Variation and association analysis in grass pea (<i>Lathyrus sativus</i> L.)

Abstract

The study was conducted on 50 landrace populations of grass pea at Adet Research Center, Ethiopia, using randomised complete Block design in three replications to determine morphological variability and character associations. Evaluations were made on plant and yield characters from eight plants tagged randomly from each plot. The data were subjected to Analysis of Variance (anova), heritability, correlation and cluster analyses. Higher mean and cv values were recorded for most of the agronomic traits for populations collected from Gondar region and the higher altitude group (>2500 m) indicating the presence of high genetic diversity in this region and altitude group. Phenotypic coefficient of variation (pcv) was slightly higher than genotypic coefficient of variation (gcv) for all the characters studied signifying that genotypic factors exerted reasonable effect in estimating the variation. The wide difference between pcv (22.4%) and gcv (13.0%) for seed yield per plant indicated the complexity of this trait and the important role of other factors such as environment in influencing yield potential in addition to the genetic factors. The strong positive association of pods per plant, 100 seed weight and primary branches per plant with seed yield indicated the possibility of selecting lines for yield improvement based on these characters at the very early stage of the grass pea breeding program. Key words/phrases: Correlation grass pea, heritability, Lathyrus sativus, variation SINET: Ethiopian Journal of Science Vol.25(2) 2002: 191-204