Impact of progeny size on genetic parameter estimation and selection gain in progeny trials of <i>Eucalyptus spp</i> .

Abstract Progeny trials are the first and most important phase of a Eucalyptus breeding program. These trials are designed to capture breeding values of parents and progenies, enabling the selection of progenies or superior individual genotypes for cloning. The influence of progeny size on the estimations of genetic parameters and the efficacy on of individual genotype selection remains underexplored. To address this question, the present study investigated the effect of progeny size on the estimation of genetic parameters and selection efficiency in eucalyptus progeny trials. Three full-sibs progeny trials were used to sample varying numbers of individuals within the progenies across blocks. Sampling ranged from one plant per block (single-tree-plot) to 16 plants per block (5 to 80 plants per progeny). For each sampling scenario, 1,000 resampling iterations without replacement were performed to fit a mixed linear model for wood volume. In each iteration, estimates of the individual narrow-sense heritability ( h a 2 ), within-family heritability ( h w 2 ), and the accuracy of selection at progeny level ( r̂gg ) were obtained. Analyses of the impact of the number of individuals per progeny showed that the r̂gg improved with increasing progeny size. A stabilization trend was observed for h a 2 when progeny size reached 50 individuals. In scenarios with fewer individuals, the variation in estimates of h a 2 , h w 2 , and r̂gg was significantly higher, with values reaching zero. Under conditions of low heritability estimates, the negative impact of sampling is more pronounced, leading to an overestimation of selection gains in smaller progeny sizes. A minimum of 50 individuals per progeny is recommended to reliably estimate genetic parameters and reduce sampling errors in the selection of superior individuals in full-sib progeny trials.