Forward vs. backwards selection for seed orchards and cooperative second-generation breeding in the US Pacific Northwest.

Abstract

Gain from various orchard strategies were modeled. The scenario tested 2,000 first-generation open-pollinated families, from which orchards of 20 selections were formed, using either parents, progeny or both. This was followed by a second-generation breeding population in which 200 full-sib families were tested followed by a second-generation orchard of 20 selections. The results showed that a 1.5 generation seed orchard (recruit from many first-generation open-pollinated testing programs with lots of parents from which to choose) using parents would give more gain than all-progeny orchards and is essentially equal to the gain from selecting both progeny and parents. However, the situation was changed in the second cycle; in many cases progeny will have the highest expected gain. Gains from a second cycle of breeding and testing 200 full-sib families (and choosing the best 20 parents or progeny) approached gains from testing 2,000 openpollinated families and selecting the top individuals from the top 200 families. This is reassuring given that the second cycle will cost only around 10% of the first cycle, if costs per planted tree remain constant. There appeared to be good justification for selecting based on age-6 data, establishing an orchard, and roguing based on age-12 data rather than waiting for age-12 data to begin building the orchard.

Citation

Johnson, R.; Jayawickrama, K. 2003. Forward vs. backwards selection for seed orchards and cooperative second-generation breeding in the US Pacific Northwest. In: Northwest tree improvement cooperative annual report, Jan-Dec 2002. Corvallis, OR: Oregon State University, Department of Forest Service: 17-23