Download pdf - Translational Genomics for Crop Breeding (Biotic Stress) || Enabling Tools for Modern Breeding of Cowpea for Biotic Stress Resistance

Chapter 10

Enabling Tools for Modern Breeding ofCowpea for Biotic Stress ResistanceBao-Lam Huynh, Jeffrey D. Ehlers, Timothy J. Close, Ndiaga Cisse, Issa Drabo, OusmaneBoukar, Mitchell R. Lucas, Steve Wanamaker, Marti Pottorff, and Philip A. Roberts

Abstract

This chapter summarizes recent advances in genomic resources, technologies, breeding platforms,and molecular markers that are being used to expedite delivery of cowpea cultivars with resistanceto important biotic stresses. The genomic resources available in cowpea include high-throughputsingle nucleotide polymorphism (SNP) genotyping platforms, a high-density consensus genetic mapwith more than 1,100 markers, a whole genome sequence assembly, bacterial artificial chromosomes(BACs), and a physical map anchored to the genetic map. Markers linked to important biotic resistancetraits, including resistance to foliar and flower thrips, Fusarium wilt, root-knot nematode, bacterialblight, ashy stem blight (Macrophomina), Striga, and viruses are described, together with toolsdeveloped to integrate SNP genotypes, genetic maps, and trait information to guide breeding. Agenetic transformation system has been developed for cowpea and breeding lines developed withresistance to cowpea weevil and Maruca pod borer. Some examples of initial work in evaluating andoptimizing marker-assisted backcross and marker-assisted pedigree breeding are described. Majorchallenges that must be overcome to allow for the adoption of modern marker-assisted breeding(MAB) in cowpea include human and precision phenotyping capacity issues. Further, as essentiallyall cowpea breeding is being conducted by the public sector with modest resources, efficient strategiesare needed to minimize running costs.

Introduction

Cowpea (Vigna unguiculata L. Walp.) is perhapsthe most widely adapted, versatile, and nutritiousgrain and fodder legume for warm and dry agro-ecologies of the tropics and subtropics. World-wide there are more than 14 million ha of cowpea

being cultivated, of which about three-fourthsof the area and two-thirds of the productionoccur in the semi-arid Sudan savanna and Sahe-lian zones of the African continent running fromwest to east, south of the Saharan desert, and inmore dispersed but extensive pockets of similaragro-ecologies in eastern and southern Africa.

Translational Genomics for Crop Breeding, Volume I: Biotic Stress, First Edition. Edited by Rajeev K. Varshney and Roberto Tuberosa.C© 2013 John Wiley & Sons, Inc. Published 2013 by John Wiley & Sons, Inc.

183

184 TRANSLATIONAL GENOMICS FOR CROP BREEDING

Substantial quantities of cowpea also are pro-duced in similar agro-ecologies in South Amer-ica, with Brazil being the second-largest produc-ing country in the world after Nigeria. Althoughof less importance today than in the recentpast, cowpea is also grown in the southern andwestern parts of North America and is a ‘tra-ditional’ crop of several indigenous tribes oflower-elevation central Mexico. A form of cow-pea (subspecies sesquipedialis) known as ‘longbean’ or ‘asparagus bean’ is cultivated as a veg-etable crop throughout East Asia for its longfleshy green pods that can be one meter in length.Long bean is fully fertile with other cultivatedand some wild subspecies of cowpea and com-parisons of genetic maps show essential iden-tity (Xu et al. 2010). Cowpea was documentedfrom before Roman times in Europe and thepresence of significant phenotypic and genotypicdiversity exists among landrace germplasm fromItaly (Tosti and Negri 2002) and other southernEuropean countries, suggesting a long history ofuse in this region. As a general rule, cowpeasare grown in hotter low-elevation equatorial andsubtropical areas, often being replaced by com-mon bean (Phaseolus vulgaris L.) at altitudesabove 1,300 m, although cowpeas are grownat altitudes up to 1,500 m in Kenya. Commer-cial production of the crop extends as far northas 40◦ latitude in California, and experimen-tal plantings of early maturing breeding lineshave been successful as far north as Minnesota(45◦ N latitude) in the United States (Daviset al. 1986).

While cowpea is both responsive to favorablegrowing conditions and tolerant to drought, hightemperatures, and poor soils (Hall 2004; Fery1990), biotic stresses from pests and diseasesinflict heavy losses and are key factors under-lying why on-farm cowpea yields of traditionalvarieties in Africa are 3- to 5-fold lower thanpotential yields (Ehlers and Hall 1997).

Development of cowpea cultivars that resistor tolerate biotic stresses would result in dra-matic yield improvements. Breeding resistant

cultivars is a particularly desirable strategy forthis crop because it is grown mostly by resource-poor farmers, many of whom are women wholack access to capital for application equipment,pesticides, and protective wear, as well as toexpertise in the efficacious and safe use of theseproducts.

Breeding for resistance to biotic stresses hasbeen undertaken by the International Institutefor Tropical Agriculture (IITA) and by AfricanNational Research Systems (NARS), in somecases supported by the USAID CollaborativeResearch Support Program (CRSP) and otherdonors. These breeding programs have achieveda number of successes in the last 20 years,having bred cultivars resistant to some of thekey pests such as cowpea aphid, cowpea wee-vil, the parasitic weed Striga gesneroides, bac-terial blight, root-knot nematodes, and cowpeaaphid-borne mosaic virus (CABMV) (Hall et al.1997; Hall et al. 2003; Timko et al. 2007) andthereby contributing to raised on-farm yields.These breeding efforts employed wholly con-ventional breeding approaches and typically tookmore than ten years from concept and crossing torelease.

Similar to breeding programs in other crops,the need to employ sequential and repeated phe-notypic evaluations and performance trials isthe most resource-intensive and time-consumingaspect of the process. In many cases, these eval-uations require complex, specialized conditions,techniques and expertise to assess phenotypesfor selection. These evaluation resources are dif-ficult to assemble and costly to operate, and inmany cases it is not possible to conduct a com-plete breeding program consisting of teams ofscientists of all necessary disciplines because ofbudget and manpower limitations for ‘orphan’crops (Delmer 2005) such as cowpea. Selec-tion based on molecular markers linked to keybiotic resistance traits or quantitative trait loci(QTL) can be used to reduce the phenotyp-ing burden typically required through conven-tional breeding efforts and thereby help speed the

ENABLING TOOLS FOR MODERN BREEDING OF COWPEA 185

delivery of improved varieties. Neutral, ‘geneticbackground’ molecular markers that are welldistributed through the genome are also partof molecular marker-assisted selection (MAS)schemes such as marker-assisted backcrossing(MABC), where the goal is to reassemble thegenetic background of the recurrent parent withthe addition of one to a few target traits. Accord-ingly, MAS provides a powerful and potentiallycost-saving avenue for increasing the rates ofgenetic gain in plant breeding programs (Xu andCrouch 2008).

In the recent past, cost and technology limi-tations meant that MAS was restricted to majorcrops and to one or a very few high priority traits,where phenotyping was costly or otherwise prob-lematic. The development of relatively low costhigh-throughput SNP genotyping platforms inmany crops and the availability of high-densitygenetic linkage maps have greatly enhanced whatis now possible with MAS. These capabilitiesmean new breeding strategies can now be con-sidered that utilize molecular marker informa-tion at tens to thousands of points in the genome,encompassing selection for multiple traits and/ormultigenic traits simultaneously. Selection tar-gets include easy-to-phenotype agronomic andgrain quality traits such as grain size, texture,and color, in addition to multiple biotic stresstolerance and resistance traits that are more dif-ficult to phenotype.

The effectiveness of MAS breeding schemesdepends on high quality phenotyping and preci-sion genotyping for assembly of robust marker-trait associations and QTL estimates, closenessof marker-trait determinant linkages, and rel-ative cost of genotypic selection compared totraditional selection protocols based on phe-notypic selection. MAS can increase the effi-ciency of breeding in several other ways, throughthe selection of desirable progeny for crossingbefore flowering, providing year-round, environ-mentally independent selection capability, andthrough simultaneous selection of multiple traitsand for a particular genetic background. In addi-

tion, it is possible to deduce useful marker-trait associations and marker effects in earlygenerations of breeding cycles that can thenbe used for selection in later generations ifgenotypic information is available at the laterstage.

Pioneering the Use of SSRMarkers for Introgressionof Striga Resistance

To date only limited deployment of molecu-lar markers in cowpea breeding programs hasoccurred. MAS using simple sequence repeat(SSR, or microsatellite) markers to develop cow-pea varieties resistant to the parasitic weed Strigagesnerioides (Striga) was initiated in 2006 bythe national research organizations of Senegal,Burkina Faso, Nigeria, and Mali, and at IITAin Nigeria in collaboration with the Universityof Virginia (Timko et al. 2007). These projectshave focused on the use of a limited numberof SSR markers with the goal of introgress-ing Striga resistance into improved or local sus-ceptible varieties. This approach eliminates theneed for phenotypic evaluation of Striga resis-tance at each stage of the breeding process.Also, by incorporating resistance genes effectiveagainst multiple races of Striga prevalent acrossnational boundaries, MAS eliminates the needfor meeting quarantine requirements associatedwith the movement of Striga seeds of differentraces across national boundaries that would berequired for centralized phenotypic evaluationsof resistance. However, the approach uses onlytrait-linked ‘foreground’ markers, meaning thatrecovery of the recurrent parent ‘background’genotype will occur at nearly the same rateas with conventional backcross breeding. SSRgenotyping with numerous background markerson multiple individuals to facilitate recovery ofthe recurrent parent would be cost prohibitive.Another major drawback of MAS using SSRmarkers in backcross breeding is that high qual-ity gels are difficult to produce consistently, thus


progress can be hampered by misclassified indi-viduals. This is in contrast to single nucleotidepolymorphisms, or SNPs, which show very highfidelity of allele calling in repeated rounds ofgenotyping.

SNPs in Cowpea Breeding forBiotic Stress Resistance

Genotyping platforms based on SNPs are pre-ferred over other marker types in plant breed-ing applications because they are amenableto high-throughput yet highly precise assays(Rafalski 2002). SNPs derived from expressedsequence tags (ESTs) are especially usefulbecause they are derived ultimately from mRNAtranscripts and thereby reflect allelic differ-ences between individuals and present oppor-tunities for cloning genes through comparisonswith annotated genome sequences of other plantspecies. SNP genotyping is highly developedfrom a technical and equipment perspectivebecause of the large investments made in thisarea in human and animal genetics that are beingleveraged for crop improvement. The privatecrop breeding sector was quick to grasp thisopportunity and develop these systems for majorcrop plants such as maize and soybean, hav-ing employed them in breeding programs formore than a decade (Eathington et al. 2007). Itappears that the private sector has found marker-assisted breeding (MAB) with SNPs to be aneffective approach, as evidenced by the factthat SNP genotyping platforms continue to bedeveloped for more and more crops bred in thissector. In cowpea, a 1536-Illumina GoldenGateSNP genotyping platform was developed in 2009(Muchero et al. 2009) and is described brieflybelow.

High-Throughput SNP GenotypingPlatform for Cowpea Breeding

Several key resources are needed for the com-prehensive implementation of modern cowpea

breeding. First among these is high-throughputgenotyping capability, for rapid and dense fin-gerprinting of parents and breeding progenies ata reasonable cost. This multiplexing approach isa critical advance over low-throughput systemswhere genotyping is conducted on a marker-by-marker, individual-by-individual basis. Forcowpea, a high-throughput genotyping platformwas developed between 2007 and 2009 as animportant outcome of a ‘Tropical Legumes I’project based at the University of California,Riverside, and funded by the ConsultativeGroup on International Agricultural Research(CGIAR) Generation Challenge Programme(GCP). The genotyping platform was developedas an Illumina GoldenGate Assay for 1536 SNPloci. This platform can genotype simultaneously96 DNA samples at 1,536 SNP loci. Thegenotypic information is provided to users withspecialized ‘Illumina GenomeStudio’ software,which allows rapid visual inspection and datasummarization. The development and technicaldetails of this platform were presented in detailin the work of Muchero and colleagues (2009).Briefly, this platform consists of 1,536 EST-derived SNPs chosen from a selected subset ofabout 10,000 SNPs identified from comparisonsof 183,000 EST sequences from 13 cowpeagenotypes. Of these, 1,375 SNPs performedtechnically well in the assay. Among the severaltechnical considerations in choosing the 1,536SNP subset, high polymorphic informationcontent among African germplasm was primary,in order to ensure that the chosen SNPs werehighly polymorphic in African breeding mate-rials. Examples of the numbers of polymorphicmarkers observed in crosses between breedinglines both within and between several Africanbreeding programs following genotyping withthis platform are shown in Table 10.1. Theseresults show a high average level of polymorphicmarkers between any two parents, ranging from134 to 391 and indicating that the SNPs selectedfor inclusion on the current 1,536 IlluminaGoldenGate platform represent an effective setfor MAS in West African breeding programs.


Table 10.1. Examples of the number of SNP markers that are polymorphic between two parental lines of potentialcombinations for intra- (e.g., IITA x IITA) and inter- (e.g., IITA x INERA) breeding program crosses.

Parent 1 Parent 2 Cross Type No. Polymorphic markers

IT84S-2246 IT93K-503-1 IITA x IITA 134IT89KD-288 Suvita-2 IITA x INERA 245KVx 525 KVx 396-4-5-20 INERA x INERA 257KVx 442-3 KVx 396-4-5-20 INERA x INERA 261Suvita-2 Melakh INERA x ISRA 274IT97K-499-35 Mouride IITA x ISRA 289Suvita-2 Calif. Blackeye No. 27 INERA x UCR 291IT84S-2246 IT00K-1263 IITA x IITA 294IT97K-499-35 IT00K-1263 IITA x IITA 309KVx 525 Bambey 21 INERA x ISRA 313Bambey 21 Melakh ISRA x ISRA 325IT84S-2246 IT82E-18 IITA x IITA 329KVx61-1 Calif. Blackeye No. 50 INERA x UCR 339Melakh Calif. Blackeye No. 27 ISRA x UCR 343Yacine Calif. Blackeye No. 50 ISRA x UCR 345IT84S-2246 Mouride IITA x ISRA 347IT89KD-288 Calif. Blackeye No. 50 IITA x UC R 372IT93K-503-1 Bambey 21 IITA x ISRA 376IT82E-18 KVx 525 IITA x INERA 381IT93K-503-1 Calif. Blackeye No. 27 IITA x UCR 391

High Quality ConsensusGenetic Map

Breeders have often conducted their markerbreeding programs based on map informationderived from a single cross. Two major limita-tions of using a single-cross map for the analysisof marker or gene segregation in a populationare (1) the limited number of traits of economicimportance that will segregate in any one popu-lation and (2) the limited number of polymorphicmarkers. Hence, a second key resource neededfor modern cowpea breeding is a high qual-ity consensus genetic map. Consensus maps areformed from merging multiple individual link-age maps. By the term ‘high quality’ we meanmarker positions determined with precision andof sufficient density across the genome. Merg-ing markers from several populations increasesmarker density and the precision of marker posi-tion estimation, as well as the quantity of mappedtraits. Consensus genetic linkage maps providebreeders with a resource for analyzing the inher-itance of traits and marker-trait associations

needed for marker-assisted breeding. The use-fulness of a genetic linkage map to plant breed-ers is determined largely by its thoroughness ofgenome coverage, with markers that are highlypolymorphic in target breeding germplasm.

The first version of the cowpea consensusgenetic linkage map was the product of mergingindividual genetic maps built from 6 recombi-nant inbred line (RIL) populations that had beengenotyped with the 1,536 Illumina GoldenGateAssay (Muchero et al. 2009). This map included928 markers spanning 11 linkage groups over atotal map distance of 680 centiMorgans (cM).Map refinement has continued since then, withthe genotyping of five additional RILs and twoF4 breeding populations, construction of indi-vidual maps for these populations, and inte-gration of these maps into an improved con-sensus map (Figure 10.1) (Lucas et al. 2011).These new map inputs resulted in improvedmarker order and density and fewer gaps, withthe addition of 179 SNP markers (bringingtotal markers to 1,107) with an average dis-tance between markers of about 0.6 cM. Further


Fig. 10.1. Consensus genetic map of cowpea and parameters depicting map characteristics (Lucas et al.2011). (A) Average distance between bins (0.25 cM). (B) Average number of markers per bin (0.5 units).(C) Number of bins (25 units). (D) Number of markers (25 units). (E) Bin locations. C and D begin at thesame radial position. For a color version of this figure, please refer to the color plate.

refinements since then have been achievedthrough comparison with the common beanwhole-genome assembly and a partially assem-bled cowpea whole-genome assembly. The cur-rent SNP-based cowpea consensus genetic link-age map is included in a publicly availablebrowser called HarvEST:Cowpea, which canbe downloaded as Windows software fromhttp://harvest.ucr.edu, or operated through anonline portal at www.harvest-web.org. This mapwill continue to be updated periodically as addi-tional genotyping and other types of informationbecome available. With this current consensusmap, discovery of useful trait-linked and geneticbackground markers in cowpea can be made with

a high probability of success by phenotyping forany trait segregating in any of the 11 RIL pop-ulations and overlaying the existing genotypicdata.

Fingerprinting of Cultivars andOther Prospective Parents

Genotyping a large sample of current varietiesand advanced breeding lines with a platformof mapped markers is valuable and convenientfor several applications. For breeding, this infor-mation supplies quick decision making for theselection of lines for crossing, while a databaseof information on parental genotypes, coupled


with historical phenotype information, can resultin the discovery of haplotypes associated withtraits. These genotyping profiles can also leadto the discovery of duplicates and help uncoverhidden, accumulated ‘off-types’ in breeding pro-grams (Lucas et al. 2013). In work conducted bythe Tropical Legumes Project, ∼600 cultivars,breeding lines, and accessions were genotypedwith the 1,536 Illumina GoldenGate Assay. ThisSNP fingerprinting information was combinedwith map coordinates from the consensus mapdescribed earlier, providing a rich resource forbreeding and haplotype analysis, as discussed inthe following sections.

Biotic Stresses and GeneticMarker Resources

Until recently, finding trait-linked molecularmarkers for cowpea breeding was slow and labo-rious. Now, with the availability of the high-throughput SNP genotyping platform and theconsensus genetic map, coupled with exten-sive phenotyping of RIL populations and pan-els of germplasm within the last five years,SNP markers or QTL have been identifiedfor a range of biotic stress-resistance traitsand examples are presented in Table 10.2.Among these are trait determinants for resis-tance to foliar thrips (Thrips tabaci) (Lucas et al.2012), flower thrips (Megalurothrips sjostedti)Trybom (M. Lucas et al., unpublished data),ashy stem blight (Macrophomina phaseolina)(Muchero et al. 2011), bacterial blight (Xan-thomonas axonopodis pv. vignicola) (Agbicodoet al. 2010), root-knot nematodes (Meloidogynespp.) (Roberts et al. unpublished data), and toFusarium wilt (Fusarium oxysporium Schlech-tend.:Fr. f. sp. tracheiphilum (E. F. Smith) W.C. Snyder & Hansen) (Pottorff et al. 2012a;2012b). Additional information on these bioticstresses and their associated SNP markers ispresented in Table 10.2 and the followingsections.

Several bacterial and fungal diseases can sig-nificantly affect cowpea production, and their

incidence and severity vary with the amount anddistribution of rainfall (Williams 1975). Sep-toria leaf spot (Septoria vignae P. Henn andS. vignicola Vasat Rao, S. Kozopolzanii Niko-lajeva), scab (Sphaceloma sp.), brown blotch(Colletotrichum capsici (Syd.) Butler and Bisbyand C. truncatum (Schwein.) Andrus & W. D.Moore), and cercospora leaf-spot are importantin the Guinea savanna, while cercospora leaf-spot, bacterial blight, and ashy-stem blight (orcharcoal rot) caused by Macrophomina phase-olina (Tassi.) Goid. are important in the Sudansavanna, and ashy stem blight in the Sahel. Scab,cercospora leaf spot, powdery mildew (Erysiphepolygoni DC.), Fusarium wilt, and ashy stemblight are significant diseases in Brazil (Lin andRios 1985). To date, molecular markers havebeen developed only for a few of these diseases,and this section will focus on those with near-term prospects for MAS.

Bacterial blight is a major disease of cow-pea affecting most production areas. Severe out-breaks occur regularly and can completely dev-astate crops when environmental conditions areconducive to disease development. Earlier inher-itance studies suggested resistance was simplyinherited (Patel 1985). More recently molecu-lar markers have become available that couldbe valuable for MAB. Agbicodo and colleagues(2010) identified three SNP-based QTLs, CoBB-1, CoBB-2, and CoBB-3, on linkage groups LG3,LG5, and LG9, respectively (Lucas et al. 2011),which co-segregated with resistance to two Nige-rian strains of bacterial blight, Xav18 and Xav19.CoBB-1, CoBB-2, and CoBB-3 explained 22.1,17.4, and 10% of the genetic variation in the RILdiscovery population. SNP markers linked to theQTL include 1_0853 on LG3, 1_0037 on LG5,and 1_1202 on LG9. Validation of these markersin an actual breeding program and across multi-ple populations remains to be done.

In semi-arid agro-ecologies, the soil-bornefungus Macrophomina phaseolina (Tassi.) Goid.causes ashy-stem blight of cowpea (known insoybean as charcoal rot). This ubiquitous dis-ease is exacerbated under drought conditions.


Table 10.2. Trait-linked SNP markers for marker-assisted breeding for biotic resistances. Marker positions are based on theconsensus genetic map (Lucas et al. 2011). Details of trait-linked SNPs are described in the text. SNP markers for virusresistance are located near AFLP markers (Ouedraogo et al. 2002a) linked to the trait.

Marker LG Position (cM) Trait

1_0277 2 18.92 Foliar Thrips1_0589 2 18.92 Foliar Thrips1_0698 2 18.92 Foliar Thrips1_0829 2 19.45 Foliar Thrips1_0492 2 19.53 Foliar Thrips1_0253 2 20.16 Foliar Thrips1_0337 2 20.35 Foliar Thrips1_0164 2 21.00 Foliar Thrips1_1086 2 22.96 Foliar Thrips1_0284 2 25.15 Foliar Thrips1_1406 2 26.12 Foliar Thrips1_1139 2 26.86 Foliar Thrips1_1061 2 29.22 Foliar Thrips1_1048 2 29.78 Foliar Thrips1_0482 2 49.92 Macrophomina1_0709 2 67.32 Macrophomina1_0071 3 9.28 Cowpea mosaic virus1_0080 3 9.91 Cowpea mosaic virus1_0853 3 10.82 Bacterial Blight1_0853 3 10.82 Macrophomina1_0604 3 64.78 Macrophomina1_0345 3 77.55 Cowpea severe mosaic virus1_0964 3 77.55 Cowpea severe mosaic virus1_1413 4 18.38 Foliar Thrips1_0774 4 20.16 Foliar Thrips1_1221 4 20.16 Foliar Thrips1_1242 4 20.72 Foliar Thrips1_1242 4 20.72 Foliar Thrips1_0678 4 27.60 Macrophomina1_0804 4 40.51 Macrophomina1_0032 5 45.28 Macrophomina1_0037 5 49.10 Bacterial Blight1_0919 6 25.68 Blackeye cowpea mosaic potyvirus1_1496 6 26.10 Blackeye cowpea mosaic potyvirus1_0897 6 47.25 Fusarium Wilt1_1363 6 47.41 Fusarium Wilt1_1202 9 0.00 Bacterial Blight1_1137 9 42.46 Striga1_0276 9 42.84 Striga1_0958 9 42.84 Striga1_0840 10 43.86 Foliar Thrips1_0161 10 44.86 Foliar Thrips1_0754 10 45.04 Foliar Thrips1_0281 10 46.25 Foliar Thrips1_1453 10 46.55 Foliar Thrips1_0354 10 46.70 Foliar Thrips1_0952 10 47.26 Foliar Thrips1_1062 10 47.26 Foliar Thrips1_1041 10 47.63 Striga1_0003 10 49.05 Striga


Muchero and colleagues (2011) identified ninequantitative trait loci (QTLs) that accounted forbetween 6.1 and 40.0% of the phenotypic vari-ance (R2). Maturity effects on the expression ofresistance were indicated by the co-location ofMac-6 and Mac-7 QTLs with maturity-relatedsenescence QTLs Mat-2 and Mat-1, respectively.The multigenic nature of resistance and inter-action with crop maturity will complicate theuse of these markers in MAB, but may lead togreater precision in identifying resistant individ-uals in segregating populations. As breeding tar-gets, combinations of pairs of QTLs, such asMac-2 and Mac-5, can provide significant pro-tection from this disease (Muchero et al. 2011).Efforts are under way to combine QTL forMacrophomina resistance and drought tolerancefor cowpea protection in rain-fed, drought-proneproduction areas (Muchero et al. 2011).

Cowpea rust caused by Uromyces vignae is asevere problem in humid ecologies across Asiaand Africa that is best managed by the useof resistant varieties (Emechebe and Shoyinka1985). Li and colleagues (2007) reported theidentification of an amplified fragment lengthpolymorphism (AFLP) marker converted to asequence characterized amplified region (SCAR)marker for resistance to Uromyces vignae inChina.

Fusarium oxysporum f.sp. tracheiphilum(Fot) is a soil-borne fungal pathogen that causesvascular wilt disease in cowpea. Fot race 3 is oneof the major pathogens affecting cowpea produc-tion in the subtropical regions where the crop isgrown. Resistance to Fot race 3 was studied in theRIL population ‘California Blackeye 27’ (resis-tant) x 24-125B-1 (susceptible). Biparental map-ping identified a Fot race 3 resistance locus, Fot3-1, which spanned 3.56 cM on linkage group 1 ofthe CB27 × 24-125B-1 genetic map. A marker-trait association narrowed the resistance locusto a 1.2 cM region and identified SNP marker1_1107 as co-segregating with Fot3-1 resistance(Pottorff et al. 2012a). Two QTLs governingresistance to Fot race 4, which attacks cowpeaswith Fot race 3 resistance, have also been iden-

tified using other cowpea mapping populationsdeveloped from crosses between resistant andsusceptible parents (Pottorff et al. 2012b).

Cowpea is attacked by a wide spectrum ofother fungal diseases. Those of major impor-tance include anthracnose (Colletotrichum lin-demuthianum (Sacc. & Magnus) Lams.-Scrib.),cercospora leaf spot (Cercospora canescensEllis & G. Martin), web blight (Rhizoctoniasolani Kuhn), ascochyta (Ascochyta phaseolo-rum Sacc.), brown blotch (Colletotrichum cap-sici (Syd.) Butler and Bisby and C. truncatum(Schwein.) Andrus & W. D. Moore), and pow-dery mildew (Emechebe and Shoyinka 1985).Provision of molecular markers for resistance tothese diseases should be an important goal forcowpea breeding programs in areas where theseorganisms are problematic.

Markers for Resistance to Viruses

Development of virus-resistant cultivars is animportant objective of several cowpea breed-ing programs in Africa and in Brazil. Themain seed-borne viruses affecting cowpea pro-duction include cowpea aphid-borne mosaicvirus (CABMV), blackeye cowpea mosaic virus(BlCMV), cucumber mosaic virus (CMV), cow-pea mosaic virus (CPMV), cowpea severemosaic virus (CPSMV), southern bean mosaicvirus (SBMV), and cowpea mottle virus (CMV).Cowpea golden mosaic virus (CGMV) and cow-pea chlorotic mottle virus (CCMV) are importantnon-seed-borne viruses. The seed-borne virusesare particularly damaging because they spreadvia infected seed. Seed-borne infections canresult in early season spread of the virus to neigh-boring plants. Sources of resistance to all of theseviruses have been identified, and cultivars withresistance to several viruses have been developedat IITA (Singh 1993). The presence in Senegaland Nigeria of different strains of CABMV thathave overcome resistance complicates breed-ing efforts, but breeding lines have been devel-oped that are resistant to multiple strains of thisvirus as well as to other viruses. IT86D-880,


developed at IITA, is exceptional in that it isresistant to four strains of CABMV, cowpea yel-low mosaic virus (CYMV), and cowpea mildmottle virus (CMMV). A severely damaging dis-ease of cowpea known as ‘stunt’ is caused bya dual infection of BlCMV and CMV (Kuhn1990). Ouedraogo and colleagues (2002a) wereable to map determinants of resistance to CPMV,CPSMV, B1CMV and SBMV in the RIL popu-lation 524B x IT84S-2049 using AFLP mark-ers. Subsequently, this population was geno-typed with the 1,536 Illumina GoldenGate Assayallowing for the localization of SNP markers forthese resistance loci (Table 10.2).

Markers for Resistance to the ParasiticWeed Striga

The parasitic weed Striga gesnerioides (Wild.)Vatke causes considerable damage to cowpeain the Sudan savanna and Sahelian regions ofAfrica. Several sources of resistance to this par-asite have been identified and resistance has beenincorporated using conventional approaches intoadvanced breeding lines by IITA (Timko et al.2007) and by Cisse and colleagues (1995) intocultivars for Senegal (Hall et al. 2004). Strigaexhibits strain variation such that cultivars thatare resistant in one location may be susceptible inanother (Lane et al. 1993). Genetic studies haveshown that three dominant, non-allelic genesconfer resistance to different Striga isolates butthat the mechanisms differ (Singh 1993).

Useful markers for implementation of MASare available only for some of the Striga resis-tance genes, and these were the first candi-dates for broad application in cowpea breed-ing programs. Ouedraogo and colleagues (2002a,2002b) found three AFLP markers linked toRsg2-1, a gene that confers resistance to StrigaRace 1 (SG1) present in Burkina Faso, and sixAFLP markers linked to gene Rsg4-3 a gene thatprovides resistance to Striga Race 3 (SG3) fromNigeria. Two of the AFLP markers were asso-ciated with both Rsg2-1 and Rsg4-3. Ouedraogo

and colleagues (2002a) were able to convert oneof these markers to a SCAR (sequence char-acterized amplified region) that has proven tobe an effective and remarkably reliable markerfor resistance to Striga SG1 and SG3 conferredby Rsg2-1 and Rsg4-3. This SCAR marker, des-ignated 61R (E-ACT/M-CAA), detects a singlepolymorphic band linked to SG1 and SG3 resis-tance in the resistant cultivars B301, IT82D-849, and Tvu 14676, and is being tested foruse in breeding trials. Subsequently, two AFLPmarkers were identified that are closely linkedto Rsg1-1, a gene that also confers resistanceto SG3 in Nigeria (Boukar et al. 2004). Oneof those AFLP markers, designated E-ACT/M-CAC115 and determined to be 4.8 cM fromRsg1-1, was converted to a SCAR marker forease of use in breeding programs (Boukar et al.2004). Because the mapping population 524B xIT84S-2049 used in the studies undertaken byOuedraogo and colleagues (2002b) and Boukarand colleagues (2004) was genotyped with the1,536 Illumina GoldenGate array, we were ableto identify SNP markers corresponding to theAFLP markers identified earlier (Table 10.1).

Markers for Resistance toCowpea Aphid

Cowpea aphid (Aphis craccivora) is one ofthe most destructive pests of cowpea in Africaand worldwide. Myers and colleagues (1996)reported the identification of an RFLP (restric-tion fragment length polymorphism) markerlinked to the resistance to cowpea aphid, butMAS for aphid resistance was not implemented.More recently we have identified SNPs associ-ated with aphid resistance in California in anRIL population developed from the cross ‘Cali-fornia Blackeye 27’ (susceptible) x IT97K-556-6 (resistant). IT97K-556-6 has also been shownto be resistant to cowpea aphid in Texas andNigeria (B. B. Singh, personal communication,2010), but its efficacy relative to possible aphid


biotype differences is unknown in other parts ofAfrica.

Markers for Resistance to Other Insects

Insect resistance is a good candidate for MASin cowpea since assessments of host plant resis-tance to insects are often difficult to conduct inthe field or greenhouse. However, other than foraphid, insect-resistance factors in cowpea dis-covered thus far do not provide immunity tothe pest and often have low heritability underfield conditions. Field screenings that rely onnatural insect infestations are subject to nat-ural fluctuations in pest pressure. When suchvariability is combined with incomplete resis-tance, field screens can lead to misclassifica-tion and selection of lines lacking the strongestresistance. For example, this has been the casewith screening cowpea breeding lines and acces-sions for resistance to Lygus bug (Lygus hes-perus) and pod sucking bugs (such as Nezaraviridula, Clavigralla tomentosicollis, Riptortusdentipes). In addition, colonies of insects may bedifficult to rear without specialized facilities andtrained entomologists to monitor insect growthand screening tests. Such resources have not beenavailable to cowpea breeding programs histori-cally. Resistance to the pod bug C. tomentosicol-lis has been identified in the wild cowpea (ssp.dekindtiana) germplasm line TVNu 151 (Koonaet al. 2002). MAS could be used to introgressresistance factors from such wild cowpea intocultivated forms using a rapid backcrossingapproach, based on simultaneous selection forthe resistance genes (markers) and against mark-ers associated with unwanted wild germplasmcharacteristics such as small seed size and podshattering.

The cowpea storage weevil (or cowpeaBruchid) (Callosobruchus maculatus Fabricius)is a devastating pest of stored cowpea in Africa,Asia, and the Americas. Resistance to the cow-pea storage weevil has been identified (Singhand Jackai 1985) and incorporated into a num-

ber of breeding lines and cultivars by IITA, usingconventional breeding approaches (Singh 1993),and by ISRA (Institut Senegalais de RecherchesAgricoles) (Cisse et al. 1995). The resistanceprovides effective protection from weevil dam-age for about two months, but levels of damagein resistant cultivars approach that found in sus-ceptible cultivars after six months of infestation.This resistance is useful in developing countriesbut when used alone provides insufficient pro-tection to ensure that the grain retains its marketvalue and seed quality characteristics. It wouldbe helpful to identify SNP markers for resistanceto cowpea weevil and to look for options thatprovide higher levels of protection.

Genetically-Modified Cowpea forControl of Maruca Pod Borer andCowpea Weevil

Cowpea was one of the last major grain legumespecies for which genetic transformation wasachieved (Popelka et al. 2006). This technologynow presents an additional option for address-ing constraints that have not yielded to otherapproaches.

Maruca pod borer (Maruca testulalis) is oneof three major insect pests affecting cowpea pro-duction in Africa and parts of Asia (Singh et al.1996). Major efforts to screen both wild and cul-tivated cowpea for resistance to this pest have notrevealed meaningful levels of resistance. Effec-tive transgenes have been identified for controlof cowpea weevil and the Maruca pod borer, andthese genes have been used to transform cowpea(Higgins et al. 2012). ‘Bt cowpea’ is currentlyundergoing yield and efficacy testing in confinedyield trials in Burkina Faso and Nigeria. Sincethe distribution of wild cowpea species overlapswith areas of cowpea cultivation in Africa andsince many of these wild cowpeas can form fer-tile hybrids with cultivated cowpeas, the escapeof the ‘Bt gene’ into wild cowpea is a proba-ble outcome of release of Bt-cowpea. An expertpanel formed to address this ‘geneflow’ issue


and other Bt-cowpea safety concerns concludedthat the gene would escape into the wild, butthat the chance of negative environmental conse-quences would be negligible (Murdock, personalcommunication, 2012). Efforts have been initi-ated in several West African countries to use themarker-assisted breeding tools cited in this chap-ter to introgress the Bt gene into locally adaptedvarieties. If this MABC effort is successful andeffective seed delivery and resistance manage-ment plans can be implemented, Bt-cowpea canhelp dramatically increase grain yields of cow-pea in regions of Africa with Maruca infesta-tions, while reducing insecticide inputs and thenegative health risks and environmental damageassociated with insecticide use.

The most important pest of stored cowpea isthe bruchid beetle, or cowpea weevil. Even lowinitial infestation rates can cause significant graindamage because of the high fertility and shortgeneration periods of this pest. As noted above,only low to moderate levels of resistance havebeen found, which provide only limited storageprotection (Murdock et al. 2008). Genetic engi-neering has been used to transfer the gene cod-ing for the α-amylase inhibitor αAI-1, a bruchidresistance factor from the common bean, intocowpea (Solleti et al. 2008). The α-amylases, thetarget of αAI-1, are key enzymes for starch diges-tion and vital for bruchid development; trans-

formed cowpea carrying αAI-1 exhibits nearimmunity to the cowpea weevil (Solleti et al.2008). However, food safety concerns have con-strained the development and release of this valu-able product.

Tools and Genetic Resources forCowpea Breeders

A computer program called ‘BreedIt C© SNPSelector’ (www.BreedIt.org) has been developedat the University of California, Riverside, whichintegrates fingerprint information, marker-traitassociation information, and the genetic map.It allows breeders to choose a set of custommarkers based on markers that are polymorphicbetween parental lines, marker interval distance,and marker association with target traits.

Eleven RIL populations have been describedin the international cowpea community. Mostpopulations were developed with the goal ofmapping specific high priority traits, and thesepopulation resources are described in Table 10.3.

Challenges for Adoption of ModernBreeding Tools in CowpeaImprovement

A suite of prospective molecular markers andgenomic tools for modern breeding are nowavailable in cowpea. Will these resources be

Table 10.3. Cowpea RIL populations available for marker-traitassociation and partial list of biotic stress resistance traits segregating.

Population Individuals Segregating biotic traits

CB27 x IT97K-556-6 92 Cowpea aphidCB27 x IT82E-18 160 Foliar thripsCB27 x UCR 779 56 Cowpea aphidCB46 x IT93K-503-1 114 Macrophomina, foliar thrips524B x IT84S-2049 85 Multiple, described in textDan Ila x TVu-7778 79 Bacterial blightYacine x 58-77 97 Flower thripsSanzi x Vita 7 122 Flower thripsTVu14676 x IT84S-2246-4 136 StrigaCB27 x 24-125B-1 87 Root knot nematodeLB30#1 x LB1162 #7 90 Rust and powdery mildew


used to develop improved varieties in a cost-effective manner? Major challenges constrainingadoption include lack of capacity in the use ofthese tools, motivation on the part of breeders tolearn these technologies and then make the nec-essary changes in the general workflow and coststructure of their breeding programs to accom-modate MAB activities, and realizing a favor-able cost/benefit ratio compared to conventionalbreeding. Part of achieving cost-effectivenesswill be to develop breeding strategies that min-imize the amount of genotyping while maxi-mizing overall marker information. Almost allcowpea breeding programs worldwide are pub-lic enterprises. Thus, the question arises as towho will absorb the costs of capacity buildingand, in some cases, infrastructure developmentfor these efforts. With the increased potential ofMAB there are costs beyond those for genotyp-ing. These include learning costs for the imple-mentation of new breeding plans and methods.More detailed planning and logistics are needed,as well as new ways of operating, greater designcomplexity, including knowledge about genotyp-ing, marker-trait association, marker platforms,and a several-fold increase in the amount ofgenotypic and phenotypic data that need to bemanaged effectively. Training in modern cow-pea breeding must be a key future activity, sinceso few breeders are prepared now to utilize effec-tively all the new tools of modern breeding. Cow-pea breeders have developed a number of culti-vars with ‘stacked’ resistance for traits such asresistance to cowpea weevil, cowpea aphid, andStriga, along with resistances to bacterial blight,CABMV, and other pathogens. The challenge isto incorporate all of these desirable traits intoindividual cultivars with acceptable grain qual-ity and adaptation to a diverse array of farmingsystems and environments. MAB can be a keyenabling tool to realize this goal.

Phenotyping

Accurate phenotyping for biotic stress resistanceor tolerance is central to success with modern

breeding as well as with conventional breed-ing. Inaccurate or failed phenotyping negativelyimpacts modern breeding programs more thanconventional breeding programs, in that poorphenotyping outcomes consume resources ingenerating genotypic information that cannot beused effectively. SNP genotyping of cowpea isnow straightforward and, from a technical stand-point, virtually error-free. For nearly all traits,phenotyping is, by comparison, much more dif-ficult. Yet each process must be conducted witha level of precision adequate to capitalize onthe information content of the other for modernbreeding to be efficient and effective. Therefore,breeders must make the necessary infrastructureand time investments in phenotyping capabilityto ensure a correct balance consistent with thegoals of the breeding program.

Improved phenotyping capability is neededin conventional as well as modern breeding pro-grams, but the way in which it is applied may dif-fer in the different types of breeding programs.A comprehensive review of classical cowpeabreeding and phenotyping for critical biotic andabiotic production constraints that is still relevantwas published by Hall and colleagues in 1997.New or modified protocols have been publishedmore recently for phenotyping for resistance tobacterial blight (Agbicodo et al. 2010), rust (Liet al. 2007), ashy-stem blight (Muchero et al.2011), and foliar thrips (Muchero et al. 2010).

MAS for Pyramiding Multiple Pest andDisease Resistance Traits

One of the major goals of most cowpea breedingprograms is to combine resistances to numer-ous pests and diseases and other desirable traitssuch as those governing maturity, photoperiodsensitivity, plant type, and seed quality. Untilrecently these programs have lacked modernbreeding tools and an array of elite parental linesthat are fixed for most of the desired resistancefactors. The elite lines can be used to recom-bine traits and generate lines with multiple pestand disease resistance, high yield, appropriate


maturity, and desirable seed quality. Now, dif-ferent cultivars and advanced breeding lines areavailable with desirable traits, such as resis-tance to cowpea weevil, cowpea aphid, andStriga, along with resistances to bacterial blight,CABMV, and other pathogens. The challenge isto incorporate suites of these desirable traits intoindividual cultivars with acceptable grain qual-ity and adaptation to a diverse array of farmingsystems and environments. The array of mark-ers now available for biotic stress resistance andother traits will be very useful in expediting thisprocess.

Outsourced Genotyping

A limitation of high-throughput genotyping sys-tems has been the initial investment required forplatform development, especially the sequenc-ing costs related to SNP discovery. However, assequencing costs have fallen dramatically overthe last several years, this is becoming less anissue. Another consideration is the high cost ofgenotyping equipment and locating dedicatedbench space and lab personnel for genotypingoperations. On the positive side, several pub-lic and private ‘fee-for-service’ SNP genotypingproviders are now available using a range of SNPgenotyping platforms, so there is little need forcowpea breeding programs to invest in in-housegenotyping infrastructure. In developing coun-tries, issues related to availability and high costsof imported consumables and repairs make thecase for out-sourced genotyping even more com-pelling. In our marker-assisted recurrent selec-tion (MARS) breeding projects, genotyping wasdone via outsourcing from two African loca-tions. Fresh leaf tissue disks from 6 plants perline were bulked, placed in 96-well plates pro-vided by the genotyping service company, theplates were enclosed with silica gel packs (todehydrate the tissue) inside a plastic bag, andmailed via a courier service to the genotypingprovider. The samples were genotyped for a sub-set of markers polymorphic in each parental pairusing the Sequenom platform after marker con-

version from the Illumina platform. Genotypicinformation was received from the genotypingprovider in about two weeks, which is considerednormal turnaround time. Since we had genotypedthe same lines with the same SNPs using IlluminaGoldenGate Assay, we compared the genotypingresults from the two assays and found identicalgenotype calls for all alleles of all individuals.

The Illumina GoldenGate Assay platform isa fixed array assay for 1,536 SNPs. This fea-ture presents advantages in that data handlingcan be standardized and time and effort are notneeded to choose subsets of markers for individ-ual populations. However, the fixed array designmeans paying for assays on a large number ofmarkers that are not informative between a givenpair of parent genotypes. As a general rule, thefixed array sets will be less expensive on largernumbers of markers, but more expensive whenfewer markers are being used, which is typicallythe case for genotyping a breeding population.However, in these calculations, the informaticscosts related to choosing and analyzing specificmarker sets for specific populations also must beconsidered.

In order to make an SNP genotyping plat-form more available, flexible, and affordable tothe broader cowpea breeding community, theCGIAR Integrated Breeding Platform has con-tracted with KBiosciences (LGC Genomics) inthe United Kingdom to convert a subset of∼1,000 of the mapped SNPs from the IlluminaGoldenGate Assay Platform for use with thesingle-plex KBiosciences KASP (KompetitiveAllele Specific PCR) genotyping platform. Thisplatform allows users to choose customized sub-sets of SNPs for genotyping on particular setsof germplasm. Hence a breeder can choose asuite of trait-linked polymorphic SNPs and well-distributed ‘background’ SNPs for genotyping,with the number chosen partially determined bybudget considerations and the trade-off betweennumbers of markers and numbers of genotypes.For example, assume the breeder is working topyramid four resistance traits for which SNPmarkers have been identified and which are


polymorphic in the target population. A suite ofSNP markers for these resistance traits (flankingmarkers) would be chosen along with a groupof evenly distributed polymorphic ‘background’markers every 5 to 20 cM along each linkagegroup, depending on the desired marker density,available polymorphism, breeding strategy, andbudget. Access to reliable, precise, and reason-ably priced genotyping services will be essentialif MAB is to become a routine methodology incowpea breeding programs.

Summary and Conclusions

In less than seven years, cowpea has moved fromthe status of an ‘orphan crop’ (Delmer 2005)with only scarce genomic resources to a situ-ation where modern breeding is being imple-mented. In the near future, other advanced breed-ing approaches such as genome-wide selection(Bernardo and Yu 2007) should be explored ifthese are shown to be effective in other crops.While cowpea breeding is poised to becomemuch more effective with the new tools that areavailable, at the same time it is increasingly morecomplex. Significant increases in training andfunding for developing country NARS cowpeabreeding programs will be required for mean-ingful adoption of new genetic tools and accel-eration of genetic improvement.

Acknowledgements

Generous funding from the CGIAR GenerationChallenge Programme and the USAID Dry GrainPulses CRSP to a group of dedicated scientistsis gratefully acknowledged.

References

Agbicodo, E.M., C.A. Fatokun, R. Bandyopadhyay, K.Wydra, N.N. Diop, W. Muchero, J.D. Ehlers, P.A.Roberts, T.J. Close, R.G.F. Visser, and C.G. van derLinden. 2010. Identification of markers associated withbacterial blight resistance loci in cowpea [Vigna unguic-ulata (L.) Walp.]. Euphytica 175:215–226.

Bernardo, R., and J. Yu. 2007. Prospects for genome-wideselection for quantitative traits in maize. Crop Science47:1082–1090.

Boukar, O., L. Kong, B.B. Singh, L. Murdock, and H.W.Ohm. 2004. AFLP and AFLP-derived SCAR markersassociated with Striga gesnerioides resistance in cowpea[Vigna unguiculata (L.) Walp.]. Crop Science 44:1259–1264.

Cisse, N., M. Ndiaye, S. Thiaw, and A.E. Hall. 1995. Reg-istration of ‘Mouride’ cowpea. Crop Science 35:1215–1216.

Collard, B.C.Y., and J.C. Mackill. 2008. Marker-assistedselection: an approach for precision plant breeding inthe twenty-first century. Philosophical Transactions B363:557–572.

Davis, D. W., D. B. Marsh, and M. N. Alvarez. 1986.MN13 and MN 150 cowpea breeding lines. HorticulturalScience 21:1080–1081.

Delmer, D. P. 2005. Agriculture in the developing world:connecting innovations in plant research to downstreamapplications. Proceedings of the National Academy ofSciences 102:15739–15746.

Eathington, S.R., T.M. Crosbie, M.D. Edwards, R.S. Reiter,and J.K. Bull. 2007. Molecular markers in a commercialbreeding program. Crop Science 47:S154–S163.

Ehlers, J.D., and A.E. Hall. 1997. Cowpea (Vigna unguicu-lata L.Walp.). Field Crops Research 53:187–204.

Emechebe, A.M., and S.A. Shoyinka. 1985. Fungal andbacterial diseases in Africa. In: Singh, S.R., and K.O.Rachie (eds.) Cowpea research, production, and utiliza-tion, Chichester: John Wiley and Sons, pp. 173–192.

Fery, R.L. 1990. The cowpea: production, utilization, andresearch in the United States. Horticultural Reviews12:197–222.

Hall, A.E., B.B. Singh, and J.D. Ehlers. 1997. Cowpeabreeding. In: Janick, J. (ed.) Plant Breeding Reviews,Vol. 15, New York: John Wiley and Sons, pp. 215–274.

Hall, A.E., N. Cisse, S. Thiaw, H.O.A. Elawad, J.D.Ehlers, A.M. Ismail, R. Fery, P.A. Roberts, L.W.Kitch, L.L.Murdock, O. Boukar, R.D. Phillips, K.H.McWatters. 2003. Development of cowpea cultivars andgermplasm by the Bean/Cowpea CRSP. Field CropsResearch 82:103–134.

Hall, A.E. 2004. Comparative ecophysiology of cowpea,common bean, and peanut. In: Nguyen, H.T., and A.Blum (eds.) Physiology and biotechnology integrationfor plant breeding. New York: Marcel Dekker Inc, pp.271–325.

Higgins, T.J.V., S. Gollasch, L. Molvig, A. Moore, C.Popelka, P. Watkins, J. Armstrong, R. Mahon, J.Ehlers, J. Huesing, V. Margam, R. Shade, and L.Murdock. 2012. Genetic transformation of cowpea forprotection against bruchids and caterpillars. In: BoukarO., O. Coulibaly, C. Fatokun, K. Lopez, and M. Tamo(eds.) Innovative research along the cowpea value chain,International Institute of Tropical Agriculture, Ibadan,Nigeria, pp. 131–137.


Koona P., E.O. Osisanya, Len. Jackai, M. Tamo, andR.H. Markham. 2002. Resistance in accessions of cow-pea to the Coreid Pod-Bug Clavigralla tomentosicollis(Hemiptera: Coreidae). Journal of Economic Entomol-ogy 95:1281–1288.

Kuhn C.W., 1990. Cowpea virus diseases in the United States:a status report. In: Miller J.C., J.P. Miller, and R.L. Fery(eds.) Cowpea research: a U.S. perspective, MP 1639.Texas Agricultural Experiment Station, College Station,TX, USA, pp. 7–23.

Lane, J.A., T.H.M. Moore, D.V. Child, K.F. Cardwell, B.B.Singh, and J.A. Bailey. 1993. Virulence characteristicsof a new race of the parasitic angiosperm, Striga gesner-ioides, from southern Benin on cowpea (Vigna unguicu-lata). Euphytica 72:183–188.

Li. G., Y. Liu, J. D. Ehlers, Z. Zhu, X. Wu, B. Wang, andZ. Lu. 2007. Identification of an AFLP fragment linkedto rust resistance in asparagus bean and its conversion toa SCAR marker. Horticultural Science 42:1153–1156.

Lin, M. T., and G. P. Rios. 1985. Cowpea diseases and theirprevalence in Latin America. In Singh, S.R., and K.O.Rachie (eds.) Cowpea research, production and utiliza-tion, New York: John Wiley and Sons, pp. 199–204.

Lucas, M.R., N.-N. Diop, S. Wanamaker, J.D. Ehlers, P.A.Roberts, and T.J. Close. 2011. Cowpea-soybean syntenyclarified through an improved genetic map. The PlantGenome 4:218–225.

Lucas, M.R, J.D. Ehlers, P.A. Roberts, and T.J. Close.2012. Markers for quantitative inheritance of resis-tance to foliar thrips in cowpea. Crop Science 52:2075–2081.

Lucas, M.R., B.L. Huynh, J.D. Ehlers, P.A. Roberts,and T.J. Close. 2013. High-resolution SNP geno-typing reveals a substantial problem among breederresources. The Plant Genome, Open Access, Vol. 6, No.1, DOI:10.3835/plantgenome2012.08.0020.

Muchero, W., N.-N. Diop, P.R. Bhat, R.D. Fenton, M.Pottorff, S. Hearne, N. Cisse, C. Fatokun, J.D. Ehlers,P.A. Roberts, and T.J. Close. 2009. A consensus geneticmap of cowpea [Vigna unguiculata (L) Walp.] and syn-teny based on EST-derived SNPs. Proceedings of theNational Academy of Sciences 106:18159–18164.

Muchero, W., J.D. Ehlers, and P.A. Roberts. 2010. QTLanalysis for resistance to foliar damage caused by Thripstabaci and Frankliniella schultzei (Thysanoptera: Thrip-idae) feeding in cowpea [Vigna unguiculata (L.) Walp.].Molecular Breeding 25:47–56.

Muchero, W., J.D. Ehlers, T.J. Close, and P.A. Roberts.2011. Genic SNP markers and legume synteny reveal can-didate genes underlying QTL for Macrophomina phase-olina resistance and maturity in cowpea [Vigna unguicu-lata (L) Walp.]. BMC Genomics 12:8.

Murdock, L.L., O. Coulibaly, T.J. Higgins, J.E. Huesing,M. Ishiyaku, and I. Sithole-Niang. 2008. Cowpea. In:Kole, C., and T.C. Hall (eds.) Compendium of trans-genic crop plants: transgenic legume grains and forages.Hoboken, NJ: Wiley-Blackwell, pp. 23–56.

Myers, G.O., C.A. Fatokun, and N.D. Young. 1996. RFLPmapping of an aphid resistance gene in cowpea (Vignaunguiculata L. Walp). Euphytica 91:181–187.

Ouedraogo J.T., V. Maheshwari, D. Berner, C.-A. St-Pierre,F. Belzile, and M.P. Timko. 2001. Identification of AFLPmarkers linked to resistance of cowpea (Vigna unguicu-lata L.) to parasitism by Striga gesnerioides. Theoreticaland Applied Genetics 102:1029–1036.

Ouedraogo, J.T., B.S. Gowda, M. Jean, T.J. Close, J.D.Ehlers, A.E. Hall, A.G. Gillaspie, P.A. Roberts, A.M.Ismail, G. Bruening, P. Gepts, M.P. Timko, and F.J.Belzile. 2002a. An improved genetic map for Cowpea(Vigna unguiculata L.) combining AFLP, RFLP, RAPDand biochemical markers and biological resistance traits.Genome 45:175–188.

Ouedraogo J.T., J.-B. Tignegre, M.P. Timko, and F.J.Belzile. 2002b. AFLP markers linked to resistanceagainst Striga gesnerioides race 1 in cowpea (Vignaunguiculata). Genome 45:787–793.

Patel, P. N. 1985. Fungal, bacterial and viral diseases ofcowpeas in the USA. In: Singh, S.R., and K.O. Rachie(eds.) Cowpea research, production and utilization. NewYork: John Wiley and Sons, pp. 205–213.

Popelka, J.C., S. Gollasch, A. Moore, L. Molvig, and T.J.V.Higgins. 2006. Genetic transformation of cowpea (Vignaunguiculata L.) and stable transmission of the transgenesto progeny. Plant Cell Reports 25:304–312.

Pottorff, M., S. Wanamaker, J.D. Ehlers, P.A. Roberts,and T.J. Close. 2012a. Genetic and physical mapping ofcandidate genes for resistance to Fusarium oxysporumf.sp. tracheiphilum Race 3 in cowpea [Vigna unguiculata(L) Walp]. PLoS ONE 7: e41600. DOI:10.1371.

Pottorff, M., G. Li, J.D. Ehlers, T.J. Close, and P.A.Roberts. 2012b. Genetic mapping, synteny and physicallocation of two loci for Fusarium oxysporum f.sp. tra-cheiphilum race 4 resistance in cowpea [Vigna unguicu-lata (L.) Walp]. Molecular Breeding (under review).

Rafalski, A. 2002. Applications of single nucleotide poly-morphisms in crop genetics. Current Opinions in PlantBiology 5:94–100.

Singh, S.R., and L.E.N. Jackai. 1985. Insect pests of cow-peas in Africa: their life cycle, economic importance andpotential for control. In Singh, S.R., and K.O. Rachie(eds.) Cowpea research, production and utilization. NewYork: John Wiley and Sons, pp. 217–231.

Singh, B. B. 1993. Grain legume improvement program.Part 1. Cowpea breeding. Archival Report (1988–1992).International Institute of Tropical Agriculture, Ibadan,Nigeria.

Singh, B.B., O.L. Chambliss, and B. Sharma. 1996.Recent advances in cowpea breeding. In: Singh B.B.,D.R. Mohan Raj, K.E. Dashiell, and L.E.N. Jackai(eds.) Advances in cowpea research. Co-publication ofInternational Institute of Tropical Agriculture (IITA)and Japan International Research Center for Agricul-tural Sciences (JIRCAS). Devon, UK: Sayce Publishing,pp. 30–49.


Solleti, S.K., S. Bakshi, J. Purkayastha, S.K. Panda, andI. Sahoo. 2008. Transgenic cowpea (Vigna unguiculata)seeds expressing a bean α-amylase inhibitor 1 conferresistance to storage pests, bruchid beetles. Plant CellReports 27:1841–1850.

Timko, M.P., J.D. Ehlers, and P.A. Roberts. 2007. Cow-pea. In: Kole, C. (ed.) Genome mapping and molecularbreeding in plants, Vol. 3, Pulses, sugar and tuber crops,Berlin and Heidelberg: Springer Verlag, pp. 49–67.

Tosti, N., and V. Negri. 2002. Efficiency of three PCR-based markers in assessing genetic variation among cow-pea (Vigna unguiculata subsp. unguiculata) landraces.Genome 45(2):268–275.

Williams, R.J. 1975. Diseases of cowpea (Vigna unguicu-lata (L.) Walp.) in Nigeria. Tropical Pest Management21:253–267.

Xu, Y., and J.H. Crouch. 2008. Marker-assisted selectionin plant breeding: from publications to practice. CropScience 48:391–407.

Xu, P., X. Wu, B. Wang, Y. Liu, D. Qin, J.D. Ehlers,T.J. Close, T. Hu, Z. Lu, and G. Li. 2010. Devel-opment and polymorphism of Vigna unguiculata ssp.unguiculata microsatellite markers used for phyloge-netic analysis in asparagus bean (Vigna unguiculata ssp.sesquipedialis. L. Verdc.). Molecular Breeding 25:675–684.