Incorporating spatial and genetic competition into breeding pipelines with the R package gencomp.

Abstract

Genetic competition can obscure the true merit of selection candidates, potentially leading to altered genotype rankings and a divergence between expected and actual genetic gains. Despite a wealth of literature on genetic competition in plant and animal breeding, the separation of genetic values into direct genetic effects (DGE, related to a genotype's merit) and indirect genetic effects (IGE, related to the effects of a genotype's alleles on its neighbor's phenotype) in linear mixed models is often overlooked, likely due to the complexity involved. To address this, we introduce gencomp, a new R package designed to simplify the use of (spatial-) genetic competition models in crop and tree breeding routines. gencomp includes functions for constructing the genetic competition matrix, fitting (spatial-) genetic competition models via the variance-component approach, and extracting key results such as variance components, heritabilities, competition classes, and total genetic values. For tree breeding, gencomp also calculates the merit of different clonal mixtures using the estimated DGE and IGE of the selection candidates. In this paper, we first present the theoretical foundation of the methods implemented in the package. We then demonstrate the use of gencomp with two datasets: one simulated from a Eucalyptus spp. trial and a real potato dataset. We used both datasets to demonstrate the influence of genetic competition in variance component estimates, heritabilities and selection. Despite the dependency on ASReml-R, a paid resource, gencomp is a user-friendly tool that can popularize genetic competition models, contributing to more informed decision-making in plant breeding.