Consistency of Yield Ranking and Adaptability Patterns of Winter Wheat Cultivars between Multi-Environmental Trials and Farmer Surveys

Abstract

Cultivar recommendation based on mean performance determined by multi-environment trials (METs) conducted on research stations could be unreliable and ineffective for assessing performance in farmers’ fields. It is important to improve the efficiency of cultivar recommendation based on METs. For this purpose, it would be useful to validate recommendations based on yield data obtained directly from farmers, i.e., through surveys. The aim of this study was to discuss the possibility and statistical methodology of assessing cultivar performance patterns based on yield data obtained through farmer surveys. We suggest that this might be accomplished by assessing the conformity of yield ranking and yield performance patterns between MET and survey datasets in the same growing regions. As an example, we compare winter wheat (Triticum aestivum L.) yield data obtained from Polish farmers via surveys with data obtained via METs. In the METs, cultivars were evaluated at two levels of crop-management, a moderate-input management (MIM) system and a high-input management (HIM) system. Based on the yield evaluations in the current study, half of the agro-ecological regions had relatively high levels of consistency in yield rankings between the MET MIM system and survey yield dataset. This indicated a relatively high efficiency of cultivar recommendations based on METs in these regions, especially for the MIM system. For the HIM system, however, with the exception of one region, we observed a poor degree of consistency in cultivar ranking.

Keywords

• crop management
• cultivar recommendation
• genotype × environment interaction
• mixed linear model