Acurácia preditiva de testes clonais de Eucalyptus spp. utilizando efeitos aditivos do parentesco e validação cruzada

Abstract

ResumoO uso de extensos plantios clonais é uma prática consenso nas empresas florestais. No entanto, os modelos estatísticos e o tamanho das parcelas utilizadas na seleção dos clones superiores ainda é uma questão entre os melhoristas florestais. Com base em 11 experimentos contendo parcelas semelhantes aos plantios comerciais e 63 clones, este trabalho objetivou avaliar o uso de informação de pedigree, heterogeneidade genética entre ambientes e a possibilidade de redução de parcelas, a fim de otimizar a predição do incremento médio anual (IMA) realizando validação cruzada entre ambientes. Os ambientes proporcionaram altas herdabilidades no sentido-restrito (entre 0,65 a 0,95) e forte alteração dos rankings entre experimentos. A inclusão de parentesco e variâncias genéticas particulares para cada ambiente ao modelo é um procedimento eficiente na melhoria das acurácias realizadas. Além disso, 4 a 16 plantas avaliadas por parcela é uma quantidade confiável para predizer a produtividade dos talhões florestais. Palavras AbstractLarge clonal planting is a practice already established by the timber companies. However, statistical models, the plot size used in selection of superior clones is still a topic of discussion between forest breeders. Based on 11 trials containing plots similar to commercial plantations and 63 clones, this study aimed to evaluate the use of pedigree information, genetic heterogeneity between environments and the possibility of plot reduction in order to optimize the predictive accuracies of the mean annual increment (MAI) performing cross-validation between environments. All trials showed high narrow-sense heritabilities (from 0.65 to 0.95) and a strong change of clonal rankings was verified. It was observed that the inclusion of pedigree information and particular genetic variances for each trial on model is an effective procedure to increase the predictive accuracies. Moreover, between 4-16 trees evaluated per plot is a reliable size for the prediction of forest stands.

Quantifying individual variation in reaction norms using random regression models fitted through Legendre polynomials: application in eucalyptus breeding

“…This implies the estimation of 14 parameters of covariance. Heterogeneous residuals were also reported by Resende et al (2017) evaluating clonal trials of eucalyptus.…”

“…The genotype × environment (G × E) interactions are characterized by the differential behavior of genotypes in relation to environmental variations (Resende 2015). These interactions can be expressed in various ways and with different intensities and can generate significant obstacles for genetic selection (Li et al 2017;van Eeuwijk et al 2016), including eucalyptus (Nunes et al 2017;Resende et al 2017;2018). Thus, the use of accurate, efficient and informative statistical methods that capture the information present in this source of variation and advantageously exploit its effects is fundamental in any breeding program.…”