Graduate Thesis Or Dissertation


Breeding Snap Beans for Organic Agriculture: Genomic Shifts under Different Agricultural Management Systems Public Deposited

Downloadable Content

Download PDF


Attribute NameValues
  • Although the concept of breeding crops for organic agricultural systems has been around for decades, few studies have been reported for vegetable crops to quantify and compare how conventional and organic systems differ in their selection pressures on genotypes advancing through the breeding cycle; especially research utilizing modern sequencing capabilities. There have been a number of studies comparing direct and indirect selection in other crops to determine the effectiveness of each in developing varieties for organic agriculture, many of which have had differing results—partially due to the wide environmental variability that is inherent in agricultural ecosystems and management methods, differences in crop types and inputs required for production, as well as the genetic diversity within crops. Many of the goals that are important in breeding for organic agriculture, are also important when considering the changing climate (i.e. nutrient use efficiency, adaptation to diverse, variable environments, above and below-ground architecture, etc.). Plant breeders must consider developing varieties that are well-suited to the changing environment, and less reliant on inputs that may be more difficult to supply in the near future. This research was designed to identify what kinds of genetic changes might be produced in snap bean populations due to natural selection in both organic and conventionally-managed production systems throughout the breeding cycle. We utilized two recombinant inbred snap bean populations, OR5630 x Black Valentine (ORBV) and Hystyle x Provider (HYPR), which were split after the F1 generation and grown in parallel organic and conventional systems from the F2 through F6 generations, resulting in four populations, two in each system. Systems treatments (organic and conventional) differed for seed fungicide treatment, fertilizers, herbicides, and other pesticides utilized. Most generations were grown in Corvallis, Oregon at Oregon State University’s Vegetable Research Farm, Lewis Brown Research Farm, and greenhouse facilities, with the exception of the F5 which was grown for seed increase in a winter nursery near Palm Springs, California. We hypothesized that selection in the systems would result in genomic shifts, and genotyped 94 families (in each population/system) in the F5 generation with the Illumina 6000 SNP BARCbean6k_3 Beadchip to give us insights into what effects these systems produced, with the goal of identifying patterns that could inform future breeding projects. Linkage maps were constructed for all four populations. The ORBV map had the most coverage with 445 SNP markers spanning 1,153 cM, followed by the conventional population with 388 SNPs spanning 798 cM (96 and 67 percent coverage, respectively, of 1,200 cM average P. vulgaris genetic map length). The linkage maps of the HYPR populations contained fewer markers as the parents were from more similar gene pools. The organic population had 272 markers and spanned 590 cM, and the conventional had 268 SNPs and spanned 566 cM (49 and 47 percent coverage, respectively, of 1,200 cM average P. vulgaris genetic map length). Phenotypic data was collected for the F6 generation from replicated trials (3 rep RCBD) at Oregon State University’s Lewis Brown Research Farm in 2018. Populations from the same cross were compared, but families selected in one system were not simultaneously grown and compared in both systems. ANOVAs of population and production system main effects revealed that seed yield and biomass were significantly higher under conventional production. Flowering and physiological maturity were significantly earlier in organic populations. Population main effect was not significant except for flowering and seed weight. Production system by population interaction was not significant for most traits, seed weight being the only one for which a significant interaction was observed. Narrow-sense heritability estimates were calculated and were similar for most traits. To estimate variation in gene frequencies between populations from the same cross, F statistics were calculated using Nei’s method of pairwise Fst. No significant differences in allele frequency variation were found between populations from the same cross. Based on expected allele frequencies, individual SNPs and regions that were biased towards one parent or the other’s alleles were identified in the genetic data of all four populations. The ORBV organic and conventional populations shared 73 distorted SNPs with multiple distorted SNPs on Pv07 and Pv10, and two SNPs that were distorted on Pv08. The organic population had 208 system-specific distorted SNPs, many of which were in regions of the genome not showing distortion in other populations. These were found on all chromosomes except for Pv04 and Pv08, with the most distorted SNPs being on Pv05 (107 distorted SNPs). The conventional ORBV population also had 53 system-specific distorted SNPs, primarily on Pv10, where there were 47. This population also had single uniquely distorted SNPs on Pv03 and Pv05, and three unique SNPs on Pv07. The Hystyle x Provider (HYPR) organic and conventional populations also had 102 shared distorted SNPs. One was on chromosome 1, with the over-represented allele being from Hystyle in organic and from Provider in conventional. There also were distorted SNPs in both populations on Pv02, Pv07, Pv08, Pv10, and Pv11. Both of these populations also had many system-specific distorted SNPs. In the organic HYPR population, there were 91 uniquely distorted SNPs on Pv01, Pv02, Pv04, Pv07, Pv08, and Pv10, with the most being 37 on Pv04. For the conventional population, there were 201 uniquely distorted SNPs on Pv01, Pv02, Pv05, Pv09, Pv10, and Pv11, with the latter chromosome having 124. Quantitative Trait Loci (QTLs) were also found for all four populations using the genetic maps that were constructed and F6 phenotypic data. For flowering and maturity traits, Pv02, Pv04, and Pv08 contained most QTLs in the ORBV populations, with the most significant in all categories being on Pv08 in the organic population. There were two unique QTLs found for flowering and maturity traits on Pv03 and Pv04 that were significant only in the organic population of ORBV. There was also a unique region for the conventional population linked to maturity on Pv06. For biomass, Pv06 contained QTLs for both populations with overlapping physical positions, and there were unique QTLs on Pv02 and Pv04 for the organic population. Seed yield and weight QTLs were mostly found on Pv07, but there were two significant QTLs associated with seed yield on Pv05 and Pv06 for the conventional population. Both organic and conventional populations had significant QTLs on Pv08 associated with seed weight. There was also a QTL on Pv11 for harvest index in the conventional population. Both populations had QTLs for colored seed and germination traits on Pv07, and on Pv04 and Pv06 for pod fiber. There were significant QTLs for maturity traits in both (organic and conventional) HYPR populations on Pv01, Pv04, and Pv05, and unique QTLs on Pv02 and Pv07 for the conventional population. Like the ORBV populations, most of the significant QTLs for seed traits (weight, yield, color) and germination were on Pv07, and for pod fiber, QTLs were found in both populations on Pv04 and Pv05. There was a unique QTL on Pv10 for pod fiber in the conventional population. QTLs for the persistent color trait were found on Pv02. The differences in map size and recombination, and unique patterns of segregation distortion and QTLs produced from natural selection in each system supports the use of direct selection (selection within the target environment) to capture the desired alleles and traits for each system.
Resource Type
Date Issued
Degree Level
Degree Name
Degree Grantor
Commencement Year
Committee Member
Academic Affiliation
Rights Statement
Funding Statement (additional comments about funding)
  • This project was made possible by a grant from the United States Department of Agriculture National Institute of Food and Agriculture (NIFA) AFRI grant #2014-67013-22420, and by the Northern Organic Vegetable Improvement Collaborative (NOVIC) project, which is funded by the Organic Research and Extension Initiative (USDA/NIFA): NOVIC 2: 2014-51300-22223 & NOVIC 3: 2018-51300-28430.
Peer Reviewed



This work has no parents.

In Collection: