Genetic variation, heritability and path-analysis in landraces of bread wheat from southwestern Iran

Abstract

Nine pure lines of bread wheat collected from landraces of southwestern Iran and one local Iranian cultivar from the same area were used to estimate genetic variation and heritability for 12 developmental and yield component characters. Path-analysis was used to partition the genetic correlations between some of the characters into direct and indirect effects. Mean values for these lines were also compared with those of five improved cultivars from Iran and California. The goal of the experiment is to breed improved landraces and/or modern cultivars for areas of low input agriculture in southern California, southwestern Iran, and regions with similar environmental and agronomic conditions. Moderate genetic variation was displayed by the number of effective heads per plant, number of grains per head, and grain weight in the landrace genotypes. The heritability estimates ranged from 43 to 97%. Expected genetic advance with selection of the highest 5%, expressed as percent of the mean, was around 20% for number of heads per plant, number of grains per head, and 1000-grain weight. Days to booting, to heading, and to anthesis were positively correlated but none of them were significantly associated with days to maturity. Plant height had a negative genetic correlation with number of grains per head, 1000-grain weight, grain yield, and harvest index. The genetic correlation between number of heads per plant and number of grains per head, 1000-grain weight, and harvest index was also significantly negative. Harvest index had a negative genetic correlation with days to booting, to maturity, plant height, number of heads per plant, and straw yield and a positive correlation with number of grains per head, 1000-grain weight, and grain yield. Days to maturity, plant height, number of heads per plant, number of grains per head, 1000-grain weight, and harvest index each had a positive direct effect on grain yield. The first two characters exhibited the highest and lowest direct effects, respectively. The positive direct effects of days to maturity, plant height, and number of heads per plant, however, were partially or completely counter-balanced by their strong negative indirect effects through number of grains per head, 1000-grain weight, and harvest index. Pathanalysis indicated that late and tall landrace genotypes tend to produce more heads per plant, but with fewer number of grains per head, smaller grains, and lower harvest index. Comparisons between the local lines and the improved cultivars revealed that, in general, the former were much taller and produced a larger number of non-effective tillers. Mean number of grains per head, grain weight, harvest index, and grain yield of local lines were smaller than those of improved cultivars. Our observations indicate that the landraces could be improved by selecting for shorter genotypes with smaller numbers of tillers per plant, but with larger numbers of grains per head and heavier grains.