AUTOFERTILITY IN A PURE LINE FABA BEAN (VICIA-FABA L) GERMPLASM COLLECTION

A pure line faba bean germplasm collection (BPL) was derived by a proc ess of cyclic single plant progenies starting in 1979 through 1985. In 1985/86, 840 of these pure line accessions from 677 original populati ons from 41 countries were evaluated for autofertility by comparing th e mean of five plants where all flowers were tripped vs. five plants w ith no tripped flowers for number of seeds produced (seed index). In a replicated trial grown at the same time, seed yield, 100 seed weight, number of pods per plant, seeds per pod, and number of flowers per in florescence were also evaluated. A high number of BPL accessions faile d to set seeds in the screen houses without tripping, the mean seed in dex measure of autofertility was 0.51 with a C.V. of 74.2%. There were marked differences among countries of origin for autofertility, with Egypt having the highest (SI = 0.94) and the USSR the lowest (SI = 0.2 6) autofertility, however, most countries expressed a large range, sug gesting that breeders should first look in their own material for this trait rather than searching in exotic germplasm. In general, there wa s a higher level of autofertility in Middle Eastern countries that may be due to natural selection for non-dependency on pollinators for hig h yield. Contrary to the findings of others, large seeded accessions h ad a high autofertility (SI = 0.704) compared to small seeded accessio ns (SI = 0.384), however, both had similar ranges. In general, there w as no significant correlations among autofertility and yield and its c omponents. The variability found in this large pure line collection fo r autofertility is of potential use for developing cultivars not depen dent on insect populations for high yield and for developing self-poll inated faba bean.