Selection of superior crosses of groundnut (Arachis hypogaea L.) in early g
enerations would increase the probability of identifying superior lines. Th
e objective of this study was to determine the potential of selecting for p
hysiological traits identified in a yield model [crop growth rate (C), repr
oductive duration (D-R) and partitioning (p)] in segregating populations. F
orty populations and nine parental lines were evaluated in replicated trial
s in 1992 (F-2, 1993 (F-3) and 1994 (F-4) at three locations in Niger. Phys
iological traits were estimated from final yield and biomass as well as dat
a on flowering and maturity. Regressions from two different parent-offsprin
g generations (F-2: F-3 and F-3: F-4) were calculated. The results were com
pared to determine if early generation performance accurately predicts the
performance of cross bulks in later generations. Differences were observed
among populations and parents for all traits. Effects of locations were sig
nificant for C, p and D-R in F-2 and F-3 but nonsignificant for yield and C
in F-4. Regression coefficients from F-3: F-2 were 0.10 +/- 0.08 for C, 0.
45 +/- 0.17 for p, 0.10 +/- 0.03 for D-R and 0.16 +/- 0.03 for pod yield. B
ased on F-3: F-4 regression, the coefficients were 0.12 +/- 0.23 for C, 0.4
6 +/- 0.17 for p and 0.57 +/- 0.17 for yield. Parent-offspring correlations
were in most cases similar to the regression values. It was concluded that
selection for yield and model components in early generation bulks may inn
effective.