Uncovering yield-related traits loci in Chinese wheat landraces using GWAS

The genetic basis of wheat yield is highly complex, being collectively influenced by key traits such as plant morphology, spike structure, and grain morphology. Therefore, exploring the genetic regulatory mechanisms underlying yield-related traits is crucial for breeding high-yield wheat varieties. In the present study, a panel of 372 wheat landraces from Gansu province, China, was phenotyped four yield-related traits: spike length (SL), spikelet number per spike (SNS), grain number per spike (GNS), and thousand-kernel weight (TKW) over four years across five field environments at two locations (Qingshui and Lanzhou) in Gansu province. Genotyping was performed using a 16K SNP array, and genome-wide association studies (GWAS) were conducted on the multi-environment phenotypic data employing the Bayesian-information and Linkage-disequilibrium Iteratively Nested Keyway (BLINK), Compressed Mixed Linear Model (CMLM), Fixed and Random Model Circulating Probability Unification (FarmCPU), Mixed Linear Model (MLM) and Multi-Locus Mixed Model (MLMM) models. In total, we identified nine loci for SL, four loci for GNS, one loci for SNS, and five loci for TKW. The phenotypic variation explained (PVE) by these loci ranged from 2.0% to 22.0%. Each locus exhibits 2-4 haplotypes and showed significant differences between haplotypes. Furthermore, 16 candidate genes associated with panicle traits were predicted and they encoded proteins such as squamosa promoter-binding protein, F-box family protein, ferredoxin-NADP-oxidoreductase 2, aspartic proteinase Asp1, Pre-mRNA splicing factor and so on. These results provide novel genetic loci for wheat yield improvement, offering valuable resources for future breeding programs.