Ecophysiological mechanisms underlying the positive relationship between seed protein concentration and yield in soybean under field heat and drought stress

A positive relationship between protein concentration and yield has been documented in different combinations of genotype and environment, often under potential conditions. However, the ecophysiological bases underlying this positive relationship under heat stress (HS) and drought stress (DS) during seed filling are still lacking. Our objective was to evaluate the relationship between seed protein content and concentration with yield in field experiments exposed to HS, DS and HS × DS interaction during the seed filling. Two field experiments were conducted and assimilates accumulation, remobilization and redistribution patterns were analysed in high and low seed protein soybean genotypes. The crop was exposed to four treatments: control (ambient temperature and soil water content near field capacity), HS (episodes above 32°C, 6 h d⁻¹) during 15 days, DS (soil water content ≤25% of field capacity) during the whole seed filling and HS × DS. Significant and positive relationships between seed protein content and concentration with yield were observed across treatments and genotypes. Under DS and HS × DS, assimilates available during the seed filling decreased, and assimilates remobilization and partition to seeds were limited, responses significantly associated with seed protein content and concentration, and yield reductions. Furthermore, we demonstrated here that the high leaf N content at the beginning of seed filling, the short early reproductive phase duration, the high source to sink ratio and the high dry matter stem remobilization capacity, as well as the low seed number and high seed weight are intrinsic characteristics of the high protein genotype that could be associated with its high seed protein content and concentration and yield under stressful conditions. This knowledge is key to develop soybean management strategies to improve seed protein level and yield under contrasting productive scenarios.