Abstract:
Multiple models have been
used to examine stability in many crops, but
little of such exits for kenaf. Relationship of
stability estimates of various models reveals
the importance of one or more estimates for
reliable predictions of cultivar behaviour
and stability. This study evaluated 33 kenaf
genotypes across six location for core and
bast fibre yield stability using four models.
Kenaf were grown in a four row plot, 5 m
each, at 0.2 m within row and 0.5 m
between rows in the trial laid out in
randomized complete block design with
three replications. Twenty plants were
randomly harvested per plot at 12 weeks
after planting and processed to fibres. Dry
core fibre weight (CFW) and bast fibre
weight (BFW) were taken. Data collected
were pooled across locations and subjected
to analysis of variance. Genotypes stability
were estimated using Finlay-Wilkinson,
Wricke’s ecovalence (Wi), Kang’s rank sum
and superiority index models. Correlations
among the weights and stability models
were performed. Significant differences
existed in the genotypes (G) (p< 0.01),
environments (E) and G×E for CFW and
BFW. Partitioning the G×E showed that
genotypes linear response and deviation
from the mean were significant for CFW
and BFW. Significant and positive
correlation existed between Finlay-
Wilkinson and Kang’s rank sum (0.570***),
Wi (0.615***) and superiority index
(0.582***) for CFW. Significant correlations
also existed between the efficacy of Kang’s
rank sum and Wi (0.569***), and with
superiority index (0.779***). Kang’s rank
sum correlated with Finlay-Wilkinson
(0.345**), while Wi model had correlation
with Finlay-Wilkinson (0.538**) and Kang’s
rank sum (0.318**) for the BFW. All the
models correlated with one another. Any of
the models is sufficient to select stable
genotypes in kenaf fibre yield breeding
programmes.
