Environment of selection for grain yield in low fertilizer input barley

Abstract

Separate breeding for high and low fertilizer inputs may be required when genotype-by-fertilizer interaction involving rank-change occurs entailing additional resource commitments. Twenty-six barley genotypes were tested in 1998 and 1999 in a factorial combination of two levels of N (0 or 11.5 g m −2) and P (0 or 2 g m −2) on a Eutric Nitosol at Holetta (9°03′N, 38°31′E), Ethiopia. The objective was to assess the efficiency of high fertilizer input selection environment for low input target environments. Grain yield averaged 303 g m −2 under high N and P (+N+P). Grain yield reduction due to lack of N and P (−N−P) was 79%. Fertility stress was negatively correlated with genotypic, genotype-by-year (G×Y) interaction, and error variances but positively correlated with genotypic and phenotypic coefficients of variation. The G×Y interaction was significant under high N and/or high P but not under low N and P. Broad-sense heritabilities for grain yield in –N–P and +N+P were equal ( h 2=0.78). The genetic correlation coefficient ( r g) between −N−P and +N+P was 0.77. Indirect selection in either environment was only 77% efficient in improving grain yield in the other environment. Broad-sense heritabilities for grain yield in −N and +N were 0.75 and 0.82, respectively. Indirect selection under +N for −N target environments was nearly as efficient (98%) as direct selection in the low N environment. Indirect selection in the low N environment for a high N target environment was 90% efficient as direct selection in the high N environment. A high N selection environment appeared to be adequate for grain yield improvement in food barley under both high and low N conditions. However, when the target environment is deficient or adequate in both N and P nutrients direct selection in each of the target environments may be considered. Nonetheless, the high inter-environment correlation for grain yield and the continuing expansion of fertilizer use in barley warrant testing of large numbers of genotypes under high input with a possible verification of the few promising genotypes under high and low N and P inputs.