Grain protein–yield relationship in wheat: Sources of variation

Grain protein concentration and yield are critical traits affecting the economic value and end-use quality of wheat. Although the literature describing a negative protein concentration-yield relationship is ample, it lacks discussion on the relevance of the source of variation. This study was designed to determine the dependency of this relationship upon specific sources of variation, i.e., year, location, and genotype We utilized multiple approaches to investigate the degree and frequency of impact of those sources of variation on the protein-yield relationship using data from the Oklahoma State University winter wheat performance trials. The database consisted of 21 years, 31 locations, and 256 genotypes (n = 9553 averages of the replicates in each experiment). There was an overall negative relationship between protein concentration and yield but with varying degrees of relevance depending upon the source of variation. Considering the significance (p value) of the relationship, it was genotype>year>location. The results suggest that changes in year, by nature an unpredictable environmental source of variation, may affect the grain protein concentration and yield relationship more frequently (higher R 2 ), and genotypes more negatively (more cases with negative slope and p value < 0.05), than location which is a more predictable environmental source of variation. The long-assumed negative relationship between wheat grain protein concentration and yield is not universal but varies with the underlying source of variation. • Protein–yield trade-off varies by genotype, year, and location. • The strongest negative relationship was obtained when analyzing the trade-off among genotypes. • Year-to-year variation affected the relationship more frequently than location. • Location had the weakest and most predictable influence on protein–yield trade-off. • Findings clarify context-dependent drivers of the wheat protein–yield relationship.