Optimal clone number for seed orchards with tested clones

Abstract

The optimal number of clones in seed orchards is discussed. A model is constructed to maximize a goodness criterion ("benefit") for seed orchards. This criterion is a function of: 1) the number of tested genotypes available for selection and planted in seed orchard; 2) the contribution to pollination from: a) the ramet itself; b) the closest neighbors; c) the rest of the orchard and sources outside the orchard (contamination); 3) variation among genotypes for fertility; 4) frequency of selfing; 5) production of selfed genotypes; 6) gene diversity (= status number); 7) influence of contamination; 8) genetic variation among candidates; 9) correlation between selection criterion (e.g. height in progeny test) and value for forestry (e.g. production in forests from the orchard); and 10) the number of clones harvested. Numeric values of the entries are discussed, and values were chosen to be relevant for scenarios with Swedish conifers (focusing on Scots pine) and for loblolly pine. Benefit was maximized considering the number of clones. The optimum was 16 clones for the Swedish scenario, while less for the loblolly pine scenario. The optimum was rather broad, thus it is not essential to deploy the exact optimum, and an approximate optimum will do. A sensitivity analyses was performed to evaluate the importance of the likely uncertainty and variation in different entries. Quantification of the benefit of gene diversity is important. Other significant considerations are the genetic variance in the goal character and the ability to predict it, as well as the impact of selfing and the variation in reproductive success between clones. Twenty clones is suggested as a thumb rule for Swedish conifers.