Genotypic competition models in loblolly pine progeny tests

Abstract

Three areas pertinent to the understanding of genetic competition were discussed in this study: 1) static competition models: 2) dynamic competition models: and 3) growth and yield models with genetic competition. Based on the relationship among phenotypic effects, genotypic effects, and environmental effects for quantitative traits, a genetic competition model was developed to depict the underlying relationship among competitive effect, competitive ability, and competitive influence. This genetic model was applied to static competition models which were able to provide a snapshot in time and allow extrapolation of yields of various genotypic combinations. The static competition models were examined for nine half-sib families common to three progeny tests in loblolly pine (Pinus taeda L.). The predicted yield of mixed family stands was compared with the predicted yield of pure stands of the same families. Three types of competitive effects (over-yielding, neutral, and under-yielding) were revealed. The competitive effects ranged from -10.07 for under-yielding and -5.0% for over-yielding. The majority of the competitive effects contrary to conventional wisdom, were neutral and under-yielding, implying that plantations of single families should over-yield mixed plantations of two or more families. Furthermore, family ranks based on the predicted performance in pure and mixed family stands were consistent with the observed family ranks revealed from actual progeny tests. The static competition models were extended to dynamic competition models by modelling the changes of growth rates of different families over time. To be realistic, the changes of growth rates due to thinning were also considered. Thus, the competitive effects at any time can be predicted by integrating the growth of individual families over time. This study revealed that families showed differences in growth rate in response to crown closure and enlarged growing space due to artificial thinning. The third aspect of this study dealt with four elements of existing growth and yield models. The corresponding observations were: 1) Area Potentially Available had the highest correlation with initial basal area and basal area growth: 2) Competition after crown closure diminished differences in the height-DBH relationship among families: 3) The two-parameter log-linear model was sufficient to describe the height-age relationship: and 4) Family competition information must be incorporated into the site index model of existing growth and yield models when simulating more than one family in a stand.