Comparison of breeding values estimated between single-tree and multiple-tree plots for a slash pine population

Forest tree breeders often use single-tree plots, rather than multiple-tree plots, to maximize the number of genotypes tested, increase precision of genetic predictions, and achieve greater genetic gains per unit of time. However, genetically improved individuals are deployed operationally on large tracts of land, as stands, more like multiple-tree plots. This raises the concern that breeding values estimated from single-tree tests are not well correlated with those from large multiple-tree (i.e., block) plots. Comparisons of breeding values from single-tree and multiple-tree plots are largely missing, due to the cost and time associated with establishing and measuring block plots. To address this need, breeding values were evaluated from a second breeding cycle of slash pine that established two polymix single-tree plot trial series and one trial series of full-sib block plots (FSBP). The first polymix series (PMX I) evaluated 140 families, and the second series (PMX II) evaluated 201 polymix families and 43 elite full-sib crosses. Each polymix series had eight sites established in Florida and Georgia, USA. The FSBP series was established at 11 locations in Florida and Georgia with an average of 63 trees per unreplicated block plot for a total of 1141 full-sib families. Breeding values were estimated for each of the families based on fitting a global linear mixed model that included all series. Heritabilities for survival, volume, and rust were 0.15, 0.35, and 0.22, respectively, in PMX I, and 0.11, 0.18, and 0.11, respectively, in PMX II. Correlations between PMX and FSBP series parental breeding value estimates were higher than 0.84, 0.61, and 0.95 for survival, volume, and rust, respectively. Hence, moderate to strong agreement exists for the genetic rankings between single- and multiple-tree plots.