Genome-wide association studies reveal stable loci for wheat grain size under different sowing dates

Abstract

Background: Wheat (Tritium aestivum L.) production is critical for global food security. In recent years, due to climate change and the prolonged growing period of rice varieties, the delayed sowing of wheat has resulted in a loss of grain yield in the area of the middle and lower reaches of the Yangtze River. It is of great significance to screen for natural germplasm resources of wheat that are resistant to late sowing and to explore genetic loci that stably control grain size and yield.

Methods: A collection of 327 wheat accessions from diverse sources were subjected to genome-wide association studies using genotyping-by-sequencing. Field trials were conducted under normal, delayed, and seriously delayed sowing conditions for grain length, width, and thousand-grain weight at two sites. Additionally, the additive main effects and multiplicative interaction (AMMI) model was applied to evaluate the stability of thousand-grain weight of 327 accessions across multiple sowing dates.

Results: Four wheat germplasm resources have been screened, demonstrating higher stability of thousand-grain weight. A total of 43, 35, and 39 significant MTAs were determined across all chromosomes except for 4D under the three sowing dates, respectively. A total of 10.31% of MTAs that stably affect wheat grain size could be repeatedly identified in at least two sowing dates, with PVE ranging from 0.03% to 38.06%. Among these, six were for GL, three for GW, and one for TGW. There were three novel and stable loci (4A_598189950, 4B_307707920, 2D_622241054) located in conserved regions of the genome, which provide excellent genetic resources for pyramid breeding strategies of superior loci. Our findings offer a theoretical basis for cultivar improvement and marker-assisted selection in wheat breeding practices.

Keywords: Breeding; Genome-wide association study; Grain size; Sowing dates; Stability; Wheat.

Figure 1. Frequency distribution and Pearson’s correlation coefficients between grain size-related traits under different sowing dates at YZ (A, C) and YC (B, D) sites.

Asterisks (* and **) represent the significance at 0.05 and 0.01 level, respectively. YZ and YC represent Yangzhou and Yancheng, respectively. stage I, stage II, and stage III represent normal, delayed, and seriously delayed sowing conditions, respectively. GL, GW, and TGW represent grain length, grain width, and thousand-grain weight, respectively.

Figure 2. Population genetic characteristics.

(A) Cross-validation error plot (K = 1 to K = 10). (B) Population structure of the 334 accessions at K = 2 and K = 3. (C) The PCA plot demonstrates the division of population into two subgroups. (D) The PCA plot demonstrates the division of population into three subgroups. (E) The linkage disequilibrium decay plots for each subgenome.

Figure 3. Population divergence (Fst) across the two subgroups (A) and nucleotide diversity (Π) of wheat population at whole genome level (B).

Red lines mark the genome-wide threshold at the top 5%.