Modern agriculture faces major sustainability challenges, including stagnating yields, dependence on fossil resources, and severe environmental impacts. Increasing intra- and interspecific diversity within plots through agroecological design is a promising method for enhancing crop productivity and stability. However, mixed-crop performance remains highly variable, and the genetic architecture of interactions within heterogeneous canopies is poorly understood. Two quantitative genetic frameworks have been proposed: trait-based models, which describe how interacting traits shape phenotypes, and variance-based models, which treat neighbor genotype effects as "black-box" social effects. However, existing variance-based models have been developed almost exclusively for intraspecific interactions and simple neighborhoods. We propose a general multispecies framework describing how a focal plant's phenotype and total breeding value arise from its own direct effects and from the indirect effects of conspecific and heterospecific neighbors. We derived analytical expressions for phenotypic variance, inter-individual covariance, total breeding value variance, and relative heritable variance, which explicitly account for spatial structure, relatedness, and environmental similarities. Using a two-species alternating-row field layout and extensive simulations based on flexible variance-covariance structures, we evaluated the statistical power and bias of joint mixed-model estimators of direct and indirect genetic and environmental effects under a wide range of parameter combinations. Our results show that accurate separation of direct and indirect effects depends on trait heritability and replication, and that modeling genetic covariances across effects and species substantially improves estimation accuracy. This framework provides a unified, individual-centered basis for analyzing complex multispecies neighborhoods and quantifying the breeding potential of plant communities.
Salas, N., Montazeaud, G., Bourke, P. M., Baranger, A., David, J.
Advertisement
Stats
- Recommendations n/a n/a positive of 0 vote(s)
- Views 12
- Comments 0