Olsson, T., Lindgren, D. and Li, B. 2001. Balancing Genetic Gain and Relatedness in Seed Orchards. ( Silvae Genetica.in press). “Printed” in Olsson (2001).  

The traditional way to avoid related mating and subsequent inbreeding depression in seed orchards is to use only unrelated clones for orchard establishment. As tree-breeding programs move to advanced generations, relatedness (coancestry) among candidates for seed orchard selections becomes more common, especially for high breeding value candidates. The traditional way of selecting the ones with the highest breeding value, provided they are unrelated, is referred to as Restricted Selection (RS). In order to consider breeding value as well as relatedness, an alternative selection method, based on a value criterion for the whole group of selected clones, is presented in this paper. The method, here called Group Merit Selection (GMS), is based on a suggestion by Lindgren and Mullin (1997), but modified for seed orchard selection by neglecting selfing and self-coancestry. The method can be regarded as the selection of a  group of clones that maximizes expected genetic value  (predicted genetic gain minus inbreeding depression).

A case study was conducted in which twenty clones for a seed orchard were selected among second-generation loblolly pine (Pinus teada L.) selections from the NCSU-Industry cooperative breeding program. Assuming an observed inbreeding depression of 40% for one unit coefficient of inbreeding, penalty constants based on estimated breeding values at age 8 was corresponding with inbreeding depression. That gave 12% more genetic value for GMS than Restricted Selection. Predictions of the penalty constant considering additional relevant factors (such as pollen contamination, breeding values based on immature trials, and unrepresentative experimental sites) resulted in selection of the same clones. Changes among the selected clones did not occur until relatedness reached  twice the penalty constant, suggesting that GMS solutions are rather robust.