Improvements in grain and fodder yield of cowpea (Vigna unguiculata) varieties developed in the Sudan savannas of Nigeria over the past four decades

Abstract

A field study was conducted to determine the rate of genetic improvement in grain and fodder yields of cowpea genotypes developed in the Nigerian Sudan savannas from 1970 to 2004. Results showed that grain yield ranged from 568 kg/ha for an old variety TVX3236 to 1851 kg/ha for a recently released variety IT04K-321-2. The average rate of increase has been 28 kg/ha/year, which corresponds to a genetic gain of 3.6%. Fodder yield ranged from 1363 kg/ha for a variety released in 1976 (TVX1836-0131) to 3346 kg/ha for IT98k-476-8 released in 1998 corresponding to a genetic gain of 1.96%. This indicated that selection for dual-purpose cowpea varieties with increased fodder as well as grain yields has been successful. Total dry matter, fodder yield, harvest index, and 100-seed weight were significantly correlated with grain yield. The strong relationship between grain and fodder yields showed the success made in selecting for dual-purpose cowpea varieties.

Full text

179
Proceedings of the Fifth World Cowpea Conference
Improvements in grain and fodder yield of cowpea
(Vigna unguiculata) varieties developed in the Sudan
savannas of Nigeria over the past four decades
A.Y. Kamara1*, S.U. Ewansiha1, H.A. Ajeigbe1, R. Okechukwu1, H. Tefera2,
O. Boukar1 and L.O. Omoigui3
1International Institute of Tropical Agriculture (IITA), PMB 5320, Ibadan, Nigeria
2IITA-Malawi, Chitedze Agricultural Research Station, PO Box 30258, Lilongwe, Malawi
3College of Agronomy, Department of Plant Breeding and Seed Science, University of Agriculture,
Makurdi, Nigeria
*Corresponding author: a.kamara@cgiar.org
Abstract
A eld study was conducted to determine the rate of genetic improvement in grain and
fodder yields of cowpea genotypes developed in the Nigerian Sudan savannas from 1970
to 2004. Results showed that grain yield ranged from 568 kg/ha for an old variety TVX3236
to 1851 kg/ha for a recently released variety IT04K-321-2. The average rate of increase
has been 28 kg/ha/year, which corresponds to a genetic gain of 3.6%. Fodder yield
ranged from 1363 kg/ha for a variety released in 1976 (TVX1836-0131) to 3346 kg/ha for
IT98k-476-8 released in 1998 corresponding to a genetic gain of 1.96%. This indicated that
selection for dual-purpose cowpea varieties with increased fodder as well as grain yields
has been successful. Total dry matter, fodder yield, harvest index, and 100-seed weight
were signicantly correlated with grain yield. The strong relationship between grain and
fodder yields showed the success made in selecting for dual-purpose cowpea varieties.
Introduction
Cowpea is a leguminous crop of vital importance to the livelihood of millions of people
in West and Central Africa. It provides nutritious grain and a less expensive source of
protein for both rural poor and urban consumers (Inaizumi et al. 1999). In addition to food
for humans, it is a valuable source of livestock fodder (Singh et al. 2003) making it very
attractive to farmers. Because of its ability to tolerate drought, cowpea is well adapted
to the semi-arid tropics. As a leguminous crop, cowpea improves soil fertility through its
ability to x atmospheric nitrogen (Sanginga et al. 2000). Together, these characteristics
have made cowpea an important component of the cropping systems of the dry savannas
of sub-Saharan Africa (Carsky et al. 2001). According to FAO statistics, cowpea is grown
on an estimated worldwide area of 14 million ha. Some 8 million ha of cowpea are grown
in West and Central Africa, especially in Burkina Faso, Cameroon, Mali, Niger, Nigeria,
and Senegal. In Nigeria, the production trend of cowpea shows a signicant improvement
with an increase of some 440% in area planted and an increase of some 410% in yield
over the period 1961 to 1995 (Ortiz 1998). According to Singh (2000), Nigeria produces
about 2 million tons of cowpea on 5 million ha of land. The production trend would have
moved upwards with the availability of more improved varieties and crop management
practices. Despite the potential for further yield increases, cowpea production faces
numerous problems including insect pest attack, Striga gesneroides parasitism, diseases,
and drought.

180 Innovative research along the cowpea value chain
Over the years, a great deal of progress has been made by IITA in breeding a range of high
yielding cowpea varieties with combined resistance to major diseases, insect pests, and
parasitic weeds and drought tolerance. Previously, from 1970 to 1988, cowpea research at
the IITA concentrated primarily on developing and distributing grain-type cowpea varieties,
which have been tested and released in many countries (Singh et al. 1997). Because
of the importance of cowpea grain and fodder in West Africa, IITA began a systematic
breeding program in 1989 to develop dual-purpose cowpea varieties (Singh et al. 2003).
This program combined breeding for high yield potential for grain as well as fodder with
resistance to major biotic and abiotic stresses. The general strategy is to develop a range
of cowpea varieties differing in growth habit and maturity, seed type, and for sole crop and
intercrop in different agroecologies (Singh et al. 2003). Since 1970, several cowpea lines
have been developed and distributed to partnership institutions in Nigeria and over 60 other
countries all over the world for evaluation and release (Singh 2000). Despite tremendous
improvement in cowpea yield over the past 40 years, there is still a wide gap between what
is obtained from the research station and that grown under on-farm conditions. However,
farmers that grow the improved varieties under well-managed conditions usually obtain
higher grain yields in the dry savannas of Nigeria. Ajeigbe et al. (2010 a, b) reported an
over 300% increase in the value of produce when improved cowpea varieties are used in
improved systems compared to local varieties under traditional systems. Although current
yield increases for most crops have resulted from the impact of both genetic and improved
crop husbandry, further improvement in yield may have to rely on genetic gains through
improvement in crop productivity per hectare or increased genetic tolerance to pests and
diseases (Francis 1991). This is because of high costs or non-availability of agrochemicals
and reduction in soil quality (Cassman 1999).
Plant breeders have been able to measure breeding progress by growing varieties
developed and released over a long period of time in the same environment (Tefera et
al. 2009). Evaluation of genetic improvement and the associated changes in agronomic
and physiological traits determining grain yield in crops may help identify traits of potential
value for future breeding. This kind of study has been undertaken for several major crops.
For example, Tefera et al. (2009) reported that grain yield potential of early maturing
soybean increased from 1117 to 1710 kg/ha during two decades of breeding in the Nigerian
savannas. Fodder yield also showed an annual increase of 22.8 kg per ha per year. Specht
et al. (1999) reported that soybean yields in the United States increased by 31.4 kg per ha
per year from 1972 to 1997. In Canada, a study on 41 early maturing soybean varieties
representing seven decades of breeding showed a yield improvement of 0.5% per year
with an associated decrease in protein content and improvement in lodging tolerance
(Voldeng et al. 1997). Kamara et al. (2004) reported a genetic gain of 0.41% per year for
maize varieties released from 1970 to 1999 in the Nigerian savannas. The increase was
associated with an increase in total biomass and kernel weight and reduction in plant
height and days to owering. No direct comparisons of grain yield potential and other
agronomic traits have been made for cowpea varieties developed in the Nigerian Sudan
savannas over the past four decades. Therefore, this study was carried out to determine
the rate of genetic improvement in grain and fodder yields and associated agronomic and
physiological changes.

181
Proceedings of the Fifth World Cowpea Conference
Materials and Methods
The eld experiments were conducted under rain-fed conditions in 2007 and 2008 at IITA
experimental station at Minjibr (12o 10’42N, 8o 39’ 33E, Alt. 453m) in the Sudan savanna of
northwest Nigeria (Table 1). A bulk soil sample was taken at the beginning of the trial and
analyzed for particle part analysis, N, P, K, and pH. Daily rainfall and minimum and maximum
temperature were also recorded.
Thirty-one medium-late maturing cowpea varieties (Table 2) representing cowpea varieties
developed for the Nigerian Sudan savannas from 1970 to 2004 were evaluated in this study.
These were best performing varieties selected to represent each decade. In both years, the trial
was laid out in a randomized complete block design with four replications. For all treatments,
prior to planting, each eld was disc-harrowed and ridged. The planting distance was 0.75 m
between rows and 0.25 m between plants. Three seeds of the cowpea varieties were sown and
later thinned to two plants per stand at two weeks after planting. Immediately after planting,
paraquat (1:1-dimethyl- 4, 4’-bipyridinium dichloride) was applied at the rate of 276 g a.i. L-1 to
control weeds. This was followed by hoe weeding three weeks later. Fertilizer at 15 kg each of N,
P and K was applied in the form of NPK 15:15:15 a week after planting. A standard formulation,
cypermethrin + dimethoate (Best Action) at the rate of 30 g +250 g a.i. L-1 was used to control
insect pests at ower bud formation, owering, and podding stages to control insect pests and
this was delivered with a knapsack sprayer.
Data were collected on days from sowing to when 50% of the number of plants per net plot
reached owering and to when 95% of the pods reached maturity. Mean number of pods per
plant, and total dry matter (leaf, stem, grain, and threshed pod) per plant were all taken from a
10-plant sample within the two central rows for each plot when the crop reached physiological
maturity. Harvest index was calculated by dividing grain weight per plant by total dry matter per
plant. Mean hundred seed weight was recorded for each plot. Grain yields were determined from
all plants harvested from two net 4-m-long central rows when the rst ush of pods was mature
and dry, and was reported on a 100% dry matter basis in kg/ha. Similarly, fodder yield was
based on all plants that were harvested from the two net central rows for each plot, sun dried to
constant weight, and calculated as kg/ha.
Analysis of variance and regression analysis were performed using PROC GLM and PROC
REG in SAS, respectively (SAS Institute 2003). Year and replications were treated as random
effects; varieties were treated as xed effects. In order to determine the annual rate of genetic
improvement for the period the varieties were released, linear regression analysis was used.
The regression analysis was based on the mean values of each trait for each variety against
the year of release for respective varieties. The ratio of genetic gain to the corresponding mean
values of the oldest varieties was used to estimate the relative gain of breeding for the period
1974–2004 (Tefera et al. 2010). Simple correlation coefcients were estimated between grain
yield and other traits.
Results and Discussion
Climate characteristics of the experimental site
In 2007, rains started in May, reached its peak in August, and ended in September with slight
shower. A total of 864.4 mm of rain was obtained in 34 rainy days (Table1). While in 2008 the
rains started in June, reached its peak in July, and ended in September with total rainfall of 698.3
mm in 41 rainy days. During the growth period, temperature ranged from a minimum of 23.05°C
to maximum of 31.48 °C in 2007 and minimum of 22.45 °C to maximum of 41.31 °C in 2008.

182 Innovative research along the cowpea value chain
Table 1. Rainfall and temperature at Mijibir during the trial period.
2007 2008
Month Rainfall Temperature Rainfall Temperature
(mm) (no) oC
(max)
oC
(min) (mm) (no) oC
(max)
oC
(min)
Jan 0 0 25.53 15.06 0 0 26.44 16.02
Feb 0 0 34.21 23.36 0 0 28.74 17.71
March 0 0 34.58 22.03 0 0 36.08 21.89
Apr 0 0 40 25.57 0 0 37.63 26.43
May 67.5 3 32.8 26.56 0 0 38.69 28.19
Jun 162.6 8 30.62 23.73 102.5 7 36.38 26.45
Jul 249.2 10 31.48 23.44 313.9 10 29.97 23.18
Aug 376 12 28.74 23.05 209.1 12 28.71 22.45
Sept 9.1 1 30.97 23.17 68.3 12 29.8 23.82
Oct 0 0 31.03 24.29 0 0 41.31 22.99
Nov 0 0 33.03 23.72 0 0 32.03 23.03
Dec 0 0 33.17 21.35 0 0 30.68 21.77
Total 864.4 693.8
Table 2. Cowpea varieties used for the evaluation.
Varieties Growth habit Year tested
IT00K-1207
IT00K-1263
IT00K-227-4
IT03K-316-1
IT03K-351-9
IT04K-217-5
IT04K-223-1
IT04K-321-2
IT04K-332-1
IT81D-985
IT81D-994
IT82D-889
IT84S-2246-4
IT86D-1010
IT86D-719
IT86D-721
IT87D-941-1
IT88D-867-11
IT89KD-391
IT90K-277-2
IT97K-1101-5
IT97K-461-4
IT97K-494-3
IT97K-499-35
IT98K-131-2
IT98K-476-8
IT98K-506-1
IT98K-628
TVx 3236 (SUVITA 4)
TVx 456-01F
TVx 66-24 (VITA 8)
TVx 1836-013J (VITA 10)
VITA 4 (TVu 1977-0D)
VITA 5 (TVu 4557)
Semi-determinate
Determinate
Determinate
Determinate
Semi-determinate
Semi-determinate
Semi-determinate
Semi-determinate
Determinate
Spreading
Semi-determinate
Determinate
Semi-determinate
Determinate
determinate
Determinate
Determinate
Semi-determinate
Semi-determinate
Semi-determinate
Determinate
Determinate
Determinate
Determinate
Semi-determinate
Semi-determinate
Determinate
Determinate
Semi-determinate
Determinate
Determinate
Determinate
Semi-determinate
Semi-determinate
2000
2000
2000
2003
2003
2004
2004
2004
2004
1981
1981
1982
1984
1986
1986
1986
1987
1988
1989
1990
1997
1997
1997
1997
1998
1998
1998
1998
1976
1975
1975
1976
1974
1974

183
Proceedings of the Fifth World Cowpea Conference
Variance analysis
Year and varietal effects were signicant for all the traits measured. Year × variety
interactions were signicant for all the traits except for harvest index which showed no
signicant interaction between year and variety (Table 3). The signicant year × variety
interaction was due to the differences in rainfall between the two years. Rainfall during
the growing period of cowpea (Table 1) was higher in 2007 than in 2008 both in amount
and distribution. The rainfall differences may have caused cross ranking of the cowpea
varieties between the two years. The signicant year effect for virtually all traits indicated
the sensitivity of the varieties to environmental factors that prevailed in the different
years. Tefera et al. (2009) reported similar year × variety interactive effects for most traits
measured in soybean in the Guinea savanna of Nigeria. Some of the varieties evaluated
have been reported to be tolerant to drought (Singh 2000) suggesting that they will perform
better in a year of low rainfall.
Table 3. Mean squares of 8 traits of cowpea varieties representing four decades of breeding
in the Nigerian Sudan savannas.
Source of
variation df Days to
owering Days to
maturity
Pod
(No./
plant)
Hundred
seed
weight
Total dry
matter
(No./plant) Grain yield
(kg/ha)
Fodder
yield
(kg/ha)
Harvest
index
Year
Variety
Year x Variety
Error
1
30
29
142
833.15***
117.86***
11.82***
1.69
2394.61***
264.61***
61.58***
3.76
89.18**
45.55***
33.68***
13.40
497.35***
33.79***
4.33**
2.36
6289.40***
794.29***
540.15***
153.11
105299.01ns
608187.04***
65638.61**
36605.94
1033735.21**
1615147.00***
1519372.85***
1519372.85
0.007**
0.012***
0.005
0.001
CV% 2.82 2.65 27.05 9.49 20.98 15.04 20.49 9.63
**Mean squares signicant at < 5% probability level
***Mean squares signicant at < 1% probability level
nsMean squares not signicant at 5% probability level
Mean performance of grain and fodder yields
Average grain yield ranged from 567.6 kg/ha for the variety released in 1976 (TVX3236)
to 1850.8 kg/ha for variety released in 2004 (IT04K-321-2) (Table 4). Grain yield showed
an increase from old to new varieties during the four decades of cowpea breeding at IITA
and this increase was signicantly associated with year of release (Figure 1). Average
grain yields of varieties released in the 2000s are 106% higher than that of varieties
released in the 1970s. The average rate of increase in grain yield was 28.2 kg/ha/yr (Fig.
1) corresponding to 3.6% (Table 5) annual genetic gain in yield. This gain is higher than
the 0.41% reported for maize (Kamara et al. 2004) and the 2.2 and 1.99% reported for
early and late maturing soybeans, respectively (Tefera et al. 2009; 2010) for the savannas
of northern Nigeria. This gain is also higher than the 1% gain reported for soybean in USA
(Wilcox et al. 1979) and the 0.5% reported for soybean in Canada during seven decades
of breeding (Voldeng et al. 1997). The result suggests that signicant progress has been
made in breeding cowpea for higher grain yield in the Nigerian savannas.
Fodder yield ranged from 1363 kg/ha for TVX1836-013J released in 1976 to 3346 kg/ha
for IT98K-476-8 released in 1998. There were signicant differences in fodder yield among
varieties and eras of release. Generally modern varieties showed higher fodder yield in
comparison to older varieties. Differences in fodder yield between the 1990 and 2000 were
however, not signicant. There was gain of 30.9 kg/ha/ yr (Fig. 2) with a genetic gain of
1.96% (Table 5). Top yielding varieties produced higher fodder yields (Table 4) suggesting
that efforts in breeding dual-purpose cowpea varieties at IITA have been a success.

184 Innovative research along the cowpea value chain
Table 4. Physiomorphological and yield characteristics of 31 cowpea varieties developed
over the past 40 years.
Year of
rst
testing
Variety
Days to
owering
Days to
maturity
Pod (No./
plant)
Hundred
seed
weight
Total dry
matter (No./
plant)
Grain yield
(kg/ha)
Fodder
yield
(kg /ha) HI
1974 VITA4 44.75 69.50 17.87 12.10 77.00 733.00 1868.75 0.29
1974 VITA5 46.50 69.00 13.88 14.00 55.50 660.00 1275.00 0.34
1975 TVX456-01F 40.67 63.17 11.41 17.53 39.70 780.98 1449.72 0.34
1975 TVX66-24 48.80 70.80 10.78 13.60 38.86 747.88 1675.00 0.31
1976 TVX1836-013J 42.60 63.60 10.58 13.60 48.54 862.64 1363.33 0.39
1976 TVX3236 46.83 69.33 14.27 11.33 43.20 567.60 1724.17 0.24
Mean 45.03 67.57 13.13 13.69 50.47 725.35 1559.33 0.32
SE (7 D.F.) 1.22 1.35 1.14 0.87 5.88 41.50 94.44 0.02
1981 IT81D-985 58.40 86.80 10.42 17.52 63.40 1199.33 2247.33 0.35
1981 IT81D-994 54.38 86.00 6.64 16.83 59.31 1107.57 2078.85 0.35
1982 IT82D-889 42.25 63.75 14.85 13.60 65.28 1034.07 1714.58 0.37
1984 IT84S-2246-4 47.00 73.00 11.58 14.25 43.61 930.69 1532.50 0.38
1986 IT86D-1010 41.50 66.83 12.15 17.40 51.68 1137.02 1779.44 0.39
1986 IT86D-719 45.43 78.57 13.08 12.51 52.59 1093.03 2177.38 0.34
1986 IT86D-721 46.25 76.88 13.56 15.60 64.92 1200.96 1880.00 0.39
1988 IT88D-867-11 45.86 74.71 12.65 19.03 59.04 1111.3 7 1837.62 0.38
1989 IT89KD-391 51.75 76.13 11.57 18.90 50.94 1195.56 1973.13 0.38
Mean 48.09 75.85 11.83 16.18 56.75 1112.18 1913.43 0.37
SE (41 D.F.) 1.88 2.55 0.78 0.78 2.49 29.46 76.52 0.01
1990 IT90K-277-2 45.63 70.63 16.74 16.60 61.66 1651.62 2202.08 0.43
1997 IT97K-461-4 37.17 64.33 15.12 18.20 53.55 1412.02 2531.39 0.36
1997 IT97K-494-3 44.63 68.00 17.42 15.50 55.73 1587.22 2155.00 0.42
1997 IT97K-499-35 42.13 70.75 12.80 16.10 57.41 1548.96 2444.38 0.39
1998 IT98K-128-3 48.67 74.33 16.37 15.60 81.40 1408.93 3213.97 0.31
1998 IT98K-131-2 46.33 71.83 18.70 14.93 69.79 1442.22 1799.44 0.45
1998 IT98K-476-8 48.63 79.13 13.78 16.45 71.63 1560.11 3346.48 0.32
1998 IT98K-506-1 42.00 66.13 13.19 17.90 49.35 1356.13 2085.21 0.40
1998 IT98K-628 39.00 65.29 14.86 15.54 51.29 1408.65 2005.95 0.41
Mean 43.80 70.05 15.44 16.31 61.31 1486.21 2420.43 0.39
SE (41 D.F.) 1.35 1.58 0.67 0.37 3.60 34.02 178.27 0.02
2000 IT00K-1207 47.86 83.43 14.19 13.83 57.46 1370.21 2048.69 0.40
2000 IT00K-1263 44.60 72.20 13.80 17.68 55.17 1398.90 2146.00 0.40
2000 IT00K-227-4 45.57 73.43 12.40 18.29 62.05 1194.39 1604.29 0.42
2003 IT03K-316-1 46.13 77.00 14.53 18.15 70.83 1707.70 2832.92 0.38
2004 IT04K-217-5 52.17 81.33 8.90 21.60 63.67 1349.49 2015.83 0.41
2004 IT04K-321-2 49.38 81.25 15.80 18.25 83.90 1850.77 2800.21 0.39
2004 IT04K-332-1 43.14 70.57 15.49 15.14 61.00 1582.42 2619.52 0.38
Mean 46.98 77.03 13.59 17.56 64.87 1493.41 2295.35 0.40
SE (28 D.F.) 1.16 1.92 0.89 0.94 3.69 86.71 175.10 0.01
Overall mean 46.00 72.83 13.53 16.05 58.69 1231.98 2078.33 0.37
Overall SE (139 D.F.)
0.79 1.18 0.47 0.42 2.01 59.05 89.86 0.01
This is important because of the increasing demand for cowpea fodder for animal feed
in the dry savannas of West Africa. Tefera et al. (2009) also reported signicant gains in
breeding dual-purpose soybean varieties in the Guinea savannas of Nigeria. Due to the
earlier emphasis on high grain yield, the improved varieties bred in the 1970s to early

185
Proceedings of the Fifth World Cowpea Conference
y = 28.193x + 772.72
R
2
= 0.782
500.00
700.00
900.00
1100.00
1300.00
1500.00
1700.00
1900.00
2100.00
0 5 10 15 20 25 30 35
Number of years since 1974
Grain yield (kg ha
-1
)
Figure 1. Relationship between grain yield of cowpea varieties and year of breeding
(expressed as number of years since 1974).
Table 5. Relative genetic gain, coefcients of determination (R2), slope and regression
coefcients (b) of grain and fodder yield from linear regression of the mean value of each
character for each variety against the year of rst testing for that variety.
Trait Relative genetic gain
(% per year) R2 a b
Grain yield
Fodder yield
3.65
1.96
0.78
0.41
772.72
1575.4
28.193**
30.873**
*signicant at 5% probability level (P < 0.05), **signicant at 1% probability level, ns = not signicant.
y = 30.873x + 1575.4
R2 = 0.405
1000.00
1500.00
2000.00
2500.00
3000.00
3500.00
0 5 10 15 20 25 30 35
Number of years since 1974
Fodder yield (kg ha-1)
Figure 2. Relationship between fodder yield of cowpea varieties and year of breeding
(expressed as number of years since 1974).

186 Innovative research along the cowpea value chain
1980s were erect and were early-medium maturing (Singh and Sharma 1996). With good
management and heavy pod load, these varieties shed most of their leaves. The grain yields
of these varieties are therefore high but fodder yields are low (Singh et al. 2003) probably
because they have a short growth period for biomass accumulation and a greater harvest
index. To increase grain and fodder yields simultaneously, breeding emphasis shifted to
developing intermediate maturing varieties (75–80 days) with semi-erect growth habit and
canopy height of 30 cm or more. Such varieties which have been bred have enough time
for biomass accumulation but short enough maturity to escape terminal drought. These
varieties have yield potential of at least 2 t/ha grain and 2 t/ha fodder (Singh et al. 2003).
Mean performance of agronomic traits and association with grain yield
Signicant differences were found among the cowpea cultivars for number of pods per
plant, 100 seed weight, total dry matter, and harvest index (Table 4). There was no clear
trend in days to owering for all the varieties because early and late owering varieties were
found among both old and recent varieties. Three varieties (IT81D-985, IT81D-994, and
IT89KD-391) released in the 1980s were found to be signicantly late owering. Among
the recent varieties released in the 2000s, IT04K-217-5, IT04K-321-2 were found to be late
owering. Tefera et al. (2010) also reported that there was no consistent trend in days to
owering among old and new soybean varieties in the Nigerian Guinea savannas. There
were signicant differences between old and new varieties for days to maturity. Recent
varieties had higher number of days to maturity than varieties released in the 1970s. For
example, three varieties (IT04K-217-5, IT00K-1207, and IT04K-321-2) released in the
2000s were particularly late in maturity. This may be due to emphasis on dual-purpose
cowpea varieties since late 1980s (Singh et al. 2003). Number of pods per plant did not
show consistent change with era. For example, VITA 4, which was released in 1974 had
number of pods not signicantly different from that of 1T98K-131-2 released in 1998 which
produced the highest number of pods per plant. One hundred seed weight was signicantly
lower in varieties released in the 1970s compared to the other eras. Seed weight of varieties
released in the 1980s, 1990s, and 2000s did not however, differ signicantly suggesting
that no further progress was made in increasing seed weight after the 1970s.
Total dry matter per plant ranged from 38.86 g for TVX66-24 released in 1975 to 83.90
g for IT04K-321-2 released in 2004 (Table 4) indicating a signicant improvement in total
dry matter in the last 40 years. Several studies have reported a signicant genetic gain
in dry matter for grain crops (Kamara et al. 2004; Tollenaar 1991; Naylor et al. 1998).
Moreover, De Bruin and Pedersen (2009) reported that new cultivars of soybean produced
higher yields due to improved total dry matter accumulation, which is in agreement with
the present study. Averaged across varieties, mean harvest index (HI) was signicantly
lower for varieties released in the 1970s than those of the other eras. HI ranged from 0.24
for an old variety (TVX-3236) to 0.45 for a new variety (IT98K-131-2) released in 1998.
There were however, no signicant changes in HI from 1980 to 2004 probably due to the
emphasis on dual-purpose cowpea where breeders select for both grain and fodder yields.
This suggests that increasing HI will be at the expense of fodder yield. The rate of increase
of HI was therefore maintained at a low level. Tefera et al. (2009) reported modest changes
in HI for soybean over a 16-year period of breeding in the Nigerian savannas because

187
Proceedings of the Fifth World Cowpea Conference
the breeders selected for both grain and fodder yields. According to Singh et al. (2003),
cowpea breeders have shifted their focus to selecting for dual-type varieties since the late
1980s because of the increasing importance of cowpea for fodder.
Grain yield showed a positive and signicant correlation with fodder yield, harvest index,
total dry matter per plant, and hundred seed weight (Table 6). The strong association of grain
and fodder yields indicated that improvement has been made for both traits. This strong
association is also expected because cowpea breeders at IITA have been continuously
selecting for dual-purpose cowpea since the late 1980s (Singh et al. 2003). The same
trend has been observed in early and late maturing dual-purpose soybean varieties in the
savannas of Nigeria (Tefera et al. 2009; 2010). The strong association of grain yield with total
dry matter and HI suggests that grain yield improvement was partly due to improvement in
dry matter and HI. Kamara et al. (2004) attributed genetic gain in grain yield of maize in the
savannas of Nigeria to improvement in dry matter accumulation. Similarly, Tollenaar (1989)
attributed genetic gain in Central Ontario maize hybrids to improvement in dry matter
production but HI did not change between eras. The positive correlation of grain yield with
HI may be due to the deliberate emphasis on breeding cowpea for both grain and fodder.
Similarly Tefera et al. (2009) reported a positive correlation of soybean grain yield with HI in
the Nigerian savannas also because of breeding emphasis on both grain and fodder yields.
Table 6. Mean values and correlation coefcients of traits with grain yield of cowpea
varieties representing four decades of breeding in the Nigerian Sudan savannas.
1974–2004
Trait Mean Correlation coefcient (r)
Days to owering
Days to maturity
Grain yield
Fodder yield
Harvest index
Total dry matter per plant
Hundred seed weight
Number of pods per plant
46.00
72.83
1231.98
2078.33
0.37
58.69
16.05
13.53
–0.011
0.273
–0.742**
0.587**
0.529**
0.498**
0.348
*Correlation coefcients signicant at 1% probability level.
Conclusion
A study of 31 medium to late maturing varieties of cowpea released in 1970–2004 showed
that grain yield ranged from 568 kg/ha for an old variety TVX3236 to 1851 kg/ha for a
recently released variety IT04K-321-2. The average rate of increase has been 28 kg/ha/
year, which matches to a genetic gain of 3.6%. Fodder yield ranged from 1275 kg/ha for
a variety released in 1974 (VITA5) to 3346 kg//ha for IT98K-476-8 released in 1998. The
genetic gain in fodder yield was 1.96% which indicated that selection for dual-purpose
cowpea varieties with increased fodder as well as grain yields has been successful. Total
dry matter, fodder yield, harvest index, and 100 seed weight were signicantly correlated
with grain yield. The strong relationship between grain and fodder yields showed the
success made in selecting for dual-purpose cowpea varieties.

188 Innovative research along the cowpea value chain
References
Ajeigbe, H.A., B.B. Singh, J.O. Adeosun, and I.E. Ezeaku. 2010a. Participatory on-farm evaluation
of improved legume–cereals cropping systems for crop–livestock farmers: Maize-double cowpea
in Northern Guinea Savanna Zone of Nigeria. African Journal of Agricultural Research 5(16):
2080–2088. Available online at http://www.academicjournals.org/AJAR.
Ajeigbe, H.A., B.B. Singh, I.E. Ezeaku, and J.O. Adeosun. 2010b. On-farm evaluation of improved
cowpea–cereals cropping systems for crop–livestock farmers: Cereals-cowpea systems in Sudan
Savanna zone of Nigeria. African Journal of Agricultural Research 5(17): 2297–2304. Available
online at http://www.academicjournals.org/AJAR.
Carsky, R.J., B.B. Singh, and B. Oyewole. 2001. Contribution of early season cowpea to late season
maize in the savanna zone of West Africa. Biological Agriculture and Horticulture 18: 303–316.
Cassman, K.G. 1999. Ecological intensication of cereal production systems: yield potential, soil
quality and precision agriculture. Proceedings National Academy of Sciences USA 96: 59525959.
De Bruin, J. L. and P. Pedersen. 2009. Growth, yield, and yield component changes among old and
new soybean cultivars. Agronomy Journal 101: 123–130.
Francis, C.A. 1991. Contribution of plant breeding to future cropping systems. Pages 83–93 in Plant
Breeding and Sustainable Agriculture: Considerations for Objectives and Methods, edited by D.A.
Sleper, T.C. Baker, and P.J. Bramel-Cox. Crop Science Society of America Special Publication
No. 18, CSSA and ASA, Wisconsin, USA.
Iniazumi, H., B.B. Singh, P.C. Sanginga, V.M Manyong, A.A Adeshina, and S.A. Tarawali. 1999.
Adoption and impact of dry-season dual-purpose cowpea in the semi-arid zone of Nigeria. Page
14 in IMPACT. International Institute of Tropical Agriculture, Ibadan, Nigeria.,
Kamara A.Y., A. Menkir, M.A.B Fakorede, S.O. Ajala, B. Badu-Apraku, and I. Kureh. 2004. Agronomic
performance of maize varieties representing three decades of breeding in the Guinea Savannas
of West and Central Africa. Journal of Agricultural Science 142: 567–575.
Naylor, R.E.L., D.T. Stokes, and S. Matthews. 1998. Biomass, shoot uniformity and yield of winter
barley. Journal of Agricultural Science, Cambridge 131: 13–21.
Ortiz, R. 1998. Cowpeas from Nigeria: A silent food revolution. Outlook on Agriculture 27(2): 125–128.
Sanginga, N., G. Thottappilly, and K. E. Dashiell. 2000. Effectiveness of rhizobia nodulating recent
promiscuous soybean selections in the moist savanna of Nigeria. Soil Biology and Biochemistry
32: 127–133.
SAS Institute. 2003. The SAS system for Windows. Release 9.1. SAS Inst. Cary NC
Singh, B.B. 2000. Breeding cowpea varieties for wide adaptation by minimizing cultvar × environment
interactions. Pages 173–181 in Cultivar × Environment Interactions Analysis of IITA Mandate
Crops in Sub-Saharan Africa, edited by I.J. Ekanayake and R. Ortiz. International Institute of
Tropical Agriculture, Ibadan, Nigeria.
Singh, B.B., H.A. Ajeigbe, S.A. Tarawali, S. Fernandez-Rivera, and A. Musa. 2003. Improving the
production and utilization of cowpea as food and fodder. Field Crops Research 84: 167–177.
Singh, B. B., O.L. Chamliss, and B. Sharma. 1997. Recent advances in cowpea breeding. Pages
30–49 in Advances in cowpea research, edited by B.B. Singh, D.R. Mohan Raj, K.E. Dashiell, and
L.E.N. Jackai. Copublication IITA and JIRCAS. IITA, Ibadan, Nigeria.
Singh, B.B. and H. Sharma. 1996. Restructuring cowpea for higher yield. Indian Journal of Genetics
56: 389–405.
Specht, J. E., D.J. Hume, and S.V. Kumudini. 1999. Soybean yield potential: A genetic and
physiological perspective. Crop Science 39: 1560–1570.
Tefera, H., A.Y. Kamara, B.Asafo-Adjei, and K.E. Dashiell. 2009. Improvements in grain and fodder
yields of early maturing promiscuous soybean varieties in the Guinea savannas of Nigeria. Crop
Science 49: 2037–2042.
Tefera, H., A.Y. Kamara, B. Asafo-Adjei, and K.E. Dashiell. 2010. Breeding progress for grain yield
and associated traits in medium and late maturing promiscuous soybean. Euphytica 175: 251–
260.
Tollenaar, M. 1989. Genetic improvement in grain yield of commercial maize hybrids grown in Ontario
from 1959 to 1988. Crop Science 29: 13651371.
Tollenaar, M. 1991. Physiological basis of genetic improvement of maize hybrids in Ontario from
1959 to 1988. Crop Science 31: 119–124.
Voldeng, H.D., E.R. Cober, D.J Hume, C. Gillard, and M.J. Morrison. 1997. Fifty-eight years of genetic
improvement of short season soybean cultivars in Canada. Crop Science 37: 428–431.
Wilcox, J.R., W.T. Schapaugh, Jr., R.L. Bernard, R.L. Cooper, W.R. Fehr and M.H. Niehaus. 1979.
Genetic improvement of soybean in the Midwest. Crop Science 19: 803–805.