ha)

Cowpea P(LER)



SoleM I, 1

Achieved maize stand

Maize Yiold

Maize P(LER)

Total LER

(kg / ha) 1 .00




1 .00 1.23













































3 1,000




(SO'I.M , SO'I.C) Ll


(1000/01.\ , 5O%C) 2, 1 (67%M,33%M) 2,1 (loo'Y.M ,67%C) 3,1 (75%M , 25%C) 3,1 (100%M ,66%C) Relay 2, 1

(IOO%M , 66%C) Mean C.Y. ('!'o)


5.E. + LSD(!')


0.46 39.6



0.96 11 .5







0 .27




C = Cowpea. M = Maize, 1: 1 = Ro\,\' arrangment, % = Plant Population

Genotype evaluation Planting same hole. The practice of planting cowpea in the same hole with other crops was examined using a number of cowpea genotypes and maize. The idea behind such a trial was to come up with cowpea genotypes that would give a reasonable grain yield while at the same time not suppress grain yields of maize, Results obtained showed significant differences in .the stands of cowpea and maize. Cowpea stands had the highest population of 60,000 plants/ha coming from KBO and the lowest of 42,000 plants/ ha coming from Muliana. Maize stands on the other hand, were between 15,000 and 22,000 as a result of cowpea, because in sole 70

Table 2. Effects of intercropping and relay planting of cowpea in sorghum Treatment

Cowpea larget stand

Cowpea adtieved stand

Cowpea yield


Sorghum target stand





achieved stand

Sorghum yield (kg/ha)


Solo sorghum Solecowea




1: 1 (50S: 50%C)













Solo cowpea




1: 1 (SO%S; 5O%q

























1: 1 (100'%5; 5O%C)

(Relay cropping)

1: 1 (750/05:


1: 1 (100째/05; 5O%C) Mean

C.V. (%)



S.E. (+-)




11 6






C = Cowpea, M = Maize, 1 : 1 = Row arrangment. %:;::: Plant Population

maize, the population reached 23,000. Grain yield for cowpea was between 340 kg/ ha and 570 kg / ha. In maize, the yields were between 1,520 kg/ ha and 2,950 kg/ha (Table 3). The reduction in both the stand and yield of both component crops demonstrated quite clearly the wastefulness of planting both crops in one hole. Leaf picking. The effect of leaf picking on grain yield was investigated using four cowpea varieties, three of them promising lines (IT 820-872, TVx 3381-02F, and IT 835875) and a released commercial variety (Muliana). Ten der leaves were plucked from these varieties beginning three weeks after emergence to shortly before flowering. The results obtained showed that leaf picking affected the grain-yielding ability of the varieties. This was manifested in the reduction of pods / plant and grain yield. Of the four varieties, the two promising lines (IT 82D-875 and IT 83S-02F) appeared to be much better than Muliana and JVx 338H12F (Table 4). 71







Moiugnin yiSd

.. Maize yield






43.0 58.0 46.0 46.0 46.0 49.0 58.0 48.0 55.0 42.0 54.0 60.0

520 520

15.0 22.0 20.0 21.0 20.0 20.0 22.0 18.0 21 .0 21.0 19.0 22.0 23.0

1520 2310 1940


Table 3. Performance of cowpea genotypes and their effedl on the yield of maize

c_ gonotype

TV" 3381.mF IT83S-872 ZVU83 TV" 4654-«E

wrutewonder Shipepo M66 IT 82D-812 ZVU237 Muliana IT82D-875 )(SO Maize (MM501) CV(%)


530 650 430 390 340 490 490 390 570 520

15.3 11.1


2040 2010 2360

115 3.3


48 3(


30 20

1940 2110




2210 2420 2n0 2950

25 18 8


24 .6


Table 4. Effect of leaf on the grain yield of cowpea Variety

IT 82D-S75 IT 82D-S75 IT 82D-S75 lT82D-S75 TVx3381~2F TVx3381~2F TVx3381~2F TVx3381~2F

IT83S.mF IT 1IJS.{)2F IT 1IJS.{)2F LT 1IJS.{)2F Muliana Muliana Muliana Muliana Mean C .V. ("'oj Sig. variety Sig. Picking Sig. var. " Picking

Piddng Ngime



Grain yield

1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4

SO.6 79.6 792 78.3 82.4 SO.1 78.3 SO.6 81.4 792 80.1 80.9 81.0 78.3 77.0 78.3

15.4 13.1 11 .8 11.8 10.0 8.8

1292 940 918

7.4 7.6 13.1 122 11.8 11 .8 10 .7 8 .1 8 .0 8 .1

390 655 1052 1010 1158 740 910 950

79.7 3.9

10.7 17.3

ns no ns


1 = No picking. 2 • One pick, 3 ~ Two pickings, 4 = Three pickings.


808 651


799 623 843 24.3



References ICaMaiyan, L MN . Mbewe, D.C. Greenberg, H .e. Klciwa, N.S. lrvin, and PH. Sohati. 1985. Cowpea production and research in Zambia. Tropical Grain Legume Bulletin 32: 144-148. ICannaiyan,J., J.M . Mulila-Mitti,D.C. Greenberg, and MN. Mbewe. 1989. Varietal improvement of beans and cowpeas in Zambia . Pages 104-116 in Proceedings of the National Workshop on Food Legume Research and Improvement in Zambia, ~11 Mar 1988. Mfuwe, Eastern Province, Zambia. Department of Agriculture, Ministry of Agriculture, Lusaka, Zambia . Reddy, M.s., K. Kanmga, I.e. Musanya, and J. Kannaiyan. 1989. Agronom ic and cropping systems research involving grain legumes in 21mbia . Pages 86-103 in Proceedings of the National Workshop on Food Legume Research and Improvement in Zambia, 9-11 Mar

1988. Mfuwe, Eastern Prov inco, Zambia. Departmentof Agriculture, Ministry of Agricultu re, Lusa ka, Zambia . Tamala Tonga . 1989. Production of certified seeds of food legumes in Zambia: problems and prospects. Pages 188-191 ill Proceedings of the National Workshop on Food Legume Research and Improvement in Zambia, 9-11 Mar 1988. Mfuwe, Eastern Province, Zambia . Department of Agricuhure. Ministry of Agriculture. Lusaka, Zambia.

Discussion Q. Mligo: Cowpea leaf plucking studies were amducted in Tanzania between 1977 and 1978. Results showed that leaf plucking did not have much adverse effect on grain yield. This is also shown in your present study. However, one of your cowpea varieties gave higher grain yield under a higher leaf plucking regime than under a lower one. What could have been the reason? A. Musanya: One of the replications might have had an extremely low yield, otherwise I do not have a ready explanation for the observation. Q. Nleya: Did the four cowpea varieties used in the leaf picking study have the same growth habit and maturity period?

A. Musanya: Yes, they are all medium-maturing. Q. Dolo: Could you describe a little more, the methods used for leaf picking, in particular, the intensity of defoliation?

A. Musanya: Leaf picking was done once a week beginning two weeks arter erne. gence and the tender second leaf on each plant was picked . Leaf yield data is not shown but during the first picking, SQO...<,()() kg/ha of fresh leaf was taken. This intenSity of picking was reduced in subsequent pickings. Comment. Kwapata: For consistency, one would have hoped to use the same persons to pick the tender leaves of cowpe~. The change in personnel picking leaves and the apparent lack of clearly defined profile of the leaves to be picked introduce some elemenl of inconsistency and make the results obtained misleading and difficult to explain, in terms of the influence of picking on grain yield . Can you elaborate on how this was handled? A. Musanra: The first picking was done by a woman selected because she knew the stage when leaves are preferred for picking. The regular field staff (men) observed her and used her technique as a reference for later pickings.


Q. Sakal Which leaves are considered tender? In tea, for example, a leaf and a bud are picked. Secondly, which leaves are physiologically important in terms of grain yield, on a cowpea plant.

A. MUllanya: We considered a leaf next to a half~eveloped young leaf. These were leaves which had just started to take on the adult color. Your next question will have to be referred to a plant physiologist.


Selection and Evaluation of Cowpea Crosses between TVx 3236 and Alectra vogelii resistant landraces from Botswana M.e. PIli/lips Selections from the F2generCloon oicrosses between TV" 3236-0-1 Gand locallandraces, 8319 and 8359, known to pos..<>ess good resistance to the parasiti c weed Attctr$l vogel;;,; were grown for three generations and further selections. made of plants with oil combination of A . vagi'li; resisbnce and good agronomic characteristics. Screening was done by growing a single plant in a plastic bag pot containing about 6.5 kg 01 soil inlo which approxima te! y 800 5E'eds of A. oogeJii had been thorough! y mixed. Counts were made of total and emerg t>d shoots, the lengths of the five longest shoots are measured, and St."ed yield r<."Corded. In the l&lSt season, 10 selections were a1so evaluated in plots in two fields know to be infested with A. vog~Jii seed. The selections combine reasonable but variable resistance and possible tole rance with the earliness of TVx 3236. The resistanc:ein the crosses has not reached. the level found in lhedonor patents. Meannxluctions in total shoot numberrecord t."Cl in four comparisons for the best selections, 86 and 125, were 76 and 72% and

correspondlng yield increases were 929 and t212%respectively.

A. vogeJii is a common parasite of cowpea and occasionally other legumes grown in Botswana, together with a range of wild species. A recent survey of arable weeds of the main crop growing areas showed that just over hall the fields planted to cowpea had the pa rasite. Cowpea population densities in broadcast crops are usually low, but levels of parasitism could be high . About half the cowpea germplasm collection in Botswana has been screened for resistance to A. vogelii and a number of lines were found that had good to almost complete resistance to the parasite. A number of crosses were made between two of these, B319 and 8359, and TVx 3236 in the 1984/85 season. In the F2 generation, selections were made of the best plants obtained when grown in soil artificially infested with A. vogeJii seed. These selections (28) have been further evaluated for resistance to the parasite and for certain agronomic characteristics.

Materials and methods The "pot in the ground" technique was used to measure cowpea response to A. vogelii . Plastic bag pots were filled with about 6.S leg soil into which had been thoroughly mixed approxima tely 800 A. vogelii seeds collected the previous season. Single superphosphate fertilizer was also added to supply 20 leg/ha P. Holes large enough to accommodate the pots were dug 1.5 m apart. Three seeds of each selection were sown and the treatments were laid out in a complete randomized block with five replicates. Once the seedlings had emerged, they were thinned to the strongest single plant. Supplementary irrigation was supplied by hosepipe so that the plants did not suffer any water stress. Insect pests were controlled by an appropriate insecticide. Observations were made on the start of flowering. start of pod formation, and the emergence of A. vogelii shoots. The plants were classified into prostrate or upright


types and where the pods were held relative to leaf canopy. At maturity, all pods were harvested and the pots lifted from the ground, cut open and the shoots carefully teased out and counted, dividing them into emerged and unemerged shoots. The lengths of the five longest shoots were measured. In the second season the individual plants showing the bestcombinationof A. vogelii resistance and desirable agronomic characteristics were evaluated. These were again screened as described earlier and grown in 5-m rows. Adjacent to each test row was one row of lVx 3236 so a direct comparison could be made with the parent that had the desirable growth habit and maturity. The rows were 1 m apart and seeds were planted at approximately lO cm spacing. The area chosen was not known to have serious A. uogelii infestation. Observations were made on flowering date and plant habit. Seed was harvested at maturity. In the third season, 1990/91, the 10 most promising selections were again tested in a pot trial and also in small plot trials in fields known to be infested with A. vogelii seed. One site was theSebele Research Station while the other was in a farmer 's fi eld at Mmamashia, 10 km from Sebele. The trial design used was a checkerboard layout where each test selection was adjacent to a plot of the susceptible parent, lVx 3236. Plots were 4 rows of I-m spacing, 4 m in length, with 20cm between plants. The Sebele site was planted on 27 December and thatat Mmamashi~un 18 January, both following Significant rain. Counts were made oftotal emerged shoots in the plot about 10 weeks after planting. After harvesting pods,S plants from each 01 the center rows were selected systematically, carefully dug up, all shoots counted, and the five longest measured. Insect pests were controlled by 路the use of an appropriate insecticide and 20 kg/ha P "pplied before planting.

Results Screening of single pl.nts grown from F3 selections. Five replicates of seed from a single plant of the F3 generation were grown. At this stage, segregation of genetic material was probably still occurring. The means of data from the replicates were calculated, but some differences in plant morphology between replicates were observed . The most desired single plants were selected for growing the following season. See Table 1 for mean results. Total shoot number was in most cases higher than that recorded in screening ofF2 selections when no plants had over 10 shoots. Thenumberof shoots that emerged was much lower than the total shoots and in most cases the mean was less than 10. There were a few attachments on B319 plants and even B359, the more resistant of the two. Grain yield of the selections varied considerably, but in most cases was well above the susceptible control. B1ackeye and lVx 3236 produced almost no seed. Flowering started between 49 and 63 days after planting. Data from individual pots were examined in order to select those plants with less than 11 total shoots and which had an upright habit with pods held above the leaf canopy. In addition some plants that had few or no A. vogelii shoots but which were more prostrate in habit, with pods held above and within the canopy, were also selected. Only 19 plants fulfilled these criteria. Seed from these was planted as before in plastic bag pots sunk in the ground containing soil infested with BOOA. vogelii seeds. There was also sufficient seed to plant a 5-m observation row. The seed from one 76

selection was badly da maged by w eevils and only one seed germinated. The means for thefivereplicates are given in Table 2. Those marked with an asterisk were selected as being morphologicaUydesirable, up right, early lIowering.and with pods held high. None of the selections was free o f shoots in this screening. This contrasts with the plants from which they were selected, nine of which had no shoots at a II and 16 had no emerged shoots. Only 8359 had no emerged shoots, although it had some attachments. 1he susceptible controls had about 25 total s hoots on average, but less emerged shoots than some o f thecrosses. However the yields of thecontrols were very low and the effect of the parasite on them appeared to have been greater than on the crosses. Table t . N umber of A . vogel;; shoots and length tmm ) of 5 Cross

Tswana parent


2 3 4 5 6 7 8 9 10 11 12 13 14

15 16 17 18 19 20 21 22 23 24 25 26 27 28

8359 8359 8359 8359 8359 8359 8319 8319 8319 8319 8319 8359 8319 8319 8359 8319 8319 B359 8359 8359 8359 8359 8359 8319 8319 8319 8359 8319 8319 8359

Blackey. TV. 3236 S.E.D. (124 d .I.) C.V. (%)










27.6 16.4 162 1.6 13.0 20.4 29.2 2\,4 9.0 5.2 32.6 5.0 22.4 39.0 26.6 8.8

6.0 2.8 2.0 2.4 0.2 5.6 8.2 6.2 0.2 5.2 11.8 0.6 5.2 12.8 3.8 2.4 2.4 6.0

4362 1422 4114 1518 1102 466 492 29 24 330 671 41 519 376 231 290 388 449 167 23 436 286 87 583 699 507 39 330 45 0 430 531

3.6 9.6 8.0 1.7 6.7 52.4


22.6 6.6 6.8 23.2 15.4 14.0 28.8 18.6 35.0 12.2 26.0 2.6 0.4 27.6 21.8


0 10.0 \.6 0.6 11 .0 8.4 13.2 0.6 5.6 0.2 0 6.4 6.4

4.34 83.2

190.1 147.2


54.9 % .0


204.4 133.0 151.7 241.4 4.8 122.8 40.8 75.9 823 65.9 47.6 81.1 207.7 29.3 77.9 57.6 40.7 39.8 123.6 9.7 49.8 82.7 122.3 0 0.6 1l0.6

Table 2. Number of shoots and length (mm) of 5 longest F4 selections Selection

48" 69 70 76" 81" 83 86" 91 103" 115 120125" 132"

]34 138" ]49 153 156

Tswana parent

Total shoots

B359 B319 B359 B319 B319 B319 B359 B319 B359 B359 B359 B359 B319 B359 B359 B359 B319 B319 B319 B359

17.8 22.6 14.6 19.4 23.2 35.2 11.4 33.0 16.8 25.4 18.2 9.0 16.2 3.2 31.0 13.2 24.0 23.6 13.0 2.4 24.0 25.0

5.6 7.2 6.8 5.2 7.4 7.6 3.4 13.8 3.8 11.2 2.6 3.0 5.2 0.8 2.4 4.8 9.8 11.2 1.2 0 1.6 9.2

6.74 61.7

3.55 112.2


TVx 3236 S.E.D


Emerged shoots

Length (mm)

Yield (g)

894 1112 902 896 1312 1191 547 1534 778 839 571 648 ]334 248 644 793 1022 1850 448 37 fIJl 688

113.3 74.9 211.8 177.9 393.5 277.6 189.5 336.9 154.0 46.1 98.8 247.3 326.1 45.4 117.2 64.3 73.1 385.8 BO.5 137.1 10.3 20.4

310.5 63.2

75.5 77.8

â&#x20AC;˘ Good agronomic characteristics

Table 3. Atuchments 10 8359 and 8319 crosses Tswana parent B359 Total shoots Emerged shoots Shoot length (mm) Yield (g)

16.1 4 .4 86.0 128.8

Tswana parent B319 24.7 8.4 1281.0 255.7

Of the two Tswana parents, B359 is more resistant. B319 has in previous screenings supported a few shoots while 8359 has usually been free of them. Of the 18 selections made, 10 had B359 as a parent and eight B319. 78

On the average the parasite grew more on 8319 crosses than on 8359. Table 3 shows the average performance of the crosses from the two Tswana parents. Thegreater number of attachments and the longer shoots were found on selections whose Tswana parent was 8319. Despite this, the average yield of the 8319 crosses was twice that of the 8359 crosses. The crosses differed in their days to flowering habit and pod pOSition (Table 4).

Table 4. Days to flowering. pod position, and plant habit (average of five plants) Selection

Days to flowering

Pod position

Plant habit

53 54 51 49 51 50 55 54 52 53 51

1 1.4 1.6 1.2 1 1.4 1.2 1.4

u/p u/p

48 76 81 86

103 120 125 132 138 156 TVx3236

1.4 1.4 15



P u/p u/p


u z upright p= prostrate 1 - Pods held above canopy 2 - Pods in and above canopy

Replicate plants varied in habit, pod position, and flowering date so means of results for five plants were presented. The number of days to flowering was similar or slightly less than that of TVx 3236 for most of the crosses. Most plants tended towards the prostrate, though nowhere near the completely prostrate tswana parent. A single plant in a pot without any neighbors tended to be more prostate than when grown in a row. TVx 3236 in some replicates was postrate, yet it is considered an upright variety. Pod position of most crosses varied, with some plants having pods in the canopy as well as held above the canopy. Because a plant's growth is influenced by its neighboring plants, the rows were grown to get a more representative picture of growth and yield in the field. The rows of each selection were assessed for days to flowering, pod position, and plant habit. At maturity, all the pods were picked and threshed and yield calculated. A rating was also given for general attractiveness of the selection. Results are given in Table 5. Majority of the selections were visually attractive when grown in rows and often more vigorous than the neighboring TVx 3236 row. Yields were high, especially those of varieties with a 2-star rating, and in most cases higher than obtained for TVx 3236. The crosses seemed to combine the early flowering and pod position of TVx 3236 with the leaf production of the Tswana parent. Many of them grew upright when grown in rows than when planted alone. 79

The same selections were screened a second time during the winter of 1990 in the greenhouse, using the same methods, with a single plant established in a po t with soil artificially infested w ith A. vogelii seed. Growth of the plants was limited by low night temperatures and theamountof seed setwas small hence no yield data were recorded. There were four replicates of each selection. The total number of shoots and the emergence number were generally higher in the greenhouse experiment than in the previous field screening. The growth of the shoots was less, which may have been a reflection of increased competition between them. The resistant controls, 8359 and B319, had no attachments at all and 1Vx 3236 was one of the worst affected by the parasite. A number of the selections, particularly 70,86, 103, 115, 125, and 149 had low numbers of emerged shoots, similar to their reaction in the field screening. Table S. Performance of F4 selections grown in rows Selection

48 69 70 76 81 83 86 91 103 115 120 125 132 134 138 149 153 156 1Vx3236

¥= upright

Days to flowe ring 51 51 51 48 51 59 51 57 51 48 51 48 48 55

55 48 48-51 55 48-51

P = pros trate plants affected by nematodes

Pod position

Plant habit

1 2 1 2 2 2 1 2 1 1 1 2


sI p u

P P P u

P u u u

P u


sIp P

2 2



1 2 1

P u u

sIp - seml-prostrate


Yield (kg/hal 2119 1781 2338 2332 3437 1882 1592 2097 1517 1842 1243 1193 1105 1244 213 1

964 1 479 1726 764

Visual rating


• • •• ••

• •• • • • •


• •

Table 6. Totlil and emerged .hoots and length of 5 longest from greenhouse e"periment Selection

Total shoots

Emerged shoots

Length (mm)

48 69 70 76 81 83 86 91 1m 115 120 125 132 134 138 149 153 156 B319 B359 Blackeye TYx3236

51.2 99.2 25 72.7 875 79.7 25 63.7 10.7 37.7 705 445 865 335 84.0 2.0 96 74.7 0 0 0 69.2

175 28.2 0.2 27.0 26.2 325 0 28.7 45 15.0 2759 7.0 16.2 14.0 38.5 0.2 38.5 34.2 0 0 22.0 33.5

780 6775 142.5 7375 7525 7375 17 7535 273 8825 51.2 630 770 545 761.2 106.2 892.5 815 0 0 715 1015

SED (74 d.f.) C.V. (%)

17.6 53.0

6.7 56.1

133.9 35.5

As in previous screenings, there was considerable variation between the repli路 cates,notonly of the test selections, but also the controls TVxandBlackeye. The parent, B319,had a significant numberofshoots in the test, more than noted before, while B359 was completely free of shoots as is usually the case with this accession. The selections with the least shoots and also the shortest lengthswere48,86, 125, and 138. Thesewere amongst the least parasitized in the previous year's evaluation. All, except 132, had B359 as the A. vogelii路resistant parent. Yields for each plant were similar to those recorded in the previous screening for most selections, though 156 was lower-yielding and B319 yielded exceptionally well . Of significance were the very low yields of Blackeye and TVx 3236. This was noted in the previous trials and indicates that in addition to supporting few attachments and limiting their growth, the crosses are also tolerant of the parasite enabling them to grow and yield well despite supporting A. vogelii shoots.


Pot screening, 1990191. The mean results from the five replicates for total and emerged shoots, length of five longest shoots, and yield per plant are given in Table 7. Numbers of days to f10weringwas slightly longer than in the previous year, except for 86, 125, and 132 which flowered earlier. As these observations were only on single plants and records were only made twice a week, some variatio n might be expected. Field trials. The level of A. vogel;; infestation at both sites was light as judged by emerged shoots. The total number of shoots in each plot was counted when all shoots were thought to have emerged. After pods were harvested, 10 plants were removed from the center two rows and all shoots counted . Table 8 shows the results for Sebele. Measured traits are expressed as a percentage of the nearest TVx 3236 plot. Table 7. Yield and number of shoots and length (mm) of 5 longest shoots Selection number

Total shoots

48 76 81 86 103 120 125 132 138 156 8319 8359 Blackeye TVx 3236

4.4 10.2 11 .6 3.0 12.0 11 .8 6.4 6.0 12.2 16.0 11.0

SED (52 d.f.) CV (%)

Emerged shootsLength (mm)

528 1560 1264 948 1259 12.80 418 652

21.4 18.0

3.2 7.4 6.8 2.6 8.8 5.4 1.4 3.6 3.0 11.4 6.0 0 8 .4 12.4

4.57 70.2

3.17 87.3

420.4 68.6


888 1794 714 0 1022 1242

Yield (g)

72.2 127.1 255.3 213.6 151.8 149.7 265.6 248.8 108.6 163.6 307.2 58.7 11.9 31.2

51.8 53 .0

The number of emerged shoots on the TVx 3236 plants was low, though there were twice as many unemerged shoots. The average number per plant of emerged shoots was 1.1 compared with 18.0 in the pot screening. Interestingly, the lengths of the five longest shoots were very similar to that in the pots, indicating a lower seed density in the field and hence fewer attachments, but those that did grow were of a similar size in both pot and field trials. Several of the selections that supported the least shoots in the pot trial also performed well in the field trial, i.e., 48, 125, 132, and 138 (Table 9). The germination and vigor of the TVx 3236 plots was noticeably poorer than that of aU other plots. Replanting was carried out with a different batch of seed wh ere necessary, but the plants did not grow well and three of the e ight plots produced no grain. Consequently the yield data have not been used. 82

Table 8. Emerged and total shools, length of five longest (mm) from 10 plonls as â&#x20AC;˘ percentage of the nearestTVx 3236 plot, mean of four replicates, Sebele Selection number

Emerged shoots

Total shoots

Length (mm)

36 153 190 16 20

41 100 73 21 15 42 22 27 28 101

67 91 82 6 38

48 76 81 86 103 120 125 132 138 156


17 35 24 488

TVx 3236 (actual)


37 51 36 173 1381



Table 9. Emerged and lolal shoots, length of livelongest(mm) and grain yield (kg! ha) from 10plants as percentage of the nearest TVx 3236plot,mean of four replicates, Mmama.hi. Selection Number

Emerged shoots

48 76 81 86 103 120 125 132 138 156 TVx 3236 (actual)

Total shoots

Length (mm)

150 109 64

0 25 6

0 0 31 378 3.9

Yield (kg/ha)




91 75

109 81 4 0 44 14 0 57 144 728



6 97 20 7 94 220 12.8

221 317 333

488 138 588

741 1033


The level of infestation at Mmamashia was less than half that of Sebele, though the performance of the selections was similar to that recorded at Sebele. Selection 48 supported more shoots at Mmamashia although the lengths were similar. Selection 156 had more parasites than TVx 3236, and also supported a large number of parasites in the pot trial. It supported similarly high shoot numbers in previous screenings but yielded well and had desirable agronomic characteristics. For these reasons it was included in the field trials . All the selections yielded more than TVx 3236, in most cases mo re than twice as much. Again, this indicates that even though the selections support parasite attachments, the effect of these parasites may be less that seen on the susceptible TVx 3236. 83

Conclusion The selections show some consistency in performance between the various pot screenings and the field trials, though there remains an unsatisfactory amount of variability. Some of this is environmental and is related to how the parasite responds to the host. Despite trying to standardize conditions by growing the plants in a homogeneous soil and seed mixture, the range of shoot numbers recorded on the controls in the pot trials has been large. Emerged shoots on TVx 3236 varied from 6.4 to 335 and totals shoots from 18.0 to (/).2. Variation between replicates within a trial has also been high. lherefore some of the variability is environmental and some genetic, .but it is not possible to distinguish between the two. In none of the screenings has the level of parasitism of the selections been as low as obtained on the better of the resistant parents, 8359. The most consistent and dosest in A. vogelii resistance seem to be selections 86 and 125, both of which have desirable agronomic traits and have yielded well. Yields of all selections have been consistently higher than that obtained from the TVx 3236 parent and thatfrom B1ackeye in the pot trials. This ind icates that the selected crosses, inaddition to supporting fewer parasites and being of shorter length, can also tolerate parasitism by A. vogelii. Selection 156 possibly has tolerance, but no resistance as shoot numbers have been similar to those obtained for TVx 3236. The work to date indicates that resistance/tolerance to A . vogelii is quite good in the best selections, but probably not good enough and further crOSSing could be worthwhile. The yields, days to flowering, and plant habits of the selections are all good and represent types that farmers in Botswana would be interested in growing.

Discussion Q. Pali-ShikhuJu: Could you comment on the possible role of Striga-tolerant cowpea lines in the control of this parasite in maize-cowpea intercropping experiments? A. Phillip.: Cowpea may encourage the germination of Striga, but this would happen with all cowpea genotypes, SO Alectra-resistant ones would offer no advantage under these circumstances. Q. Phillips: Could representatives from other countries indicate if they have seen

Aieclra growing on cowpea and whether they consider it a significant constraint on plant growth and yield? A. Dashiell: A/ectra is present in West Africa. Presently we believethatStriga isa more serious and widespread pest. However, no detailed surveys have been conducted to determine the extent of Alectlll in West Africa. A. Doto: A/tetra is present in both northern and southern Mozambique. A. MuJila-Mitti: Alectra has been observed in Zambia, however, we do not consider it an important constraint. Insect pests and diseases are more important than Aiectra at the moment. A. Saka: Striga asiatica in common on late-planted maize in some marginal areas of Malawi. A/tetra vogel;i may be present on cowpea,however, there is the need for closer survey.


A. Kwapata: The problem of A. voge/ii in cowpea exists in some parts of Malawi but little survey has been done to assess the extant of the problem countrywide. No screening for resistant lines has as yet been initiated. A. M1igo: A. voge/ii is present in Tanzania. It has been spotted at one of the testing sites, Hombolo, in Dodoma Region. Infestation at this site is very high and threatens the cowpea evaluation trials. The cowpea germplasm is being screened for resistance to this weed. In thecomingseason, the line B301 reported to be resis tant, will be included . We would also be happy to have the resistant lines 8319 and B259 used in your study, for testing in Tanzania. Presence of this weed has also been reported from the southern part of the country in Mtwara region . Q. Doto: How much variation do you observe in A. vogelii in Botswana? How do you collect and homogenize A. vogelii seed for your resistance screening studies.

A. Phillips: Severaldifferent populations from within Botswana have been tested and all showed a similar reaction to the genotypes tested. Aleetra from West Africa is different. Different populations may well exist within the region. Enough seed can easily be collected from one field with natural infestation. About 800 seeds are mixed with enough soil to fill a pot. This is done by bulking up and mixing in a concrete mixer.


An Evaluation of Maize-Cowpea Intercropping Systems in Swaziland John Pali-Shikhulu and Zodwa Mamba

An experiment was conducted during 1989190 and 1990191 cropping seasons to evaluate and identify interuopping syslems that are productive and acceptable to Swazi Nation Land (SNL) fanne .. who cultivate small fann holdings of < 2 ha per household. Maize (u. ""'Y' l .) was intercn>ppeci wiIh a new determinate cowpea (Vigna ungUiCldllt. (L) Walp.) line IT82D889, seiectedfrom UTA materials and due for release in SWil7.iland. There was an incre<lse in total grain yield and relative gross output when maize and cowpea were inlercropped. Maize and cowpea yields were affected by inlercropping systems and by environmental factors. partkuJary rainfall. Farmers considered grain yields inadequate. Suitable crop combinations fcx maize-cowpea intercropping systems are also proposed.

Most Swazi Nation land (SNL) farmers practice intercropping to insure against adverse environmental conditions, diversify diets and farm activities, and improve productivity with limited resources. Research has shown that intercropping can increase output per unit land area through the production of two or more crops on the same piece of land within a single season (Mandai et aI. 1987). [ntercropping can also reduce pest and disease damage, compensate for damage to a component crop, and through higher land equivalent ratios provide better returns for the same labor input (Lightfoot et al. 1987a). Crop production stability resulting from reduced probability for crop failure, is dted as another advantage of intercropping (Rao and Willey 1980). However, intercropping has its shortcomings. Shorter components suffer greater yield reduction as a result of shading from the taller crops. A more productive intercrop combination would result if this influence was reduced. Adaptation of suitable crop cultivars and manipulation of plant population and times of planting of component crops are commonly used to achieve this goal (Wahua et aI . 1981). Maize, the staple food in Swaziland, is grown by every SNL farmer. Maize is intercropped with many other crops, particularly pumpkin, bean, cowpea, and melon (Anon 1989). The usual objective of this intercropping systems is to produce an additionad crop while maintaining maize output. However, there is clearly the need for a more complete understanding of these systems, so as to identify the most productive practices and to develop cultivars compatible with farmers' cropping practices. Many different spatial arrangements are encountered in intercropping. One system in Swaziland involves planting the component crops randomly within the same row. Neither the efficiency of this system nor the compatibility of its components have been established.


This study was undertaken to assess the perfonnanceof maize<owpea intercrops planted within-row and in alternate-row arrangements in three agroecological zones of Swaziland.

Materials and methods Two similar sets of trials were planted in 1989/90 and 1990/91 seasons at Luve Experimental Plot (LEP), NhIangano Experiment Farm (NEF), Lowveld Experiment Station (LES), and Malkems Research Station (MRS). Unfortunately trials at LES and MRS were destroyed during the two seasons before harvest. n.e Luve experimental plot is located at an altitude of 350 meters above sea level (masl) and is characterized by unreliable rainfall distribution, and shallow sandy soils (Uthosols) with a depth of less than 50 em which easily become waterlogged under continuous rainfall. Summer temperatures at this site are high and range between 25 and 30 "C. NhIangano on the other hand is located 1000 masl, receives adequate and reliable rainfall (> 850=), and iscoolerwith summer temperatures ranging between 20 and 25 oc. n.e soils at this site are deep loam clays. A maize hybrid (R 201) and cowpea line IT 820-889 were used in this study. R 201 is anearly maturing three-way hybrid that originates from Zimbabweand is particulary suitable for the more marginal rainfall areas because of its long pollination period and its effective pollen-shed to silking synchronization process. The cowpea line on the other hand is determinate, extremely early maturing and thus can be grown twice in the dry middle veld and lowveld regions of Swaziland. The experiment was conducted in a randomized complete block design with five replicates at each site. Treatments were: ICl: 1 row maize alternating with 2 rows cowpeas. IC2: 2 row maize alternating with 3 rows cowpeas. 1C3: 3 hills maize to 1 hill cowpea within the same row. IC4: 4 hills maize to 2 hills cowpea within same row. 1be trials were planted during the normal maize planting season (November I December). Trial plot sizes were 3.6 x 6 m (ie four rows 0.9 mapart and 6 m long). Both crop species were planted on the same day at each location at a spacing of 90 cm by 25 cm to yield a plant density of 4.4 plants/m2. At planting. all plots received a basal fertilizer application of 25 kg N lha, 36 kg P /ha and 25 kg/ha. An additional 35 kgl ha of nitrogen was applied to all plots as a top dressing at about four W AP. Weed control was effected by hand-hoeing while pest control was carried out using a1phamethrin (100 rnl/ha) as and when necessary. A univariate analysis of variance was conducted for each crop separately to identify intercropping treatments that maximize production and income. 'These involved comparing maize and cowpea yields to determine intercropping advantages, particularly the financial value of output under each system. A combined analysis of variance was subsequently carried out on grain yield to determine the relative importance of environments and treatments and their interactions on the performance of the two crops.


Results Grain yield In all the four treatments, maize grain yield was reduced with increasing proportions of cowpea as intercrops. Yield reduction due to intercropping depended to some extent on whether the intercrops were planted in alternate rows or within row . Combinations ICI and IC2 affected maize yields most. 1C3 and 1C4 on the other hand maximized maize yields with yields ranging between 3 and 4.5 t/ha (Table 1). Table la. Effect of cropping systems on maize and cowpea grain yield. during 19891 90 season Luve System


Nhalangano Cowpea



COraiD )dell! (kg lilal M/Maize M/Cowpea ICI 1C2 1C3 IC4



1676.4 896.8 2998.8 1743.2

522.8 991.8 192.6 458.4

1328.4 1893.4 4474.6 4506.0

1601.0 741.2 1042.6 165.4 485 .4

Mean SE(dift)

1828.8 268.1

541 .4 106.4

3500.0 174.7

608.7 95.4


Table lb. Effectofcroppingsystemson maize and cowpea grain yields during19901 91.eason Luve System


Nhalangano Cowpea



Grain yield 'kg /hal

M/Maize M/Cowpea



ICI 1C2 1C3 IC4

3939.6 20583 6283.7 47675

3725 520.61 144.; 219.2

3139.6 16732 4616.2 3578.4

266.0 536.2 80.4 211.0

Mean SE (dift)

42623 377.7

3143 37.2

3251 .9 437.4

273.0 81.2

684 .2



At Luve, cowpea growth and development was more rapid and vigourous than at NhIangano where cooler conditions prevailed . A second crop of cowpea however was not successful in both years owing to wa terlogging. Combina tion 1C2 produced the highest cowpea grain yields across the seasons and locations while 10 was the least productive. Maximum yields for cowpea as intercrops (IC2) were about 70% of the potential in both seasons. The best cowpea yields, however, were obtained during the 1989/90 season (> 1 t/ha). Because of the variations in grain yields across locations and seasons, pooled analyses of variance were carried out to identify possible interactions between cropping systems, location, and seasons. The analyses revealed that (i) the main effects due to intercropping system (5) and the interaction (5 x E) between treatments and environments were both highly significant (P < 0.001), suggesting that the effects of cropping systems on grain yield for both crops differ and that their response to the cropping systems was confounded with environments, (ii) the main effects of cropping systems and the three-way interaction were highly Significant (P < 0.001) for maize grain yields while for cowpea yields only the seasons x intercropping systems interaction and the three-way interaction were found to be Significant (P < 0.001 and

P < 0.05). Not all the four systems are likely to be acceptable to SNL farmers. An average household size of ten persons requires 2-25 tons of grain/year (Simelane 1991). This suggests that for household self-sufficiency an acceptable intercropping system must ensure a minimum production level of this magnitude. Intercrop system IC2 does not satisfy this criterion and is therefore unlikely to be accepted by farmers as it compromises more maize in all the four environments. At Nhlangano, the remaining three systernsappear satisfactory. At Luve, however, the same three systems are less stable which is probably a reflection of rainfall distribution in this agroecological zone. It would appear therefore that at Luve the only satisfactory combination is 10. This intercrop system unfortunately produces negligible quantities of cowpea which may not satisfy the farmers' household consumption or provide a surplus for sale.

Financial analysis While grain yield and other agronomic measure of advantages of sole crops over inteKropsmaybeadequate, they may not appeal to the peasant farmer. Hence, the use of relative prices for the two crops alongside agronomic determinants was adopted . The official price of maize in Swaziland over the last two seasons averaged EO.45 per kg (E1.00 = USS 0.36). the price of cowpea during the same period fluctuated between EO.9O and El.80, giving a maize: cowpea price ratio of between 1 : 2 and 1 : 4. 1hese price ratios were used to assess gross income from each combination. Thedata in Table 3 show that intercropping maize and cowpea was beneficial and that higher incomes were associated with the combinations that yielded higher cowpea yields: ICl, 1C2, and IC4 and that changes in income were rather inelastic in 1C3 which produced the least amount of cowpea. In an abnormally wet year high maize yields suppressed cowpea yields so that the income generated from monocropped maize dominates regardless of the location or even the intercropping combination. In abnormally wet seasons, high rates of leaf area development and dry matter accumulation which provide for high maize grain yields are expected. Under such 89

conditions cowpea growth and development is bound to be suppressed by maize through greater shading which according to Lightfoot and Taylor(1987)and Wein and Nangju (1979) reduces cowpea plant size, leaf area, and ultimately yield potential. These conditions, coupled with disease and insect infestation that appear to be weU associated with wet environments, combined to depress cowpea yields. On the other hand in normal seasons maize growth particularly in the drier areas is less vigourous, leaf area formation is low and therefore the level of shading is low. The infl uence of biotic factors also appear to be less pronounced resulting In greater cowpea production. This suggests that maize-mwpea intercropping practices are likely to remain beneficial in the drier areas where the potential for maize is low. Gross returns from the various systems tested were not stable for the two seasons. Nevertheless the benefits to be derived from intercropping are evident. SNL farmers, like rnostother resource-limited farmers elsewhere according to Warland etal. (1991), require cash returns to purchase inputs and equipment, and to carty out other investments. This suggests that those intercropping systems that satisfy household maize grain requirements and produce surplus for sale are likely to attract farmers' interest. In this regard combinations ICl, 1C3, and IC4 are possible candidates for release. It has been suggested that high input conditions under which on-station trials are conducted do not represent actual conditions found on peasant farms (Shaner 1991). Lightfoot and Tayloremphasized that attempts must be made toacquiredala from real fanning conditions instead of depending on artificial environments. Kass (1978) (cited in Lightfoot and Taylor 1987a) suggest that economic advantages from intercropping are more likely when management is limited. The results of this study are in agreement with these observations and indeed reveal that intercroppingadvantages are not so spectacular under good crop management. Maize in these trial s dominated in most systems most probably because of the high level of management applied to the trials. Cowpea on the other hand was suppressed by the more vigourous maize growth and development. Thus, while no firm conclusion can be drawn from these results, the data suggest that the most promising intercropping combinations are ICl, and IC4. IC3 may be s uitable but because it produces limited quantities of cowpea it may not be so attractive. IC2 on the other hand compromises too much maize grain so much so that household self-sufficiency is also comprised. However the four intercropping systems need to be further evaluated under typical small holder crop management conditions to ascertain their productivity (Shaner 1991), and their acceptability to farmers. References Anon. 1990. Annual survey of Swazi Nation Land: 1988-1989. Central Statistical Office, Mbabane, Swaziland. Lightfoot, CW., and RSTaylo r.19S7a, lntercropping sorghum withrowpe.a in dryland farming systems in Botswana. I. Field experiments and relative advantages of intercropping. Experimental Agriculture 23: 42>-434.


Lightfoot, C .W., K.B.C . Dear, and R Mead. 1987b. lntettropping sorghum with cowpea in dry\and farming systems in Botswana. 2. Comparative s"'bility of alternative cropping systems. Experimental Agriculture 23: 43S-ÂŤ2. Mandai, B.K. RIC QtQsh, N.C. [)as, and A.K. Som Choudhury. 1987. Studies on cotton-based multiple cropping. Experimental Agriculture 23: 443-449. Rao, M.R. and RW. Willey. 1980. Evaluation of yi.,ld stability in intercropping studies on sorghum/pigeon pea. Experimental Agriculture 16: 10~116 . Shaner, W.W. 1990-91. Agricultural Research Division. Ma1kems, Swa ziland. Sim.,lane, A.V. 1991. Optimal crop enterprises f"" farms on Swazi Nation Land in Swaziland. MSc Di... College of Agriculture and Forestry. West Virginia University, West Virgia, USA. Wahua, T.A.T., O. Babalola and M.E. Akenova. I981. lntercropping morphologically different types of maize with cowpea : LER and growth attributes of associated crops. Experimental Agriculture 17: 407-413. Warlanct.R.H ., S.M. Dlamini, KR. Hsieh and M.T. Malaza. 1991. Swaziland Cropping System Researmand Extension Training Project Impact Study. ThePennsylavania Stat., University Department of Agricultural Economics and Rural Sociology, Weaver Building, University Park, PA 16802, USA . Wein, H .C. and D. Nangju. 1976. Cowpea as an intercropping under cereals. In Symposium on lntercropping in Semi-arid areas. Morogoro, Tanzania : University of Dar-es-Saloam.

Discussion Pali-Shikhulu: Could the lITA representatives comment on the way their accessions are named? Dashiell: Take for example IT 82E-16. The "IT' indicates that the variety was develo ped by DTA, "82" indicates that a progeny row was bulked in 1982 to form the source of seed, "E" indicates that this was done during the early season in 1982 and '16' is the row number. The four letters that are used are F, S, E,and D. They stand for the four seasons in which material are advanced each year at lbadan-first, second, early and dry, respectively. MosarwelDashielUAmable: lV" stands for 'Tropical" (as in UTA) Vigna unguiculata crosses made at Ibadan. TVu stands for an accession in the germplasm collection .


Cowpea Pathology in Botswana Baikabile Matilo

The paper presents a review of cowpea digeases and. root knot nematode under local conditioos in BoIsw_and concludes that (I) most palho1ogicalSludies and screerUng work conducted on cowpea aphid-bome mosaic virus have yielded very few sources of resistance~ (2)root路knot nematodes are a potential problem in sandy soils and even though. lOIi&e5 have not been estimated. observations dearly indicate that serious infestation can lead 10 reduction in plont stand and yield loss, (3) very litllework has been done on soil-bome fungal pathogens and. future studies wiU investigate the role of FlISOri"", and MIIcrupltomi .. in premature death and wilting of plants (a phenom-

enon that has been observed under drought corditions).

Cowpea [Vigna unguicula/ll (L.) Walp.), is the most widely cultivated pulse crop in Botswana mainly because it can withstand drought cor.... itions. The yields vary from zero to 200 kilograms per hectare depending on various factors among which are diseases. According to a survey conducted recently (Anon 1987) in Botswana diseases affecting cowpea include: 1. Cowpea aphid-borne mosaic virus 2. Bacterial blight ()(Qnthomonas vignicola) 3. Leaf spot (Cercospora canesuns) 4. Rust (Uromyces appendiculatus) 5. Ashy stem blight (Mncrophomina phaseolina) 6. Root-knot nematodes (Meloidogyne in cognita and M. javanica) During the 1989/90 to 1990/91 growing seasons, wilting disease symptoms were observed at Sebele and Goodhope and Mncraphomina spp and Fusarium were isolated from diseased plants.indicating a possible complex. Another Fusarium complex was observed on cowpea infested with root-knot nematodes. The economic importance of any of the diseases mentioned earlier have not been determined under local conditions.

Cowpea aphid-borne mosaic virus Cowpea aphid-borne mosaic virus (CAMY) disease is the most documented disease of cowpea in Botswana. Although past reports indicate first observations in 1974 (Molefe 1983), a severe case of mosaic was reported in one field in the Barolong farms earlier (Nilsson 1972). In recent years the disease has been prevalent at agricultural research stations probably due to increased cowpea cultivation and introduction of new cultivars. Symptoms of the disease are typically a chlorotic mosaic on the primary and trifoliate leaves but complete chlorosis of the leaves may be observed. The virus infection is systemic and infected plants set fewer than normal pods depending on the severity of infection.


Virus isolates from Sebelewere found tocontain two strains of CAMV (Anon 1981) and a third strain was reported by Burke. Further studies showed the virus to be seedborne, sap transmissible, and in a non-persistent manner by the groundnut aphid, Aphis c:rw:cioor~ (Anon 1981). Previous studies have indicated that the virus is more severe in late planted cowpeas (Anon 1988). This is poSSibly due to a build-up of populations of the vectors during the season. The virus has also been found to be more severe on irrigated land than on dryland (Bruke unpublished). It is possible therefore, that the virus could reach epiphytotic levels in particularly wet years. The pathologyprogramhas been involved predominantly in saeeningforCAMV resistance in collaboration with the cowpea improvement program. LocaIand exotic cowpea varieties have been evaluated under fJeld conditions. The test entries were grown in l(}.m rows and three plants in each row were mechanically inoculated at the Heaf stage using sap obtained from infected plants. Appearance of viral symptoms was noted as early as possible but the incidence was assessed six weeks after planting. (See Tables 1 and 2 for results of 2 seasons) Although most local cowpea accessions are susceptible to the virus, some resistant varieties have been identified:TYu 410, UCR237, 8027, B306,IT82D885, IT 82D889, and TYX 3236. Resistance screening work has been based on symptoms, and the three strains of the virus cannot be separated based on this. Asaresult, resistance does not hold across locations or may vary with seasons. Attempts to investigate the effect of virus infection on the cowpeas and to determine the seed-bome transmission of the released and promising varieties have not yet yielded conclusive results.

Ashy stem blight This disease, caused by the fungus Macrophomina phaseolina, was found at epiphytotic levels at Sebeleduring the 1984/85 growing season (Broke, unpublished). Ashy stem blight has been found mainly on plants at or nearing maturity but early infection might have been undetected. Lesions appear on the stem near the soil line and extend to the upper portion of the stem resulting in death of the affected tissue. The lesions are grey and numerous small black pycnidia may be seen scattered throughout the surface of the lesion. Plants eventually wilt and may die if conditions remain favorable for disease development. The sclerotia of the fungus survive in plant debris and soil and these act as initial inoculum. Plants are particularly predisposed by high temperature and low moisture conditions. Supplemental irrigation to avoid moisture stress will minimize thedisease. Under the extremely dry conditions of Botswana the disease is likely to remain endemic. Susceptibility of local varieties must be evaluated in order to determine the possibility of using resistant or tolerant as a means for control.


T.ble 1. Oi""."" incidence on cowpea lines .t Sebele % Early Infection

# Inoe. plants with symptoms

Incidence of virus (%)

0 0 1 1 0 0 0 3 0 2 3 0 3 0 2 0


0 0 0 0 0 0 0 0 0 0 0 5.34 0 0 0 0 0 0



Magnolia Blackey<

0 2.47 (.26 6.25 0 0




0 0 0 0 1 3 3 3.5 4.5 4.5 4.5 5.5 6.0 6.5 6.5 6.5 6.5 7.0 7.5 7.5 8.0 10.0 10.5 11.0 12.5 14 14 15.5 16 16 16.5 19.0 19.5 205 215 21.5 21.5

Entry OCR 737 lVu410 8027

Black:eye Vita 90 lVu 1000 lVuMS BIllA UCR236 lV.I999-OOF lVu2755 UCOM-857 TT 820 881 TT 820 885 TT 8101l31 lVu4111 TT 820 889

B097 lVu 1185 TT709 TT880 UCR2M BI63 IT 82拢路 71) B 127 IT 820 785 UCR201 ~pplWhite IT8~

801M UCR 183 P1293Sai lVu8O! B 171 TT82OMO lVu6520 lV.32J6.alC OCR 194 TT82D755 8055 lVuW 8051 lVx3236

11232 TT 820 Ml ER7 P1471521 B 201

I 2 0 0

3 1 4





0 2.23 3.53 1.52 0


0 1

0 5

0 4.06 0

0 1 6 1

0 6.251

0 3.113 0

0 0 0 5.75 3.34 5.41 16.0



2 3 2 2 4 3 3 2 6

73.5 26.5 27.0 28.0 30.5 31.5 32 35.5 35.5 35.5 37.5


Hll UI 9.84


4 2






Although the causal organism has not been confirmed, symptoms fit those described for bacterial blight caused by XIlnthomonas uignicola. Irregular necrotic spots with yellow halos have be noticed on the leaves of mature plants. The diseases is spread by rain splashes and is worsened by overhead irrigation. Since no research has been done on this disease, control measures have not bl!en detennined . The disease is believed to be seed borne and therefore use of clean seed is a possible control measure. T~ble


CAMV incidence on cowpe~ lines ~t Sebele Virus incidence ("!o)


TVu410 !T82D889 !T82D 831 !T82D 709 TVx 3236-01G UCD84-857 TVu 1185 TVu645 !T82D885 Blackeye TVu408 !T81D 1137 BIllA TVu2755 UCR237 0097 0005C Vita 9 ER7 !T82E-9 Tswana TVx 199901F

Early Planting

Late Planting

0 2.0 4.1 5.3 5.5 5.9 73 8.7 9.8 11.4 3.5 .6 0.5 9.4 29.7 1.7 3.6 53 36.7 0.2 63 52.2

1.8 0.0 28.9 13.1 1.6 11 .4 5.8 63.4

.3 8.2 10.1 51.2 65.0 56.3 48.8

71.8 90.0 91.0 40.7 77.2 88.2 85.0 SO.O

UCR236 Date of planting: 8/1/88 and 19/ 2/88 resp.

Root knot nematodes These have been found to be a problem particularly in sandy soils. The damage due to these pests is probably underestimated due to symptoms being below ground. Cowpeas have been found with root knot nematode symptoms both on research plots and in farmers' fields (Anon 1980). Although the nematodes have been implicated in damage on cowpea, there is no data to support this. 95

Studies to investigate the effectof nematode infestation on yield of cow peas have, so far, not been fruitful.

References Anonymous. 1980. Annual Report for the Division of Arabi. Crops Research 1979/80. Department of Agricultural Research, Ministry of Agriculture Gaborone, Botswana. Anonymous. 1981. Annual Report for the Division of Arable Crops Research 1980/81. Department of Agricultural Research, Ministry of Agriculture Gaborone, Botswana. Anonymous. 1988. Annual Report for the Division of Arable Crops Research 1987/88. Department of Agricultural Research, Ministry of Agriculture Gaborone, Botswana. Nilsson, G.1. 1972. Plant diseases: report to theCovemmentof Botswana . Food and Agriculture Organization of the United Nations, Bulletin no. TA 3057. 34 pp. Molefe, T.L. 1983. Preliminary investigations of the cowpea aphid-borne mosaic virus on cowpea in Botswana . The Bulletin of Agricultural Research in Botswana (1): 42-51 .


Evaluation of Cowpea [Vigna unguiculata (L.) Walp.1 Cultivars Against Root-Knot Nematode Meloidogyne javanica V.W. Saka Eighteen cultivars of cowpeajVig .. IUlguicU/.,. (L.) Walp.l were evaluated in. field natura1ly infested with root路knol nematode Mdoi4ogyrt< i<""'""'" Using. scale of 1)..5 (where 0 = highly resistant and 5 = highly susceptible), culUvor CB 5D w.. highly resistant, and 1-8-5 was susceptible while the others w .... highly susceptible raising questions as to whether these cuJtivan; should be included in agronomic trials.

In Malawi, cowpea is intercropped with maize, cassava, sorghum, and other annuals. The cowpea is generally used as a cheap source of protein. However, yields are low (about 300-500kg/ha), under smallholder farming systems due to a number of major constraints among which are pests and diseases. The major diseases that reduce yields includeAscochyta blight cau sed by Ascochyta philseolorum and root galling caused by root-knot nematodes particularly Meloidogyne incognilll and M. javaPlica. Root-knot nematodes have been reported to cause an estimated crop loss of about 43% in cowpea in West Africa (Sasser 1979). M . incognita and M. javanica are widely distributed in many SADCC countries particularly Malawi, Zambia, and Zimbabwe (Martin 1955; Saka and Siddiqi 1979). The objective of this screening exercise was to identify resistant cultivars among local landraces and exotic lines with the hope of increasing yields.

Materials and methods Eighteen cultivars of cowpea (Vigna unguiCIIla ta (L.l Walp.l were planted in a field naturally infested with Me/oido8Yne javanica at the Bunda College of Agriculture, Crop Production Department Research Farm, during the 1990/91 season. Three weeks before planting, the presence of the nema tode in the soil was ascertained by randomly collecting soil into six 20cm plastic pots. Water was added to the pots and then a twoweek-old seedling of a susceptible tomato cultivar, Money Maker, was transplanted into each pot. The seedlings were analyzed for galling after two weeks. The eighteen cultivars were obtained from the Chitedze Research Station and from Dr. Kwapta at Bunda College of Agriculture. Eachcultivarwas replicated twenty times with a spacing between plants of 20 cm x 91 cm x one seed per hill. Nematode analysis was done three times: two weeks after planting, at flowering, and at physiological maturity. Plants were uprooted and roots were washed free of soil under running tap water and stained with lactophenol anniJine blue. A scale of ~5 was used to measure the degree of susceptibility where 0 =0 galls, 1 = 1- 2 galls, 2 = ~10 galls, 3 = 11- 30 galls, 4 =30-100 galls, and 5 = 100 galls. Therefore ~1 was regarded as highly resistant, 2 = resistant, 3 = susceptible, 4 = very susceptible, and 5 = highly susceptible. 97

Results Results (Table 1) revealed that cultivarCB5Dwas highly resistant judging from its gall index of zero. Cultivars 1-8-5 and CBS-E} were susceptible while 1-2-5, IT82E-16, 1T82D-716, lYXI94-{)lF, JT835.68O-9, IT83S-818, and G124 were very susceptible. IT835-24, lYX66-2H, 1-2-5, IT85F-2264, Ncheu Local, IT835-680-9, and Bombe 21 were highly susceptible to M . javaniCJl . Table 1. Reactiona of eighteen cowpea cultivaralo M. javanica Cultivar 1 2 3 4

5 6 7 8 9 10 11

12 13 14 15 16 17 18


Gall Index â&#x20AC;˘

Highly resistant Highly susceptible " " " Very susceptible Suceptible


IT835-686 lYX66-2H 1-2-5

8006 1-8-5 CB5-El IT85F-2264 IT82E-16 IT82D-716 lYX194-{)1F Ncheu Local JT835.68O-9 IT840-448 IT835-814 Bombey21 Sudan GL24

Reaction b

5 5 4 3 3 5 4 4 4 5 4 4 4 5 5 4


Highly susceptible Very susceptible " " " " Highly susceptible Very susceptible u

u u

Highly s usce ptible " Very susceptible


â&#x20AC;˘ = Gall index, mean of15 replic.... using a scale of O-S, 0 = 0 galls 1 = 1-2 galls, 2 = 3-11 galls, 3 = 11-30 galls. 4 = 3O-100gallsand 5 = 100 galls. b= Reaction wh<>", ~ 1 = highl y r<Sistan~ 2 = resistant, 3 = susceptible, 4 = very susceptible, S = highly susceptible

Discussion Although results were collected for only one season, the amount of galling which was observed on most of these cultivars indicated that they could not be grown where M . javanica is present. CB5D was resistant and therefore, depending on its other agronomic characteristics, it can possibly be evaluated in agronomic trials . Screening of cowpea cultivars will continue in order to identify as many resistant cultivars as possible for the smallholder farmers who are the primary users of this crop.

References Martin, C .c. 1955. Plant and soil nematodes of the Federation of Rhodesia and Nyasa1and. Preliminary investigations. Nematodes catalogued under host or aS6OCiation plants. Rhodesian Journal of Agricultural Research 52: 346-361 .


Saka, V. W., and MA. Siddiqi. 1979. Plantparasitk nematodes associated with plants in Malawi. Plant Disease Reporter 63: 945-948. Sasser,J.N. 1979. Economic importance of M.,o~spp . in tropical countries. Pages 359-374 in Root-knot nematodes (Meloidogyne species) systematics, biology and control, edited by F. Lamberti and C.E. Taylor. Academic Press, London, UK .

Discussion Q. Amable: he your reaction ratings descriptions of the gaU index? Secondly, I think the entry named "Bombe 2" in both Malawi papers is actually

"Bambey 21" from Senegal_ In view of the classification of this entry as highly susceptible to root-knot nematode which makes it a promising candidate for use as a susceptible check, could you please confirm the name of this entry? A. Sakal Yes, zero gall index means resistance. The entry is Bambey 21. Q_ Haciwa: Is there any variation between genotypes for gall size and could gall size be an indicator of resistance?

A_ Sakal Absence of galling is certainly an indicator 01 resistance and could be due to failure of penetration by second stage larvae or biochemical breakdown of the enzyme produced by the larvae, by the cultivar. Different genotypes react differently and therefore the size of galls produced [in response to infestation) will vary.


Adverse Effect of a Yellow Mosaic Disorder on Growth Nodulation and Seed Yield of Two Cowpea Varieties at Umbeluzi in Southern Mozambique A.Z. Matangue, AL. Doto, and D.M. Naik

A field study was conducted to determine the effect of a yellow mosaicdisord"'(YMO) on growth. nodulalioo. and seed yield of cowpea varieties ms and INIA 36 at the Umbeluzl Experiment Slotion during the 1990-1991 croppins.....,.,路 YMD reduced _ and dry weights of shoots of both varieties. However, root growth was not affecled . effects on growth and nodulation of both varieties were more pronounced with iii longer durati()l\ at lnhction. Alter 21 days weights of shoots and roots of w.cted crops were reduced by almost half in INIA 36 and by about 3 to 4 times in IT 18. Although seed yield per plant was reduced irrespective of the growth stage at which inlection occurred in both varieties, the reduction was less when plants were infected at later growth stages.

Symptoms of a yellow mosaic on cowpea (Vigna unguiculalsl (L.) Walp.1 were observed in 1987 in southern Mozambique, where in parts of Maputo Province, the incidence of yellow mosaic in farmers' fields ranged between 11 percent and 85 percent (Doto personal communication). In 1989, symptoms linked with the presence of numerous whiteflies (Bemisia spp) among the cowpeas were considered very similar to those described for cowpea golden mosaic virus (CGMV) (Rossel 1989). CGMV has been described in Nigeria (Anno-Nyako, Vetten, Allen and Thottappilly 1983) and similar diseases have been reported to occur in Niger, Kenya, and Tanzania (Rossel and Thottappilly 1985;Singh and Allen 1979). lnlection byCGMV has been reported 10 reduce yield in cowpea by 77 percent (Santos and Freire-Filho 1984, quoted by Lima 1988)and by about 60 percent in southem Mozambique (Doto personal communication). After groundnuls [Arachis hypogttlf (L.1I, cowpea is considered among the most important grain legumes inMozambique, where leaves and green pods as well as fresh and dry seed are consumed (Heemskerk 1985). This work was undertaken to determine the effect of the yellow mosaic disorder (YMD) on twocowpea varieties recommended for production in soulhem Mozambique.

Materials and methods The experiment was conducted on 21 February 1991 using two varieties of cowpea. Each variety was planted next to the other in one large plot, 10 m x 20 m separated by a space of 2 m . Plots were not fertilized, and a mixture of 15 ml Cypennethrin 20EC and 30mI dimethoate40EC (in 20 I of water), was applied along the row at the seedling stage to prevent loss of plants to cutworms. Plants infected with YMD in each plot were randomly selected and tagged weelcly from the onset of infection 23-37 days after planting (DAP). All infected but untagged plants were rouged weelcly. At the commencement of infection, 100 infected plants were labelled A. And at the beginning of the second and third weeks, 70 and 60 100

infected plants were labeled "B" and "CO respectively. A week prior to harvest, all remaining healthy plants were tagged ''0''. In order to determine the effect of YMD on growth and nodulation at different stages of growth, 10 healthy and 10 infected plants from group "A" were carefully uprooted a week after tagging and brought to the laboratory. TIle shoots and roots of the plants were separated and the nodules were removed and counted. Fresh weight of the shoots and roots were measured, and the shoots and roots were dried to constant weight at 8O"C in an oven in order to obtain the dry weight. The same procedure was repeated for plants labeled B and C. To elucidate the influenCe of duration of infection on the fresh and dry weight of shoots and roots and nodulation of IT-18 and INIA36, data was obtained as described above on eachoflOinfected and 10 healthy plants from groups A, B,and Cat37DAP. To determine the influence of YMD at three stages of growth on some seed yield components, tagged plants were harvested separately according to their groups at maturity between 15 and 21 May 1991. The number of harvested plants and the total number of pods in each group were recorded (Table 3), and the average number of pods per plant calculated. The number of seeds in two random pods per plant in each group was also counted to obtain an estimate of the number of seeds per pod for each group at plants. All pods in each group were threshed and the seed yield determined. Data on growth, nodulation, and yield attributes for the two varieties were analyzed using the unpaired t-test with equal sample size between groups.

Results Effect of YMD infection ~t different growth .uges Fresh anddryweightof the shoots of both varieties were reduced in infected compared to healthy plants (Table 1). However, in IT-18, late infection did not affect the dry weight of shoots. Also root growth of both varieties was generally unaffected except in the earlier infection of IT-IS, when the fresh and dry weights of roots of healthy plants were greater (Table 1). Influence of duntion of infection Adverse effects on growth and nodulation at both varieties were more pronounced with a longer duration of infection. After 21 days, dry weights of shoots and roots of infected INIA 36 were reduced by almost half, and by about three to four times respectively in IT-IS, when compared to healthy plants. The YMD infection generally reduced nodulation in both varieties but INlA 36 was affected more (Table 2). Effect of time of infection on yield Although seed yield per plant was reduced irrespective of the growth stage at which infection occurred in both varieties, the reduction was less when plants were infected at later growth stages (Table 3). The reduction in yield was reflected by lower pod production per plant and less seeds per pod, particularly when infection occurred earlier (Table 3).


Table 1. Effect of time of i nfection by a yellow mosaic disord e r on the fre sh and dry weight (g) of shoots and roots of two cowpea varieties at Umbeluzi in Southern Mozambi que, 1991

Time of Infection (Days after Planting)

Va riety INlA 36

IT 18 Fresh Wt Shoot



Fresh Wt

Dry Wt











1 (23)











2.25 NS





1 (30)



11 .6






0.97 NS

121 .51-

4.70 NS


1 (37)










7.34 NS

20.78 NS2.14 NS


8.09 NS


0.35 NS 0.92 0.89 NS 1.26 1.27 NS

- Sign ifica nt at p= 0.05 (t-test) All figu res represent a mean of 10 p lan ts

Table 2.

Influence of duration of infection of a yellow mosaic dis order o n the fresh and dry weight (g) of shoots and roots and nodulation of two cowpea varieties 37 days after planting at Umbeluzi in Southern Mozambique


Dura tion (days) -]

INlA 36

IT 18 Fresh Wt Shoot

Roo t


Fresh Wt Shoot


DryW t Shoot





7.95 •



83.84 •





13.75 b

\.0\ b

1.34 b





5.30 b 13.94 b

1.34 b


104.30 b


I (7)

134.78 b


20.75 c

1.92 c


127.71 c

7.95 be 14.87 c

1.26 c 15.40c


242.24 c

7.34 c

20.78 c

2.14 ,


159.99 c

8.09 c








4.10 a



- Significant at P = 0.05 (Host) 1. Days after commencement of infection All figures represen t a mea n of 10 plants



20.49 d 1.51


9.60d 0.59

Table 3. Effect of time of infection by a yellow mosaic disorder on pods per plant, leed. per pod and ned yield in g per plant of two cowpea varieties at Umbeluzi in Southern Mozambique, 1991 Time of Wection (Oaysafter planting)

Pods/Plant IT 18





Seed Yield/P1ant !T18


A (23)

4.1 a

2.1 a

12.3 a

11.2 a

3.3 a





IB.l b

11.3 a

4.4 a

3.2 b



6.7 c


17.1 b

B.4 b

9.9 c




23.4 c











'Significant at P = o.os (I· test) AD Iigwe IEjA"-' • mean ol10planlo

References Anno-Nyako, F.O ., H .J. Vetten, D.J. Allen, and C . Thottappilly. 1983. The n!lationohip between cowpea golden mosaic and its vera, &mil., IIIbrio. Annals of Applied Biology 102: 319-

323. Heemskerk, W. 1985. Regionalism do Feijao Nhemba, INIA, Maputo, SerieAgronomica no. 8, 16 p. Mocambique. RoooeI, H.W. 1989. Report of a virus disease aurvey of cowpea and other crops in various countries of southern and eastern Africa . Unpublished. Rosoel H.W. and C . Thottappilly. 1985. Virus diseases 01 Important food crops in tropical Africa. lbodan, Nigeria: Intemationallnstitute 01 Tropical Agriculture. dOl Santos, A.A. and F.R. Freire Filho. 1984. Reducao na producao defeijao massacar causada peIo virus do mosqueado amarelo. Cowpea Research in Brazil, edited by Earl Eugene Watt ana Joao Pr.tagH Pereira d. Araujo. ITrA/EMBRAPA. 1988: 21!>-232. Singh, S.R., and DJA1\en. 1979. Cowpea pests ..d diseases. IDA Manual Series No. 2. ITrA, Ibadan Nigeria. 113 pp.


Importance and Extent of Yield Losses Caused by Major Insect Pests on Cowpea in Zambia P.H. SoiulU and S. Sitiulnant/zam

The study set out to iden tify the various arthropod pests affecting cowpea in the fieJd and their effect on cowpea. Aphids, leafhoppers, defoliating caterpillars, grasshoppers, leaf beetles, flower thrips, flower beetles, 搂J>Otted pod borers, apian weevils, and pod sucking bugs are commonly found attacking the crop. Surveys across severaJ districts in three provinces of Zambia during 1987-89 revealed that gent.'l'ally. pests attacking flowers or pods cause more damage than defoliators and lor early stage sucking pESts: On-station trials at Msekera during 1987-91 showed a wide range in extent of avoidable yield losses due to pests. The percent loss due to all major pests averagoo 24, with a monetary villue of about ZKlS,248 (USS21 S) per hectare. Estimated losses due to individual pest groups are also provided. Related futu re research needs are discussed and a schematic map of the pests occurring at different stages of development and an updated list of insect pests affecting cowpea in Zambia arc provided.

Insect pests are regarded as a major constraint to cowpea production in Zambia (Mbewe et a!. 1985; Kannaiyan et a!. 1986). However, the information available on insect pests in the country has been largely limited to their biOlogy, ecology, and symptomatology (Angus 1%2-66; Irving 1984). Recently, Sithananthan et al. (1989) reviewed the research information available on cowpea entomology in the country, and indicated that the quantitative data on relative occurrence of different insect pests and extent of loss they cause in different agroecological regions are lacking. This paper provides an updated list of insect pests so far recorded in the country and recent information collected on the relative importance of the major insect pests, based on surveys of farmers' crops and yield loss studies during 1987-91.

Insect pests at different crop stages Insect pests affect cowpea at different stages of growth. During the seedling stage, mortality may be caused by the stem maggot (Ophiomyia sp.), aphids (with possible build -ups even up to late podding depending on drought spells occurring in the season) and leafhoppers and whiteflies (up to flowering or podding). Leaf beetles (flea beetles, weevils) also tend to be damagi ng during the same period. Other defoliators such as grasshoppers and caterpillars commence their attack during the路 vegetative stage up to the mid-podd ing stage. Flower thrips are active in the late vegetative stage and continue their infestation till late flowering or podding . Flower damage by blister beetles is very common, extending from budding up to late flowering. Infestation by the spotted pod borer begins at bud formation and can extend up to the end of podding. attacking buds, flowers, and pods. Other pod borers and the apion weevil commence attack from flowering or early podding. Pod sucking bugs (several genera) can infest throughout flowering or podding extending up to maturity stage, while field infestation by bruchids occurs during the late maturity stage of pods. 104

Among the insect pests so far identified as occurring in the country (Table 1) are some which have been recorded in Zambia for the first time. In addition, we are yet to identify the scientific name for stem girdling weevils, which are commonly found infesting cow peas around Kaoma in Western province. Several species of leaf beetles, grasshoppers, leaf caterpillars, flower beetles, and pod sucking bugs infest cowpeas and we need to make more extensive collections of these, to enable a comprehensive listing of the species infesting the crop across the major growing areas in the country. Table 1. List of insed p"st recorded on cowpea Zambia Common name

Scientific name


Pest status

Stink bug

Nc""ra viridu'. (L.) Ntzara robusta var. vircscens Freeman Halyomorplta viridiscens (Walker) PieZMOTliS ltybeJlcTi (Gm"lin) Piezodorus mbrofosciatus (F.) 路Vctema? mimica Distant Anop/ocne7nis curvipcs (F.) C/Avigra/Ia to"",ntosicollis (Stal) Mirpcms jaeulus (Thurn.) Empoasca doriclli Paoli EmpoasCR fu rcatula Chauri 路Butracomorpllu5 sp. Megalurothrips sjostedti (frybom) Aphis craccivora Koch Bemisia tal1aci (Cennadius) Coryna sp. My/Doris dichlcta Bertoloni Ootheca m"tabilis (SahJberg) Oo tlteal betllligseni Weise OphiomyiR ccntrosematis (de Meijere) 9 plu'moyia spenccrclla Systafes sp. Piezotrachelus varium (Wagner) Alcidodcs Icuc()granIMJUs (Erichon Heliotllis armigera (Hubn.) Maruca test"lalis (Geyer) Etidla z;lJckcllclia (Treitschke) CaUosobruclllls nU2cuiatlls (F.) Acollllmscelidcs o/llcctus (Suz:)

Pentatomidae Pentatomidae

Minor Minor

Pentatomidae Pentatomida e Pentatomidae Penlatomidae Coreidae Coreidae


Alydidae Cicadellidoe

Minor Minor

Cicadel1idae Cicade11idae


Giant coried bug Spiny brown bug Long-legged bug Leafhopper Flower th rips Aphid Whitefly Blister

Beetle Leaf beetle Stem maggot Leilfweevil

Pod weevil Striped bean weevil Bollworm Spotted pod borer Limo bean pod Bruchid

Mino r Minor

Minor Majo Major ~tin or

Thripidae Aphididae Aleyrodidae

Mojor Major

Meloidaae Meloidaae


Chrysomelidoe Chrysomelidae Agromyzidae-

Agromyzida. Curculionidile Curculionidae Cwculionidae

Noctuidae Pyralidtle

Phycitidae Bru-:hidsea Bruchidae

Minor Minor Minor Minor Minor Mino r Minor Minor Minor Minor

M.1jor Mojor Mojor Minor

.. Association to be confirmed

Relative severity of insect in different regions The information on the relative severity of different insect pests in different regions is critical for determining priorities in research. A brief survey made during 1982-.'l3 had indicated that aphids were common around Magoye (southern province), leaf beetles around Mpika (northern province), flower/pod pests around Masumba (eastern province-valley) and pod borers in southern and western provinces. 105

Pilot surveys were undertaken during 1987-89 seasons in the eastern, southern and western provinces to assess pest damage severity on farmers' crops. TIle surveys were conducted during April, when much of the cowpea is normally in flowering or podding stage. Fields chosen were about 15-20 kin apart. In each field, four plots (1 m by 1 m) were chosen at one per quadrant, and the severity of damage by the major groups of pests was recorded on a 1-9 rating scale: 1 =no apparent damage; 3 =light (up to 100/0) damage; 5 = moderate (11-25%) damage; 7 = severe (26-50%) damage; 9 = very severe (above 50%) damage. Insect groups included tor this rating included sucking pests (aphids, leafhoppers, whiteflies), defoliators (beetles, grasshoppers, caterpillars), and flower /pod feeders (thrips, blister beetles, spotted pod borer, pod sucking bugs, and apion weevil). the presence/absence of individual insects was also recorded. Information on cropping situation and crop stage was also collected in each field . TIle results of damage severity ratings for the major pest groups as obtained by these surveys are summarized in Table 2. In the eastern province, the severity by defoliators was greater than that of sucking pests in the districts of OUpata-North, Chipata-South, and Chadiza while the sucking pests were equally, or more severe in Lundazi, I<atete, and Petauke districts. In the western province, flower and pod feeders were apparently rnoreimportant, followed by defoliators and sucking pests in that order. In the southern province, flower/pod feeders were nearly as severe as sucking pests, while defoliators appeared to be less important. T~ble

2- Extent of d~mage severity by major pes' group. in farmen' cowpu crops in Zambia, 1987--89


Mean pest _aily ralillg (1-9 sale)'

Number of field surveyed


Sucldng pesos




Chadiza Chipata North Chipata South Ka_ LWldazi Petauke Southem Choma

2 2 4


1.0 2.0 3.0

(6) (6) (7) (1) (3) (3)

5.0 (8)





4.0 (5) 3.0 (7) 4.0 (5)

5.0 5.0 4.0


5.0 5.0 5.0

(7) (7) (5)

(5) (3) (5) (3) (4) (3)

5.8 5.0


3.9 3.0 32 3.0 3.3 3.0


10 7

1 4





Kaoma Mongu

2 â&#x20AC;˘ Figures in parenthesis show maximum ratings in individual fields.


(8) (5)

NR=No. record

TIle limitation in the survey in the eastern province was that damage by flower/ pod feeders was not assessed as the crop stage during the survey was not appropriate.


However, visits during non-survey trips have shown that damage is as severe as obtained in the western and southern provinces. In general, the pest severity in most growing areas was high for flower/pod feeders; moderate to high for defoliators, and low to moderate for sucking pests.

Estimation of avoidable loss due to major insect pests Direct loss in yield. During 1987-91, a number of replicated trials were conducted at Msekera, in which there were treatments which provided protection against all or selected major pests. The difference in yield from these plots and the unprotected plots in the same trial constituted the avoidable loss, worked out as percentage over yields in the protected plots. Themonetaryvalueof the loss (based on the market price in local currency and in US d o llars) has also been worked out (Table 3). Table 3. Estimates of avoidable losses due to majo< pests on cowpea, Msekera, Zambia, 1987-91

Pest group/yea r

Number of genotypes observed

Mean extent of quantitative seed yield loss % kg / ha

Monetary value of loss/ha â&#x20AC;˘â&#x20AC;˘



19495 12425

279 178 129 158 244 321 218

All major pests

87-88 8IHl9 89-90 89-90 90-91 90-91

1 10 17 12 7 11

39 27 20 25 14 18 24

557 355 258 315 488 641 435

16 10 1 3 1

21 18 32 20 26 23

229 222 408 348 399

8015 7770 14280 12180



25 25

99 147

3465 5145






11025 17080 224335 15248

Flower/pod feed ers

87-88 8IHl9 8IHl9 89-90 90-91 Overall


115 III 204

174 200 161

Early sucking pest

87...a8 8IHl9

10 2


50 74 62

â&#x20AC;˘ Farmers' sale price estimated at K3S.00 per kg and exchange rate 01 K70.00 per USS in JuneJuly 91. The avoidable loss due to all major pests together has been found to range from

14 to 39 (average of 24%). In monetary terms, this works out to an average of ZKI5,248 107

(US$ 218) per hectare, while the maximum was up to ZK22,435 (U5$ 321) per hectare. Losses attributable to flower and pod feeders were found to range from 18 to 32%, with a mean of 23%. Monetary loss per hectare due to these pests averaged ZK11,242 (US$ 161). For early sucking insect pests, the mean loss was estimated as 25 percent, the mean being ZK4,305 (US$62) per hectare. These data tend to confirm that insect pests cause substantial yield losses of cowpea in Zambia. Those which damage flowers and pods appear to be the most important. This confirms an earlier estimate arrived at during 1983--84 at the same location, which reported about 48% loss in seed yield attributable mainly to this group of pests (Sithananthan et aJ. 1989). As the present results are all from on-station trials at Msekera, it would be useful to verify the losses in some other sites, including onfannassessments. Since the farmers consider leaf yields equally important (fo r relish), the extent of loss caused directly by defo liato rs should also be quantified . Indirect los.e.. The insects which damage the seeds, especially bruchids, apion weevils, and pod sucking bugs have been found to cause appreciable reduction in the germinability of seeds (Sithanantham et aJ. 1989; Sithanantham and Kannaiyan unpublished). This loss of germinability can lead to reduced plant stand s and SO affect the yield. Hence, there is need to determine such losses due to seed damage. Aphids are known to transmit cowpea aphid-borne mosa ic virus (CAMV). Agronomic trials at Msekera have shown that (AMV Inc;dence differs with planting date (Reddy et aJ. 1989; Kannaiyan 1991) which is attributable to diffe rences in aphid vector activity. It would be useful to q uantify the extent of vecto r·bome vers us seed borne spread, so as to develop suitable strategies to minimize the losses caused.

Future research needs The information so far ga thered on the occurrence and impo rtance of different insects on cowpea in Zambia needs to be supplemented in the following areas: • crop damage surveys of farmers' crops should include at least two stages of crop growth. • Regionwide sampling for assessing the species composition of flower beetles, pod borers, and pod sucking bugs should be augmented. • The quantitative and qualitative losses in foliage yield, as it relates to use of leaves for relish should be quantified . • Theextentof seed damage by insects as affecting plant stand and prod uctivity in onfarm situations should be estimated. • The role of aphids in causing losses by spreading (AMY should be evaluated.

References Angus, A . 1962-66. Annotated list of plant pest~, disei"lses ;md fung i in z..amb i,l, p.uts ) ·7 <l nd supplement. Zambia: Mount Mil kulu Centr.,} Research ion (Mimeo.) Irving, N.5. 1984. Insects collected fro m legume crops in Zambia , 1 ~~2 -84 . M, ekera Region al Research Station, Chip,t" Zambia. 7pp. (Mi rna,.) K.lnnaiyan, ]., MN . Mbewe, D.C. Creenbe rg, H .C. Haciwa. NS. Irving, and P.H . Sohati. 1986 . Cowpea production and research in Za mbia. World Cowpea Research Co nference(co untry papers) 5-9 Nov 1984, In terniltionili Institute of Tropical A gricu lt ure, lbad,)n, Nigerill . Tropical Crain Legume Bulletin 32: 144-148. (special issue)


!<ann.iyan, ] ., H .e. Haciwa, S. Sithananthan, I' .H. Sohati, and J.M. Mulila-Mitti. 1991. Recent research on cowpea aphid-borne mosaic virus in Zambia . Paperpresentoo at the SADCCI illA Cowpea Workshop lor Southern Africa, 2&-27 Sep 1991, Harare, Zimbabwe. Mbewe, MN., D.e. Greenberg, J. !<annaiyan, H.e. Haciwa, N.S. Irving, and P.H. Soh.ti. 1985. Cow peas in Zambia. P.perpresented at the EEC / SAFGRAD Joint Planting and Evaluation Workshop (Maize and Cowpea), 1&-20 Sep 1985, Cotono u, Republic of Benin. Reddy, M .S., K. Kanenga, J.e. Muso nya, and J. Kanna iyan. 1989. Agronomic and cropping systems research involving grain legumes in 'b.rnbia. Pages 104-1 16 ju Proceedings of the National \OVorkshop on Food Legumes Research and Improvement in 2'..ambia, 9-11 Mar, Mfuwe, Eastern Province. Zambia . Department of Agriculture, Ministry o f Agriculture, Lusaka, Zambia .

Sithananthan, 5., N.S. Irving, and P.H. Sohati. 1989. Recent entomological on bean and cowp ea in Zambia. Pages 135-156 in Proceedings of the National Workshop of Food Legume Research and Improvement in Zambia, 9-11 Mar 1988, Mfuwe, Eastern Province, Zamb;" . Department of Agriculture, Ministry of Agriculture, Lusaka, Zamb;" .

Discussion CommenL Uriyo: Table 3 has the results of surveys conducted in several districts in Zambia and in some of the districts, only one farm has been surveyed. The sampling errorinsurveys is very large and thus there is the need to increase the number of farms surveyed so as to reduce this error. In the light of scarce resources, there might be the need to focus the surveys to limited areas so that the data obtained can be considered to be reliable. A. Sohali: The comment has been well-noted. In future we intend to survey limited areas, as long as enough funds are avaiJal>le. Q. Kwapata: Regarding the insects which were observed as visitors and no tdamaging

cowpeas; were they beneficial or predatory insects which could be manipulated in biological pest control? A. Sohali: They were not beneficial insects I>ut pests of o ther species, such as grasses. Comment. Doto: I would like to support the previous observation regarding the usefulness of concentrating the limited resources available on a given area . Tha t way one may be able to gather deta iled informatio n and obtain a mo re comprehensive picture. The paper was stimulating and thought provoking. In addition, the information being collected is likely to be of considerable value in the development of a I>alanced integrated pest management program. I wish you luck. Q. Barrata: Can you estimate the percent [of seed) attacked by bruchids in the field?

A. Sohali: No. There may be information available in the literature but I have not come across it. Q. Barata: If you have in the same place of storage, one bruchid resistant and one susceptible variety, for how long can the resistant variety escape attack by bruchids?

A. Sohali: It is possible for the resistant check to be attacked by bruchids at the same time as the susceptible one, butthedamage and loss will bedifferent. The resistant line will incur less damage.


Comment. MUgo: Resistance 10 bruchids is rather relative and resistant lines have shown less bruchid attack and damage as compared to susceptible lines. I am not sure whether inforrna tion as to the time i1takes for resistant lines to be attacked by bruchids, is available.


Field Losses Caused by Cowpea Pests at Umbeluzi, Southern Mozambique Richard A. Amable

_ _ tments w .... used to control theexposureoffour cowpea growth stages (ooodIlng _en~ Bower bud initiatlOl\. full /lawerins. and pocIdins) to inoect . _ and tod-.rune thee/fectoflmect pesGon grain yield. Lor8"piotocomprising fuIIyp_ and """",,,,,,ted Weft used to _ yield I_on. field. &caIe. Pl04sunprotectedat Dowerbudiritlatlon (3Sdays afterplanting IDAP]) hod high Dower thrips f.M<to/urotlrri,.. .joIst4h) and legume pod borer (MoruaI t<Slw./is) w .... talklnby 49 DAP. At/ull80wering(S3 DAP). the pod borer was nota significantpesl but unprotected plots hod very high flower thrips infestation .t 59 DAP. Pod bugs. especially A â&#x20AC;˘ .", tum"... w .... observed at full Dowering and poddir1g (65 DAI'). and probably contributed to the high proportion of shrivelled pods In pl04s unprotected ., these stag... The critical stages fur grain yield are Dower bud initiation and full flowering. Plots unprotected at these stages suffored yield MlKtiDns of 17 and 24 percent respectively . Unprotected controls had yield Iosr;eo ranging from 63 per"",t to complete crop failwe in the Jar8" plots. The moot impo<lanl pests obo<rved wen Dower thrips and A. cum"... The pod 00"" and the striped bean weevil (04/._ ,.,.",..........) were secondary pests. It is concluded that uNess flower thrips levels are high enough to completely suppress Jlowering. a &ingle insecticide spray at full DOWering would. probably poduce ecC>r\Cll\kall y acceptable grain yields. This would be buttresoed by Incorporationofthemoderalelevelsof ftower thrips and pod 00= resisI.aIlcethat arerr-ndy .vailable, into elite lines.


Insect pests are an important biotic constraint to cowpea production. Severe pest pressures can cause total crop failure in some locations. Several factors. including availability, cost, handling precautions, and problems with field application, limit the use of insecticides by smallholders. As a result, much emphasis has been placed on host plant resistance. While high levels of resistance are available against bruchids (OUlosclmu:hus mJlcuJalUS) and aphids (Aphis craccivora), only moderate levels of resistance have been identified to date in cultivated cowpea species against other important pests, particularly the legume pod borer (Maruca testulalis), fI.o wer thrips (MegaluTOlhrips sjostedti) and the complex of pod-sucking bugs. Thus, even the best elite lines can suffer significant yield reduction when grown without insecticide protection. The principal objective of the SAOCC Cowpea Project is to develop cowpea cultivars adapted to the region and production methods for the cropping systems of the SADCC region. Understanding the nature of the pest problem in this area is crucial in the development of an effective cowpea crop management strategy and in the seIection of adapted breeding lines. It is important to determine which specific pests cause greatest yield reduction in the different agroecological zones of the region. This study was conducted to identify key cowpea pests and to quantify associated yield 111

losses in the mesothermal subhumid ecology of the main project field station in southern Mozambique.

Materials and methods Two trials were conducted at the INIA Research Station, Umbeluzi, during the 199091 growing season. The cultivar "IT Oesoito路, which is the most widely available improved cultivar in Mozambique was used in both trials. Supplementary irrigation was provided using sprinklers. Experiment 1: Identification of critical pests. To determine the critical pests that reduce grain yield in cowpea, insecticide treatments were used to control theexposure of each of four growth stages seedling establishment, flower bud initiation, full flowering, and podding-to pest attack. Eachof thesegrowth stages is associated with a particular group of pests. The design was a randomized complete block with three replicates. Plots consisted of 10 rows, 5 m long with 75 em between rows and 20 cm between plants. Planting was done on 11 January 1991. Hills were thinned to one plant/hill at 13 OAP. Insecticide treatment regimes were as follows: T1 - unsprayed control 1'2 - unprotected at the seedling stage 1'3 - unprotected at flower bud initiation T4 - unpro~ted at full flowering T5 - unprotected at podding T6 - full protection . Treatments 1'2 to T5 were sprayed thrice and T6 four times. A mixture of dimethoate 40EC and cypermethrin 20EC, at 20 ml and 10 ml in 20 I of water was used for the first two treatments and 40ml and 15ml in 201 for subsequent treatments. These corresponded to approximately 40g ai dirnethoateand 10gai /ha cyperrnethrinatthe first two treatments and approximately 110 g ai dirnethoateand 20 g ai cyperrnethrin ai/ha for the later treatments. Spray dates were 13, 39, 53, and 61 OAP. Plant stands in the two center rows were counted at 13 and 42 OAP. Counts of plants with stems or branches damaged by the striped bean weevil (Alcidcdes /mcogrammus) were made at41 OAP, as were visual scores of thrips damage to stipules (scale 1-5) and observations on the presence of whiteflies and leafhoppers. Cowpea flowers normally open soon after sunrise and close a few hours later. The corolla abscises the next day. Ten freshly closed flowers were sampled from each plot at 49 and 59 OAP for counts of adult flower thrips and MilruCJIlarvae. The number of pods per plant was determined using 5路plant samples taken at 61 and 88 OAP. Pod bug damage was assessed by determining the proportion of shrivelled pods from 5-plant samples taken at first harvest (88 OAP). A second harvest was made at 96 OAP. Grain yield was determined from six bordered rows, an area of 225 m 2. Experiment 2: Yield loss assessment. The trial consisted of unreplicated large plots of cowpea with two treatments: full insecticide protection and an unsprayed check. Two separate plantings were made on 12 Dec 1990 and 16Jan 1991. Each plotconsisted of 25 rows, 13.5 m long at 75 em spacing. Two seeds were sown 20 cm apart within rows. Stands were not thinned. 112

TIle fully protected plots were monitored and sprayed when pest incidence was judged to be high enough to warrant control. A mixture of dimethoate 40EC and cyperrnethrin 20EC was applied as in Experiment l. Plots were sprayed at 17, 30, 40, and 70 DAP in the first planting, and 15,35, and 58 DAP in the second. Harvests for yield determination were taken from three bordered areasof24 m 2 (8consecutive rows each 4 m long) selected at random. Mature pods were harvested twice. Pods were threshed by hand.

Results and discussion Seedling emergence in all trials was excellent and nearly full stands were established inall plots by 13 DAP. However an attack of Sclerotium wi It beginning around 26 DAP resulted in stand losses of between 10 and 33 percent in Experiment 1 and the second planting date in Experiment 2. Growth in all plots was affected by a high incidence of cowpea golden mosaic virus (CGMV) infestation, to which IT Desoito is highly susceptible. First symptoms were observed as early as 21 DAP. Critical pesb. There was very little pest attack during the seedling and early vegetative stages. In Experiment 1, slight leaf puckering, a symptom of leafhopper (Empoasm sp) or foliar thrips (Sericothrips OCCipitalis) damage, occurred at 17 DAP. Between 6and 12 percent of plants were damaged by the striped bean weevil and there were no significantdifferences between treatments. Damagewas rnainlyon the lower branches, with no associated stand losses and very little branch breakage. In experiment 2, leaf miner damage was seen at 29 DAP in the first planting and was the only significant pest incidence during the vegetative stage. In experiment 1, 50 percent of plants had flowered by 45 DAP. All plots showed browning of stipules, a symptom of flower thrips damage, at 41 DAP. Severity was rated 3 on a 1-5 scale. Large numbers of leafhoppers and whiteflies were observed in plots which were unprotected during flower bud initiation (Tt and T3). These treatments also had significantly higher flower thrips (Fig. 1) and legume pod borer (Fig 2) infestation by路49 DAP. In Experiment 2, there was flower thrips damage on stipules at flower bud initiation in both protected and unprotected plots. This was probably due to the low rates of insecticide applied during the early growth stages. At full flowering (53 DAP), the most prominent insects in Experiment 1 were leafhoppers, adult 路Iegume pod borer moths, and pod-sucking bugs especially Anopiocnernis curvipes. The incidence of flower thrips and the legume pod borer were assessed againat59 DAP,aftersprayingat full flowering. The unprotected treatments at this stage, Tl and T4, had more than 9 adult thrips per flower compared to 1 in protected plots (Fig. 1). The legume pod borer was not a significant pest at this stage. A Significantly higher proportion of dry pods in treatments n , T4, and T5 were shrivelled at 88 DAP suggesting higher pod bug damage in these plots (Fig. 3). Treatments nand T5 were unprotected at the podding stage. Damagein T4correlates with the high incidence of pod bugs observed at full flowering when T4 was unprotected . During pod development at around 65 DAP, large numbers of leafhoppers, whiteflies, and blister beetles (Mylabris sp) were observed in unprotected plots in Experiment 2. There was heavy pod bug damage in unsprayed plots, the dominant species being A. curoipes. 113

No. of thrips per Oower 14,-------------------------------------------,


SE@49 DAP =0.6 thrips/Oower SE@59 DAP = 1.7 thrips/Oower








T1 T2











Insecticide spray regime Fig. 1. Flower thrips infestation at 49 and 59 DAP Plots that were unprotected at Oower bud initiation and at full Oowering (T1 , 1'3, and T4) had fewer pods per plant at 61 DAP (Fig. 4). This could have been due to the effects of Oower thrips and the legume pod borer on Oowering, and pod bug damage to developing young pods. Differences in podding were not significant at 88 DAP (Fig. 4). The incidence of CGMV resulted in high variability of plant samples taken for the assessment of podding. However, when compared to the fully protected control, podding in T4 appeared better at 88 DAP than at 61 DAP while that in T5 seemed to decline. These effects could be ascribed to the abscission of young pods resulting from severe pod bug pressure in unprotected treatments. The apparent recovery of T4 was probably due in part to the protracted Oowering and podding period of IT Desoito. About55 percent of pods in all plots were green at 88 DAP. 5egeren and vanderOever (unpublished) observed peak legume pod borer and Oower thrips populations at Chokwe, southern Mozambique between early January and mid-February. Similar population dynamics were observed at Umbeluzi. Reproductive structures produced 114

Percent shrivelled pods ~r---------------------------------------~







Insecticide spray regime

Fig. 3. Pod bug d amage al88 DAP

Table 1. Grain yield from cowpea plots unprotected and protected againsl insect pests at Umbeluzi, 1990-91 Treatment

Planting dale 16 Jan 1991

12 Dec 1990

11 Jan 1991

-·-·----·---··Grain yield (g.m-2 ± 50)-·---·---Unprotected Fully protected

Yield loss (%)

5.1 ± 4.0 SO.S± 0.6

0.3 ± 0.3 128.0± 6.8


90 116

38.l± 5.7 103.3± 263 63

Pods per plant

W,-~~----------------------------------, _ 61 DAP 88 DAP SE @ 61 DAP = 4.4 pods


SE @ 88 DAP = 65 pods 50





0 '---- T1






Insecticide spray regime Fig. 4. Pod production at 61 and 88 DAP

The critical stages for cowpea grain yield were full flowering and flower bud initiation (Fig. 5). Unprotected treatments at these stages suffered yield reductions of 24 and 17 percent, respectively, compared to the fully protected control. At full flowering. flower thrips, the legume pod borer, and pod bugs were all present in significant numbers, exerting heavy pest pressure on flowers and young pods. Yield losses from lack of protection at the seedling and late podding stages were only 10 and 3 percent respectively, and not significantly different from the control. Seedling pests cause yield losses through severe stand reduction .and defoliation. However, these pests tend to be sporadic. For instance, cutworms (Agrotis sp.) caused up to 100 percent stand losses in some plots of a cowpea trial within 100 m of these trials. Although incidence of the striped bean weevil is fairly predictable at Umbeluzi, 117

Gain yield (Mg/ha) 1.2 SE = 0.15 (Mg/ha) 1





o '--_....L T1






Insectkide spray regime Fig. 5. Final grain yield (Mg/ha)

the levels of infestation which occurred were too low to affect yields significantly. Similar results were obtained in the control of early cowpea pests at Morogoro, Tanzania (IITA 1988)_

Conclusion TIle most important cowpea insect pests observed at Umbeluzi during the 1990-91 season were flower thrips and thecoreid pod-sucldngbugA. curoipes. Control of these pests at full flowering and flower bud initiation is necessary for maximum cowpea 118

grain yield. Unless flower thrips populations were high enough to severely supress flowering, a single insecticide spray at full flowering would probably produce economically acceptable yields. An overwhelming proportion of flowers on fullyprotected cowpea plants fail to develop into mature pods (Ojehomon 1968; Sinha and Savithri 1978; Wien and Tayo 1978). Secondary pests at Umbeluzi (such as the legume pod borer and the striped bean weevil) and sporadic pests (such as cutworms) constitute potential threats to grain and leaf production. Reliablemethods of forecasting outbreaks would behelpful in control of these pests.

References DTA. 1988. Grain Legume Improvement Programme, Annual Report. Ojehomon, 0.0. 1968. Flowering. fruit production, and abscission in cowpea, VignRunguicu/Q/R (L.) Walp. Journal of the West Africa Science Association 13: 227-234. Sinha, S.K.â&#x20AC;˘ and K.S. Savithri. 1978. Biology of yield in food legu mes. Pages 237-240 in Pests of grain legumes: ecology and control, edited by S R Singh, HF. van Emden, and TA. Taylor. Academic Press, London, UK . Wien, H.C,and Tayo T.0.1978. Theeffectofdefoliationand removal ofreproductive structures on growth and yield of tropical grain legumes. Pages 241-252 in Pests of grain legumes: ecology and control. edited by S.R. Singh, H.F. van Emden, and T.A. Taylor. Academic Press, London, UK.

Discussion Q. Kwapata: What was the flowering pattern of the indeterminate cowpea? Was it continuous or did it occur in flushes? If it was continuous, one would expect that the insect pests attacking flowers would persist. How would you then reconcile your statement that the indeterminate cowpea has an advantage over the determinate cowpea in producing some yield after the peak period of insect pest occurrence with this observation.

A. Amable: Flowering occurred continuously over a protracted period and supplemental furrow irrigation was provided throughout the trial. The presence of flowers or pods on which a pestfeeds is not the only determinant in population dynamics. The pest population could still crash in spite of the abundance of these structures, due to the influence of other factors such as rain and temperature. Q. PhiUi ps: The yield losses in the large plots were greater than in the small plots. Do you think that this could be due to the spaces between plots to some extent mimicking conditions in intercropping and that intercropping may be very valuable in reducing the effects of insect pests. A. Am.ble: ) cannot explain these yield differences at the present. There was some insecticide drift in the small plots in spite of the alley separation because of the windy nature of the experimental site. The large plots were situated in a different part of the site and could have had different pest pressure. For example, a trial situated about 50 m away from the small plots was devastated by cutworms while the small plots escaped attack entirely. All I can conclude at present is that completely unprotected plots suffered between 60 to 100 percent yield loss.


Intercropping reduces the incidence of some pests, e.g., Oower thrips through shading while other pests, e.g., blister beetles and other pollen feeders could increase, especially in maize intercrops. 1he alleys would only restrict insects which have low mobility. Q . Dashiell: Are there any insect pests that cause a reduction in leaf yield or quality?

A. Amable: Foliage feeders were not important in this study. However, there was some leaf miner damage atearly stages 01 growth in the large plots. This reduced leaf quality, but we would have to determine what degree of damage is unacceptable.


Exploratory Field Testing of Plant Extracts Against Some Pests on Cowpea in Zambia P.H. Sohati and S. Saithananthan Field trials conducted during 1989-91 at Msebra, Zambia, have shown that proper comol ol pod damage by boftrs (including spotted borer, MIl""'" ,sullll"), and seed damage by ap;on weevil (PinDI,.tlodus ..';um) can be achieved by spraying cowpea plots with leaf extraets ol neem (Md;' 'ud<roch~ Neem sprays also teduted seed damage by pod suclting bugs. Significant seed yield increase was observed due 10 neem spray in one trial . Leaf.xtractsol sisal (Ag.wsp.) and ububo (T.,,1rrosio oogdi.) were also tested. although their effectson damage by the insects mentioned earlier and on seed yie1d are not so distinct The potential for the use of neem in pest management on cowpea is discussed.

Cowpea is subjected to substantial damage by insect pests in Zambia (Sithanantham et al. 1989). While resistant varieties are the most durable and economical means of minimizing losses due to pests on cowpea, short-term strategies can be a useful supplement. The use of plant derivatives, wh ich have insecticidal properties, is one approach which appears to bea locally adoptable method ata small farm level. Among a large number of such plants researched upon isneem(Melia azaduach L.) (Schmulterer 1990). Another plant which has received interestis the legume, ububa, Tephrosia voge/ii Hook, which is a known source of rotenoids (Gaskins et al. 1972). 1ltese two plants, along with another locally available fibre-yielding plant, sisal (Agave sp), were evaluated against major pests of cowpea in Zambia recently and the results are discussed in this paper.

Materials and methods Ouring 1989-90, a field trial was conducted at Msekera with cowpea (11'820-889) in a randomized complete block design with seven treatments (Table 1) and four replications, using plots of 4 rows of 3 m, with a spacing of 50 by 10 em. No fertilizer was applied. The treatments included sprays of leaf extracts of two locally occurring plants-f\eem and sisal. The extracts were prepared with freshly plucked leaves, chopped and soaked in equal weight of waterovemight and filtered . The extracts were later stored ina refrigerator (8°C) and used within 3 months. Theextractswerediluted at a dose of 20 kg leaf per hectare, mixed with about 500 liters of water and sprayed wi th a high volume spray equipment (manual Knapsack sprayer). A1phamethrin at 50 g/ ha was used as a chemical insecticide check. Sprays were either applied three times at intervals of seven days or twice at intervals of fourteen days, commencing at midflowering or early podding. Observationson pod damage severity (by the spotted pod borer, Maruca testulalis (Geyer» and ,seed damage by all borers (including the spotted pod borer) as well as by aplon weevil (Pinotrachelus varium (Wagner» were made on a visual rating scale of 1-9 (1 = no apparent damage; 3 = light (I-lOOk) damage; 5 = 121

moderate (11-25%) damage; 7 =severe (26-50%) damage; 9 =very severe (aboveSO%) damage). The seed yields were also recorded. Table 1. Evaluation of two plant extracts in comparison with alphamethrln against pest damage in cowpea, Msekera 1989-90"


Pod damage by spotted borer (1-9 scale)

Seed damage b~ bOrers Apion (1-9 scale) weevil

Seed yield (kg/ha)

Neem· 3 sprays Neem - 2 sprays Sisal - 3 sprays Sisal- 2 sprays Neem+Sisal-2sprays A1phamelhrin-2 sprays No sprays (control)

2.8 3.0 25 33 2.8 2.0 4.0

2.8 3.8 33 4.0 35 3.0 45

33 4.8 45 55 6.0 53 6.8

712 570 500 338 432 524 364






12.8 0.2 05

12.8 .2 0.7

13.7 0.4 1.1

C.V.(%) S.E. ±(m) LSD (0.05)

• ..

295 69.0 202.9

Design: RBD; Replications: 4: Plot size: "' rows x 3 m Spacing: 50 x 10 em Plant extracts at 20 kg/ha and aJphamethtin 50 kg/ha for each spray

During 199(}-91, two trials were conducted, using cowpea genotypes 1T82D-889 (early maturing) and ZVu 83 (medium maturing). Each trial consisted of eight treatments (Table 2), with four replications in a randomized complete block design, in plots of 4 rows of 3 m. The spacing and other practices were the same as for the previous year's trial. One more locally occurring plant (ububa) was included along with neem and sisal, and these were compared alone as well as in mixtures (of two plants) along with triazophos (Hostathion 40%EC) as a check. Plant extracts when sprayed alone were applied twice a week (3-4 day intervals) and once a week when applied as mixtures. Spraying was started at early flower stage and continued till late podding. Observations were on pod set, pod damage by borers, seed damage by bugs (mainly Anop/ocnemis and C/nvigra//a) and apion weevil following the 1-9 scale. Seed yields were recorded only in one trial, since the second trial (ZVu 83) had problems of poor plant stand.

Results Ouring 1989-90, the results (Table I) showed that pod damage by the spotted pod borer was least with alphamethrin sprays, followed by sisal sprays. The other spray treatments also brought about significant reduction in pod damage. 122

Seed damage by all borers (including the spotted pod borer) was least in plots receiving three sprays of neem. All other trea tmen ts except two sprays of sisal res ulted insignificant reduction in seed damage by borers. Three sprays of neem also resulted in least damage to seeds by apion weevil. Two sprays of neem and three sprays of sisal were next in rank, followed by alphamethrin and two sprays of sisal. A mixture of neem and sisal, however, did not cause significant reduction in seed damage by apion weevil. Seed yield was greatest (712 kg/hal in plots sprayed thrice with neem, while two sprays of neem resulted in yields similar to wha twas obtained with alphamethrin. The other spray treatments had little or no effect on seed yield. In the 1990-91 season trials, the extent of pod set (as affected by flower damaging pests especially thrips and beetles) was significantly improved by triazophos sprays alone and not by any of the plant extract sprays (Table 2). Pod damage by borer was significantly reduced in plots receiving sprays of neem, triawphos, sisal, ububa, and neem plus ububa. Table 2. Performance of extracts from three plants in comparison withhostathion against rowpn insect pests, Me.. kera, 1990-91 Seed damage by Treatment路

Neem 1 Sisal I Ububa 1

Neem ... Sisal 2 Neem ... Ububa2 Sisal ... Ububa 2 Hostathion3 No spray Mean

C.V.% S.. E. 卤 (m)

LSD (0.05)

Pod set Pod damage by borers score (1-9) (1-9)

bugs (1-9) B




6.0 55 5.0 4.8 6.0 4.8 7.8 5.3

4.0 4.8 5.5 6.8 5.8 6.0 4.3 7.3

3.5 4.5 4.3 5.3 4.5 4.3 4.8 6.0

3.3 4.0 4.0 4.5 4.0 4.5 3.0 4.5

5.6 13.2 0.4 1.1

5.5 18.7 0.5 1.5

4.6 14.2 0.3 1.0

4.0 17.3 0.3 1.0

apion weevil B 2.3 2.8 2.8 3.0 2.8 3.33 2.8 4.0 2.9 22.8 0.3 1.0

Seed yield (kg/ha) A

1323 1330 1199 1187 1197 1151

1558 1159 1263 14.9 94.2 277.1

120 kg Jeaf/ha; Iwo sprays a week; 2 20 kg leaf/ha: one spray a week; 3 0.04 percent; one spray a week A =Trial! (cv. JT82D-889; B =Tri.l 2 (cv. ZVu 83); Design: RBD; Replications: 4: Plot size 4 rows x 3 m;

spacing: SO x 10 em

Seed damage by pod sucking bugs was also least with neem spray, while all the other spray treatments except neem plus sisal, also resulted in significant reduction in damage. Data from the two trials on seed damage by apion showed that neem and Hostathion sprays resulted in significant reduction indamage in both trials.1he other treatments except sisal plus u buba, caused significant reduction in damage only in one trial. Seed yields were greatest for plants sprayed with Hostathion. High yields were 123

also recorded for plants sprayed with neem and sisal. However these yields were not significantly more than the yields Irom the unprotected plots.

Discussion In general, the results have shown that leal extracts 01 neem sprayed during the reproductive stage 01 cowpea brought about significant control 01 seed damage by lepidopteran borers (including the spotted pod borer) and apion weevil. In one trial, seed damageby pod sucking bugs was also significantly reduced by neem sprays. This resulted in enhanced seed yield, but the increase was significant in only one of the two trials. The insecticidal effects 0/ constituents lrom M. awderach (Kraus et al. 1987) and its related genus, Azadirachta indica A. Juss (Schmutterer 1990) are well documented. Saxena (1989) has indicated that neem derivatives are reported to be active against several genera of caterpillar pests (lepidoptera), beetles, weevils (Coleoptera), and bugs (Heteroptera) among some other insect orders so lar researched upon. It therefore appears useful to pursue further research towards optimizing the dosage and frequency of sprays of neem against some of these cowpea pests. Among the other two plants tested , sisal appears somewhat promisingcompared to ububa and so could be further evaluated. Ububa b a l-.nown sou rce of rotenoids, which have insecticidal properties (Gaslcins et al. 1972), but apparently at equal quantities as neem or sisal on fresh weight basis, the extracts from this plant have not shown promise against the cowpea pests concerned. The use of locally occurring plant extracts offers an economically viable and attractive alternative in African agriculture, as plants such as neem and sisal are handy and easily propagated. They are also environmentally more acceptable, compared with the use 01 synthetic, broad-spectrum insecticides. It therefore appears important to further evaluate neem leal extract as a potential alternative lor the control of major insect pests on cowpea in the SADCC region. Screening 01 other plants should be also encouraged where appropriate.

References Coskins, MH., C .A. White, FW. Martins. N.E. Delfel, E.G. Ruppel, and OK Barnes. 1972. To/hrosiD vogeUi:01 sourceo( rotenoids (o r insec ticidal and piSCicidal use . Uni ted Stntes Drug Agency A.R.S., Techni",,1 Bulletin No. 1445. Kraus, W., S. B'lUmon, M. Bakel, I. Keller, A Klen k, M. Kingle, H . Pohnl, and M. Schwinger. 1987. Control of insect feeding and developmen t by constituents of Melia azodcracIJ and AYldi rocilla i"dico. In Natural pesticides from the neem tree and other tropical plants, edited by H . Schm utterer and K.R.5. Asdler Rossdorf:TZ-Verlagsgesellschaft. Departmen t of Agriculture, Ministry of Agriculture, Lusaka, Z,mbia. Saxena, R.c. 1989. Insecticides from neem. Insecticides from plants edited by J.T. Amason, B.).R Philongene, and P. Morand. ACS Symposium Series No. 387, Washington, D.C. Schumtterer, H.l990. Properties and potentiill of natural in.secticidesfrom neem tree,Azadirachhl indica . Annual Review of Entomolology 35: 271-297. Sithonanthon, S., N.5.Irving, and P.H . Sohati. 1989. Recententomological research on bean and cowpea in Zambia . Pages 135-156 in Proceedings of the National Worksh op of Food Legume Research and Improvement in Zambia, !>-11 March, 1988, Mfu we, EasternProvince, Zambia . Deportment of Agriculture, Ministry of Agriculture, Lusaka, Zambia. 124

Discussion Q. Mligo: Neem is also used as a medicinal plant in human ailments. 1he question in most cases is the effective dose. What rates of application of your neemextractsdid you use against cowpea pests and have you done studies to determine the best rates for control.

A. The rate applied was 20 kg leaf/ha. No studies have been done as regards the best rate for cowpea pest control. This and the frequency of application of the leaf extracts will constitute the next step in our research. Q. Kwap.t.: It is reported thatgrowingsorne plants together, forexampleAmalllnlhus sp. with cabbage, tends to reduce insect pest infestation. Is this a viable option in cowpea production systems in African countries where resources are limited?

A. Sohali: Of course it is feasible but only on a small scale. Amaranlhus is not the only candidate but also any plant or weed which provides food or refuge for natural enemies of the crop. Q. Matilo: In your tests with neem, why did you not consider using seed extracts?

A. Sohali: The neem tree Ml!lia azaderach, used in our study does not produce seed in the ecosystem inwhich itwasgrown. M. azadf!l'achis not like the related genusAzadirachla indicA. We would have liked to use the seed extract for our study had it been available.


Prevalence of Cowpea Diseases and Extent of Yield Losses Caused by Scab in Zambia H.C. Haciwa and J. Kannaiyan Roving surveys were carried out during 1987-98 to determine the motive prevalence

01 cowpea diseases in theDU1jor crop growing aJeaSin thecountry.Su.veyofI3disbids infiveprovinces revealed the wideprevil1enceof diseases suchascet cospora leaf spots. scab, ascochyta bligh~ web blight, dactu1iophora leal spot. powdery mildew, bacterial blight, and cowpea aphid路borne mosaic virus (CAMV). An updated check list of diseases recorded on cowp<8 in Zambia is Wo presented. lhe extent of avoidable seed yield lass due to scab disease and CAMV was also in\路estigated. Scab disease cauwd. a yiekl loss 01 up to 4S percent in a susceptible

gmotype. lT82D-87:i.

Cowpea [Vigna unguiculata (L.) (Walp.)] is an important food legume crop. grown mostly in the low to medium rainfall regions of Zamb:".!t is cultivated in a complex of cropping systems by small farmers for grain, fresh pods. and leaves, Diseases are a constraint to cowpea production in the coun try (Kannaiyan and Haciwa 1989). Angus (1962-66) reporl1!d the occurrence of several diseases including scab (Sphaceloma sp = Elsinoephaseoli Jenkins), cercospora leaf spot (Pseudocercosporu cruenta (Sacc.) Deigthon and CnrosporaamescensEll. andMart.),powdery mildew(ErysiphepolygoniDC),septoria leaf spot (Septoria vignae p, Henn),ascochyta blight (Phoma e:tigua var. diversisporu (Bub,) boerema), dactuliophora leaf spot (DYlctuliophoru larrii Leaky). brown rust (Uromyces appendiculatus (Pers.) Unger), anthracnose (Colletotrichum lindemuthiallum (Sacc. and Magn.) Bri. and Cav.) and root knot (MeJoidigyne jauallica (Trenb.) Otitwood) on the crop in Zambia. Kannaiyan et al. (1987) reported the outbreak of cowpea aphid-borne mosaic virus (CAMV) and false rust (Synchytrium do/ichi (Cooke) Guam) in the 198384 cropping seasons at Msekera and Masumba research stations in the Eastern province. Subsequently, some other diseases-web blight (Rhizoctonia soiani Kuhn = Thanatephorus cucumeris (Frank) Donk,) fusarium (FUSIlTium oxysporum ScI. f. sp. Tracheiplrilum (Smith sny. and Hans.), brown blotch (Colletotrichum capsid Sny. (Butl. and Bisby)and C.lruncatum(Schw.)Andrus &: Moore),lamb's tail pod rot(Choanephora cucurbitarum (Berk. and Rav.) 1haxt. and C. infundibulifera (Currey) Sacc.), sclerotium stem rot (Sclerotium ro!fsiiSacc.), bacterial blight (Xanthomonas campestris pv. vignicola (Bruk.) Dye) and a parasitic weed (Altetra sp.}--have also been observed and identified on the crop. Since 1988-89 onwards, 'Yellow speckle', a suspected virus symptom, has been recorded on a few genotypes at Msekera as well as in other research stations. However, its causal agent has not been established. An outbreak of systemic necrosis (similar to blackroot symptoms of BCMV on several CAMV -resistant cowpea genotypes) was also reported at Msekera during 1989-91 cropping seasons. It is suspected to be caused by a necrotic strain of CAMV. See Table 1 for an updated check list of all the diseases recorded on cowpea up to 1990 in Zambia. This paper also p.r ovides information !Xl the prevalence of cowpea disease in the farmers' fields and research stations and on the extent of seed yield losses caused by the important diseases, scab andCAMV. 126

Prevalence of cowpea diseases Farme..' field.. In order to study the relative occurrence of cowpea diseases in farmers' field, roving surveys were conducted in the Eastern, Luapula, Northern, Southern, and Western provinces during the 1987-S9 crop seasons. The surveys were carried out during the podding stage (March-April). Observations were recorded in cowpea fields located about 20 Icm from each other. The occurrence and severity of various ctiseases were scored on a 1-9 visual rating scale (1 = no disease and 9 = very severe ctisease). The mean score for each district was calculated and classified as low (1-3 score), moderate (4-5 score), and severe (6-9 score). The prevalence of important cowpea diseases in eachdistrictis summarized in Table 2. Results show thatcercospora leaf spots (CLS) were observed in most field and the overall severity was moderate. Scab diseases were next in importance and were moderately severe in Kasama and Mongu districts. Ascochyta blight was severe in Kasama district where rainfall is usually high. Prior (1979) reported the destructive nature of ascochyta blight in association with scab on most cowpea trials at Misamfu (Northern province) in the 19605. Dactuliophora leaf spot was also moderately severe in the districts of o.ipata, Katete, Lundan (Eastern province) and Mansa (Luapula province). Powdery mildew was recorded only in some fields in Eastern province. The survey also revealed the low severity of CAMV in a few fields in Eastern, Southern and Western provinces. This may be due to sparse cowpea population and low aphid,vector activity under the predominant cereal intercropping system. Low severity of bacterial blight and a parasitic weed, alectra, was also recorded in a few fields. Research .tationa The importance of cowpea diseases under on-station conditions as observed over several seasons is relatively differentfrom what is obtained on-farm .In most seasons,

CAMV caused severe damage to the crop at Msekera and other stations (Masumba, Kaoma. Mochipapa, and Misamgu). Both scab and ascochyta blight are important not only in high rainfall regions (Misamfu), but also at Msekera, Kaoma, and Mochipapa under favorable high rainfall crop periods. All the three diseases cause severe yield losses to the crop on-station. On the other hand dactuliophora leaf spot. bacterial blight. and cercospora leaf spots only cause moderate damage.

Seed yield 1 _ due to cowpea scab To estimate avoidable seed yield losses to scab, comparisons were made between 'protected' and 'unprotected' treatments in a replicated field trial. The trial was conducted with three genotypes IT82D路875 (scab susceptible), 1Vx 4654-44E (moderately scab susceptible), and ZVu '137 (scab resistant) at Msekera, Eastern province in the 1989-90 season. The protected treatment received five sprays of Benlate (500 g/ha) + Dithane M45 (1.5 kg/hal at 40, SO, 60, 70, and 80 days after emergence. The unprotected treatment received only water spray. The extent of loss was estimated from the differences in seed yield between protected and unprotected treatinents. 1here were significant ctifferences between the sprayed treatments (p < 0.01) and between genotypes (p <0.01). The avoidable seed yield loss due to scab disease across 127

three genotypes ranged from 1810 45 percent (500 10 1118 kg/hal (Table 3). The extent of seed yield I05S d ue to scab was higher (45%) in scab susceptible genotype IT82D-875 than in moderately susceptible genotype 1Vx4654-44E (19%) and resistant genotype ZVu 237 (18%). Emechebeand ShoyinJca (1985) also reported up 1060% crop lo5s due to scab in farmers' fields and in experimental plots in Nigeria.

Table 1. Diseases of cowpea in Zambia Pathogen 1. Fungal ctiseooeo Scab

Sp/Ulu""" 'P. (Elsi""" phostolj Jenkins)


Cercospo<a leaf spols

P""'*'=co<ponr trunot. (Sacc.) Deighton and Cercc<ponr â&#x201A;Źlin"""" EO. and Mart


Powd .. y mildew

Erysiphe po/ygOllj DC


SeplOria leaf spot

5<pIon. tng ... P. Henn


Astochyto bUght

"""'"" nig ... Vllt. dim-sisponr (Bub.) Boerema (Ascochyt. pIuIworum Sacc.)


RlU2Ddonill soI#ni Kuhn (lM ...",1IDrw C"""","" (Frank) Donk).

PotentiaUy ilnpotant

Doctwiopllon 'am; LeKy

PotentiaUy important

Syochytrium dc/ichi (CooIee) Gaum



Urvmyus .ppmdiculotw; (pen;.) Unger


Fusarium wilt

FUSQrium o.zysporwm Sch1. f.sp. ''''''''phi/um (Smith Soy. ""d Hans



Collmrichum lindtmulhumu.". (Sacc. and Bri. and eav.


Brown blotch


_y ..


Lamb'. taiJ pod rot

o.o.nqmo... """,bitl,.."" (Beric. and Rav.)


Web b light

Folse ntsl

",,,.ici (Sny.) Bull and d C. truncatum (Schw.) Andrus 4. Moore


Thaxt. C. i. ftllldibwiforo (Currey) Sacc. SdmJ/ium roIftii Sacc.


Mosaic/ black root

Cowpea aphklbome mosaic virus (CAMY)


Yellow spodde?

Virus? (unidentified)


x..tlo>mo.... annpeslri. pv. rng";roIo (Butk.) Dye

Potentially ilnporlant

Me/oidogyne jooa.i", (Trerb) Chitwood


..11m,. 'P.


Sclerotium stem rot

2. Viral diseases

3. Bacterial disease Blight

4. Nematode disease Rootlalol

5. P..asitic weed Alectra






Luapula Man ..

L(J-S) M(1-a) Nil


M(5) Nil

5: s.v.... (6-9), M = Moder.te (4-5) and. L = Low (2-3)

M(l-<;) MI-S) M(4- 5)


L(3) Nil


Nil L(I-J) L(3) L(I-J) Nil Nil



L(I-a) Nil Nil





L(2-J) M(H) L(2) M(3-5) L(2-J) L(2)

Cerrospo" leaf spots

Scored on 1-9 scale, wherel~ no disease and 9= very severe disease


10 8


J 1


L(I-J) L(2-J) L(2) L(2-J) Nil Nil

Ascochyta blight


Mongu Senanga








2 1






N limber offield.

Eastern Chadiza Chipata kateta Lundazi Mambwe Petauke

Province/ district

L(1-a) Nil Nil


Nil Nil


L(2-J) M(3-6) M(4) M(3-7) Nil L(3)

Nil Nil Nil


Nil Nil


Nil L(1-3) Nil L(1- 3) L(J-J) Nil

D.ctuliophor. Powdery leaf spot mildew

L(1-S) L(1-4) Nil


Nil Nil


Nil L(I-J) L(2) L(I-J) Nil Nil

L(I- S) M(1-6) Nil


Nil Nil


L(I-J) L(I-J) Nil L(J-J) Nil Nil

Nil Nil Nil


Nil Nil


Nil L(I-J) Nil L(J-J) Nil Nil





Mosaic (CAMV)

Mean di .... se severity I and ,ange 2

Table 2. Relative severity of some cowpea disea.esln five provinces of Zambia, 1987-89

Cowpea aphid-borne mosaic virus During the 1988--89 cropping season, a preliminary seed yield loss study was carried outon four mosaic disease susceptible genotypes (Muliana, ZVu 83, ZVu 237, and lYx 4654-44E) at Msekera, Eastern province. In each genotypes, 100 plants showing mosaic and another 100 apparently healthy were tagged. At harvest the number of pods per plant, seed yield, and 100 seed weight were also recorded. The results are summarized in Table 4 . In all four genotypes, plants that had mosaic produced significantly fewer number of pods as well as sma ller-sized seeds in comparison to healthy plants_Overall mean seed yield loss due to mosaic disease was 52 percent. The loss in seed yield was greater (60%) in ZVu 237 than in others (48--50%). Kaiser and Mossahebi (1975) also reported seed yield losses of 15 to 87 percent due to infection by CAMV in Iran. Table 3. Effect of scab on seed yield and its components in three cowpe. genotypes, Msekera, Zambia, 1989--90 No. of pods I plant

100 seed

Seed yield


(kg / ha)







Percent Joss in seed yield

ZVu 237(R) TVx 4654-44E(M) 1T82D-875 (S)

12.0 123 14.8

103 10.4 103

10.8 14.9 13.9

9.4 12.9 12.7

2767 2759 2508

2267 2233 1390

18 19 45

Mean C.V. ('Yo) SE ± (Genotypes) SE ± (p lit UP) SE ± (Interaction)






1963 7.9 132 54 94


Geno types

18.6 0.56 0.62

6.4 0.14 0.23 039


P = Protected (fiye sprays Benlate .. Dithane M45) up:; UnprotE.'('ted (only water spray) Split plot d ~ign with 4 rt>pliciltions, main plot (Cenotypes) Sub plot (Sprays treatment") R= Resistant to scab, M = Mooerately resis tant to scab 5 = Su~ptible to s<:ab.

References Angus, A. 1962-66. Plant pests and diseases in Zambia, parts 1-7 and supplements,MT. Makulu Research Station, Zambia (mimeo). Emechebe, A.M., and SA ShOyink.l. 1985. Fungal and bacterial diseases of cowpeas in Africa. Pages 173-1 97;" Cowpea Research Production and utilisation, edited by S.R. Singh and K.O. Ramie. John Wiley and Sons Chichester, U.K. Kaiser, W.J., and H. Mossahebi. 1975. Studies with cowpea aph id-borne mosaic virus and its effect on cowpea in Iran. FAO Plant Protection Bulletin 27: 27-30. Kannaiyan, J., D.C. Greenberg, H.C. Haciwa, and M.N. Mbewe. 1987. Screening cowpea for resistance to major diseases in Zambia. Tropical Grain Legume Bulletin 34: 23-26.


ICannaiyan, J. and Haowa H .C. 1989. Bean and cowpea pathology research in Zambia. Pages 117-134 in Proceedings of the National Workshop on Food Legumes research and Improvement in Zambia, 9-11 Mar 1988, Mfuwe, Eastern Province, Zambia . Prior, A.J. 1979. Cowpea. in Zambia. A review paper, Mt. Makuku Research Station, Zambia (Mimeo).

Discussion Q. Phillips: Do you have any explanation for some genotypes being resistant in two seasons and susceptible in the other two seasons?

A. Hadwa: [do not know the reason but suspect that a new strain of the virus appeared during the last two seasons. Q. Data: Do you have agronomic performance data on the lines identified to possess

high levels of resistance to CAMV? A. Haciwl: We do have data On the high-yielding ones which have been included in the national variety trials. Q. Matilo: Do you think weeds have any role in the incidence of CAM V, ie. could they be alternate hosts?

A. Haciwl: 'That is a possibility. However, some preliminary experiments we have carried out at Msekera indicate that small cowpea plots located more than 700 m away from the main cowpea fields never got any CAMV.


Recent Research on Cowpea Aphid-borne Mosaic Virus in Zambia J. Kannaiyan, H.C. HaciWil, S. Sithananthan, P.H. Sohati, and J.M. Mulila Cowpea aphid路borne mosaic virus (CAMV) is one 01 the importantdiseaseoof cowpe.1 in Zambia where it causes yield losses of up to 60%. The virus is seedborne in cowpea and oocoadary spread is by aphid vectors. Several cowpea genotypes were screened for CAMV over many seasons and thirty路five genotypes have been identified as resislant. Ten CAMV resistant genotypes (lT82D-889. IT831).32~S.IT8SF-867路S. ITSSF-26B7, IT83e-.42, ITSSF-13BO.1T82D-87S. ITSSF-1517, JT84S-2221-2 and TVx 3381-C2F) which aIso.howed high yield potential .... being evaluated in nationol variety trials .t severlllloc.tiono in Zambia. Some of these genotypes are being used in the cowpea hybridization program in the country.

Cowpea aphid-bome mosaic virus (CAMV) disease occurs in Zambia as well as in Uganda, Kenya, and Tanzania (COPR, 1981). A severe outbreak of CAMV was recorded on both exotic and local 1andraces/genotypes of cowpea at Msekera and Masumba research stations, eastem province. in the 19~ season (Kannaiyan, etal. 1987). Since moderate to servere CAMV has been observed in almost all cowpea testing sites in the country. Natural field infections of CAMV caused 48 to 60'0/. seed yield losses at Msekera (Haciwa and Kannaiyan 1991). Unlike exp_ station site the prevalence of the disease is low in farmers' fields. 'This may be due to the low cowpea plant population and less aphid vector activ i ty in the traditional, intercroping systems in the country_ The virus is primarily seed bome. secondary spread is by several aphid species including Aphis cnu:ci~ Koch. The diseased plants show varying degrees of mosaic, mottle, vein banding, interveinal chlorosis, leaf distortion, blistering, and stunting. The development of resistant varieties is the only simple and economic solution toCAMV. A screening program for identifying sources of resistance and transferring such resistance to the high-yielding varieties has been started at Msekera (Kannaiyan et at. 1987; Kannaiyan and Haciwa 1989). The present paper reports on (1) identification of cowpea genotypes resistant to CAMV the (2) seedbome nature of this virus (3) and on the incidence of a hypensensitive response in some cowpea genotypes.

Seedbome CAMV The primary means of spread of CAMV is through seeds. A very low percent (1-5) of seedbomeCAMV is sufficient to cause epidemics in susceptible genotypes by secondary spread of aphid vectors as observed in the warmer low to moderate rainfall regions in Zambia. To estimate percent seedbome CAMV. seed of 22 cowpea genotypes was sown in the field at Msekera and the number of seedlings showing seedbome CAMV were recorded during 1989-90. Seedbome CAMV was recorded in 14 genotypes, representing 64 percent of the total samples_ The mean percentage of seedlings with 132

CAMV in individual genotypes varied from 0.2 to 6.3 with an average of 2.8. Several workers have reported that levels of seed transmission vary with genotypes (Bock and Conti 1974; Ladipo 1977). During 1989-90, typical systemic necrosis symptoms, similar to were observed for the fust time on several genotypes that were described as resistant to CAMV mosaicresistant at Msekera.1hese symptoms were also recorded the following season (199091). The systematic necrosis is characterized by initial appearance of necrotic local lesions on leaves, fo\1owed by apical dieback, vascular necrosis, and finally plant death. In a few cases part of the plant died, while the other living branches produced new leaves with typical mosaic symptoms. Systemic necrosis was recorded in the following genotypes (ZVu 259, ZVu 590, ZVu 593, ZVu 594, lT845-2135, IT86D-801, and IT86D-795). Thediseased samples sent to the Horticulture Research International, Warwick, UK, showed positive serological reactions to blackeye cowpea mosaic virus (serologica\1y related to CAMV) necrotic ring spot isolate (personal communication from N. Spence 1991).

Sources of resistance to CAMV In order to induce high CAMV pressure,two cowpea varieties, 1Vx 4654-44E and muliana, highly susceptible to CAMV were sown two weeks before the test material CAMV severity on te5t lines was scored onascale ofa1t09where 1 =9no disease and 9: very severe disease. Test lines which registered scores between 1 and 3 ronsidered resistant (Table). CAMV was very severe in all the four seasons, as evidence by highscoresof7 and 9A1l 35 entries were resistant (1-3 scores) to the local isolate at Msekera all field screenings. The isolates were ZVu路273, 746, 1Vu401,41O, 1948,2480, 11710, 12369, 1Vx 338t-()2F, 11'820-875, 1T82D-889, IT830路328-1, 11'83D-442, IT835-68(}.9, lT83S-818, IT840-1008, IT845-2221-2, IT85F-867-5, lT85F-l380, lT85F-1517, lT85F-2687, lT85F2716, lT85F-2805, lT85F-3139, IT860-314, IT860-409, IT86D-440, IT860-520, IT86D1010, IT86D-I035, lT86D-1048, lT860-1052, IT86D-1086, and lT860-1124. Nine oi these are liTA differentials which are known to be resistant to three isolates (Onne, IT16,and FAR-13)ofCAMV atiITA, Nigeria (H.W. Rossel and G.1hottapi\1y, Personal Cooununication 1991). 1Vu 2480 has also been reported resistant to diverse isolates of CAMV at lITA (Ladipo and Allen 1979).

Cowpea aphid-borne mosaic virus yield loss studies During 1988-89 seed yield loss study due to CAMV was studied with four CAMVsusceptible genotypes (Muliana, ZVu 83, ZVu 237, and TVx4654-44E) at Msekera. 1n each genotype, 100 plants shOwing mosaic and 100 appearing healthy were tagged at harvest the number of pods per plant, seed yield, and 100 seed weight were recorded. In all four genotypes, plants that had mosaic produced significantly fewer number of pods as well as smaller-sized seed in comparison to healthy genotypes. Overall mean seed yield loss due to CAMV was 52 percent of healthy plants. The Io6s in seed yield was greater (600/.) in ZVu '237 than in others (4S..S00/0). Kaiser and Mossahebi (1975) also reported seed yield losses of 15 to 87 percent due to infection by CAMV in Iran.


References Bock, KR., and M . Conti. 1974. Cowpea aphidbome mosaic virus. CMII AAB Descriptions of plant viruses no. 134. CORP (Centre forOverseasPestResearch) . l981. Pest contro l in tropical grain legumes, London, UJ<,COPR lIT A. 1980. Annual Report for 1979. IITA, !badan, Nigeria, 152 pp. Haciwa, H .e., and J. Kannaiyan . 1991. Prevalence of cowpea diseases and extent of seed yield losses due to scab and CAMV in Zambia . Paper presented at SAOCC / IITA Cowpea Research Workshop, 26路27 Sep 1991, Harare, Zimbabwe. Kannaiyan, L D.e. Greenberg, Haciwa, RC., and MN. Mbewe. 1987. Screening cowpea for resistance to major diseases in Zambia . Tropical Grain Legume Bulletin 34: 23--26.

Kannaiyan,]., and Re. Haciwa. 1989. Bean cowpea pathology research in Zambia. Pages 117134 in Proceeding of the National Workshop on Food Legumes Research and Improvement in Zambia, 9路11 Mar 1988, Mfuwe, Eastern Province, Zambia: Ministry of Agriculture. LadiJX>, J.L. 1977. Seed transmission of cowpea aphidbome mosaic virus in some cowpea cultiva". Nigeria Journal Plant Protection 3: ~10. Ladipo, JL, and DJ Allen. 1979. Identification of resistance to cowpea aphidborne mosaic virus. Tropical Agriculture (Trinidad) 56: 35~59 . Mulila路Mitti, J.M. 1991. Development of cowpea Improvement Programme in Zambia, Paper presented at SAOCC/lITA Cowpea Research Workshop, 26-27 Sep, 1991 , Harare, Zimbabwe. Reddy, M.s., K. Kanenga, J.e. Musanya, and J. Kannaiyan . 1989. Agronomic and cropping s ystems research involving gra in legumes in Z,mbia . Pages 104-116 in Proceedings of the National Workshop on Food Legumes Research and Improvement in Zambia, 9-11 Mar 1988, Mfuwe, Eastern Province, Zambia, Ministry o f Agriculture. Ta iwo, M.A. 1978. An evaluation of advanced cowpea selections for resistance to cowpea aphidbome mosaic andcowpe3 mosaicviruse5. MS thesis, Cornell University, Ithaca, USA.


List of Participants

R.A.AmabIe SAOC/IITA Cowpea Research Project, CP 2100 INIA, Maputo, Mozambique K. E. Dashiell CUP, lITA, PMB 5320, Ibadan, Nigeria A.LOoto INIA, Maputo, Mozambique R.Oube Crop Breeding Institute, PO Box 8100, Causeway, Harare H.CHadwa Food Legumes Research Team, Msekera Regional Research Station, P. O. Box 510089, Chipata, Zambia J.CB. ~blssa llonga Research Institute, Ilonga, Kilosa, Tanzania

J. Kannai yan Food Legumes Research Team, Msekera Regional Research Station, Chipata, Zambia M.B. Kwapata University of Malawi, Bunda College of Agriculture, PO Box 219, Lilongwe A.Z. Matangue INIA, Maputo, Mozambique

B. Matilo Plant Pathologist, Department of Agricultural Research, Private Bag 0033, Gaborone, Botswana G.A. Mitti Adaptive Research Planning Team, Meskera Research Station, Chipata, Zambia K.M1lgo Crain Legume Research A.R.I.路IIonga, lIonga, !Glosa, Tanzania S.S.Moima Agricultural Research Division, P. O. Box 829. Maseru, Lesotho


E.S.Mourwe Department of Agricultural Research, Private Bag 0033, Gaborone, Botswana

J.M. MuJiJa-Mitti Food Legumes Research Team, Msekera Regional Research Station, PO Box 510089, Chipata, Zambia

J.C. Musanra Food Legumes Research Team, Msekera Regional Research Station, PO Box 510089, Chipata, Zambia F.A.Myaka Crain Legume Research A.R.I .â&#x20AC;˘ nonga Research Institute, PO lIonga, Kilosa,

Tanzania D .M.Naik

INIA. Maputo, Mozambique T. Nleya Agronomy Institute, Box 8100, Causeway, Harare B.I. Nyoka Crop Breeding Institute, PO Box 8100, Causeway, Harare

J. Pali-Shikhulu Malkems Research Station, P. O. Box 4, Malkems, Swaziland

M.e. Philli pa Dept. of Agricultural Research, Gaborone, Botswana S. Saithananthan Food Legumes Research Team, Msekera Regional Research Station PO Box 51()(w!9 Chipata, Zambia V.W. Saka Crop Production Department, University of Malawi, Bunda College of Agriculture, PO Box 2119, Ulongwe, Malawi

A. P. Uriyo CUP, lITA, PMB 5320, Ibadan, Nigeria


ISBN 9781310936

Trends in Cowpea Research  

This report shows different trends in Cowpea's Research in order to improve on its production

Read more
Read more
Similar to
Popular now
Just for you