Genotype by environment interaction and resource optimization in sugarcane variety evaluation in Swaziland

[摘要] English: The primary aim of this study was to assess the genotype by environment interaction (GEI) in the sugarcane Variety Evaluation Programme (VEP) of the Swaziland Sugar Association (SSA). This was achieved by pursuing the specific objectives: to evaluate the adaptability and phenotypic stability of imported sugarcane genotypes in Swaziland; to determine the optimum combination of locations, replications and crop-years necessary to provide an adequate level of discrimination among genotypes within the SSA VEP; and to undertake a combined data analysis of irrigated sugarcane variety trials in Swaziland and South Africa. Data from replicated trials established across locations, seasons and crop-years were used, and the GenStat® 17th Edition software was used for the analyses.The study indicated that the main effects of variety, location and crop-years were significant, showing genetic diversity amongst the tested material and the complex environments under which they are tested. Consequently, these resulted in significant genotype by environment interactions (GEI). For sucrose yield, variety x location and variety x crop-year interactions were significant, while the higher order interaction (variety x location x crop-year) was not significant. Consistently, environments accounted for a larger portion of yield variation. The ratio of variety:GEI was largely dependent on the source of varieties. Varieties imported from Mauritius had a larger GEI effect than variety effect, while South African varieties had larger variety than GEI effects. This is attributed to the fact that conditions at which South African varieties are tested are similar to those in Swaziland. The grouping of test environments according to season clearly indicated that seasonal effects are stronger than location effects, as a result more emphasis should be put on testing across seasons than locations/soil types. Variety M1176/77 proved to be high yielding, stable and widely adapted to the sugarcane growing conditions of Swaziland, while M1400/86 was high yielding but specifically adapted to good draining soil conditions.To model the optimum number of locations, replications and crop-years necessary to provide adequate discriminating ability among varieties, the broad sense heritability was used. The results indicated that four locations, four replications per trials, and four crop-years were sufficient for variety testing within the industry. While the data used for the combined analysis of irrigated sugarcane variety trials of Swaziland and South Africa werelargely unbalanced, the analysis indicated potential for SSA to utilise results from SASRI post-release trials to inform the industry's variety recommendation. Both industries should consider pooling together testing resources and conduct variety evaluations at regional level. It is envisaged that making these adjustments will ensure that SSA derives maximum benefit from the resources used to test imported varieties.

[发布机构] University of the Free State