Multi-environment Clonal Selection Using Ideotype-Design Derived From Factor Analytic Linear Mixed Models: An Application on Eucalyptus Breeding

Abstract The growing demand for raw materials in the forestry sector and the imperative to conserve native forests have spurred the expansion of forest areas planted globally, particularly in Brazil. Eucalyptus stands out as the primary genus employed in the Brazilian forestry industry. Addressing the need for new clones due to expanding exploitation and changing climate conditions poses a considerable challenge. The evaluation of genotypes in multi-environment trials (MET) is complicated by their varying behavior. Factor analytic mixed models (FAMM) are presently employed for MET analyses, enabling the modeling of genotype-by-environment interactions without a substantial increase in parameters. However, existing tools for assessing genotype stability and adaptability within FAMM are unsuitable for scenarios involving a large number of eucalyptus clones, typical in intermediate clonal testing stages. To address this gap, a methodology for clonal selection in eucalyptus is required. FAMM offers a way to summarize clone responses across environments through scores. Utilizing the genotype-ideotype (GI) distance based on these scores facilitates the ranking of clones. This study aims to propose and assess the use of genotype-ideotype distance in conjunction with factor analytic mixed models for selecting eucalyptus clones in multi-environment trials. The combined use of factor analytic mixed models and genotype-ideotype distance enables the ranking of clones based on broad adaptability. This approach balances the simplicity of a single parameter's interpretability with the capability to handle a large number of clones effectively.