Response of cowpea genotypes to Alectra vogelii parasitism in Kenya

Cowpea is popular in Eastern Kenya where it is attractive to farmers because of its high economic value and the belief that it does not require many external inputs. Farmers are however discouraged to grow the crop in this region due to massive attack by a parasitic weed Alectra vogelii (Benth). Yield losses due to A. vogelii have being estimated to range from 50 to 100% in Mbeere, Kitui and Makueni districts. No single method however is available to farmers in these regions in control of the parasitic weed. Combining several control methods, as in the management of Striga spp in Western Kenya should be a sustainable option. Field studies were conducted in 2010 and 2011 at Kenya Agricultural Research Institute (KARI), Kiboko farm to determine the response of 143 cowpea genotypes to Alectra infestation. The aim for the study was to identify resistant genotypes that could be used in breeding programme. Significant differences were observed amongst cowpea genotypes in days to first Alectra emergence, number of Alectra shoots emerged at 6, 8, 10 and 12 week after planting and grain yield. Cowpea genotypes Kir/Nya-005 and Mbe/Mach-022 showed complete resistance to Alectra while Ken-Kunde, M66 and K80 (all commercial varieties) supported the highest number of Alectra shoots. Grain yield loss in the three susceptible varieties was 80, 79 and 50% respectively. On the other hand, Sia/Cia-004, Mbe/Mach-014 and Kib-006 had high grain yields despite the high number of Alectra shoots present. There was a strong correlation (r = -0.57) between grain yield and number of Alectra shoots emerged at 12 weeks after planting. A significant negative (r = -0.37) correlation was also obtained between pod number per plant and number of emerged Alectra shoots at 12 weeks after planting. This negative correlation proves the high accumulation dry matter in the cowpea roots at the expense of the pods thus decreasing grain yield. This information showed that there is sufficient genetic variability in the cowpea genotypes studied, which can be exploited in breeding improved cowpea varieties for resistance to A. vogelii in Kenya. A great progress towards developing improved cowpea variety that meets farmer’s preferences with durable resistance to A. vogelii can be achieved if the genes from the resistant and tolerant local cowpea cultivars identified in this study could be introgressed into the adapted susceptible improved varieties. This will increase the potential impact of adoption of resistant cowpea varieties in the zones.


INTRODUCTION
Cowpea (Vigna unguiculata L. (Walp.) is an herbaceous warm-season annual crop that is similar in appearance to common bean except that leaves are generally darker green, shinier, and less pubescent. Early maturing cowpea cowpea varieties provide the first food from the current harvest sooner than any other crop (in as few as 55 days after planting), thereby shortening the "hunger period" that often occurs just prior to harvest of the current season's crop in farming communities in the developing world. The relatively high protein content (22%) of cowpea makes it an important supplement to the diet of many African people (Bressani, 1995) who consume cereals, roots, and tuber high in carbohydrate and low in protein.
Cowpea is the second most important grain legume in Kenya after common beans. The area under cowpea is estimated at 1800 hectares excluding the area under the crop in home gardens (Kimiti et al., 2009). About 85% of the total area under the cowpea is in arid and semi-arid lands (ASALs) of Eastern province and 15% in the Coast, Western, and Central provinces (Kimiti et al., 2009). Despite its importance in the dry regions of Eastern Kenya, its potential, growth and yield are constrained by several abiotic and biotic factors. Among them include low soil fertility, inadequate farm inputs, noxious weeds, pest and diseases and lack of seeds during planting times. This has decreased in the yield potential of 1500 to 239 kg/ha (Kimiti et al., 2009). A parasitic weed Alectra vogelii (Benth) an obligate root-parasitic flowering plant of the family Scrophulariaceae is one of the major concerns in lowering cowpea yields. In 1929, one report estimated a 20% loss in yield for cowpea crops in Embu district Eastern Kenya (Bagnall-Oakley et al., 1991).
A. vogelii has also being observed in Bambara (vigna subterranean (L) Verdc.), soyabean (Glycine max (L) Merr.), mung bean (Vigna radiate (L.) Wilczek.), ground nut (Arachi hyogaea L.) and common bean (Phaseolus vulgaris L.) (Botha, 1948;Visser, 1978;Salako, 1984;Riches, 1989;Riches et al., 1992;Lagoke et al., 1993). Although A. vogelii is autotrophic, its photosynthetic rate is half that of host leaf on per gram dry mass basis (Gouws et al., 1980). It is said that it is not able to fix carbon at a daily rate in excess of its diurnal requirements (Harpe et al., 1981). This means A. vogelii depends on the host photosynthate as it induces the formation of lateral roots of the host plant , also for water and nutrients. Information on cowpea yield losses resulting from A. vogelii infection is very scarce, but ranges from 41% (Lagoke et al., 1993) and 80 -100% in Botswana (Riches,1989) in a highly susceptible cultivar Black-eye. In Tanzania yield losses of up to 50% have been reported (Mbwaga et al., 2000). Yield losses of up to 15% have been reported in groundnut in Nigeria (Salako, 1984), while in South Africa 30 -50% reductions in yield of bambara were reported (Beck, 1987). The negative effect of the weed in reducing the vegetative and grain yield of cowpea has been well researched (Botha, 1948;Visser et al., 1977Visser et al., , 1990Okonkwo and Raghavan, 1982;Rambakudzibga et al., 2002). Reports show that yield reduction is mediated through the delayed onset of flowering, reduced number of flowers and pods, and reduced mass of pods and grain (Mugabe, 1983).
From a recent survey of the level of A. vogelii infestation on farmers' field, Karanja et al. (2010) observed that more than 80% of the fields grown to cowpea in Mbeere district of Eastern Kenya were infested with A. vogelii leading to serious crop losses. The threat of A. vogelii to the crop is increasing with farmers reporting up to 100% yield loss under severe infestation in these regions (Karanja et al., 2010). It is expected that the rapid spread of this parasitic weed and the enormous yield reduction caused would constitute a severe threat to cowpea production in the region.
Despite the scourge of Alectra on cowpea, relatively less work has been done on its control in Kenya. Several control measures, including hand weeding, chemical control, biological control, trap crops and host plant resistance have been suggested (Boukar, 2004;Riches, 1993). Of all these methods, host plant resistance appears the most effective, economical and environmentally friendly method in controlling the Alectra and affordable to farmers (Rubiales et al., 2006;Mainjeni, 1999;Riches, 1989). While a number of improved, high yielding striga/Alectra resistant cowpea genotypes have been developed and are fast becoming popular with farmers in Nigeria, Bukinafaso, Malawi, Tanzania and South Africa Kabambe et al., 2008), the same cannot be said for Kenya.
Cowpea as a crop of resource-poor households been affected by A. vogelii, it imposes an additional stress with which farmers, who have little capacity for investment in crop production, have to cope in an environment characterized by marginal rainfall for cropping and declining soil fertility. Cowpea has traditionally been grown in multiple cropping systems in which low populations of landraces are planted in mixtures with cereals. An increase in the importance of A. vogelii in the last 20 years in Kenya has often been associated with a change to sole cropping of introduced, potentially higher yielding susceptible cultivars, an increase in the area and frequency of cultivation (Farms' own communication, Mbeere district). Based on the foregoing, it is clear that there is need to screen for Alectra resistance among existing local cowpea cultivars or varieties. This would aid in identifying A. vogelii resistant genotypes that can be exploited in breeding improved cowpea varieties for resistance to A. vogelii in Kenya. A great progress towards developing improved cowpea *Corresponding author. E-mail: jakakah79@gmail.com. variety that meets farmer's preferences with durable resistance to A. vogelii can be achieved if the genes from the farmer's local cultivars are introgressed into the adapted susceptible improved varieties in the area. This would increase the potential impact of adoption of resistant cowpea varieties in the zones. The objectives of the studies in this paper were to collect and screen collected accessions of cowpea for their resistance/tolerance/ susceptibility to A. vogelii in Kenya.

MATERIALS AND METHODS
Pot and field evaluations were conducted at the research farm of Kenya Agricultural Research Institute, Kiboko sub-centre (11° 11'N; 7° 38'E; 686 m above sea level). The pot screening was conducted in 2009 while field trials were conducted in 2010 and 2011 under irrigation. A total of 143 cowpea genotypes were collected from Mbeere, Kitui, Makueni and Makindu ( Figure 1). The 143 cowpea genotypes and two commercial checks were planted in pots. The experiment was arranged in a randomized complete block design with three replications. Alectra inoculum stock was obtained by thoroughly mixing 30 g of Alectra seeds with 500 g of sieved sand. About 10 g of A. vogelii from the stock (about 7,000 Alectra vogelii seeds) were mixed with 1.8 kg of top soil (4-10 cm) and transferred into 2 kg polythene pot. Four seeds were sown in each pot and thinned to two plants two weeks after planting. The pots were watered everyday and maintained free from any other weed other than A. vogelii in the course of the research through hand weeding. Observation was made on days to cowpea flowering, days to A. vogelii first emergence, number of emerged Alectra shoot at 6, 8, 10 and 12 weeks after planting (WAP) and grain yield. Analysis of variance (ANOVA) was performed on all data with the general linear model (GLM) procedure of SAS (version 8.0; SAS Institute, Cary, NC). The GLM procedure of SAS uses the method of least squares to fit data to a general linear model. The accessions were rated as resistant/tolerance and susceptible depending on number of Alectra plant/s emerged and grain yield.
Twenty eight cowpea accessions were selected from the pot results for further screening. The 28 cowpea genotypes and two commercial checks; M66 and K80 were further evaluated to re-validate their reaction to A. vogelii in 2010 at Kiboko research farm. The soils of the area are broadly classified as alfisols. The soil of the experimental site was loam with pH of 6.7. The trials were conducted under irrigation in a field that had previously been observed to be heavily infested with A. vogelii. The land was ploughed with disc harrow in order to get a fine tilth. Cowpea cultivars were arranged in a randomized complete block design with 3 replicates. Each plot consisted of four rows 5m long. Two seeds were sown at intra-row spacing of 0.20 m between hills and 0.75 m between rows. About 10 g of A. vogelii inoculum from the stock (as explained above) were added to each planting hill and thoroughly mixed before sowing to add the pool of Alectra seeds in the soil. The trials were always surrounded with three rows of Alectra susceptible check (M66). All the weeds other than A. vogelii were controlled through hand weeding. Cyper-methrin and dimethoate were applied with a knapsack sprayer fortnightly at the rate of 1.01/ha 4 WAP until harvest to control insect pests.
At 8 and 10 WAP, the numbers of emerged Alectra shoots were recorded to assess the host support for growth in the two central rows of each plot (3.75m 2 ). All the plots were harvested at maturity and grain yield was measured as the weight of threshed grain from a plot.
In 2011, three cowpea genotypes; Kir/Nya-005, Mbe/Mach-022 and Mnk/Wot-003 identified to support zero Alectra emergence in 2010 were further evaluated alongside three widely grown cowpea genotypes; M66, K80 and Ken-Kunde to assess their yield potential under artificial, natural Alectra infestation, and Alectra-free condition. The trials were arranged in a randomized complete block design and in three replications. Annual weeds, except Alectra were controlled by hoe-weeding at 2 and 5 WAP and hand pulling during the last two weeding to avoid tampering with the un-emerged Alectra shoots.
Data on the number of days to first Alectra emergence and number of emerged Alectra shoots were collected from the two middle rows of each plot at 6, 8, 10, and 12 WAP from the artificial and natural Alectra infested treatment, while number of days to cowpea flowering, number of seeds for 10 randomly selected pods and grain yield at physiological crop maturity were collected from the three treatment. Analysis of variance (ANOVA) was performed on all data with the general linear model (GLM) procedure of SAS (version 8.0; SAS Institute, Cary, NC). The GLM procedure of SAS uses the method of least squares to fit data to a general linear model. Duncan Multiple Range (DMR) Tests were performed to compare treatment means at the 5% level.

Cowpea germplasm collection
Cowpea germplasm were collected late in the season, 2009 from farmers in Mbeere, Kitui, Yatta, Machakos, Makueni and Makindu districts of Eastern Kenya ( Figure  1). Other than Machakos district, A. vogelii incidences were found to have severely affected cowpea production in Mbeere, Kitui, Yatta and Makueni districts. A total of 143 cowpea landraces were collected and were found almost all of them to have mixed seed colours ranging from white to black, with cream and red colours dominating. Also they were of different seed sizes from small to large, but large was dominating. The collections were sor-sorted according to seed colour and screened against A. vogelii resistance. From farmers' views, mixed seed colour was not a problem as the majority grow the crop for green vegetables and home consumption and local markets. However, those found to grow the crop for other markets reported said that traders prefer uniform seed colour.
In both pot and field trials, Alectra shoots emerged on susceptible cowpea genotypes 44 days after planting (Table 2). However, the genotypes varied significantly in their support for Alectra shoots in the year. High number of Alectra shoots was observed at 10 and 12 WAP during the screening period. Genotypes Kir/Nya-005, Mbe/ Mach-022 and Mkn/Wot-003 exhibited zero support for Alectra shoots, both in pot and field screenings. Under pot experiment, 81% of the collected cowpea genotypes supported A. vogelii emergence while 19% showed resistance for the parasitic weed (data not shown). After prescreening the selected cowpea genotypes from the pot experiment were planted at Kiboko field to confirm their resistance against two checks (M66 and K80), Kir/Nya-005, Mbe/Mach-022 and Mkn/Wot-003 confirmed their resistance (Table 1). From Table 1, Kir/Nya-005, Mbe/ Mach-022 and Mkn/Wot-003 supported no Alectra emergence while Mbe/Mach-004, Gac/Kar-003, Sia/Wit-004 and Ki-006 recorded the highest Alectra shoots at 10 WAP compared to M66 and K80. At 10 WAP, the number of emerged Alectra shoots ranged from 0 to 127 per 7.5 m 2 under artificial infestation.
In 2011, there was significant differences in the number of emerged Alectra shoot per plot in both artificial and natural infestation. The results showed that Mbe/Mach-022, Mkn/Wot-003 and Kir/Nya-005 completely supported zero Alectra weed emergence. However, Alectra first emergence coincided with 50% days to host flowering in susceptible genotypes (Table 2). There was no significant difference on 50% days to cowpea flowering and number of seeds per pod under Alectra infestation and Alectra free condition. Alectra infestation significantly reduced grain yield of the susceptible cowpea genotypes (Table  3). Yield losses were statistically highly significant with Ken-Kunde and M66 recording 80 and 79%, respectively. Resistant genotype Kir/Nya-005 was the most stable recording insignificant reduction of 6.5% (Figure 1).
Mkn/Wot-003 which recorded zero number of emerged Alectra shoots recorded grain yields reduction of 63.5% (Table 3 and Figure 2).

DISCUSSION
For seed size, large seeded types were more preferred Means followed by the same letter (s) in a column are not significantly different at 5% level of probability using Duncan multiple range test.
at the local markets than the small seed. These are good guidelines that are to be considered when breeding for Alectra resistance to select what the farmers and market like. This study showed significant differences exist amongst cowpea genotypes in their performance under Alectra infestation. These differences occurred in the number of emerged Alectra shoots and grain yield. Lower cowpea grain yield with number of emerged Alectra shoot, shows that Alectra infestation reduced grain yield for cowpea genotypes susceptible to Alectra. This is well signified by Ken-Kunde which is the highest yielder under un-infested conditions but lowest under Alectra infested conditions. Ken-Kunde, M66 and K80 which had high grain yield reduction (over 50%) could be regarded as being susceptible. The symptoms displayed these susceptible cultivars were that of stunted growth, chlorosis and premature defoliation as earlier reported by Magani (1994). Longe et al. (2002) observed that resultant chlorosis could be due to chlorophyll degradation which result to reduction in photosynthetic site hence, yield reduction. A combination of high yield, growth potentials and no support for Alectra shoots by Kir/Nya-005, and Mbe/ Mach-022 cultivars make them suitable candidates for use to improve the genetic base of existing cultivars. The   failure of Alectra emergence may be due to low production of germination stimulants by these two genotypes or host-plant-parasite incompatibility whereby the initiation of haustoria, and subsequent attachment and penetration are inhibited (Okonkwo and Raghavan, 1982). This confers resistance of the genotypes as earlier reported in cereals to striga resistance (Ejeta, 1993a, b). In 1983, Mugabe showed delayed onset of flowering, reduced number of flowers and pods and reduced weight of pods and seeds in cowpea due to Alectra infestation. This indicates that Sia/Wit-004, Sia/Cia-005, Kib-021, Mbe/Mach-004 and Sia/Wit-001 are either moderately resistance or tolerant to A. vogelii. This might have been achieved through low production of germination stimulant which in one of the resistance mechanism advocated by Yohanna et al., 2010. Kurech andAlabi (2003) considered tolerance as a horizontal resistance which is polygenic in contrast to vertical resistance which breaks down faster with time. Since the tolerant genotypes can produce high yield in spite of high parasitism, it implies they have to be very efficient in the production of assimilates to support the parasites and still have enough to give high yields (Musell, 1980). Several authors (Atokple et al., 1995;Adetimirin et al., 2000; have indicated that the use of moderately tolerant varieties in combination with other control measures help in the depletion of Alectra/Striga soil seed bank. The decrease in the yield of varieties Mnk/Wot-003 (variety that was hardly found with any attached Alectra) by Alectra could have been partly due to the reduction in its root growth and root nodulation by the Alectra. This could result in inadequacy of nitrogen and nutrient absorption for adequate shoot growth and thereby reduced yield production (Dart and Mercer, 1965). It is also likely that, the seeds of Alectra contain certain toxins which leaked into the soil and hindered Mkn/Wot-003 root growth, but this may require further investigation as also suggested by Omoigui et al., 2007. In addition to the export of nutrients, water and metabolites from host to the parasite, Alectra prevents nodulation (Kurech and Alabi 2003).

Conclusion
These results could be used as a preliminary basis for choosing cowpea genotypes in encouraging farmers to grow and assembling management packages for A. vogelii to ensure reduced risk to farmers. Compared to Ken-Kunde, M66 and K80, it was interesting to note that the Kir/Nya-005 and Mbe/Mach-022 had zero emergences Alectra recorded throughout the evaluation period hence suggesting a possible absolute resistance. This information would be valuable for breeding effort to develop or select Alectra resistant cowpea varieties. However, there is need to evaluate the genotypes with farmers.

ACKNOWLEDGEMENT
This publication is part of the output from a project funded by the European Union through Kenya Arid and Semiarid lands program in Kenya Agricultural Research Institute (KARI) to whom we are very grateful.