Abstract

A total of 389 accessions and a local cultivar Duorado precoce were evaluated in a simple 30 × 13 alpha lattice design with two replications under four experimental environments (N^-P^-, no N or P application; N^-P⁺, P applied; N⁺P^-, N applied and N⁺P⁺, both N and P applied) at the rate of 60 kg P and 90 kg N ha^-1. Data was recorded on Days to heading, anthesis and maturity (days), P and N tolerance, plant height (cm), above ground biomass (g), number of panicles (absolute numbers per ten plants), days to maturity (days), 1000 grain weight (g), and grain yield (kg ha^-1). The genotypes and environments were highly significant for all the traits studied. The degree of genetic determination (H²) ranged from 6.8% for P tolerance to 36.5% for above ground biomass. The phenotypic coefficient of variation of genotypes ranged from 14.3% for days to maturity to 159.7% for top biomass. The genetic advance (GA) ranged from 0.2 for phosphorous tolerance to 1080.5 for grain yield, while the genetic advance expressed as percent of the mean was 5.7% for days to maturity and 87.9% for top biomass. The top biomass seems to be highly heritable trait and simple phenotypic selection is possible. The ten characters studied had wide variability under the four environments with days to maturity ranging from 188 for genotype ARCCU1Fa1-L4P3-HB under N⁺P⁺ to 177 for genotype CT16333(1)-CA-1-M under N-P- condition. The highest yielding genotype was CT16328-CA-18-M under N-P- with 5916 kg ha^-1. The germplasm showed variability for low soil N and P adaptation, and hence improvement was possible to take advantage of the vast unexploited upland environments for increased rice productivity. There was high variability in the genotypes to warrant rice improvement for yield.

Key words: Soil fertility, genetic advance, genotypes, heritability, Kenya, rice.