Tree Genetics & Genomes

, Volume 9, Issue 1, pp 1–17

Accounting for competition in genetic analysis, with particular emphasis on forest genetic trials

Original Paper


Available experimental evidence suggests that there are genetic differences in the abilities of trees to compete for resources, in addition to non-genetic differences due to micro-site variation. The use of indirect genetic effects within the framework of linear mixed model methodology has been proposed for estimating genetic parameters and responses to selection in the presence of genetic competition. In this context, an individual’s total breeding value reflects the effects of its direct breeding value on its own phenotype and its competitive breeding value on the phenotype of its neighbours. The present study used simulated data to investigate the relevance of accounting for competitive effects at the genetic and non-genetic levels in terms of the estimation of (co)variance components and selection response. Different experimental designs that resulted in different genetic relatedness levels within a neighbourhood and survival were other key issues examined. Variances estimated for additive genetic and residual effects tended to be biased under models that ignored genetic competition. Models that fitted competition at the genetic level only also resulted in biased (co)variance estimates for direct additive, competitive additive and residual effects. The ability to detect the correct model was reduced when relatedness within a neighbourhood was very low and survival decreased. Selection responses changed considerably between selecting on breeding value estimates from a model ignoring genetic competition and total breeding estimates using the correct model. Our results suggest that considering a genetic basis to competitive ability will be important to optimise selection programmes for genetic improvement of tree species.


Competition Direct and indirect genetic effects Linear mixed model Genetic (co)variance components BLUP Response to selection 

Supplementary material

11295_2012_521_MOESM1_ESM.pdf (26 kb)
Online Resource 1(PDF 26.2 kb)
11295_2012_521_MOESM2_ESM.pdf (23 kb)
Online Resource 2Number of likelihood-ratio tests that were not statistically significant at the 5% level from a total of 100 comparisons between the AR and GC-AR models in a given simulation scenario. (PDF 22.9 kb)
11295_2012_521_MOESM3_ESM.pdf (47 kb)
Online Resource 3(PDF 46.9 kb)

Copyright information

© Springer-Verlag 2012

Authors and Affiliations

  1. 1.Centro de Estudos Florestais, Departamento dos Recursos Naturais, Ambiente e Território, Instituto Superior de AgronomiaUniversidade Técnica de LisboaLisboaPortugal
  2. 2.PlantPlan Genetics Pty Ltd, School of Plant ScienceUniversity of TasmaniaHobartAustralia

Personalised recommendations