GENETIC VARIABILITY AND CORRELATIONS AMONG METRIC TRAITS OF SOME HIGH PROTEIN PEA LINES

PNL Volume 12 1980

RESEARCH REPORTS .35

GENETIC VARIABILITY AND CORRELATIONS AMONG METRIC TRAITS OF SOME HIGli

PROTEIN PEA LINES

Kaul, M.L.H. Institute of Genetics, University of Bonn, West Germany

Most of the 19 high pea protein lines tested (PNL 10: 27-28) proved

to be low yielders. Grain yield of 16 such genotypes was also low (Fig. 1)

but most of them possessed considerable genetic variation for various yield

components, the maximum being in grain yield and the minimum in seed protein

content (Table 1). Since selection was directed towards higher seed protein

Fig. 1. Seed protein content and grain yield of 16(1-16) high protein pea genotyj

36 RESEARCH REPORTS

PNL Volume 12 1980

content, its variation was narrowed. The trait "seed number per plant"

possessed the maximum phenotypic variability, genotypic coefficient of

variation, heritability, and expected genetic advance (Table 1). Therefore,

its selection value and breeding significance is high and the enhanced

variability within the genotypes observed in this trait is due to a high

degree of additive genetic effect. As shoot height possesses a high herita-

bility and genetic advance, its control by major genes is inferred. Total

grain yield (Y) is related to the various metric traits as follows:

Y = 5.67 - 0.01 X (X = shoot height); Y = 5.39 - 0.034 X (X = nodes

per plant); Y = 5.73 - 0.262 X (X = internode length); Y = 0.95 + 0.290 X

(X = pods per plant); Y = 2.77 + 0.702 X (Y = seeds per pod); Y = 1.22 +

1.07 X (Y = seeds per plant).

These regression equations imply a strong dependence of grain yield

upon seed number per plant. This relationship is substantiated by the

phenotypic, genotypic, partial and multiple correlation values (Table 2).

Approximately 65-67% of the variance (R²) in grain yield was accounted for

by its association with seed and pod number per plant, the remaining being

due to action and interaction of other variables including the environment.

Significant positive phenotypic and genotypic correlations between

grain yield and seed number per plant, and the insignificant genotypic

correlations between grain yield and shoot height indicate that selection

of plants with increased seed number would produce plants higher in grain

yield without influencing their stature significantly. Since sufficient

genotypic variability exists in the "seed number per plant", it appears

that the trait can be improved through recurrent selection without substan-

tially affecting other traits.