Associated SNPs, Heritabilities, Trait Correlations, and Genomic Breeding Values for Resistance in Snap Beans (Phaseolus vulgaris L.) to Root Rot Caused by Fusarium solani (Mart.) f. sp. phaseoli (Burkholder)

Abstract

Root rot is a major constraint to snap bean (Phaseolus vulgaris) production in the United States and around the world. Genetic resistance is needed to effectively control root rot disease because cultural control methods are ineffective, and the pathogen will be present at the end of one season of production on previously clean land. A diversity panel of 149 snap bean pure lines was evaluated for resistance to Fusarium root rot in Oregon. Morphological traits potentially associated with root rot resistance, such as aboveground biomass, adventitious roots, taproot diameter, basal root diameter, deepest root angle, shallowest root angle, root angle average, root angle difference, and root angle geometric mean were evaluated and correlated to disease severity. A genome wide association study (GWAS) using the Fixed and random model Circulating Probability Unification (FarmCPU) statistical method, identified five associated single nucleotide polymorphisms (SNPs) for disease severity and two SNPs for biomass. The SNPs were found on Pv03, Pv07, Pv08, Pv10, and Pv11. One candidate gene for disease reaction near a SNP on Pv03 codes for a peroxidase, and two candidates associated with biomass SNPs were a 2-alkenal reductase gene cluster on Pv10 and a Pentatricopeptide repeat domain on Pv11. Bean lines utilized in the study were ranked by genomic estimated breeding values (GEBV) for disease severity, biomass, and the root architecture traits, and the observed and predicted values had high to moderate correlations. Cross validation of genomic predictions showed slightly lower correlational accuracy. Bean lines with the highest GEBV were among the most resistant, but did not necessarily rank at the very top numerically. This study provides information on the relationship of root architecture traits to root rot disease reaction. Snap bean lines with genetic merit for genomic selection were identified and may be utilized in future breeding efforts.

Keywords: best linear unbiased prediction; common bean; disease resistance; genome wide association studies; genomic prediction; genomic selection; root morphology.

FIGURE 1

Manhattan and corresponding Q-Q plots from a GWAS analysis of disease severity (A) and biomass (B) in the BeanCAP Snap Bean Diversity Panel evaluated in 2 years for Fusarium solani reaction at the Oregon State University Vegetable Research Farm. The Bonferroni cutoff based on effective marker number (-log₁₀ 4.68, α = 0.05) is shown as a solid line. For the Q-Q plots, the null distribution is shown as a red line.

FIGURE 2

Linkage disequilibrium (LD) heat map of common bean chromosomes Pv03, 07, 08, 10, and 11 showing all possible pairwise comparisons of SNPs arranged along the chromosome. R² values are displayed above and right of the diagonal and corresponding probabilities below and left of the diagonal. Color scales show corresponding R² and probabilities where red for each would indicate strong and highly significant LD. SNPs associated with disease severity (black ^∗) or biomass (yellow ^∗) are indicated along the diagonal.

FIGURE 3

Correlation of predicted and observed values for training vs. testing populations at four ratios (60:40, 70:30, 80:20, and 90:10%) and compared with observed vs. predicted for the entire population (100%) of the BeanCAP Snap Bean Diversity Panel for Fusarium solani root rot disease severity and plant and root traits. Correlation coefficients were generated by rrBLUP using 10 K iterations and 10 repetitions per trait-level combination. (A) Adventitious roots, (B) Biomass, (C) Basal stem diameter, (D) Disease severity, (E) Deep root angle, (F) Shallow root angle, and (G) Taproot diameter.

FIGURE 4

Effect of number of SNPs on predictive accuracy for Fusarium root rot disease severity and root traits of a snap bean diversity panel. SNPs were first filtered for MAF < 0.05, then sorted from smallest to largest P value and arranged in nine subsets approximately doubling in size with each step. Full set of SNPs for biomass was 7,082 while for all other traits totaled 8,036. Number of SNPs is plotted on a logarithmic (base 10) scale.