Response of white clover genotypes to evaluation environments of dense andspaced planting, and implications for selection

Twelve genotypes were grown for two years as monocultures in the evaluation environments (EE) of spaced planting (SP) and dense planting (DP). Dry mat ter (DM) yield, seed yield and their contributing traits were recorded. The environments were compared for: i) genotype performance, and selection eff iciency for the target environment of DP, and ii) estimates of genetic corr elations between traits. DP was obtained by close transplanting of clonal c uttings. EE x G interaction occurred for DM yield over harvests, and stolon density, petiole length, inflorescence production and number of florets pe r inflorescence averaged over repeated measurements. Interaction effects le ading to relatively higher DM yield in SP tended to be associated positivel y with those for the other traits and featured the genotypes with greater a bility to colonize bare soil by stolon production. The interaction effects decreased with time and generally disappeared (p < 0.05) by the end of the experiment, when the environments reached a similar DM yield and stolon den sity. EE x G interaction also occurred for seed yield (p < 0.08), recorded at the second year. Indirect selection in SP for performance in DP was pred icted to be less efficient for DM yield (-40%), seed yield (-23%) and the o ther traits subject to EE x G interaction, compared with direct selection i n DP. Evaluation under SP led to identification of several genetic correlat ions that were not confirmed under DP. DP compared with the traditional SP evaluation technique is more labour-demanding but allows for i) increasing the selection efficiency and/or decreasing the time needed for a reliable e valuation, and ii) obtaining a more reliable estimation of genetic correlat ions for ordinary agricultural environments.