GENETIC DIVERSITY AND ASSOCIATION OF TRAITS IN ETHIOPIAN GRASSPEA (Lathyrus sativus L.) ACCESSIONS

Abstract

Grasspea is one of the principal grain legumes in Ethiopia adapted to adverse climatic and edaphic conditions. It is high yielding and the only available food source during drought times. However, the grain contains a toxin, ODAP (Oxalyl diamino propionic acid) which causes paralysis of legs when consumed as a stable food for more than three months. In addition, the crop has got less research attention compared to other legumes. As a result, very limited information is available on morphological, biochemical and molecular diversity of Ethiopian grasspea accessions. Therefore, the general objective of the present study was to evaluate genetic diversity and association of traits in Ethiopian grasspea accessions. The field experiment was conducted in 2013/14 cropping season at DZARC; and the laboratory studies were conducted at ILRI Ethiopia and BecA-ILRI in Kenya in 2014/2015 and 2015/2016, respectively. Four hundred and twenty six accessions were evaluated in a randomized complete blocks design with two replications using 15 qualitative and 18 quantitative traits. The results showed that most of the diversity on qualitative traits was found within regions. Accessions of adjoining regions were grouped together. The traits showed different level of polymorphism across regions. The analysis of variance showed highly significant difference for most of the quantitative traits. High heritability coupled with high genetic advance was observed for plant height, number of pods and effective pods per plant, days to 50% flowering and biomass yield per hectare. The first three and two principal components accounted for 76.34 and 88.85% of the total variation among accessions and regions, respectively. More genetic similarity was found between adjoining regions. The accessions were grouped into nine clusters. From biochemical analysis, ODAP and protein contents showed very highly significant (P < 0.0001) differences among 426 accessions. Nine accessions with lowest ODAP content (< 0.2) were identified as the most promising accessions. The first two principal components explained 83.65 and 92.7% of the total variation among accessions and regions, respectively. Most of the adjoining regions were grouped together by cluster analysis. The correlation analysis showed that grain yield was positively and highly significantly xvi correlated with most of the traits both at genotypic and phenotypic levels. On the contrary, ODAP content showed negative and highly significant correlation with harvest index and seeds weight at genotypic and phenotypic levels. From the path analysis, all the traits, except days to flowering, ODAP and dry matter contents showed positive direct effects on grain yield at genotypic and phenotypic levels. For ODAP content, grain yield, plant height, pods per plant and seed weight had direct negative effects at genotypic and phenotypic levels. The molecular analysis revealed that the 12 simple sequence reapts (SSR) markers were polymorphic with average polymorphic information content (PIC) of 0.54. The populations were highly polymorphic ranging from 83.3 to 100% with an average of 97.4%. The other measures of diversity like Shannon’s information index, observed heterozygosity, and the average Nei’s genetic diversity also showed the presence of high genetic diversity though most of the variation was found within populations. Though cluster and principal coordinate analyses did not show clear pattern of relationships between diversity and regions, populations of the adjoining regions showed more genetic similarity than distantly located regions. The analysis of molecular variance showed the presence high differentiation among accessions though most of the variation (63%) was found within populations. The structure analysis identified two population groups. However, geographic populations failed to show distinct population structure, indicating the presence of high admixtures within the populations. Generally, the information generated on magnitude and distribution of grasspea diversity can be useful for planning future collection, conservation, utilization and research programs.