Abstract
Genetic gains from between- and within-family selection were predicted for a trait of low (0,1) medium (0,2) and high (0,4) narrow-sense heritability, under five different breeding-population options and for 12 plant-production population options. The breeding populations (BP) included open-pollinated (OP), control-pollinated (full-sib) and cloned open-pollinated BP; and the production populations included four different open-pollinated seedling and six clonal orchards, as well as two schemes of selecting clones for clonal forestry. Selection intensities (selection ratios) were chosen throughout to be practical and realistic. It was assumed that between-family selection ratios in the BP would be very low (two out of three families) to minimise reduction of effective population size. Gain from the OP BP was only slightly inferior to the full-sib BP under these constraints. Cloning the or BP with 4 and 10 ramets per clone and 10 clones per family resulted in gains that were considerably higher than from the seedling BPs with the same total number of trees per family. Among the production-population options the forwards selection clonal orchard with roguing showed high gains (but required an extra cycle of testing). The clonal orchard from backwards selection was even better, but with a still-longer cycle. A forwards selection clonal orchard that derived from clonal selection in the cloned OP BP gave higher gains than the equivalent forwards selection orchard, and converting the cloned BP gave a seed source of similar gain to the forwards-selection clonal orchard. The cloned BP could also form a source of clones for clonal forestry, and gains immediately available after bulking-up the best clones were higher than most seed-based options. An alternative method to produce clones for clonal forestry was by crossing the best forwards-selected clones from the BP, and gave highest gains but had a very long cycle. Comparative genetic gains from various breeding-population and production-population options were not very sensitive to heritability of the selection trait, from 0,1 to 0,4, though cloned options were most effective at low heritability. Combining cycle length and genetic gain as gain per year, showed highest values for the cloned BP, both for clone-within-family selection for the BP and also for clonal selection for clonal forestry, reflecting the high gains and short cycles of these options. Costs were sensitive to the number of plants involved in the tests. Highest gains % per year per dollar were found for the cloned breeding population with the lowest number of ramets per clone (four), both as a cloned OP breeding population and as clonal selection for clonal forestry.

This publication has 5 references indexed in Scilit: