A comparison of the effect of genetic improvement, seed source and seedling seed orchard variables on progeny growth in Eucalyptus nitens in South Africa

An investigation of assumptions made in estimating genetic parameters and predicting genetic gain in a Eucalyptus nitens breeding programme in South Africa

“…These trials were subsequently thinned to seed orchards, based on the results of the final measurements, and seed collected over several years to establish progeny trials (F2) of this material (Swain 2013). In addition, three genetic gain trials were established on temperate forestry sites in South Africa early in 2001 to test the improved material (Swain et al 2013a). The genetic gain trials also allowed for comparison of realised gain in the F2 with predicted genetic gain, as well as the calculation of realised heritabilities in the F1 (Swain 2013).…”

“…Although levels of relatedness and inbreeding are not available from ICFR seed orchards until DNA genotyping/genetic marker studies have been completed, percentage of trees flowering has ranged from 15 to 47% over orchard and year, in terms of the number of individual trees flowering (Swain et al 2013a), which could decrease genetic gain achieved. Thus this scenario considered an additional higher level of selfing of 25% (and, by extension, outcrossing rate of ≈ 75%) for the family (provenance) parameters estimated above, using a coefficient of relationship of 0.4 as a surrogate for the selfing rate.…”

“…A possibility is that the similar predicted gains reflect the low additional selection possible in the orchards due to the low proportion of trees flowering. Further investigation, however, showed that although the proportion of individuals flowering in the ICFR orchards was relatively low in the year prior to the seed collections for the genetic gain trials (15 to 47% (Swain et al 2013a)), the total number of families flowering in each seed orchard was generally high (59 to 91%). As genetic gains are most sensitive to the family selection intensity (as opposed to within family selection) (Shelbourne 1992), this, together with the self-incompatibility mechanisms that exist in E. nitens, may explain the similar predicted gains for the actual flowering and roguing scenarios.…”

“…However, these flowering windows do overlap (Gardner 2003;Jones 2002), which may allow for panmictic pollination some of the time. In addition to asynchronocity, flowering in E. nitens is subject to seasonal influences (Gardner and Bertling 2005;Jones 2002;Moncur and Hasan 1994;Tibbits 1989) and, as the stability, or greater influence, of female effects in reproductive success in some eucalypts has been noted (Leal and Cotterill 1997;Suitor et al 2009;Tibbits 1989), it is possible that only a few effective males may contribute to pollination of all the other trees in a seed orchard (Hodge et al 1996;Suitor et al 2009;Swain et al 2013a). Conversely, pollen from a few heavy-flowering individuals may have too large a genetic influence on adjacent female trees (Moncur and Boland 2001).…”

“…ii) Actual selection intensity -in reality, stratified selection based on flowering could have taken place in the ICFR seed orchards, as flowering of individual trees ranged from 15 to <50% in the year prior to seed collections (Swain et al 2013a). This means that up to ten trees may have contributed seed from some families and as little as one tree contributed from other families, i.e., the selection intensity in the orchard could have been higher than the figures used in the predictions, resulting in higher realised genetic gains.…”