Are genotypes’ single-year YREMs and BLUPs good predictors of their performance in future years? An empirical analysis in dolichos bean [Lablab purpureus (L.) Sweet var. Lignosus

Abstract

Crop breeders conduct trials to assess performance of elite breeding lines/germplasm accessions across locations and/or years to identify promising ones for use as cultivars and/or for use in breeding programmes. It is hypothesized that in a given location, genotypes’ multi-year performance are more predictive of their performance in the next/future years. However, only a few researchers have examined this hypothesis. To test this hypothesis, we predicted genotypes’ present-year’s performance based on their previous 1-, 2-, 3- and 4-years’ performance using fresh pod yield data from 53 dolichos bean genotypes evaluated across 5 years. We used two statistics, namely yield relative to environment (year) maximum (YREM) and best linear unbiased predictors (BLUP) as measures of genotypes’ performance for fresh pod yield in dolichos bean. Correlation coefficients of genotypes’ present-year’s performance with previous 1-, 2-, 3- and 4-years’ performance were used as measures of predictive power. The results of our study provided adequate evidence to show that that single-year dolichos bean genotypes’ YREM and BLUP have sufficient power to predict their future 1-year/multiple-years’ performance and to identify superior genotypes. These findings adequately support the common practice by breeders/cultivar sponsors of withdrawing their cultivars from the tests based on a single-year trial.