Cowpea (Vigna unguiculata L. Walp.) is an edible legume crop grown for its grain while leaves may be used as a vegetable, green manure or animal  feed  or  as  a  silage crop. Cowpea grain contains 24 to 32% protein (Neil et al., 1992), and the crop is adaptable to drought (Hall, 2004). In Malawi cowpeas are important in the dry areas where legumes such as beans (Phaseolus vulgaris L.) do not do well (MoAFS, 2012). In the 2011 to 2012 season average cowpea yields were 348 kg ha^-1, against potential yields of 2000 to 2500 kg ha^-1 (MoAFS, 2012). Like many legumes, cowpeas are complimentary to cereals and other non-legume crops through fixation of atmospheric nitrogen in a process called biological N fixation (Nyemba and Dakora, 2010; Snapp, 1998; Ojiem et al., 2000; Myaka et al., 2006). In Burkina Faso, Bado et al. (2006) reported that cowpeas accumulated 2.5 to 3.4 t ha^-1 of stover containing 50 to 115 kg ha^-1 N of which 52 to 56% was derived from the atmosphere. In soils low in phosphorus, the roots of cowpea develop effective mycorrhizal association improving the soil’s available phosphorus content (Valenzuela and Smith, 2002). The roots and fallen leaves of cowpea make significant differences to subsequent cereal yields (Carsky and Vanlauwe, 2002; Bado et al., 2006). The crop forms an excellent crop cover for suppressing weeds while at the same time providing soil erosion control (Valenzuela and Smith, 2002). Cowpeas are also considered suitable as cover crops in conservation agriculture systems (Mupangwa et al., 2012). Like many legumes, cowpeas are also important in the rotation or intercropping cereal crops where cowpea serves as a trap crop by helping to reduce the seed bank of striga spp, parasitic weed of cereals (Berner et al., 1996; Kabambe et al., 2005; Carsky et al., 1994). In Malawi, several efforts have been undertaken to promote legumes for the above reasons (Ngwira et al., 2012;   ICRISAT/MAI, 2000), including the national Farm Input Subsidy Programme (FISP), which included seeds of grain legumes (MoAFS, 2007; MoAIWD, 2012). 

 

From the foregoing, it is evident that most of the non-grain uses of cowpea are related to its above ground biomass and its nitrogen fixation. Thus in order for cropping systems planners to take full advantage of the same, it is important that information on biomass and nodulation is known for pre-release or released varieties. Many screening programs do not evaluate for this (Mazuma et al., 2008; Kabambe et al., 2014).  The objective of this study was to evaluate eight elite cowpea lines in an on-going cowpea project supported by the McKnight Foundation of USA for yielding ability, nodule formation and effectiveness and biomass production. 

Study location, treatments, experimental design and plots sizes

 

This study was conducted at Bunda College Crop and Soil Sciences Student’s Research Farm during the 2011/2012 growing season. The site is 1158 m above sea level, latitude 14° 35 S’ and longitude 33° 50 E’. The soil type varies from clay loam to sandy loam textural classes with medium fertility. The mean annual rainfall is approximately 1031 mm with coefficient of variation (cv %) of 16.6% indicating adequate reliability of total rainfall (Jones and Kanyama, 1975).  The  total  rainfall l for  the  season  under review (2011 to 2012) was 1001.7 mm. The crop received rainfall amounting to 816.5 mm between January and mid-April 2012 which was 81.5% of the season’s total rainfall. This was adequate rainfall for cowpea production and distribution was also good throughout the crop growing period.

 

The study evaluated eight ex-International Institute of Tropical Agriculture (IITA) elite cowpea lines in comparison to two released varieties Sudan 1 and IT82E-16, out of only three released cowpea varieties in Malawi (Mviha et al., 2011). Table 1 provides a list of the treatments and their descriptions. The lines were in advanced stages of evaluation by a cowpea project at Lilongwe University of Agriculture and Natural Resources, supported by the McKnight Foundation of USA. A randomized complete block design with three replications was used. Plots comprised of 4 ridges x 4.0 m long x 0.75 m apart. The net plot areas was comprised the middle 2 ridges x 4.0 m long x 0.75 m (area = 6 m²). Two seeds of cowpea were planted per station, spaced at 20 cm apart. The expected plant count was 13.3 plants m^-2 or 133,000 ha^-1.  There was no fertilizer of any kind applied to the plots. All these agronomic practices were in line with current recommendations (MoAFS, 2012). The experiment was planted on 3^rd January 2012. Nodulation was based on resident bacteria in the soil, not supplementary inoculation.

 

 

Data recording and collection

 

Data was collected and reported on the following parameters: Plant count at harvest, plant biomass, pods per plant, 100 seed weight, number of nodules, effective nodules, canopy width and plant height. Data on plant counts and grain yield were recorded from the net plot area of 6.0 m². Recorded grain yield was converted to 12% storage moisture. Pods per plant were recorded from the plants harvested and expressed as pods m^-2. Cowpea plant biomass or fodder was the sum of periodic cowpea leaf fall collections and above ground fodder from plants at harvest. Dry leaves from free fall were collected from litter traps covering an area of 2 m x 0.75 m (1.5 m²) in the net plot area. The traps were simply plastic sheet laid in between plant rows at 2 weeks after planting. Three collections were made at 8, 11 and 14 weeks after planting (WAP) and oven dried at 72°C for 48 h (Koide et al., 2000). Fodder at harvest was taken from all above ground plant matter from net plot area and oven-dried as above. Harvesting was done in mid-April after 100 days from planting (DAP). Average N content of fodder was determined by taking sub-samples of combined fodder from leaf fall and standing plants in all plots in each of the three replications. The sample analyzed for N using the Kjeldhal method as modified by Anderson and Ingram (1993).

 

Canopy width and plant height (cm) was an average five readings within the net plot. Canopy width was recorded by as linear measurement of the width of the canopy at 49 DAP (mid-flowering) only. A maximum distance of 0.75 m reflected full canopy closure, as this was the row width. Plant height was linear measurement from base to tallest upright growing point (tip) and was taken at 49 days after planting only. Number of nodules per plant was recorded as an average from four plants that were dug out per treatment and nodules found on the plants were counted. This was done when the cowpeas had just started flowering and sampling was done on borders of ridges because it was destructive sampling. Nodule effectiveness was assessed by looking at the inside color of nodules. Ten nodules from each variety/elite line were sampled, opened and color-examined. The ten nodules were randomly collected from the four plants dug out in a treatment. Those nodules with red/brown color were deemed effective and those with different colors such as green or white (lacking in leghemoglobin) were considered not effective for nitrogen fixation according to Gobat et al. (2004). Harvest index was calculated by as: [total grain yield per  hectare/(total above  ground  biomass  per hectare)] that is (above ground biomass + grain yield + pod weights).  

 

Data analysis and mean comparison

 

Data were analyzed using the analysis of variance and regression procedure using Genstat 16 statistical package, and means were separated by Least Significant Difference (LSD) at 5%. 

Cowpea growth, yield and yield components

 

The analysis of variance showed that there were significant treatment differences (<0.05) on plant height, podsm^-2, pod length, seeds per pod, 100 seed weight and grain   yield  (Table 2).  There  were  no  significant  entry effects on harvest count (mean 130,556 = plants ha^-1), harvest index (mean = 0.52) and canopy width (mean = 50.7 cm). The absence of significant difference on canopy width and harvest index is surprising considering that there were significant yield differences. The observed variability in plant height is expected as these are distinct lines with varying growth habits. Valenzuela and Smith (2002) noted that cowpea grows rapidly, reaching a height of 48 to 61 cm when grown under favorable conditions. On seed weight, the seed sizes of all entries was generally small in comparison to sizes of up 21 g/100 seeds reported by Omokonye et al. (2003).

 

 

On grain yield, the released variety Sudan 1 gave joint highest yield with the elite line IT98K-205-8 of about 3.0 t ha^-1. Only one other elite entry yielded above 2.5 tha^-1, along with released IT82E-16.  The high yielding entries also had higher pods m^-2.  The high yield of Sudan 1 was associated with long pod length,  seed  weight  and  plant height, but this was not true for the other entries with high grain yield. Only two of the elite lines compared well with potential yields of 2.0 to 2.5 tha^-1 for Malawi (Mazuma et al., 2008; MoAFS, 2012; Kabambe et al., 2014). Mazuma et al. (2008) also reported that Sudan 1 and IT82E-16 were amongst high yielding entries giving yields of 1789 and 1509 kgha^-1 respectively. Results are also comparable to those of Akande and Balogun (2009) who reported highest cowpea yields of 1737 kg ha^-1 in a multi-locational programme in Nigeria. However, average grain yield in this study was much higher than the average yield of 750 to 1000 kg ha^-1 in Nigeria (Omokanye et al., 2003; Kamara et al., 2008).

 

Leaf fall and total fodder yield

 

The results of leaf fall show significant entry differences for leaf fall at 8 WAP, 11 WAP, cumulative leaf fall and total biomass but not for leaf fall 3 (Table 3).  The results show that Sudan 1 and IT93K-693-2 shed most leaves at leaf fall 1 while IT99K-491-7 shed most leaves at leaf fall 2. Leaf fall declined with sampling time. It is therefore likely that some of the fallen leaves could contribute to improved fertility status and yield in cropping systems involving intercropping with long season crop types. However, early leaf may not be compatible with systems aiming to use residues for animal feed or as mulch to prevent evaporation in subsequent dry months.

 

 

The highest cumulative leaf fall (1834 kgha^-1) came from Sudan 1, followed by 1553 kg ha^-1 from entry IT99K-491-7.  Notably, for Sudan 1 the cumulative leaf fall was 94 % of total biomass, while for the later this was 51 %. The highest fodder biomass  was  recorded  from  IT99K-491-7 (3021 kg ha^-1) followed by IT98K-503-1, IT99K1060 and Sudan 1 (2041-1931 kg ha^-1). These fodder yields are in same range as reported by Omokanye et al. (2003) and Bado et al. (2006). Of the entries in the study only two (IT99K-1060 and Sudan 1) gave yield and fodder of about 2000 kg and could be considered dual purpose.  The average N content of the stover was 2.4%, thus giving potential contributions of 24 to 72 kg ha^-1 nitrogen. The N content compares well with 2.5% reported in pigeon pea (ICRSAT/MAI 2000), but below 3.6% reported by Bado et al. (2006) in cowpeas. In general a minimal green manure rate of 2.0 t ha^-1 is considered as agronomic threshold (ICRISAT/MAI 2000). While these entries are not purpose-bred, the results show the potential benefit of screening for the same at later stages of research programs to enrich uses of released varieties.

 

Nodulation

 

There were significant treatment differences on number of nodules per plant (Table 4), but not on % effective nodules (mean = 95.0%).  The highest nodule number came from Sudan 1 (21.7 nodules/plant), followed by IT8282E-16 (18.9), IT95-1090-12 (14.42) and 1T99K-1060 (14.0) Only two entries (Sudan 1 and IT99K-1060) appear to give grain and forage yields of about 2 tha^-1 or more and high nodule numbers. These results concur with the ranges of 3-21 nodules per plant reported by Bhuvaneswari et al. (1998). Ndor et al. (2012) showed that nodules per plant in cowpeas increased from 20 to 32 when P application was raised from 0 to 40 kg ha^-1 P. The authors reported cowpeas yield increase from  about 1.0 to 1.5 t ha^-1 with associated with the same P application rates. 

 

 

General discussion

 

A number of regression relationships were explored, and only the grain yield and pod m^-2 yielded a significant regression (Figure 1). This is a fitting result since pods m^-2 is an aggregate expression of many yield components and management factors including pest control. The results from this study have also shown that elite lines IT99K-491-7, IT98K-503-1 and IT99-1060 were outstanding fodder producers at over 2 t ha ^-1. The best fodder yield  of  3 t ha^-1  gave  1.8 t  ha^-1  grain  yield.  The results show that various combinations of biomass or grain yield options exist and it might be good for researchers to release a range of lines to provide for options in use. 

 

The authors have not declared any conflict of interest.

Authors would like to acknowledge the Lilongwe University of Agriculture and  Natural  resources  and  the McKnight Foundation for financial support to this work, and all staff of Crop and Soil Sciences of LUANAR who assisted in conducting this work.