Presentation of a method for improving our understanding of varietal characteristics, the value of models, and the need for further development

The knowledge gained in agronomy and statistics now enables us to make much greater use of the results from variety evaluation networks, and to deepen our understanding of growing conditions and varieties. With regard to varieties, it is possible to estimate their tolerance to factors limiting production whose effects are not readily apparent or are difficult to measure. This article presents a method for interpreting the results of variety trials, based on the description of factors limiting production. This method relies on the collection of environmental data, the observation of control varieties and the application of agronomic models. Statistical tools (linear regression, factorial regression), which allow for rapid and in-depth interpretation, are also used to establish links between variations in a production variable and the variables describing the limiting factors. The foundations required for implementing this method (data, observations) and its current limitations are presented. Ongoing work on improving agronomic and statistical models, which involves all stakeholders in variety evaluation, should help to reduce the additional workload involved in monitoring trials and achieve greater convergence between statistical interpretation and agronomic expertise in variety trials.