AGE TRENDS IN GENETIC PARAMETERS AND EARLY SELECTION OF LODGEPOLE PINE PROVENANCES WITH PARTICULAR REFERENCE TO THE LAMBETH MODEL

摘要

This study of time trends in genetic parameters is based on height growth from 50 provenance test plantations involving 140 provenances Total heights were measured at plantation ages 1, 3, 6, 10, 15 and 20 We used correlative heritability (h_(jm)~2) to measure the response of early selection; h_(jm)~2is the product of age-age genetic correlation (r_(Gjm)), and square root of provenance heritability at early (h_j~2) and target (h_m~2) age of selection (h_(jm)~2 =h_jh_mr_(Gjm). Weexpect h_(jm)~2 to bea realistic measure of early selection since it reflects the collective effect of all determinant parameters. We estimated efficiency of early selection (E_(jm)) as the ratio of h_(jm)~2 and provenance heritability at target age of selection We employedthe logarithm of early-target age ratio (log(t_j/t_m)) as predictor to extrapolate E_(jm) to different target ages of selection beyond age 20 (LAMBETH 1980). Cumulative growth indicated that provenance variation in mean height became stable after age 6and experienced only minor shifting in rankings after age 10. Heritability remained high throughout the testing period (h~2 = 0.70 to 0.85) and increased with age except for a slight decline from ages 1 to 3. Both age-age phenotypic and genetic correlations increased steadily with age, and coefficients of the latter were higher than those of the former Genetic correlation reached 0 95 between age 6 and 20 on average h_(jm)~2 showed an age trend similar to that of genetic correlation suggesting the latter drove early selection. E_(jm) increased exponentially and reached over 80% at age 6 and 90% at age 10 if target age of selection was 20. We found high predictability of log(t_jt_m) to project E_(jm) (r~2 = 0.92), indicating high robustness of log(t_j/t_m) as a proxy for age-age correlation as expounded in the Lambeth model. However, we have some reservations about the soundness of gain per unit time as a measure of optimum age of selection Instead we propose E_(jm) as an alternative to assess optimumage of selection We discuss its advantages in both genetic and non-genetic considerations.