Breeding Values of Parental Trees, Genetic Worth of Seed Orchard Seedlots, and Yields of Improved Stocks in British Columbia

• Published in: Western journal of applied forestry Vol. 18; no. 2; pp. 88 - 100
• Main Authors: Xie, C.Y, Yanchuk, A.D
• Format: Journal Article
• Language: English
• Published: Bethesda Society of American Foresters 01.04.2003; Oxford University Press
• Subjects: breeding value; genetic correlation; genetic improvement; growth models; parents; pollen; pollination; prediction; reforestation; Seed orchards; trees; variance; British Columbia
• ISSN: 0885-6095; 1938-3770
• DOI: 10.1093/wjaf/18.2.88

This report describes the procedures that are currently used in British Columbia for predicting the breeding values of parents, estimating the genetic worth of orchard seedlots, and projecting the yields of genetically improved stocks. Breeding value is a measure of the genetic quality of an individual as a parent. There are several procedures available for its estimation/prediction. Among those, the best linear prediction (BLP) relaxes most of the assumptions required by the others and minimizes the error variance of prediction. In most situations in British Columbia, it should provide predictions with satisfactory accuracy and precision with greatly reduced computational complexity. In this province, breeding value for growth potential is expressed as percent gain of stem volume over the unimproved population at a designated rotation age. Genetic worth is an important attribute of the genetic quality of a seedlot. It represents the average level of genetic gain expected for the trait of concern at a designated rotation age when a seedlot is used for reforestation. Currently, the genetic worth of a seedlot is estimated by the mean breeding value of all the parents, including those that contribute to pollen contamination and supplemental mass pollination, weighted by their proportional gamete contributions The yield of a genetically improved plantation is projected by incorporating the genetic worth of the seedlot into the existing growth model developed based on extensive data from managed unimproved stands. The current approach not only takes account of the stand dynamics determined by site conditions and silvicultural regimes but also the declining nature of expected gain over time because of imperfect age-age genetic correlation. Because of errors from genetic and environmental sampling, measurement, and analysis, as well as possible violation of model assumptions, estimates/predictions may still be subject to errors and/or biases. Various conservative measures have been taken to minimize any possible upwards biases. As more matured data and advanced analytical technologies become available, both the accuracy and precision will be improved. The advancement made in the procedures described in this document should contribute to superior decisions in many aspects of forest management.