P. Martre: Model aided phenotyping: from ecophysiological knowledge to phenotype prediction

Pierre Martre is an ecophysiologist and crop modeler at INRAE Montpellier (France). His research focuses on cereal adaptation to climate change. His group develops and integrates a combination of ecophysiological, phenomics, and modeling approaches to predict the responses of genotypes to heat and drought scenarios and identify traits that can be used by breeders, and that are ultimately integrated in genomic prediction pipelines. He is the director of the INRAE – Institut Agro Montpellier Joint Research Unit LEPSE, co-leader of the AgMIP-Wheat team and a PI of the French Convergence Institute #DigitAg.

Plant growth results from multiple feedback mechanisms between processes of various nature (e.g. morphological, physiological, biochemical) acting at different scales and physiological traits or genes that are favorable in an environmental scenario may be negative in another one. A purely experimental approach cannot explore the effects on plant performances (yield, plant resilience, environmental balance) of each combination of traits under all possible environmental scenarios. A combination of observational, experimental and modeling studies is thus required to define the best combinations of traits in a target environment.

By formalizing the relations among processes and associated traits, crop growth models provide a platform for integrative analyses of the impact of a combination of traits on whole-plant and crop phenotype. The application of crop growth models in the context of phenotype analyses relies on their capacity to quantify the effect of individual traits within a trait network from plant measurements made on a large number of genotypes thanks to new developments in phenomics.

The talk presents a novel integrated approach where jointly developed phenotyping methods and models are used to convert phenotyping data into knowledge and information useful for breeders. Case studies are presented to support the idea that crop growth models can help breeders to transition from statistical approaches in analyzing genotype-by-environment interactions to a knowledge-based view that emphasizes crop responses to specific environmental factor. While breeders have traditionally favored broad adaptation, model aided phenotyping opens new avenues to develop genotypes specifically adapted to target climate and weather scenarios.