Benefit–cost analysis of DNA marker-based selection in progenies of Pinus radiata seed orchard parents

The financial viability of DNA marker-based within-family selection (MBS) compared with full-sib family forestry was evaluated for Pinus radiata Donn. ex D. Don. Two traits were investigated: wood density (WD) and diameter at breast height (DBH, 1.4 m). Assuming 20 biallelic loci of equal additive effect controlling trait variation in 15 unrelated top full-sib families of P. radiata, marginal costs of quantitative trait loci (QTL) detection and selection were estimated based on an average of slightly less than five loci per family. We assumed a program where 10 genotypes per family per year were deployed over a 5-year period, and each replicated 100 000 times via fascicle cuttings methods. Estimated marginal costs were NZ$32 and NZ$72 per 1000 plants for WD and DBH, respectively. Genotyping costs were the single largest component for both traits. Genetic gains were estimated by modifying predicted log volumes (DBH) or proportion of structural-grade timber (WD) with and without pruning. Estimated genetic gains ranged from 3.2 to 3.4%. Net present values (assuming a 9.5% discount rate) ranged from an average of NZ$51 to NZ$621/ha. Results showed that MBS for DBH was more profitable than for WD, despite markedly higher costs of QTL detection. All trait-silviculture combinations showed financial gains with internal rates of return of 9% or greater, even when estimated revenues were decreased 70% from forecast revenues. While this analysis is based on a large number of assumptions, it is robust and the results show that significant financial gains from MBS are possible even when selection is based upon DNA markers linked to a few loci each of relatively small effect.