bioRxiv2021
Wheat is highly sensitive to elevated temperatures, particularly during pollen meiosis and early-to-mid grain filling. The impact of heat stress greatly depends on the plant developmental stage. Thus, germplasm ranking for heat tolerance in field trials may be confounded by variations in developmental phase between genotypes at the time of heat events. A photoperiod-extension method (PEM) was developed allowing screening of 35 diverse genotypes at matched developmental phase despite phenological variations. Paired trials were conducted to compare the new PEM against conventional field screening in plots. In the PEM, plants were sown in single rows or small plots. Artificial lighting was installed at one end of each row or plot to extend day length, inducing a gradient of flowering times with distance from the lights. Individual stems or plot quadrats of each genotype were tagged at flowering. Late-sown plants received more heat shocks during early to mid grain filling than earlier sowings, suffering reductions in both individual grain weight (IGW) and yield. IGW was reduced by 1.5 mg for each additional post-flowering day with temperature > 30{degrees}C. Significant genotypic differences in heat tolerance ranking were observed between PEM versus conventional plot screening. Strong correlations between trials experiencing similar degree of heat were found both for IGW and for total grain weight with the PEM either with individual-stem tagging (e.g. average r of 0.59 and 0.54, respectively for environments with moderate postflowering heat) or quadrat tagging (r of 0.53 and 0.47). However, correlations for IGW and yield in these environments were either poor or negative for conventional trials (e.g. average r of 0.11 and 0.12, respectively for environments with moderate postflowering heat). Accordingly, a PCA grouped genotypes consistently for heir performance across environments with similar heat stress in PEM trials but not in conventional trials. In this study, most consistent genotype ranking for heat tolerance was achieved with the PEM with tagging and harvesting individual spikes at matched developmental phase. The PEM with quadrat sampling provided slightly less consistent rankings but appears overall more suitable for high-throughput phenotyping. The method promises to improve the efficiency of heat tolerance field screening, particularly when comparing genotypes of different maturity types.