|
African Crop Science Journal
African Crop Science Society
ISSN: 1021-9730 EISSN: 2072-6589
Vol. 7, Num. 4, 1999, pp. 327-331
|
African Crop Science Journal
African Crop Science Journal, Vol. 7. No. 4, pp. 327-331, 1999
A PATH COEFFICIENT ANALYSIS OF SOME YIELD
COMPONENT INTERACTIONS IN COWPEA
C.K. Nakawuka and E. Adipala
Department of Crop Science, Makerere University, P. O. Box 7062, Kampala, Uganda
Code Number: CS99022
ABSTRACT
Correlation and path-coefficient analyses were done to find associations among
cowpea characters and assess the indirect and direct contribution of each character
to grain yield. Branch number, pod number and seeds/pod were the major direct
contributors to grain yield. These characters should therefore receive the highest
priority in developing high yielding cowpea varieties.
Key Words: Correlation, indirect and direct effects, yield determinants,
Vigna unguiculata
RÉSUMÉ
Lanalyse du coefficient de chemin et de correlation a été
faite pour étudier des associations entre caractères et évaluer
la contribution directe et indirecte pour chaque caractère au rendement
en grains. Le nombre de branches, le nombre de gousses et les graines par gousse
ont été les majeurs caractères contribuant au rendement
en grains. Ces caractères devraient alors recevoir une priorité
dans le developpment des variétés du niébé.
Mots Clés: Correlation, effets direct et indirect, déterminants
du rendement, Vigna unguiculata
Introduction
Cowpea (Vigna unguiculata L. Walp) is considered the second most important
grain legume in Africa (Purseglove, 1974). In Uganda, the crop is grown predominantly
in the northern and north-eastern parts of the country, i.e., in Tororo, Pallisa,
Kumi, Soroti, Lira, Nebbi and Arua (Sabiti et al., 1994). However, the
few cowpea varieties grown in the country are landraces with low yield potential
(Adipala et al., 1997). Before yield improvement can be realised, selection
criteria must be established. One way to do this is to carry out correlation
analysis which quantifies the relationship between any given pair of traits
(Obisesan, 1985). However, yield is a complex trait, and it is difficult from
correlations alone to determine which traits contribute more to grain yield.
Therefore, it is important to carry out other analyses so as to establish the
direct and indirect contribution of each trait on grain yield. Path coefficient
analysis, developed by Wright (1921, 1923), determines the significance of correlations
between yield components and assigns relative importance to yield relations
(McGiffens et al., 1994). The coefficients generated by path analysis
measure the direct and the indirect influence of a variable upon another (Dewery
and Lu,1959). This study determined the phenotypic and genotypic correlations
and evaluated the direct and indirect effects of yield components of cowpea.
Materials and Methods
The study was conducted at Makerere University Agricultural Research Institute,
Kabanyolo, and involved 25 cowpea lines. Plantings were done on 14 April and
16 October 1994 and 21 March 1995. All experiments were laid out in randomised
complete blocks replicated three times. Rows, 4.5m long were spaced 75cm apart
with 30cm between plants. There was no fertiliser application but Dimethoate
(200 g a.i ha-1) and Chlorpyrifos (Dursban), at 600 g a.i ha-1 were applied
to control insect pests. Data were collected from five randomly selected plants
in each row on the following parameters; 1) branches per plant, 2) pods per
plant, 3) pod length, 4) seeds per pod, 5) 100 seed weight, 6) plant height,
and 7) grain yield per plant. For characters 3 and 4, 20 pods were taken from
each sample.
Analysis of variance (ANOVA) found no significant differences between the three
seasons data, and as such data were pooled for other statistical analyses.
Variance and covariance analyses for each variable and each pair of variables,
respectively, were carried out. Path-analysis was also carried out to determine
the relationships among the yield components. Phenotypic and genotypic correlation
were calculated from the variance and covariance components using the formula:
r(x,y) = |
Cov(x,y)
_______
|
Ösx2
sy2 |
where r(x,y) is either genotypic or phenotypic correlation between
variables x and y;
Cov(x,y) is the genotypic or phenotypic covariance between the two
variables;
sx2 is the genotypic or the phenotypic
variance of the variable x; and
sy2 is the genotypic or the phenotypic
variance of the variable y.
Correlations and path-analyses were done to determine the interrelationships
among yield components and their direct and indirect contribution to the grain
yield. Basing on the genotypic correlations, path coefficients were calculated
by solving a series of simultaneous equations as suggested by Dewery and Lu
(1959).
The residual effect, which determines how best the causal factors account for
the variability of the dependent factor, yield in this case, was obtained using
the formula;
1 = P2R7 + å Piyriy
where P2R7 is the residual effect
Piyriy = the product of direct effect of any variable
and its correlation coefficient with yield.
Results
Correlations. The genotypic and phenotypic correlation coefficients
for yield and its various components are presented in Table 1. Except for 100-seed
weight, all other characters showed significant genotypic correlations with
grain yield. Number of pods per plant had the highest positive genotypic correlation
with grain yield (rg=0.894, P=0.001), and number of seeds per pod was highly
correlated with pod length (rg=0.678, P=0.01). However, plant height and 100-seed
weight showed negative correlations with most traits including grain yield.
Seeds per pod and plant height had the highest negative genotypic correlation.
The phenotypic correlations were significant and positive for grain yield per
plant, branches per plant, pods per plant, and seeds per pod. Correlations between
grain yield and pod length, and between plant height and 100-seed weight were
not significant (P>0.05). In all significant cases, genotypic correlations
were higher than the phenotypic correlations.
TABLE 1. Genotypic (G) and Phenotypic (P) correlation coeffiecients for relations
among yield componentsof cowpea
Character |
Branches/plant |
Pods/plant |
Pod length |
Seeds/pod |
100-seed weight
|
Grain yield plant
|
G
P |
0.860**
0.677** |
0.894**
0.657** |
0.219
0.173 |
0.833**
0.293 ** |
-0.046
0.119 |
Branches/plnat |
G
P |
|
0.775**
0.589 ** |
0.259*
0.170 |
0.042
0.252* |
0.205
0.100
|
Pod length |
G
P |
|
|
0.002
0.068 |
0.277*
0.127 |
-0.718**
-0.141 |
Seeds/pod |
G
P |
|
|
|
0.678**
0.420** |
-0.342**
0.180 |
100-seed weight |
G
P |
|
|
|
|
-0.828**
0.069 |
*,** significant at P=0.05 and 0.01, repectively
Path-coefficient analysis. Because of the presence of significant associations
of grain yield with other characters, the genotypic correlations of six causal
variables were partitioned into direct and indirect effects (Dewery and Lu,
1959). The path-coefficient analysis based on grain yield as a dependent variate
revealed that the contribution of all the six characters towards variation in
grain yield was 75.3% (Table 2). All the characters exhibited direct effect
on grain yield (P>0.100) but the estimates were higher and positive for branches
per plant (0.592), pods per plant (0.270), seeds per pod (0.257) and pod length
(0.213); plant height and 100-seed weight had the highest negative direct effect
on the yield of cowpea.
TABLE 2.The direct and indirect effects contributions of various characters
to grain yield in cowpea
|
Branches/plant |
Pods/plant |
Pod length |
Seeds/pod |
100-seed weight |
Plant height |
Correlation with yield |
Branches/plant |
0.592a |
0.210 |
0.055 |
-0.011 |
-0.067 |
0.087 |
0.866** |
Pods/plant |
0.459 |
0.270 |
0.000 |
0.071 |
0.235 |
0.001 |
0.894** |
Pod length |
0.153 |
0.001 |
0.213 |
-0.174 |
0.112 |
-0.085 |
0.219 |
Seeds/pod |
0.025 |
0.075 |
0.145 |
0.270 |
-0.257 |
0.575 |
0.833** |
100-seed weight |
0.121 |
-0.194 |
-0.073 |
0.213 |
-0.327 |
0.214 |
-0.046 |
Plant height |
-0.089 |
-0.001 |
0.032 |
0.255 |
0.121 |
-0.578 |
-0.260* |
*,** significant at P=0.05 and 0.01, repectively
a Direct effects on grain yeild
Discussion
Genotypic correlation coefficients provide a measure of the genotypic associations
between characters and give an indication of the characters that may be useful
as indicators of more important ones under consideration (Johnson et al.,
1955). Generally, genotypic correlations were higher than the corresponding
phenotypic ones. This implied that the traits under consideration were genetically
controlled. Similar observations were obtained on soybean by Weber and Moorthy
(1952) and Johnson et al. (1955). The closeness of phenotypic and genotypic
correlation coefficients in some instances suggest that environment had some
effect on the yield correlated characters.
The significant positive genotypic correlations between grain yield and its
components suggest that these characters contributed positively towards yield
and should be considered when selecting for high grain yield in cowpea. Similar
results were reported by Ombakho and Tyagi (1987) .
The highly positive correlation between number of seeds per pod and pod length
indicates that with longer pods more space is provided for seeds. However, the
highly negative significant genotypic correlation between seeds per pod and
100-seed weight indicated that with many seeds per pod the seed size was reduced.
The decrease in seed size could lead to unacceptability of some cultivars, especially
where large seeded cultivars are preferred. For example, in parts of eastern
Uganda (Kumi and Pallisa) the cultivar Ebelat is preferred because of
its large seeds (Sabiti et al., 1994).
Although pod number had the highest correlation coefficient with yield, its
direct effect was lower than that of branch number. Therefore, its high correlation
(r=0.894) could be attributed to the substantial positive indirect effect through
branch number. One hundred-seed weight had a negligible correlation (-0.046)
with yield, and its direct effect was negative (-0.327) but significant (P=0.01).
This suggests that 100-seed weight was indirectly influenced by the positive
effects of seeds per pod, plant height and branch number. However, this effect
was offset by the negative direct and indirect effects through pod number, ultimately
resulting in the small negative correlation.
The residual effect of 0.247 explains that the variables in consideration accounted
for 75.3% of the variability in cowpea yield. The highest contributers to seed
yield appear to be branch number, pod number and seeds per pod (Jackai, 1995).
Although earlier research showed that seed size is a primary determinant of
yield in cowpea (Imrie and Bray, 1983; Obisesan, 1985), this was not the case
in the present study. This discrepancy may have been due to the different genotypes
used. Thus for cowpea yield improvement, branch and pod numbers, and seeds/pod
should be part of the selection criteria.
Acknowledgement
The research was jointly funded by USAID and the Rockefeller Foundation Forum
Grant RF 93040 #13.
References
Adipala, E., Obuo, J.E. and Osiru, D.S.O. 1997. A survey of cowpea cropping
systems in some districts of Uganda. African Crop Science Conference Proceedings
3:665-672.
Dewery, D.R. and Lu, K.H. 1959. A correlation and path-coefficient analysis
of components of crested wheatgrass seed production. Agronomy Journal
51:515-518.
Imrie, B.C. and Bray, R.A. 1983. Estimates of combining ability and variance
components of grain yield and associated characters in cowpea. Proceeding
of Australian Plant Breeding Conference, pp. 202-204.
Jackai, L.E.N. 1995. The legume pod borer Maruca vitrata, and its
principal host plant, Vigna unguiculata (L.) Walp. Use of selective
insecticide sprays as an aid in the identification of useful levels of resistance.
Crop Protection 14:299-306.
Johnson, H.W., Robinson, H.F. and Comstock, R.E. 1955. Estimates of genetic
and environmental variability in soybean. Agronomy Journal 47:314-38.
McGiffen, M.E., Pantone, D.J. and Masiunas, J.B. 1994. Path analysis
of tomato yield components in relation to competition with black and eastern
black nightshade. Journal of American Society for Horticultural Science
119:6-11.
Obisesan, I.O. 1985. Associations among grain yield components in cowpea
(Vigna unguiculata L. Walp). Genetical Agriculture 39:377-386.
Ombakho, G.A. and Tyagi, A.P. 1987. Correlation and path-coefficient analysis
for yield and its componets in cowpea (Vigna unguiculata (L.) Walp.).
East African Agriculture and Forestry Journal 53:23-27.
Pursegrove, J.W. 1974. Tropical Crops. Dicotyledons.Vol. 1 and 2.
English Language Book Society and Longman, England. pp. 321-328.
Sabiti, A.G., Nsubuga, E.N.B., Adipala, E. and Ngambeki, D.S. 1994. Socioeconomic
aspects of cowpea production in Uganda: A rapid rural appraisal. Uganda
Journal of Agricultural Sciences 2:29-36.
Singh, K.B., Bejiga, G. and Malhota, R.S. 1990. Association of some
characters with seed yield in chickpea collections. Euphytica 49: 83-88.
Weber, C.R. and Moorthy, B.R. 1952. Heritable and non-heritable relationships
and variability of oil content and agronomic characters in the F2 generations
of soybean crosses. Agronomy Journal 44:202-209.
Wright, S. 1921. Correlation and causation. Journal of Agriculture Research
20:557-585.
Wright, S. 1923. Theory of path co-efficients. Genetics 8:230-255.
©1999, African Crop Science Society
|