Inheritance of reproductive characters in selected lines of pea

Abstract

The feasibility of improving pea (Pisum sativum L.) yields by decreasing ovule, flower or pod abortion was investigated. A full diallel experiment and an evaluation across varied environments were carried out. The full diallel used crosses between Green shaft, Ichiban, Huka, Maro and Combi to provide information on heritability of and correlations and path coefficients between yield, yield components and reproductive characters. To conduct this investigation a set of pea genotypes was chosen on the basis of diverse ovule abortion and included garden, white and marrowfat peas. Phenotypic stability and genotype X environment interactions of these traits were estimated by growing eight genotypes (Ichiban, Huka, Maro, Birte, Progreta, Combi, Green golt and Carpo) in nine environments (three planting densities X three seasons). Broad sense heritability, correlation and path-coefficients were also determined. Conclusions were also drawn with respect to the best sources of genetic advance among the tested genotypes for the characters listed above. In the case of Combi and Green shaft general combining ability (GCA) effects were outstanding for high ovule number and for low reproductive abortion. Although Combi had a positive GCA effect for pod abortion, GCA effects for ovules aborted per pod and ovule abortion percentage were highly negative (causing low abortion). Huka and Maro had intermediate GCA effects. Combi and Huka showed superior phenotypic stability for ovule number per pod (which is, however, only slightly influenced by environment). Ovule abortion percentage was inversely related to flower and pod abortion percentages (which were of minor importance to selection). Hence Combi, Huka and Green goIt would be good parents for breeding for low ovule abortion. Breeding for high number of ovule initials per pod would initiate the maximum reproductive potential of a pea pod. Selection for low ovule abortion could greatly increase pea yields by ensuring more seed reached maturity. Flower and pod abortion did not cause major yield loss and appear of little use as selection parameters for increased yield.