LEVERAGING QUANTITATIVE GENETICS TO BREED FOR DURABLE FUNGAL DISEASE RESISTANCE IN DRY BEAN (PHASEOLUS VULGARIS L.) By Molly Joy Irvin A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Plant Breeding, Genetics and Biotechnology - Crop and Soil Sciences - Master of Science 2023 ABSTRACT Dry beans (Phaseolus vulgaris L.) are an important legume for human consumption worldwide, providing key dietary nutrients such as carbohydrates, protein, and fiber. Michigan is the second largest dry bean producer in the United States, producing over 400 million pounds of dry edible beans per year and contributing to a farm gate value of over $139 million. However, biotic stress caused by fungal disease infection is among the main constraints that limit yield and increase management cost of dry bean production. Two of the most devastating fungal diseases in Michigan are white mold and root rot, which can cause up to 80-100% yield and seed quality decline in susceptible cultivars under heavy disease pressure. Resistant breeding lines are the ideal solution to managing these diseases as chemical control is expensive, harmful to the environment, and doesn’t ensure complete protection against infection. Unfortunately, resistance to both diseases is controlled by complex quantitative inheritance methods, with a lack of large effect resistance genes, laborious screening protocols, and historic difficulty pyramiding genes into one durably resistant phenotype. Therefore this research aimed to assist dry bean breeders in developing resistant lines by i) evaluating genomic prediction and GP + de novo GWAS as a tool for white mold (Sclerotinia sclerotiorum) resistance screening, ii) evaluating a diverse set of lines for resistance to root rots conferred by Rhizoctonia solani and Fusarium oxysporum, iii) investigating correlations between Fusarium oxysporum inoculated field and greenhouse screens and natural infection conditions and, iiii) investigating correlations between Fusarium oxysporum root rot resistance and major agronomic/ aboveground traits as an alternative non- destructive, high-throughput phenotyping method. The findings from these studies will ultimately assist dry bean breeders in the development of resistant lines for these important diseases. I would like to dedicate this work to my family: My fiancé Mark, for his constant support and humor during the challenges of graduate school and life; My parents, sister, and brother for their love and encouragement in pursuing my passion; and my grandparents for always believing in me. iii ACKNOWLEDGEMENTS I would like to express my gratitude to those who helped me achieve my master’s degree. I would like to thank my major advisor, Dr. Francisco Gomez, for giving me the opportunity to pursue this degree, and for his guidance and support throughout its completion. I would like to thank my guidance committee, Dr. Martin Chilvers, Dr. Gustavo de los Campos, and Dr. David Douches for their contributions to my research projects. I am also grateful to everyone in the dry bean breeding lab, Evan Wright, Halima Awale, Dr. Leonardo Volpato, Madison Whyte, and John Hawkins for their support. Without them, the work and the experience would be far less. iv GENERAL INTRODUCTION ....................................................................................................... 1 REFERENCES .......................................................................................................................... 15 TABLE OF CONTENTS CHAPTER ONE: POTENTIAL OF GENOMIC PREDICTION TO SCREEN FOR WHITE MOLD RESISTANCE IN DRY BEAN ....................................................................................... 25 REFERENCES .......................................................................................................................... 84 CHAPTER TWO: SCREENING DRY BEAN BREEDING LINES FOR RESISTANCE TO COMMON MICHIGAN ROOT ROT PATHOGENS ................................................................. 91 REFERENCES ........................................................................................................................ 179 v GENERAL INTRODUCTION Common bean (Phaseolus vulgaris L.) is one of the most important legumes for human consumption worldwide, providing an important source of key dietary nutrients such as carbohydrates, protein, fiber, iron, and zinc (Uebersax et al., 2023). There are two common bean markets, dry shelled seed (dry bean) and green pods (snap bean), often split into individual breeding programs (Singh & Schwartz, 2010). Dry beans are a dietary staple in Latin America and sub-Saharan Africa (Leterme & Carmenza Muũoz, 2002; Paparu et al., 2018) and an important agricultural commodity in the US (primarily North Dakota, Minnesota, and Michigan), Latin America, and other regions. Most dry bean breeding programs aim at improving yield, biotic resistance, nutritional quality, and abiotic stresses (Kelly and Cichy, 2012). However, there are many diseases that greatly affect dry bean production including anthracnose, common bacterial blight, viral mosaics, root rots, and white mold (Singh & Schwartz, 2010). There are two major gene pools within Phaseolus vulgaris resulting from two centers of domestication in the Andes and Central America (Gepts, P., and Debouck D., 1991; Mensack et al., 2010). Within these gene pools there are over 10 major market classes of dry bean with unique seed color and morphology, growth habit, disease reaction, and adaptation. Common dry bean market classes include black, navy, red, pink, yellow, pinto, great northern, kidney, and cranberry beans. The wide genetic diversity of dry bean across market classes offers an opportunity to incorporate many novel traits within a variety development program. However, the stringent requirements for seed traits within market classes along with other traits of economic importance have made genetic progress more challenging when compared to other crops. Michigan is the second largest dry bean producer in the United States, producing over 400 million pounds of dry edible beans per year and contributing to a farm gate value of over $139 1 million (USDA Crop Production Summary 2022). It also leads the nation in the black, small red, and navy bean market classes in terms of acreage. Michigan's temperate climate, with warm humid summers, is highly conducive to fungal diseases. Among the fungal diseases present in Michigan, white mold and root rots are ranked by growers as the 1st and 2nd most important diseases limiting production respectively (MBC, 2022). Agronomic management is often inefficient in controlling fungal diseases, primarily due to the durability of overwintering structures that persist in the soil and crop residue (Bolton et al., 2006; Katan, 2017). Breeding for resistance is often the most sustainable, durable, and cost- effective method to protect against infection. However, many fungal diseases (including white mold and root rots) have complex genetic inheritance methods and cumbersome phenotypic screening that limit breeding progress (Fuller et al., 1984; Hagerty et al., 2015; Miklas et al., 2004; Nakedde et al., 2016; Park et al., 2001; Schwartz & Singh, 2013; Singh et al., 2014; Wang et al., 2018). Given the complexity of breeding for resistance, more research is needed to improve screening methods, identify, and pyramid resistance genes, and produce robustly resistant varieties. WHITE MOLD White mold, conferred by the fungal pathogen Sclerotinia sclerotiorum (Lib.) de Bary is one of the most destructive fungal diseases of dry bean, resulting in poor seed quality and reduced yield (del Río et al., 2004; Vasconcellos et al., 2017). In temperate regions it is considered the most yield limiting disease of dry bean production, causing up to 100% yield loss and may also result in significant seed quality decline in susceptible cultivars (Schwartz, 2011; Singh & Schwartz, 2010). Consequently, economic losses from white mold in the United States have exceeded 200 million dollars yearly due to yield reduction and fungicide costs (Bolton et al., 2006). 2 Sclerotinia sclerotiorum, a necrotrophic fungal pathogen from the phylum Ascomycota, has a broad host range of over 400 species of plants, primarily dicotyledonous species (Boland & Hall, 1994). Sclerotinia sclerotiorum can be seed transmitted. However, the bulk of inoculum comes from melanized sclerotia that can overwinter in soil for five or more years (Schwartz & Singh, 2013). Around dry bean flowering time, S. sclerotiorum sclerotia germinates and forms a fruiting structure, a small cup shaped mushroom called an apothecia (Miklas et al., 2013). Once the apothecia has developed, asci can forcefully release ascospores into the plant canopy. These ascospores can infect the plant through wounds on the stem left by the senescent flowers and on other foliar parts, by penetrating the host cell walls (Bolton et al., 2006). Once colonized, infected plants often exhibit wilted leaves due to reduced vascular function. As disease development progresses, stems become brittle, and have a bleached appearance due to oxalic acid and other enzymes produced by the fungi, often leading to severe plant lodging (Schwartz & Singh, 2013). Yield and seed quality decline occurs indirectly due to plant stress that reduces seed development or lost pods that abort entirely and directly due to colonization of bean pods that results in moldy, discolored, or shriveled seed. In severe infections, white mycelium can be observed on affected plant parts and whole plant death can occur (Schwartz & Singh, 2013). Sclerotia form inside infected tissue in the stem pith or pods and provide inoculum for subsequent infection (Bolton et al., 2006). CONTROL Managing white mold can be challenging due to the durability and extended viability of sclerotia in the soil and broad host range (Bolton et al., 2006). White mold can be spread from field to field by infected seed, contaminated soil on farm equipment, and wind-blown ascospores (Steadman and Boland, 2005). Control of white mold using fungicides, biopesticides, and other 3 disease management strategies has been difficult. Common strategies for controlling white mold include crop rotation, fungicide sprays, limiting nitrogen fertilizer and irrigation to reduce vegetative growth, wide row spacing, low density planting, and the development of low biomass, upright, and open canopy cultivars (Ender & Kelly, 2005; Miorini et al., 2017; Schwartz & Singh, 2013). Biomass reduction and upright architecture in particular focuses on reducing sclerotial germination and ascospore development by altering the microclimate to reduce canopy humidity and temperature and therefore colonization ability (Schwartz & Singh, 2013). However, these strategies are often not sustainable or economically feasible and reduce productivity of dry bean fields. Control of white mold through biomass management, wide row spacing, and reduced fertilizer and irrigation compromises yield per acre. Additionally, fungicide sprays are expensive and have adverse effects on the environment. Furthermore, these practices do not fully eliminate white mold infection due to its virulence and ability to overwinter in the soil. HOST RESISTANCE AND SCREENING Host resistance to white mold in dry bean can be obtained through two means: physiological resistance and avoidance mechanisms, although low resistance has been found to date. Physiological resistance (Miklas et al., 2013; Schwartz & Singh, 2013) is associated with pathogen recognition, including reactive oxygen species (oxidative burst) as an initial defense mechanism and synthesis of pathogenesis-related protein (Mamidi et al., 2016). Conversely, disease avoidance mechanisms are related to plant architecture traits that confer a tall, upright growth habit, reduced lodging, and a porous plant canopy less ideal to the pathogen. Due to the complex infection method of the pathogen, a combination of both traits is considered ideal to ensure robust white mold resistant lines. 4 Multiple screening methods have been proposed, including the greenhouse straw test, cut stem oxalate test (Kolkman & Kelly, 2000), and field rating systems such as the 1-9 plot wise visual scale (Miklas et al., 2001). Previous studies have failed to show a strong correlation between field and greenhouse screening methods (Terán et al., 2009). The greenhouse trial is preferred by many researchers due to higher heritability, ease of screening, and less environmental dependance but this test does not screen for resistance conferred by avoidance mechanisms in field conditions (Chung et al., 2008; Mkwaila et al., 2011; Soule et al., 2011). Many highly resistant lines in the greenhouse exhibit low expression in field conditions. Field testing is necessary to screen for plant avoidance traits that reduce infection, but physiological resistance and architectural avoidance is confounded in field conditions (Miklas et al., 2013) False positives for physiological resistance may result due to reduced disease severity in open plant canopies (Kolkman & Kelly, 2000; Miklas et al., 2013). No completely resistant lines to white mold in dry bean have been identified to date. A multi-faceted screening and selection method that incorporates both aspects of resistance would greatly improve breeding progress for this complex trait. BREEDING FOR WHITE MOLD RESISTANCE Developing white mold resistant cultivars in dry bean has proven difficult primarily due to the complex inheritance method of resistance, low levels of innate resistance in breeding populations (Kolkman & Kelly, 2000; Terán & Singh, 2009; Vuong et al., 2004), low heritability (Fuller et al., 1984; Park et al., 2001), and variable screening methods (Ender & Kelly, 2005). Previous studies have identified that both physiological resistance and architectural avoidance to white mold is quantitatively inherited (Fuller et al., 1984; Miklas et al., 2004; Park et al., 2001). Hundreds of germplasm accessions, cultivars, and breeding lines of common bean and related 5 Phaseolus species have been screened for their reaction to white mold. Unfortunately, completely resistant varieties are unavailable and most known sources of genetic resistance to white mold are of Andean origin, usually from unadapted landraces and wild relatives, and from secondary gene pools (Terán and Singh, 2009, Singh et al., 2014; Vasconcellos et al., 2017). While low levels of resistance have been identified in small Middle-American germplasm (Ender and Kelly, 2005; Mkwaila et al., 2011; Hoyos-Villegas et al., 2015), progress to develop white mold resistant cultivars has been hindered by difficulty in pyramiding resistance genes given the quantitative inheritance of disease avoidance mechanisms and physiological resistance (Singh et al., 2014). Over 35 quantitative trait loci (QTL) with minor effects have been identified conferring resistance to white mold in dry bean by either physiological resistance or architectural avoidance which highlights the complexity of this trait (Singh et al., 2014; Vasconcellos et al., 2017). QTL conferring physiological avoidance and architectural traits have been mapped to most dry bean linkage groups with common traits being oxalate resistance, determinate growth habit, increased internode length, plant height, days to flowering, branching pattern, lodging, seed size, and yield (Ender & Kelly, 2005; Miklas, 2007; Miklas et al., 2001, 2003, 2013; Mkwaila et al., 2011; Soule et al., 2011). Miklas et al. (2007) identified two QTL located on PV2 and PV3 conditioning physiological resistance associated with the stay green stem characteristic and disease avoidance traits including late maturity and the stay-green stem characteristic, which are undesirable agronomic traits for dry bean production (Miklas et al., 2006). These traits are important to account for when screening for resistance to white mold because while they confirm a resistant genotype and can be used as a parent for breeding, they are deleterious traits for dry bean production. Recently, Vasconcellos et al. 2017 compiled 37 individual QTL across 14 recombinant inbred bi- parental populations developed previously and condensed these QTL into 17 meta-QTL found on 6 chromosomes 1,2,3,5,6,7, and 8. These meta QTL in particular are potential targets for marker assisted selection (MAS) of partial resistance to white mold. Unlike breeding for qualitative resistance, breeding for quantitative resistance is more challenging because it requires multiple cycles of breeding and screening, leading to a gradual improvement of resistance in a breeding population over time. With the sheer number of QTL identified previously conferring many environmentally dependent resistance traits, it is no surprise that there has been difficulty pyramiding genes into one robustly resistant genotype with the desired agronomic traits. An integrated screening and recurrent selection method that considers all aspects of resistance and agronomic performance would greatly assist future cultivar development. Multiple breeding tools such as recurrent selection and the use of alternative populations such as multiparent intercrosses are one way to assist the identification and pyramiding of resistance genes (Escobar et al., 2022; Osorno et al., 2018). Recently, Escobar et al. explored integrated multi-parent crosses and gamete selection using a Multiparent Advanced Generation Inter-Cross (MAGIC) population, to facilitate mapping and breeding efforts, resulting in multiple partially resistant lines. This method crosses together multiple founder lines and cycles through several additional generations of crossing, resulting in offspring with multiple recombination events, often leading to improved results that maximize diversity. This and other emerging breeding tools will assist future breeding for complex traits. GENOMIC PREDICTION Genomic prediction creates a unique opportunity to select and pyramid major and minor alleles conferring resistance to complex diseases (Merrick et al., 2021; Poland & Rutkoski, 2016; Tiede & Smith, 2018). This established tool in Marker Assisted Selection (MAS) utilizes genome wide markers and phenotypic data to train a linear statistical model to predict and make selections 7 based on genomic estimated breeding values (GEBVs) for a trait. In contrast to MAS, genomic prediction does not identify significant markers and estimate individual marker effects. Instead, all marker effects are estimated simultaneously and used to accurately predict overall breeding values for a trait (Jannink et al., 2010). When properly implemented, genomic prediction has the potential to improve breeding program efficiency and reduce phenotyping costs when screening for complex traits. Genomic prediction has assisted breeding efforts for quantitative disease and pest resistance traits in many crop species including wheat: (Arruda et al., 2016; Juliana et al., 2017, 2022; Larkin et al., 2021; Merrick et al., 2021; Odilbekov et al., 2019; Rutkoski et al., 2012; Sarinelli et al., 2019) , dry bean (Diaz et al., 2021; Shi et al., 2021a), soybean (Bao et al., 2015; de Azevedo Peixoto et al., 2017; Đorđević et al., 2019; Duhnen et al., 2017; Hemingway et al., 2021; Shi et al., 2021b; Wen et al., 2018), and other crop species. Specifically, when compared to MAS or phenomic selection alone, genomic prediction has shown to be more effective for pyramiding quantitative traits controlled by many small effect QTL (Heffner et al., 2010, 2011; Massman et al., 2013; Merrick et al., 2021; Zhang et al., 2016). Many previous studies of the genetic control of white mold in dry bean fail to account for the many small effect QTL unable to be detected by genome wide association studies (GWAS) that are likely controlling resistance to this trait. To date, there have been no major resistance genes identified for white mold resistance, unlike some disease traits such as Fusarium head blight in wheat (Larkin et al., 2021). Since genomic prediction utilizes all markers spread across the entire genome to make selections, it has the power to detect small effect QTL conferring resistance that would otherwise be overlooked by GWAS (Jannink et al. 2010; Meuwissen et al. 2013). Genomic selection for white mold resistance has been studied in soybean diversity panels (de Azevedo Peixoto et al., 2017; Wen et al., 2018) with moderate to high prediction accuracies (0.4-0.7) 8 suggesting that this warrants further research into the validity to assist breeding for resistance in dry bean. Another benefit of genomic prediction in breeding for complex traits controlled by many major and minor QTL is that major QTL controlling a trait of interest can be identified using a GWAS and implemented in genomic prediction as fixed effects, allowing for further accuracy in selection (Merrick et al., 2021). Fixed QTL can be selected based on previous research or identified through de novo GWAS of the training population. GS + de novo GWAS involves two main stages where in the first stage GWAS is conducted on individuals in the training set to identify fixed QTL to be implemented in genomic prediction of the testing set (Bian & Holland, 2017; Haile et al., 2021; Sarinelli et al., 2019; Spindel et al., 2016). For example, Bian and Holland 2017 and Spindel 2016 both found that GS + de novo GWAS outperformed all other models tested. A simulation study by (Rice & Lipka, 2019) evaluated the effect of the number of QTL implemented across multiple genetic architectures and trait heritability and observed increased, decreased, and mixed (increase then decrease or decrease then increase) effect on prediction accuracy depending on the trait and number of fixed QTL. This study emphasizes that a varying number of fixed effect QTL should be tested for each trait before implementation. ROOT ROTS Root rot is conferred by a soilborne disease complex of multiple fungal and oomycete pathogens including members of Fusarium sp., Rhizoctonia sp., Pythium sp., Macrophomina sp., and Alternaria sp. (Bilgi et al., 2011; Harveson et al., 2005; Schwartz, 2011; Sendi et al., 2020; Singh & Schwartz, 2010). Pathogen composition is variable by region and usually involves multiple species. In Michigan, the most common root rot pathogens belong to the Fusarium and Rhizoctonia species as well as multiple oomycetes (Jacobs et al., 2019). Root rots conferred by 9 these pathogens cause up to 84% yield loss in susceptible dry bean cultivars through damage of root biomass, reducing vigor, and sometimes whole plant death (Jacobs et al., 2019). Root rots greatly affect overall plant health because a damaged root system limits nutrient and water intake, greatly affecting yield potential. The life cycle of root rot complex pathogens are similar. First, primary inoculum (resting spores, sclerotia, or mycelium) in the soil or crop residue germinates and infects seedling roots. The pathogen spreads from plant to plant using secondary spores (conidia or sporangia) or mycelium. Finally, it produces new primarily inoculum at the end of the season that overwinters until the next growing season (Gossen et al., 2016). Root rot disease symptoms vary depending on the pathogen(s) involved, but are primarily characterized by root lesions, root and foliar biomass loss, and reduced stand. Fusarium and Rhizoctonia root rots of dry bean are both primarily characterized by water-soaked, dark brown to rust colored lesions on the root and death of lateral roots (Hall et al. 1991, Hagedorn et al. 1994). Fusarium root rot develops slowly in some genotypes and tends to lead to an overall root biomass loss, vigor loss, and yield decline. Entire plant death is rarer and usually occurs later in the season in only severe infections. Aboveground symptoms are difficult to see, but plants may be stunted and yellowed, exhibit premature leaf drop and have poor pod fill (Schwartz, 2011 ). Conversely, Rhizoctonia root rot tends to affect plants early in the season primarily through reduced stand resulting from seed rot, seedling blight, and pre/post emergence damping off (Gossen et al., 2016). Rhizoctonia root rot also tends to result in the formation of deeper sunken lesions at the soil level (Conner et al., 2014). Root rot disease severity is highly variable with environmental conditions. In particular, cool wet weather, soil compaction, or flooding events can increase infection severity in both infection scenarios due to additional plant stress and increased virulence of soilborne pathogens (Kumar and Kudada, 2018, Cichy, 2007). 10 CONTROL Current agronomic management practices assist with lowering root rot disease severity but are not sufficient for the complete control of root rot. Fungicidal seed and soil treatments, reduced irrigation, crop rotation, cover crops, seedbed preparation, and other agronomic practices are currently used to combat yield loss from root rots (Abawi and Pastor Corrales, 1990; Gossen et al., 2016; Harveson et al., 2005; Rubiales et al., 2015). A 3-5 year crop rotation with a non-host crop (alfalfa, barley, wheat, oats, and corn) is recommended for fields with root rot disease pressure to reduce inoculum load (Schwartz, 2011). If possible, planting in warm moist soil is ideal for quick germination and emergence. High-quality treated, certified seed is also recommended to maximize plant vigor and stand. One major limitation of many root rot control methods is that they are often not sustainable or economically feasible and reduce productivity of dry bean fields. For example, control of root rots through reduced irrigation and fertilizer limits yield per acre. Fungicidal seed and soil treatments are expensive and have adverse effects on the environment. Furthermore, these practices are not complete guarantees against root rot infection due to its virulence and ability for thick-walled spores, hyphae, or sclerotia to overwinter in the soil for multiple years (Schwartz, 2011). The development of resistant dry bean cultivars would ensure high levels of protection against this disease. Therefore, the most sustainable, durable, and cost-effective method to ensure protection against infection would be to develop resistant dry bean cultivars. HOST RESISTANCE AND SCREENING Resistance to root rot in dry bean appears to be a combination of physiological mechanisms and root system avoidance due to architecture traits such as root dry weight, root length, and root mass (Snapp et al., 2003; Kamfwa et al., 2013; Wang et al., 2018; Haus et al., 2020). These root 11 traits in particular are identified as targets for the screening and development of root rot resistant varieties. Previous research has established susceptibility to root rot in both dry bean gene pools, but susceptibility to individual fungal and oomycete pathogens differ. Large-seeded dry bean cultivars from the Andean gene pool are generally more susceptible to Fusarium species and small seeded cultivars from the Middle-American gene pool are generally more susceptible to Rhizoctonia species, with Andean types being more susceptible to root rots overall (Conner et al., 2014; Schneider et al., 2001). Previous research has also established the importance of high root biomass cultivars with high density of lateral roots, high basal root number, and many adventitious roots (Haus et al., 2020; Román-Avilés et al., 2004; Snapp et al., 2003). Andean genotypes in particular, tend to have less robust root systems which is thought to be why they have lower levels of resistance to root rots (Cichy et al., 2007; Román-Avilés & Kelly, 2005; Schneider et al., 2001). There have been multiple phenotyping methods proposed for the greenhouse and field evaluation of Fusarium root rots including the liquid inoculum method (Schneider & Kelly, 2000), inoculum layer method (Chaudhary et al., 2006), and nutrient culture (Boomstra et al., 1977). Inoculated grain planted alongside the crop seed is a common method for field evaluation (Haus et al., 2020; Pandey et al., 2020; Wang et al., 2018). The presence of multiple resistance mechanisms through both physiological resistance and architectural avoidance complicates screening. Greenhouse trials are preferred by researchers because they provide a controlled environment free of natural root rot pathogens, have higher heritability, and ease of screening, but they do not screen for some avoidance mechanisms that can only be observed in the field. Conversely, field screens allow for measurement of both architectural traits and physiological resistance, but are often hampered by low heritability, high coefficient of variation (CV), high error variances, and natural pathogen presence (Guzman, 2016.; Hagerty et al., 2015; Nakedde et al., 2016; Román-Avilés et al., 2004; 12 Román-Avilés & Kelly, 2005). Furthermore, these destructive phenotyping methods require the whole plant to be dug up with a shovel in the field or removed from the pot in the greenhouse to be evaluated for root rot symptoms, which is laborious and prevents additional measurements of yield and other traits throughout the season. Establishment of aboveground traits related to root rot disease severity would enable high-throughput methods of phenotyping such as UAS imaging to assist breeders in screening for root rot resistant varieties (Guo et al., 2021; Lu et al., 2019; Manganiello et al., 2021; Marzougui et al., 2019). BREEDING FOR ROOT ROT RESISTANCE Genetic resistance to root rots in dry bean is quantitatively inherited, primarily through root architecture traits such as root weight, root length and root mass (Haus et al., 2020; Kamfwa et al., 2013; Snapp et al., 2003; Wang et al., 2018) and physiological avoidance (Kamfwa et al., 2013; Mukankusi et al., 2011; Román-Avilés et al., 2004; Snapp et al., 2003). The quantitative inheritance of these traits has been demonstrated by the low narrow-sense heritability estimates (26%-44%) reported and many small effect quantitative trait loci (QTL) found for root rot (Román- Avilés et al., 2011; Nakedde et al., 2016; Kamfwa et al., 2018; Wang et al., 2018; Zitnick-Anderson et al., 2020). Previous research has established the importance of high root biomass cultivars with high density of lateral roots, high basal root number, and many adventitious roots for root rot avoidance (Haus et al., 2020; Román-Avilés et al., 2004; Snapp et al., 2003). These traits have been identified as significant targets for improving resistance. Resistance to root rot is primarily found in the small-seeded Middle-American bean germplasm compared to the highly susceptible large-seeded beans of Andean origin and has served as the only source of resistance (Cichy et al., 2007; Mukankusi et al., 2011; Román-Avilés & Kelly, 2005; Schneider et al., 2001) 13 Although few studies to date have evaluated genetic resistance to Fusarium oxysporum or Rhizoctonia solani, multiple studies have established genomic regions associated with resistance to root rots conferred by other Fusarium sp. (Hagerty et al., 2015; Kamfwa et al., 2013; Nakedde et al., 2016; Román-Avilés & Kelly, 2005; Schneider et al., 2001; Wang et al., 2018; Zitnick- Anderson et al., 2020). These studies often identify many different QTL which highlights the highly quantitative inheritance method of resistance. QTL studies of resistance to Fusarium sp. in dry bean have identified genomic regions associated with root traits, plant immune/defense mechanisms, and other disease resistance genes (Hagerty et al., 2015; Nakedde et al., 2016; Wang et al., 2018; Zitnick-Anderson et al., 2020). One study to date has evaluated genetic resistance to Rhizoctonia solani- Oladzad et al. 2019. In this study, researchers evaluated the Andean and Middle American diversity panels and observed multiple QTL associated with gene models encoding proteins like known disease resistance genes. Other Rhizoctonia solani resistance studies have focused on screening and ranking breeding lines and establishing trait correlations (Adesemoye et al., 2018; Conner et al., 2014; Peña et al., 2013). 14 REFERENCES Abawi GS, Pastor-Corrales MA (1990) Root rots of beans in Latin America and Africa: diagnosis, research methodologies, and management strategies. Cali Colombia. CIAT. Adesemoye, A. O., Orrell, T., & Kodati, S. (2018). Effect of Virulence of Root Rot Pathogens and Cultivar Resistance on Disease Occurrence in Dry Beans. Plant Health Progress, 19(3), 237–241. https://doi.org/10.1094/PHP-06-18-0034-RS Arruda, M. P., Lipka, A. E., Brown, P. J., Krill, A. M., Thurber, C., Brown-Guedira, G., Dong, Y., Foresman, B. J., & Kolb, F. L. (2016). Comparing genomic selection and marker- assisted selection for Fusarium head blight resistance in wheat (Triticum aestivum L.). Molecular Breeding, 36(7), 84. https://doi.org/10.1007/s11032-016-0508-5 Bao, Y., Kurle, J. E., Anderson, G., & Young, N. D. (2015). Association mapping and genomic prediction for resistance to sudden death syndrome in early maturing soybean germplasm. Molecular Breeding, 35(6), 128. https://doi.org/10.1007/s11032-015-0324-3 Bian, Y., & Holland, J. B. (2017). Enhancing genomic prediction with genome-wide association studies in multiparental maize populations. Heredity, 118(6), 585–593. https://doi.org/10.1038/hdy.2017.4 Bilgi, V. N., Bradley, C. A., Mathew, F. M., Ali, S., & Rasmussen, J. B. (2011). Root Rot of Dry Edible Bean Caused by Fusarium graminearum. Plant Health Progress, 12(1), 14. https://doi.org/10.1094/PHP-2011-0425-01-RS Boland, G. J., & Hall, R. (1994). Index of plant hosts of Sclerotinia sclerotiorum. Canadian Journal of Plant Pathology, 16(2), 93–108. https://doi.org/10.1080/07060669409500766 Bolton, M. D., Thomma, B. P. H. J., & Nelson, B. D. (2006). Sclerotinia sclerotiorum (Lib.) de Bary: Biology and molecular traits of a cosmopolitan pathogen. Molecular Plant Pathology, 7(1), 1–16. https://doi.org/10.1111/j.1364-3703.2005.00316.x Boomstra, A. G., Bliss, F. A., & Beebe, S. E. (1977). New Sources of Fusarium Root Rot Resistance in Phaseolus vulgaris L.1. Journal of the American Society for Horticultural Science, 102(2), 182-185. Retrieved Sep 24, 2023, from https://doi.org/10.21273/JASHS.102.2.182 Chaudhary S, Anderson TR, Park SJ, Yu K (2006) Comparison of screening methods or resistance to fusarium root rot in common beans. J Phytopath 154:303–308. https://doi.org/10.1111/j.1439- 0434.2006.01097.x Chung, Y. S., Sass, M. E. and Nienhuis, J. 2008. Validation of RAPD markers for white mold resistance in two snap bean populations based on field and greenhouse evaluations. Crop Sci. 48: 22652273 15 Cichy, K. A., Snapp, S. S., & Kirk, W. W. (2007). Fusarium root rot incidence and root system architecture in grafted common bean lines. Plant and Soil, 300(1–2), 233–244. https://doi.org/10.1007/s11104-007-9408-0 Conner, R. L., Hou, A., Balasubramanian, P., McLaren, D. L., Henriquez, M. A., Chang, K.-F., & McRae, K. B. (2014). Reaction of dry bean cultivars grown in western Canada to root rot inoculation. Canadian Journal of Plant Science, 94(7), 1219–1230. https://doi.org/10.4141/cjps2013-416 de Azevedo Peixoto, L., Moellers, T. C., Zhang, J., Lorenz, A. J., Bhering, L. L., Beavis, W. D., & Singh, A. K. (2017). Leveraging genomic prediction to scan germplasm collection for crop improvement. PLOS ONE, 12(6), e0179191. https://doi.org/10.1371/journal.pone.0179191 del Río, L. E., Venette, J. R., & Lamey, H. A. (2004). Impact of White Mold Incidence on Dry Bean Yield Under Nonirrigated Conditions. Plant Disease, 88(12), 1352–1356. https://doi.org/10.1094/PDIS.2004.88.12.1352 Diaz, L. M., Arredondo, V., Ariza-Suarez, D., Aparicio, J., Buendia, H. F., Cajiao, C., Mosquera, G., Beebe, S. E., Mukankusi, C. M., & Raatz, B. (2021). Genetic Analyses and Genomic Predictions of Root Rot Resistance in Common Bean Across Trials and Populations. Frontiers in Plant Science, 12, 629221. https://doi.org/10.3389/fpls.2021.629221 Đorđević, V., Ćeran, M., Miladinović, J., Balešević-Tubić, S., Petrović, K., Miladinov, Z., & Marinković, J. (2019). Exploring the performance of genomic prediction models for soybean yield using different validation approaches. Molecular Breeding, 39(5), 74. https://doi.org/10.1007/s11032-019-0983-6 Duhnen, A., Gras, A., Teyssèdre, S., Romestant, M., Claustres, B., Daydé, J., & Mangin, B. (2017). Genomic Selection for Yield and Seed Protein Content in Soybean: A Study of Breeding Program Data and Assessment of Prediction Accuracy. Crop Science, 57(3), 1325–1337. https://doi.org/10.2135/cropsci2016.06.0496 Ender, M., & Kelly, J. D. (2005). Identification of QTL Associated with White Mold Resistance in Common Bean. Crop Science, 45(6), 2482–2490. https://doi.org/10.2135/cropsci2005.0064 Escobar, E., Oladzad, A., Simons, K., Miklas, P., Lee, R. K., Schroder, S., Bandillo, N., Wunsch, M., McClean, P. E., & Osorno, J. M. (2022). New genomic regions associated with white mold resistance in dry bean using a MAGIC population. The Plant Genome, 15(1). https://doi.org/10.1002/tpg2.20190 Fuller, P. A., Coyne, D. P., & Steadman, J. R. (1984). Inheritance of Resistance to White Mold Disease in a Diallel Cross of Dry Beans 1. Crop Science, 24(5), 929–933. https://doi.org/10.2135/cropsci1984.0011183X002400050025x 16 Gepts, P., & Debouck, D. (1991). Origin, domestication, and evolution of the common bean (Phaseolus vulgaris L.). Common beans: research for crop improvement, 7-53. Gossen, B. D., Conner, R. L., Chang, K.-F., Pasche, J. S., McLaren, D. L., Henriquez, M. A., Chatterton, S., & Hwang, S.-F. (2016). Identifying and Managing Root Rot of Pulses on the Northern Great Plains. Plant Disease, 100(10), 1965–1978. https://doi.org/10.1094/PDIS-02-16-0184-FE Guo, W., Carroll, M. E., Singh, A., Swetnam, T. L., Merchant, N., Sarkar, S., Singh, A. K., & Ganapathysubramanian, B. (2021). UAS-Based Plant Phenotyping for Research and Breeding Applications. Plant Phenomics, 2021, 1–21. https://doi.org/10.34133/2021/9840192 Guzman, J. E. V. (2016). A Dissertation Submitted to the Graduate Faculty of the North Dakota State University of Agriculture and Applied Sciences. 86. Hagedorn, D. J. (1994). “Rhizoctonia root rot,” in Compendium of Bean Disease, ed. R. Hall (St Paul, MN: American Phytopathological Society Press), 9–13. Hagerty, C. H., Cuesta‐Marcos, A., Cregan, P. B., Song, Q., McClean, P., Noffsinger, S., & Myers, J. R. (2015). Mapping Fusarium solani and Aphanomyces euteiches Root Rot Resistance and Root Architecture Quantitative Trait Loci in Common Bean. Crop Science, 55(5), 1969–1977. https://doi.org/10.2135/cropsci2014.11.0805 Haile, T. A., Walkowiak, S., N’Diaye, A., Clarke, J. M., Hucl, P. J., Cuthbert, R. D., Knox, R. E., & Pozniak, C. J. (2021). Genomic prediction of agronomic traits in wheat using different models and cross-validation designs. Theoretical and Applied Genetics, 134(1), 381–398. https://doi.org/10.1007/s00122-020-03703-z Hall, R. 1991. Fusarium root rot. In: R. Hall, editor, Compendium of bean diseases. Am. Phytopathol. Soc. Press, St. Paul, MN. p. 9–10. Harveson, R. M., Smith, J. A., & Stroup, W. W. (2005). Improving Root Health and Yield of Dry Beans in the Nebraska Panhandle with a New Technique for Reducing Soil Compaction. Plant Disease, 89(3), 279–284. https://doi.org/10.1094/PD-89-0279 Haus, M. J., Wang, W., Jacobs, J. L., Peplinski, H., Chilvers, M. I., Buell, C. R., & Cichy, K. (2020). Root Crown Response to Fungal Root Rot in Phaseolus vulgaris Middle American × Andean Lines. Plant Disease, 104(12), 3135–3142. https://doi.org/10.1094/PDIS-05-20-0956-RE Heffner, E. L., Jannink, J., & Sorrells, M. E. (2011). Genomic Selection Accuracy using Multifamily Prediction Models in a Wheat Breeding Program. The Plant Genome, 4(1), plantgenome2010.12.0029. https://doi.org/10.3835/plantgenome2010.12.0029 17 Heffner, E. L., Lorenz, A. J., Jannink, J., & Sorrells, M. E. (2010). Plant Breeding with Genomic Selection: Gain per Unit Time and Cost. Crop Science, 50(5), 1681–1690. https://doi.org/10.2135/cropsci2009.11.0662 Hemingway, J., Schnebly, S. R., & Rajcan, I. (2021). Accuracy of genomic prediction for seed oil concentration in high‐oleic soybean populations using a low‐density marker panel. Crop Science, 61(6), 4012–4021. https://doi.org/10.1002/csc2.20607 Hoyos‐Villegas, V., Mkwaila, W., Cregan, P. B., & Kelly, J. D. (2015). Quantitative Trait Loci Analysis of White Mold Avoidance in Pinto Bean. Crop Science, 55(5), 2116–2129. https://doi.org/10.2135/cropsci2015.02.0106 Jacobs, J. L., Kelly, J. D., Wright, E. M., Varner, G., & Chilvers, M. I. (2019). Determining the Soilborne Pathogens Associated with Root Rot Disease Complex of Dry Bean in Michigan. Plant Health Progress, 20(2), 122–127. https://doi.org/10.1094/PHP-11-18- 0076-S Jannink, J.-L., Lorenz, A. J., & Iwata, H. (2010). Genomic selection in plant breeding: From theory to practice. Briefings in Functional Genomics, 9(2), 166–177. https://doi.org/10.1093/bfgp/elq001 Juliana, P., He, X., Poland, J., Roy, K. K., Malaker, P. K., Mishra, V. K., Chand, R., Shrestha, S., Kumar, U., Roy, C., Gahtyari, N. C., Joshi, A. K., Singh, R. P., & Singh, P. K. (2022). Genomic selection for spot blotch in bread wheat breeding panels, full-sibs and half-sibs and index-based selection for spot blotch, heading and plant height. Theoretical and Applied Genetics, 135(6), 1965–1983. https://doi.org/10.1007/s00122-022-04087-y Juliana, P., Singh, R. P., Singh, P. K., Crossa, J., Rutkoski, J. E., Poland, J. A., Bergstrom, G. C., & Sorrells, M. E. (2017). Comparison of Models and Whole‐Genome Profiling Approaches for Genomic‐Enabled Prediction of Septoria Tritici Blotch, Stagonospora Nodorum Blotch, and Tan Spot Resistance in Wheat. The Plant Genome, 10(2). https://doi.org/10.3835/plantgenome2016.08.0082 Kamfwa, K., Mwala, M., Okori, P., Gibson, P., & Mukankusi, C. (2013). Identification of QTL for Fusarium Root Rot Resistance in Common Bean. Journal of Crop Improvement, 27(4), 406–418. https://doi.org/10.1080/15427528.2013.786775 Katan, J. (2017). Diseases caused by soilborne pathogens: biology, management, and challenges. J. Plant Pathol. 99, 305–315. Kelly, J.D., & Cichy, K.A. (2012). Dry Bean Breeding and Production Technologies. Dry Beans and Pulses. Kolkman, J. M., & Kelly, J. D. (2000). An indirect test using oxalate to determine physiological resistance to white mold in common bean. Crop Science, 40(1), 281–285. https://doi.org/10.2135/cropsci2000.401281x 18 Kumar, M.S., & Kudada, N. (2018). Effect of sowing dates and weather conditions on Rhizoctonia root rot disease incidence and green pod yield of French bean. Journal of Pharmacognosy and Phytochemistry, 7, 893-899. Larkin, D. L., Mason, R. E., Moon, D. E., Holder, A. L., Ward, B. P., & Brown-Guedira, G. (2021). Predicting Fusarium Head Blight Resistance for Advanced Trials in a Soft Red Winter Wheat Breeding Program With Genomic Selection. Frontiers in Plant Science, 12, 715314. https://doi.org/10.3389/fpls.2021.715314 Leterme, P., & Carmenza Muũoz, L. (2002). Factors influencing pulse consumption in Latin America. British Journal of Nutrition, 88(S3), 251-254. doi:10.1079/BJN/2002714 Lu, N., Zhou, J., Han, Z. et al. Improved estimation of aboveground biomass in wheat from RGB imagery and point cloud data acquired with a low-cost unmanned aerial vehicle system. Plant Methods 15, 17 (2019). https://doi.org/10.1186/s13007-019-0402-3 Mamidi, S., Miklas, P. N., Trapp, J., Felicetti, E., Grimwood, J., Schmutz, J., Lee, R., & McClean, P. E. (2016). Sequence‐Based Introgression Mapping Identifies Candidate White Mold Tolerance Genes in Common Bean. The Plant Genome, 9(2). https://doi.org/10.3835/plantgenome2015.09.0092 Manganiello, G.; Nicastro, N.; Caputo, M.; Zaccardelli, M.; Cardi, T.; Pane, C. Functional Hyperspectral Imaging by High Related Vegetation Indices to Track the Wide-Spectrum Trichoderma Biocontrol Activity Against Soil-Borne Diseases of Baby-Leaf Vegetables. Front. Plant. Sci. 2021, 12, 630059. Massman, J. M., Jung, H. G., & Bernardo, R. (2013). Genomewide Selection versus Marker‐ assisted Recurrent Selection to Improve Grain Yield and Stover‐quality Traits for Cellulosic Ethanol in Maize. Crop Science, 53(1), 58–66. https://doi.org/10.2135/cropsci2012.02.0112 Marzougui, A., Ma, Y., Zhang, C., McGee, R. J., Coyne, C. J., Main, D., & Sankaran, S. (2019). Advanced Imaging for Quantitative Evaluation of Aphanomyces Root Rot Resistance in Lentil. Frontiers in Plant Science, 10, 383. https://doi.org/10.3389/fpls.2019.00383 MBC 2022. Michigan Bean Commission Production Practices Survey. 2022. https://michiganbean.com/https-michiganbean-com-wp-content-uploads-2022- production-practices/ Mensack, M. M., Fitzgerald, V. K., Ryan, E. P., Lewis, M. R., Thompson, H. J., & Brick, M. A. (2010). Evaluation of diversity among common beans (Phaseolus vulgaris L.) from two centers of domestication using “omics” technologies. BMC Genomics, 11(1), 686. https://doi.org/10.1186/1471-2164-11-686 19 Merrick, L. F., Burke, A. B., Chen, X., & Carter, A. H. (2021). Breeding With Major and Minor Genes: Genomic Selection for Quantitative Disease Resistance. Frontiers in Plant Science, 12, 713667. https://doi.org/10.3389/fpls.2021.713667 Meuwissen, T., Hayes, B., Goddard, M., 2013. Accelerating improvement of livestock with genomic selection. Annu. Rev. Anim. Biosci. 1 (1), 221–237. https://doi.org/ 10.1146/annurev-animal-031412-103705. Miklas, P. N., Hauf, D. C., Henson, R. A., & Grafton, K. F. (2004). Inheritance of ICA Bunsi‐ Derived Resistance to White Mold in a Navy × Pinto Bean Cross. Crop Science, 44(5), 1584–1588. https://doi.org/10.2135/cropsci2004.1584 Miklas, P. N., Johnson, W. C., Delorme, R., & Gepts, P. (2001). QTL Conditioning Physiological Resistance and Avoidance to White Mold in Dry Bean. Crop Science, 41(2), 309–315. https://doi.org/10.2135/cropsci2001.412309x Miklas, P.N., Kelly, J.D., Beebe, S.E. et al. Common bean breeding for resistance against biotic and abiotic stresses: From classical to MAS breeding. Euphytica 147, 105–131 (2006). https://doi.org/10.1007/s10681-006-4600-5 Miklas, P. N., Porter, L. D., Kelly, J. D., & Myers, J. R. (2013). Characterization of white mold disease avoidance in common bean. European Journal of Plant Pathology, 135(3), 525– 543. https://doi.org/10.1007/s10658-012-0153-8 Miorini, T. J. J., Raetano, C. G., & Everhart, S. E. (2017). Control of white mold of dry bean and residual activity of fungicides applied by chemigation. Crop Protection, 94, 192–202. https://doi.org/10.1016/j.cropro.2016.12.023 Mkwaila, W., Terpstra, K. A., Ender, M., & Kelly, J. D. (2011). Identification of QTL for agronomic traits and resistance to white mold in wild and landrace germplasm of common bean: QTL for resistance to white mold in bean germplasm. Plant Breeding, 130(6), 665–672. https://doi.org/10.1111/j.1439-0523.2011.01876.x Mukankusi, C. M., Melis, R. J., Derera, J., Buruchara, R. A., & Mark, D. (2011). A screening technique for resistance to Fusarium root rot of common bean. Afr. J. Plant Sci. Nakedde, T., Ibarra-Perez, F. J., Mukankusi, C., Waines, J. G., & Kelly, J. D. (2016). Mapping of QTL associated with Fusarium root rot resistance and root architecture traits in black beans. Euphytica, 212(1), 51–63. https://doi.org/10.1007/s10681-016-1755-6 Osorno, J. M., McClean, P. E., North Dakota State University, USA, Close, T., & University of California (Riverside), USA. (2018). Advanced breeding techniques for grain legumes in the genomics era. In formerly International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), India & S. Sivasankar (Eds.), Burleigh Dodds Series in Agricultural Science (pp. 57–86). Burleigh Dodds Science Publishing. https://doi.org/10.19103/AS.2017.0023.03 20 Odilbekov, F., Armoniené, R., Koc, A., Svensson, J., & Chawade, A. (2019). GWAS-Assisted Genomic Prediction to Predict Resistance to Septoria Tritici Blotch in Nordic Winter Wheat at Seedling Stage. Frontiers in Genetics, 10, 1224. https://doi.org/10.3389/fgene.2019.01224 Oladzad, A., Zitnick-Anderson, K., Jain, S., Simons, K., Osorno, J. M., McClean, P. E., & Pasche, J. S. (2019). Genotypes and Genomic Regions Associated With Rhizoctonia solani Resistance in Common Bean. Frontiers in Plant Science, 10, 956. https://doi.org/10.3389/fpls.2019.00956 Pandey, A. K., Burlakoti, R. R., Rathore, A., & Nair, R. M. (2020). Morphological and molecular characterization of Macrophomina phaseolina isolated from three legume crops and evaluation of mungbean genotypes for resistance to dry root rot. Crop Protection, 127, 104962. https://doi.org/10.1016/ j.cropro.2019.104962 Paparu, P., Acur, A., Kato, F., Acam, C., Nakibuule, J., Musoke, S., Nkalubo, S., & Mukankusi, C. (2018). PREVALENCE AND INCIDENCE OF FOUR COMMON BEAN ROOT ROTS IN UGANDA. Experimental Agriculture, 54(6), 888–900. https://doi.org/10.1017/S0014479717000461 Park, S. O., Coyne, D. P., Steadman, J. R., & Skroch, P. W. (2001). Mapping of QTL for Resistance to White Mold Disease in Common Bean. Crop Science, 41(4), 1253–1262. https://doi.org/10.2135/cropsci2001.4141253x Peña, P. A., Steadman, J. R., Eskridge, K. M., & Urrea, C. A. (2013). Identification of sources of resistance to damping-off and early root/hypocotyl damage from Rhizoctonia solani in common bean (Phaseolus vulgaris L.). Crop Protection, 54, 92–99. https://doi.org/10.1016/j.cropro.2013.04.014 Poland, J., & Rutkoski, J. (2016). Advances and Challenges in Genomic Selection for Disease Resistance. Annual Review of Phytopathology, 54(1), 79–98. https://doi.org/10.1146/annurev-phyto-080615-100056 Rice, B., & Lipka, A. E. (2019). Evaluation of RR‐BLUP Genomic Selection Models that Incorporate Peak Genome‐Wide Association Study Signals in Maize and Sorghum. The Plant Genome, 12(1), 180052. https://doi.org/10.3835/plantgenome2018.07.0052 Román-Avilés, B., & Kelly, J. D. (2005). Identification of Quantitative Trait Loci Conditioning Resistance to Fusarium Root Rot in Common Bean. Crop Science, 45(5), 1881–1890. https://doi.org/10.2135/cropsci2005.0028 Román-Avilés, B., Snapp, S. S., & Kelly, J. D. (2004). Assessing root traits associated with root rot resistance in common bean. Field Crops Research, 86(2–3), 147–156. https://doi.org/10.1016/j.fcr.2003.08.001 21 Rubiales, D., Fondevilla, S., Chen, W., Gentzbittel, L., Higgins, T. J. V., Castillejo, M. A., Singh, K. B., & Rispail, N. (2015). Achievements and Challenges in Legume Breeding for Pest and Disease Resistance. Critical Reviews in Plant Sciences, 34(1–3), 195–236. https://doi.org/10.1080/07352689.2014.898445 Rutkoski, J., Benson, J., Jia, Y., Brown-Guedira, G., Jannink, J.-L., & Sorrells, M. (2012). Evaluation of Genomic Prediction Methods for Fusarium Head Blight Resistance in Wheat. The Plant Genome, 5(2), 51–61. https://doi.org/10.3835/plantgenome2012.02.0001 Sarinelli, J. M., Murphy, J. P., Tyagi, P., Holland, J. B., Johnson, J. W., Mergoum, M., Mason, R. E., Babar, A., Harrison, S., Sutton, R., Griffey, C. A., & Brown-Guedira, G. (2019). Training population selection and use of fixed effects to optimize genomic predictions in a historical USA winter wheat panel. Theoretical and Applied Genetics, 132(4), 1247– 1261. https://doi.org/10.1007/s00122-019-03276-6 Schneider, K. A., Grafton, K. F., & Kelly, J. D. (2001). QTL Analysis of Resistance to Fusarium Root Rot in Bean. Crop Science, 41(2), 535–542. https://doi.org/10.2135/cropsci2001.412535x Schneider, K. A., & Kelly, J. D. (2000). A Greenhouse Screening Protocol for Fusarium Root Rot in Bean. HortScience, 35(6), 1095–1098. https://doi.org/10.21273/HORTSCI.35.6.1095 Schwartz, H. F., & Singh, S. P. (2013). Breeding Common Bean for Resistance to White Mold: A Review. Crop Science, 53(5), 1832–1844. https://doi.org/10.2135/cropsci2013.02.0081 Schwartz, H. F. (2011). Root rots of dry beans. Dissertation, ColoradoState University, Fort Collins. Sendi, Y., Pfeiffer, T., Koch, E., Mhadhbi, H., & Mrabet, M. (2020). Potential of common bean (Phaseolus vulgaris L.) root microbiome in the biocontrol of root rot disease and traits of performance. Journal of Plant Diseases and Protection, 127(4), 453–462. https://doi.org/10.1007/s41348-020-00338-6 Shi, A., Gepts, P., Song, Q., Xiong, H., Michaels, T. E., & Chen, S. (2021b). Genome-Wide Association Study and Genomic Prediction for Soybean Cyst Nematode Resistance in USDA Common Bean (Phaseolus vulgaris) Core Collection. Frontiers in Plant Science, 12, 624156. https://doi.org/10.3389/fpls.2021.624156 Singh, S. P., & Schwartz, H. F. (2010). Breeding Common Bean for Resistance to Diseases: A Review. Crop Science, 50(6), 2199–2223. https://doi.org/10.2135/cropsci2009.03.0163 Singh, S. P., Schwartz, H. F., Terán, H., Viteri, D., & Otto, K. (2014). Pyramiding white mould resistance between and within common bean gene pools. Canadian Journal of Plant Science, 94(5), 947–954. https://doi.org/10.4141/cjps2013-321 22 Spindel, J. E., Begum, H., Akdemir, D., Collard, B., Redoña, E., Jannink, J.-L., & McCouch, S. (2016). Genome-wide prediction models that incorporate de novo GWAS are a powerful new tool for tropical rice improvement. Heredity, 116(4), 395–408. https://doi.org/10.1038/hdy.2015.113 Soule, M., Porter, L., Medina, J., Santana, G. P., Blair, M. W., and Miklas, P. N. (2011). Comparative QTL map for white mold resistance in common bean, and characterization of partial resistance in dry bean lines VA19 and I9365-31. Crop Sci. 51, 123–139. doi: 10.2135/cropsci2010.06.0356 Snapp, S., Kirk, W., Román-Avilés, B., & Kelly, J. (2003). Root Traits Play a Role in Integrated Management of Fusarium Root Rot in Snap Beans. HortScience, 38(2), 187–191. https://doi.org/10.21273/HORTSCI.38.2.187 Steadman, J.R., and G. Boland. 2005. White mold. In: H.F.Schwartz, J.R. Steadman, R. Hall, and R.L. Forster, editors, Compendium of bean diseases. 2nd ed. Am. Phytopath. Soc., St. Paul, MN. p. 44–46. Terán, H., & Singh, S. P. (2009). Gamete selection for improving physiological resistance to white mold in common bean. Euphytica, 167(3), 271–280. https://doi.org/10.1007/s10681-009-9885-8 Tiede, T., & Smith, K. P. (2018). Evaluation and retrospective optimization of genomic selection for yield and disease resistance in spring barley. Molecular Breeding, 38(5), 55. https://doi.org/10.1007/s11032-018-0820-3 Uebersax, M. A., Cichy, K. A., Gomez, F. E., Porch, T. G., Heitholt, J., Osorno, J. M., Kamfwa, K., Snapp, S. S., & Bales, S. (2023). Dry beans ( PHASEOLUS VULGARIS L.) as a vital component of sustainable agriculture and food security—A review. Legume Science, 5(1). https://doi.org/10.1002/leg3.155 USDA Crop Production 2022 Summary 01/12/2023. (2022). Crop Production. Vasconcellos, R. C. C., Oraguzie, O. B., Soler, A., Arkwazee, H., Myers, J. R., Ferreira, J. J., Song, Q., McClean, P., & Miklas, P. N. (2017). Meta-QTL for resistance to white mold in common bean. PLOS ONE, 12(2), e0171685. https://doi.org/10.1371/journal.pone.0171685 Vuong TD, Hoffman DD, Diers BW, Miller JF, Steadman JR, Hartman GL (2004) Evaluation of soybean, dry bean, and sunflower for resistance to Sclerotinia sclerotiorum. Crop Science 44, 777–783. doi:10.2135/ cropsci2004.7770 Wang, W., Jacobs, J. L., Chilvers, M. I., Mukankusi, C. M., Kelly, J. D., & Cichy, K. A. (2018). QTL Analysis of Fusarium Root Rot Resistance in an Andean × Middle American Common Bean RIL Population. Crop Science, 58(3), 1166–1180. https://doi.org/10.2135/cropsci2017.10.0608 23 Wen, Z., Tan, R., Zhang, S., Collins, P. J., Yuan, J., Du, W., Gu, C., Ou, S., Song, Q., An, Y. C., Boyse, J. F., Chilvers, M. I., & Wang, D. (2018). Integrating GWAS and gene expression data for functional characterization of resistance to white mould in soya bean. Plant Biotechnology Journal, 16(11), 1825–1835. https://doi.org/10.1111/pbi.12918 Zhang, J., Song, Q., Cregan, P. B., & Jiang, G.-L. (2016). Genome-wide association study, genomic prediction and marker-assisted selection for seed weight in soybean (Glycine max). Theoretical and Applied Genetics, 129(1), 117–130. https://doi.org/10.1007/s00122-015-2614-x Zitnick-Anderson, K., Oladzadabbasabadi, A., Jain, S., Modderman, C., Osorno, J. M., McClean, P. E., & Pasche, J. S. (2020a). Sources of Resistance to Fusarium solani and Associated Genomic Regions in Common Bean Diversity Panels. Frontiers in Genetics, 11, 475. https://doi.org/10.3389/fgene.2020.00475 24 CHAPTER 1: POTENTIAL OF GENOMIC PREDICTION TO SCREEN FOR WHITE MOLD RESISTANCE IN DRY BEAN ABSTRACT Dry bean (Phaseolus vulgaris L.) production in the U.S. suffers severely from white mold (Sclerotinia sclerotiorum (Lib.) de Bary) infection. Dry bean cultivars lack high levels of resistance, and progress to breed new cultivars with durable levels of resistance has been slow due to the quantitative inheritance of this trait, difficulty pyramiding resistance, and screening dependence on the presence of the pathogen under suitable environmental conditions. Genomic prediction provides an alternative method to pyramid resistance genes by utilizing genome-wide marker coverage to predict genotypic values for quantitative traits. This study evaluated the efficiency of different genomic prediction models given the complex population structure of multiple market classes present in dry bean breeding programs. A panel of 303 Middle-American breeding lines were genotyped with 3,026 markers and evaluated for white mold in the field. Prediction accuracy across models and subsets was moderate (0.3 - 0.36) given the population size. Furthermore, when fixed effect QTL were identified and implemented through GP + GWAS, 1-3 QTL increased prediction accuracy (0.36 - 0.4). These results indicate that genomic prediction is a promising screening tool in dry bean breeding for white mold resistance. INTRODUCTION Dry bean (Phaseolus vulgaris L.) is one of the most important grain legumes for human consumption worldwide, providing an important source of key dietary nutrients such as protein and fiber (Uebersax et al., 2023). However, biotic stress caused by fungal disease infection is among the main constraints that limit yield and increase management cost of dry bean production. White mold, conferred by the fungal pathogen Sclerotinia sclerotiorum (Lib.) de Bary is one of 25 the most destructive fungal diseases of dry bean, resulting in poor seed quality and reduced yield (del Río et al., 2004; Vasconcellos et al., 2017). In temperate regions it is considered the most yield limiting disease of dry bean production, causing up to 100% yield loss and may also result in significant seed quality decline in susceptible cultivars (Schwartz et al., 1987; Singh & Schwartz, 2010). Consequently, economic losses from white mold in the United States have exceeded $200 million yearly due to yield reduction and fungicide costs (Bolton et al., 2006). Sclerotinia sclerotiorum is a necrotrophic fungus with a broad host range of over 400 species of plants (Boland & Hall, 1994). Sclerotinia sclerotiorum can be seed transmitted. However, the bulk of inoculum comes from melanized sclerotia that can overwinter in soil for five or more years (Schwartz & Singh, 2013). Around dry bean flowering time, S. sclerotiorum sclerotia germinates and forms a fruiting structure, a small cup shaped mushroom called an apothecia (Miklas et al., 2013). Once the apothecia has developed, asci can forcefully release ascospores into the plant canopy. These ascospores can infect the plant through wounds on the stem left by the senescent flowers and on other foliar parts, by penetrating the host cell walls (Bolton et al., 2006). Once colonized, infected plants often exhibit wilted leaves due to reduced vascular function. As disease development progresses, stems become brittle, and have a bleached appearance, often leading to severe plant lodging (Schwartz & Singh, 2013). Yield and seed quality decline occurs indirectly due to plant stress that reduces seed development or lost pods that abort entirely and directly due to colonization of bean pods that results in moldy, discolored, or shriveled seed. In severe infections, white mycelium can be observed on affected plant parts and whole plant death can occur (Schwartz & Singh, 2013). Michigan is the second largest dry bean producer in the United States, producing over 400 million pounds of dry edible beans per year (USDA Crop Production Summary 2022). It also leads 26 the nation in the black, small red, and navy bean market classes in terms of acreage (USDA Crop Production Summary 2022). Michigan's temperate climate, with warm humid summers, is highly conducive to fungal diseases. Among the fungal diseases present in Michigan, growers consider white mold the most serious disease impacting dry bean production. In the 2022 annual production practices survey, Michigan farmers and crop advisors regarded white mold as the number one disease of dry bean production that needs further research to increase yield and reduce fungicide costs (MBC, 2022). Managing white mold can be challenging due to the durability and extended viability of sclerotia in the soil and broad host range of the pathogen (Bolton et al., 2006). Common strategies include crop rotation, fungicide sprays, limiting nitrogen fertilizer and irrigation to reduce vegetative growth, wide row spacing, low density planting, and the development of low biomass, upright, and open canopy cultivars (Ender & Kelly, 2005; Miorini et al., 2017; Schwartz & Singh, 2013). These strategies focus on reducing sclerotial germination and ascospore development by altering the microclimate to reduce canopy humidity and temperature and therefore colonization ability (Schwartz & Singh, 2013). However, they are often not sustainable or economically feasible and reduce productivity of dry bean fields. For example, control of white mold through biomass management, wide row spacing, and reduced fertilizer and irrigation compromises yield per acre. Additionally, fungicide sprays are expensive and have adverse effects on the environment. Furthermore, these practices do not fully eliminate white mold infection due to the fungal virulence and overwintering ability. Therefore, the development of resistant cultivars is the most cost- effective, sustainable, and durable approach to combat white mold (Kolkman & Kelly, 2000; Miklas et al., 2001) along with appropriate cultural management practices as needed. 27 Breeding for white mold resistance in dry bean is challenging due to the quantitative inheritance method of resistance, which can be obtained through both physiological defense mechanisms and architectural avoidance. Disease avoidance mechanisms (Hoyos‐Villegas et al., 2015; Miklas et al., 2001, 2013) are associated with plant architecture traits that impart a tall, upright growth habit, and porous plant canopy. Physiological resistance (Miklas et al., 2001; Terán & Singh, 2009) is associated with pathogen recognition, including utilizing reactive oxygen species as an initial defense mechanism and synthesis of pathogenesis-related proteins (Mamidi et al., 2016) Given the complex infection process of white mold, a multi-faceted quantitative resistance mechanism is needed for dry bean (Mamidi et al., 2016). Unfortunately, completely resistant varieties are unavailable and most known sources of genetic resistance to white mold are of Andean origin, usually from unadapted landraces and wild relatives, and from secondary gene pools (Schwartz & Singh, 2013; Singh et al., 2014; Vasconcellos et al., 2017). While low levels of resistance have been identified in Middle-American germplasm (Ender & Kelly, 2005; Hoyos‐ Villegas et al., 2015; Mkwaila et al., 2011), progress to develop white mold resistant cultivars has been hindered by the lack of high levels of resistance (Schwartz & Singh, 2013), difficulty in pyramiding resistance genes (Singh et al., 2014), low heritability for the trait (Fuller et al., 1984; Miklas et al., 2004; Park et al., 2001), and environmental dependency for field evaluation requiring highly managed disease nurseries under frequent overhead irrigation (Ender & Kelly, 2005). Furthermore, greenhouse studies provide an alternate method of measuring physiological resistance, but previous studies have failed to show strong correlation between field and greenhouse screening methods (Terán & Singh, 2009). The complex quantitative inheritance mechanism of this trait has been demonstrated by the identification of at least 35 quantitative trait loci (QTL) with minor effects for resistance to white 28 mold in dry bean (Singh et al., 2014). This genetic architecture has posed challenges to dry bean breeders aiming to deploy marker-assisted selection (MAS) for QTL conferring partial resistance and low accuracy of genomic intervals resulting in negative linkage drag for yield and other traits (Miklas, 2007; Vasconcellos et al., 2017). To provide a better resolution, Vasconcellos et al., (2017) conducted a meta-QTL analysis and identified a total of nine-QTL as potential targets for MAS for partial resistance. While this study determined 9 major gene candidates that contribute to white mold resistance, the integration of an adequate level of white mold resistance in adapted dry bean germplasm is difficult because backcrossing many genes is inefficient and takes time even with the assistance of molecular markers (Lee, 1995). Unlike breeding for qualitative resistance, breeding for quantitative resistance is more challenging because it requires multiple cycles of breeding and screening, leading to a gradual improvement of resistance in a breeding population over time. With the sheer number of QTL identified previously conferring many environmentally dependent resistance traits, it is no surprise that there has been difficulty pyramiding genes into one robustly resistant genotype with the desired agronomic traits. An integrated screening and recurrent selection method that considers all aspects of resistance and agronomic performance would greatly assist future cultivar development. Multiple breeding tools such as recurrent selection and the use of alternative populations such as multiparent intercrosses are one way to assist the identification and pyramiding of resistance genes (Escobar et al., 2022; Osorno et al., 2018). Recently, Escobar et al. explored integrated multi- parent crosses and gamete selection using a Multiparent Advanced Generation Inter-Cross (MAGIC) population, to facilitate mapping and breeding efforts, resulting in multiple partially resistant lines. This method crosses together multiple founder lines and cycles through several additional generations of crossing, resulting in offspring with multiple recombination events, often 29 leading to improved results that maximize diversity. This and other emerging breeding tools will assist future breeding for complex traits. One such tool is genomic prediction/selection (GP/GS). Genomic prediction creates a unique opportunity to select and pyramid major and minor alleles conferring resistance to complex diseases (Merrick et al., 2021; Poland & Rutkoski, 2016; Tiede & Smith, 2018). This established tool in Marker Assisted Selection (MAS) utilizes genome wide markers and phenotypic data to train a linear statistical model to predict and make selections based on genomic estimated breeding values (GEBVs) for a trait. In contrast to MAS, genomic prediction does not identify significant markers and estimate individual marker effects. Instead, all marker effects are estimated simultaneously and used to accurately predict overall breeding values for a trait (Jannink et al., 2010). When properly implemented, genomic prediction has the potential to improve breeding program efficiency and reduce phenotyping costs when screening for complex traits. Genomic prediction has assisted breeding efforts for quantitative disease and pest resistance traits in many crop species including wheat: (Arruda et al., 2016; Juliana et al., 2017, 2022; Larkin et al., 2021; Merrick et al., 2021; Odilbekov et al., 2019; Rutkoski et al., 2012; Sarinelli et al., 2019) , dry bean (Diaz et al., 2021; Shi et al., 2021a), soybean (Bao et al., 2015; de Azevedo Peixoto et al., 2017; Đorđević et al., 2019; Duhnen et al., 2017; Hemingway et al., 2021; Shi et al., 2021b; Wen et al., 2018), and other crop species. Specifically, when compared to MAS or phenomic selection alone, genomic prediction has shown to be more effective for pyramiding quantitative traits controlled by many small effect QTL (Heffner et al., 2010, 2011; Massman et al., 2013; Merrick et al., 2021; Zhang et al., 2016). Many previous studies of the genetic control of white mold in dry bean fail to account for the many small effect QTL unable to be detected by genome wide association studies (GWAS) 30 that are likely controlling resistance to this trait. To date, there have been no major resistance genes identified for white mold resistance, unlike some disease traits such as Fusarium head blight in wheat (Larkin et al., 2021). Since genomic prediction utilizes all markers spread across the entire genome to make selections, it has the power to detect small effect QTL conferring resistance that would otherwise be overlooked by GWAS (Jannink et al. 2010; Meuwissen et al. 2013). Genomic selection for white mold resistance has been studied in soybean diversity panels (de Azevedo Peixoto et al., 2017; Wen et al., 2018) with moderate to high prediction accuracies (0.4-0.7) suggesting that this warrants further research into the validity to assist breeding for resistance in dry bean. Another benefit of genomic prediction in breeding for complex traits controlled by many major and minor QTL is that major QTL controlling a trait of interest can be identified using a GWAS and implemented in genomic prediction as fixed effects, allowing for further accuracy in selection (Merrick et al., 2021). Fixed QTL can be selected based on previous research or identified through de novo GWAS of the training population. GS + de novo GWAS involves two main stages where in the first stage GWAS is conducted on individuals in the training set to identify fixed QTL to be implemented in genomic prediction of the testing set (Bian and Holland, 2017; Haile et al., 2021; Sarinelli et al., 2019; Spindel et al., 2016). For example, Bian and Holland 2017 and Spindel 2016 both found that GS + de novo GWAS outperformed all other models tested. A simulation study by (Rice and Lipka, 2019) evaluated the effect of the number of QTL implemented across multiple genetic architectures and trait heritability and observed increased, decreased, and mixed (increase then decrease or decrease then increase) effect on prediction accuracy depending on the trait and number of fixed QTL. This study emphasizes that a varying number of fixed effect QTL should be tested for each trait before implementation. 31 Genomic prediction is more complex to implement when the crop of interest has multiple sub-classes. Dry beans are one such crop because their breeding programs are funneled into many smaller programs based on market standards for over 10 distinct seed classes. While there is some cross breeding between classes (ex: black by navy), this is discouraged due to the possibility to transmit deleterious seed traits. Due to distinct market class traits, the number and genetic composition of market classes added to the training population may have a significant effect on prediction accuracy. Since this is the first study to date evaluating genomic prediction as a tool to screen for white mold resistance in dry bean breeding lines, a major objective was to determine the ideal population composition to optimize prediction accuracy. Given these factors, the primary objective of this study was to evaluate the potential of genomic prediction as a screening tool to breed for white mold resistance and determine how it could be efficiently used in a dry bean breeding program. To do so, we evaluated advanced breeding lines from the Michigan State University (MSU) dry bean breeding program belonging to three distinct major market classes in Michigan. Our specific objectives were to i) evaluate genomic prediction across and within major market classes of dry bean, ii) identify the optimal training population composition to increase prediction accuracy in the presence of distinct market classes and population structure, and iii) compare predictive abilities between GP alone and GP + de novo GWAS. MATERIALS AND METHODS Plant material A set of 303 lines were used to evaluate the potential of using genomic prediction for white mold resistance. This panel consisted of advanced breeding lines from four market classes (black, navy, small red, and pink) developed by the MSU dry bean breeding program, 7 commercially 32 available lines grown commonly in Michigan, and the partially white mold resistant lines USPT- WM-12 (pinto) and G122 (cranberry) (Phillip N. Miklas, USDA). Detailed line information can be found in (Table 1.5). All lines were evaluated for white mold resistance in the field. During the 2021 field season 176 lines were evaluated and during the 2022 field season, 127 lines were evaluated. Finally, 55 lines were added to the study that were previously evaluated in the National Sclerotinia Initiative multi-state national trials and 43 lines were evaluated in multiple years. The final training set consisted of 144 black, 117 navy, 32 red, 7 pink, 1 pinto, 1 cranberry, and 1 great northern line. Phenotype data collection All lines were grown under natural white mold infestation in a disease nursery at Montcalm Research and Extension Center in Montcalm County, MI. This location has consistent yearly white mold disease pressure and an overhead pivot irrigation system to promote disease infection. The lines were planted in an alpha-lattice design with three replicates using four row plots 6.1 m in length with 50cm row spacing. The outer two rows were planted with the white mold susceptible black bean line (Black Bear) to encourage white mold infection. The resistant, moderately resistant, and susceptible checks, G122, Bunsi, and Beryl, were added to adjust for local disease variation. Plants were evaluated for white mold multiple times per season between the R7-R9 growth stages, beginning when white mold disease was first observed (around pod fill) and continuing through dry down. The best white mold rating to be used for genomic prediction was chosen based on heritability and variance for disease score. A plot-wise visual disease rating system was used to screen for white mold resistance in the field. The visual rating consisted of incidence and severity on a scale of 1 to 9, where a rating of 1 indicates no diseased plants and a rating of 9 33 indicates 80 to 100% diseased plants or 60 to 100% infected tissue as described in Miklas et al., 2001. Plots were trimmed to 4.9m prior to harvest and the center two rows were harvested with a Wintersteiger Classic plot combine (Wintersteiger AG, Austria). Yield data was obtained utilizing a Harvestmaster H2 Classic Graingage (Juniper Systems, Logan, UT) to record plot weight and moisture. Prior to data analysis, seed yield was standardized to 18% moisture. Standard agronomic traits including plant height, lodging score, days to flower, and maturity were also collected in the field. Genotypic data collection and processing Single-nucleotide polymorphism (SNP) data was collected for all lines in the training population. The tissue collection, DNA extraction, and genotyping procedures were as follows. Ten first trifoliate leaves from each line were collected and bulked. Leaves were desiccated using liquid nitrogen, placed in a -80 ○C freezer, and lyophilized. DNA was extracted using the Qiagen DNeasy plant mini kit following the manufacturers protocols and concentration standardized to at least 50 ug/ul. SNP chip sequencing using the Illumina Infinium BARCBean12k Bead chip was performed at the USDA-ARS, Soybean Genomics and Improvement Lab in Beltsville, Maryland. SNP calling for 11,929 markers was performed in the Illumina Genomestudio software (Genomestudio Software, 2023). Markers were filtered in the TASSEL trait analysis software, which removed any SNP with a minor allele frequency of less than 0.10 or more than 50% missing individuals (Bradbury et al., 2007). A principal component analysis (PCA) and genomic relationship matrix were developed utilizing the remaining markers. Imputation was performed in the rrBLUP package in R using the A.mat function and the impute=mean parameter to input missing data according to the population mean at each marker (Endelman, 2011). Realized 34 relationship matrices were also calculated using the A.matrix function within rrBLUP. These steps resulted in a matrix of ~3,026 polymorphic markers used in this study. Phenotypic statistical analysis A linear mixed model was fit using the ASREML package in the R coding language to calculate Best Linear Unbiased Estimators (BLUEs) for white mold disease score (Butler et al., 2023). The model was as follows: Yijk = μ + Linei + Envj + Rep(Env)kj + (Line x Env)ij + eijk Eq. 1 Where Yijk is the observed phenotype. μ is the overall mean, Linei is the fixed effect of the i-th line, Envj is the random effect of the jth environment, Rep(Env)kj is the random effect of the k-th rep nested within the j-th environment, (Line x Env)ij is the random effect of interaction between the i-th line and j-th environment,, and eijklm is the random residual term. To estimate broad sense heritability (H2) for white mold resistance on an entry means basis, variance components were extracted fitting Equation 1 with all terms random. Heritability was estimated as follows: 𝐻2 = 2 𝜎𝐺 2 𝜎𝐺𝐸 𝑛 + 2 𝜎𝑒 𝑟𝑛 2 + 𝜎𝐺 Eq. 2 Where σ2 G ,σ2 GxE ,and , σ2 e are the genotype, genotype by environment interaction, and error variances, n is the number of environments and r is the number of reps. Genomic heritability for white mold resistance was also calculated by first using the marker matrix to generate a kinship matrix using the A.mat function in rrBLUP and then inputting the kinship matrix and the BLUEs for white mold disease score into the function marker_h2 means in the 35 AGHmatrix package (Amadeu et al., 2023; Endelman, 2011). Genomic heritability is the proportion of variance of phenotypes explained by a regression on a set of markers (De Los Campos et al., 2015). Genomic predictions for white mold disease The accuracy of genomic prediction was evaluated using two models (rrBLUP and GBLUP) implemented in the rrBLUP package in the R coding language (Endelman, 2011). All genomic prediction models and subsets were evaluated using 5-fold cross-validation (CV) repeated 20 times. For each round of CV, the datasets were divided into equal sets of five. Four of the sets were used as the training set, while the remaining set was used as a validation set. BLUEs were predicted for individuals in the validation set and prediction accuracy was evaluated as the correlation between the observed BLUEs and predicted genomic BLUEs. The rrBLUP model was expressed as follows: Y = Xb + Zu + e Eq. 4 Where y is the vector of BLUEs for disease score; X is the incidence matrix for the fixed effects b is the vector of fixed effects; Z is the incidence matrix for the genotype effects; u is the vector of marker effects (BLUEs), assumed to follow multivariate normal distribution such that u∼N(0,Iσ2u). Finally, e is a vector of residual errors assumed to be independent and identically distributed according to N(0,σ2e). GBLUP utilizes the same formula with the following differences. In GBLUP u is the vector of additive genotype effects (breeding values), assumed to follow multivariate normal distribution such that u∼N(0,Kσ2a) where K is the realized kinship matrix and σ2 a is the additive genetic variance. 36 Population Subsets Dry bean breeding programs are unique because they consist of over 10 separate market class subsets with distinct characteristics (and therefore separate breeding populations). To simulate a breeding scenario of employing genomic prediction to screen for white mold disease resistance in a dry bean breeding program, multiple market class subsets were tested. For simplicity, red and pink beans are considered the same market class, both referred to as red. Four subsets (entire population, black, navy, and black + navy) used the respective market classes as the training and validation population for cross validation. The entire population subset included all lines in the training set (n=303) whereas the black, navy, and black + navy subsets included the respective market classes only. Red and pink beans were not tested as subsets as there were only 39 lines total. Two subsets (black predict navy and navy predict black) used one market class as the training set and another as the validation set. Genome-Wide Association Analysis Association analysis was performed on the entire training set to evaluate the effectiveness of including fixed effect markers on the accuracy of the genomic prediction through GP + de novo GWAS. With GEBVs for white mold disease score, GWAS was conducted in the Farm CPU model of the Genome Association and Prediction Integrated Tool package (GAPIT) version 3 in R studio (Wang & Zhang, 2021). The same SNP marker dataset and phenotype dataset was used for GWAS as for genomic prediction. To avoid bias in the calculation of prediction accuracy, the identification of markers to use as fixed effects was based on 20 cycles of 5-fold cross validated GWAS using each training set of lines only following a similar protocol to Haile 2021 and Sarinelli 2018. The Fixed and Random Circulating Probability Unification (Farm CPU) model was chosen because it uses the bin approach to avoid selecting markers from the same location and it accounts for 37 population structure using two principal components and kinship among individuals (Liu et al., 2016). P-value inflation was assessed using a QQ-plot. Significance was calculated using the Benjamini-Hochberg FDR procedure at p-value threshold of p=0.05. The average effect on white mold score for each significant QTL over all calls was also calculated within GAPIT. The top 1-7 QTL for each run of GWAS were called as fixed effects in each corresponding run of genomic prediction implemented in the rrBLUP model and mean prediction accuracies were obtained resulting from CV (100 runs for each QTL threshold). RESULTS White Mold Disease A continuous distribution of BLUEs for white mold disease score was observed when evaluating all years of data combined, which ranged from 2.07 to 8.41 (Figure 1.1a). Disease scores ranged from 3.07 to 7.34 for the black market class, 2.07 to 8.41 for the Navy market class, and 2.29 to 7.95 for the Red and Pink market classes (Figure 1.1b). The two resistant, one moderately resistant, and two susceptible checks, USPT-WM-12, G122, Bunsi, Beryl, and Black Bear had scores of 4.83, 4.17, 6.16, 7.62, and 6.41, respectively (Table 1.4). Broad sense heritability (H2) estimates for disease severity were moderate at 0.52. A significant correlation (- 0.27) was identified between white mold resistance and yield in the combined dataset. Variance components for white mold and yield can be found in (Table 1.6, Table 1.7). Training Population Structure The population structure of the entire training set was evaluated with a principal component analysis (PCA) and a genomic relationship matrix (GRM). The first two principal components accounted for 15% of the phenotypic variation (Figure 1.2a). The PCA was able to separate three distinct clusters and showed admixture between two clusters. The black and navy market class 38 clusters overlap on the right side of the PCA, whereas the red/pink cluster is isolated to the left- hand side. This is supported by the GRM, where three darker color clusters correspond to the three major market classes used in the training population (Figure 1.2b). Genomic heritability was low- moderate at 0.31. Genomic Prediction Accuracy The accuracy of genomic prediction models was evaluated using rrBLUP and GBLUP across and within multiple market class subsets. Overall mean prediction accuracy ranged from 0.12-037 for rrBLUP and 0.03 to 0.38 for GBLUP when using the entire training set and across all subsets (Figure 1.3, Table 1.1). In the four sets utilizing the respective market classes as the training and testing set (Entire Population, black, navy, black + navy) prediction accuracy ranged from 0.30-0.38 and prediction accuracy was similar between rrBLUP and GBLUP. The navy bean subset had the highest prediction accuracy overall, followed by the entire training set. In the two subsets where different market classes were used for the training and testing sets (black predict navy and navy predict black) prediction accuracy ranged from 0.03-0.25. The black predict navy subset had higher prediction accuracies than the navy predict black subset. The black predict navy subset had higher prediction accuracy using GBLUP and the navy predict black subset had a higher prediction accuracy using rrBLUP. Variance in prediction accuracy across individual runs of genomic prediction was highest when the navy subset was used to predict the black subset and lowest when the entire population was used as the training and testing sets. Genomic Prediction with Fixed Effect QTL Association analysis was performed on the entire training set (n=303) to evaluate the effectiveness of fixed effect markers on the accuracy of the genomic prediction through GP + de novo GWAS. To avoid bias in the calculation of prediction accuracy, the identification of markers to use as fixed 39 effects was based on 20 cycles of 5-fold cross validated GWAS using the corresponding training set only. The top 1-7 QTL for each run of GWAS were called as fixed effects in each corresponding run of genomic prediction implemented in the rrBLUP model. Significant QTL were identified on all chromosomes across GWAS cycles. The top 10 most frequently called QTL were identified on chromosomes 2,4,6,7,8, and 11 (Table 1.2). The most frequently called QTL was on chromosome 11 (sc00007ln1695141_1432256_A_C_13462529) with 44 calls total across cycles of CV. Prediction accuracy rose from 0.36 to 0.40 when 1-3 QTL were implemented as fixed effects in genomic prediction (Figure 1.4, Table 1.3). When 4-7 QTL were implemented, prediction accuracy decreased from 0.4 to 0.22. Mean effect of the top 10 most identified fixed QTL on phenotype over all cycles ranged from -0.81 - 0.7 with a negative sign referring to the minor allele being favorable. The most called SNP on chromosome 11 (sc00007ln1695141_1432256_A_C_13462529) also had the largest mean effect (-0.81) with the minor allele being favorable. DISCUSSION In this study, advanced dry bean breeding lines from distinct market classes of the MSU dry bean breeding program were used as a training population to evaluate the potential of genomic prediction as a screening tool for white mold resistance. Specifically, we aimed to understand whether random sets of four-fifths of the lines could be used to predict white mold disease resistance in the remaining one-fifth. Our results indicated moderate cross-validated prediction accuracies (0.3-0.38) given the small sample size in the entire training population and population subsets when the same market class was used as both the training and testing population, implying that genomic prediction for white mold resistance within Middle-American breeding panels is 40 promising and can be implemented by dry bean breeding programs. Although prediction accuracy is lower than previous studies utilizing the same cross validation methods, these studies had a larger training population size, different population composition, and were focused on different traits and crops. As the size of the training population increases, we expect prediction accuracy to rise as has been confirmed by multiple studies (Edwards et al., 2019; Fernández-González et al., 2023; Sarinelli et al., 2019). Diversity and Population Structure We observed moderate to severe white mold disease severity in the field yearly, which was aided by overhead irrigation and the presence of a spreader genotype. Broad sense heritability of white mold resistance under disease pressure in the field was moderate which suggests selections for white mold resistance on this population would be effective, considering the environmental dependance of a complex quantitative disease trait such as white mold. Our population had significant diversity for white mold resistance in the field, within and among market classes ranging from about 2-8 on a 1-9 scale. Three distinct clusters were observed via PCA which corresponded to the three market classes present in the population with red beans being the most distinct from black and navy beans. This relationship between individuals is likely due to distinct characteristics among the market classes and due to the crossing scheme used to develop the population. Red, pink, pinto and great northern beans belong to the Jalisco sub-race and black and navy beans belong to the Mesoamerican sub-race within the Middle-American gene pool (Mensack et al., 2010). Additionally, the MSU dry bean breeding and genetics program frequently performs black by navy crosses to develop new breeding lines, but red/pink beans are not commonly crossed with the two other classes, which could have contributed to the observed structure. 41 Genomic Predictions Across Models and Subsets A secondary objective of this study was to identify the optimal training population to increase prediction accuracy in the presence of distinct dry bean market classes. When evaluating genomic prediction accuracies among market class subsets, we found that sample size and population structure both affected prediction accuracy. This is supported by the fact that the entire population subset and the navy subset had the two highest prediction accuracies across both the rrBLUP and GBLUP models. Population size was highest in the entire population subset and genetic relatedness was higher in the navy subset vs the entire population which included three market classes. Overall, population structure affected prediction accuracy more than population size, as reflected by the navy subset having the highest prediction accuracy, followed by the entire training population. This result is like previously published manuscripts, for example (Duhnen et al., 2017) in soybean, observed higher prediction accuracies when early and late genotypes were considered separately rather than when the entire population was considered. Another future direction would be to utilize a training population optimization algorithm or sparse selection index to select the optimal genotypes to increase prediction accuracy in the presence of multiple market classes (Fernández-González et al., 2023; Isidro et al., 2015; Lopez-Cruz & de los Campos, 2021). Furthermore, across most subsets the difference in prediction accuracy between rrBLUP and GBLUP was minimal. These genomic prediction models are similar, with the main difference being the incorporation of the genomic relationship matrix in GBLUP in replacement of the pedigree relationship matrix. If the pedigree relationship matrix is equal to the genomic relationship matrix, then the two models should have the same accuracy. The findings from this research show that the pedigree relationship and genomic relationship are similar in this population. 42 To this effect, when one market class was used to predict another market class, prediction accuracy dropped significantly. This indicates that to obtain a high genomic prediction accuracy in a dry bean breeding program, the training population and validation population should consist of the same market classes. Even in the case of black and navy beans, where the two market classes are frequently crossed, there are distinct characteristics of each market class that lead to insufficient predictions. Integration of Fixed Effect QTL Another secondary objective was to compare prediction accuracies between genomic prediction and GP + de novo GWAS. We observed increased prediction accuracy when 1-3 QTL were integrated as fixed effects through GP + de novo GWAS and a decrease in prediction accuracy when 4+ QTL were integrated. Previous studies have observed an increase in prediction accuracy when utilizing fixed effect QTL in GP + de novo GWAS such as in rice (Spindel et al., 2016) and wheat (Haile et al., 2021). Similar to results observed in the simulation study by Rice and Lipka 2019, prediction accuracy initially increased then decreased when 4+ QTL were added. Interestingly, although significant QTL were identified, all had a relatively low contribution to overall variation in resistance (Table 1.2) highlighting the likelihood that this trait is controlled by many small effect QTL rather than large effect R genes. Limitations One primary limitation in our study to note is the small size of the training population. Many public breeding programs are limited by sample population due to the number of lines moving through the breeding pipeline in a generation. In the MSU dry bean breeding and genetics program ~100-150 lines are advanced to the preliminary and advanced trial stage yearly. This training population will be updated with these lines and prediction accuracies are expected to 43 increase as a result. Additionally, our training population contained the three most important market classes in Michigan, prediction accuracies and genetic architecture for other market classes require further investigation. The low to moderate prediction accuracy in all population subsets could have been due to multiple factors including sample size, marker density, and low heritability of the trait. One other important aspect of the fixed effect QTL identified using GWAS is the high possibility that our current population of breeding lines isn’t sufficient to cover the breadth of resistance sources for white mold in dry bean. It is also likely that rare alleles in our population were filtered out. As the population size increases, more QTL conferring resistance will likely be identified and prediction accuracy will increase. The genotyping method used was a high throughput, cost-effective whole genome method that would allow the program to process all individuals in each year's preliminary yield trial stage. Any reduction in prediction accuracy due to marker density was likely outweighed by the ability to develop a larger training population. Applications to Dry Bean Breeding Genomic selection is a useful tool for breeding and can outperform phenotypic selection and MAS when screening and pyramiding complex traits. White mold in dry bean is one such trait where genomic selection could be extremely beneficial to breed for resistance if implemented properly. Progress to develop white mold resistant cultivars has been greatly hindered by difficulty in pyramiding resistance genes and complexity of field evaluation. A major advantage of implementing genomic selection in a breeding program is the reduction of phenotyping. This is an important aspect for white mold resistance in dry bean especially because field screening requires a specialized disease nursery with overhead pivot irrigation and natural white mold infection. If genomic prediction accuracy in the training population is sufficient, genomic data could be 44 collected for members of the validation population and breeding values could be predicted, thus allowing for more lines to be screened per year. Another advantage of genomic prediction is it allows for earlier accurate selection of parents for the next generation of crosses therefore reducing breeding cycle time and increasing gain from selection. As selection for quantitative traits occurs in the advanced testing stages for many crops (including dry bean) this can have a large effect on genetic gain, effectively skipping 1-4 years of testing. CONCLUSION Complex fungal diseases such as white mold are a major constraint in dry bean production and the identification of QTL and development of resistant cultivars for white mold resistance in dry bean has been historically slow due to many factors. This study aimed to test genomic prediction as a novel method for the development of resistant cultivars. Implementation of genomic prediction in a plant breeding program relies on the development of a robust training population which requires consideration of multiple factors including size, marker density, and population structure. Results from this analysis indicate that it is possible to obtain moderate prediction accuracy in dry bean in the presence of population structure and multiple market classes, if sample size is adequate. The training population will continue to be updated, modified, and validated yearly to further increase prediction accuracy. The results from this research will assist plant breeders in the development of training populations for genomic prediction of highly quantitative disease resistance traits. 45 TABLES AND FIGURES Figure 1.1a: Phenotypic distribution of genomic estimated breeding values (GEBVs) for the entire training set. 46 Figure 1.1b: Phenotypic distribution of genomic estimated breeding values (GEBVs) for the entire training set by market class. 47 Figure 1.2a: Principal component analysis. The first two principal components accounted for 15% of the phenotypic variation. The PCA was able to separate three distinct clusters and showed admixture between two clusters. The black and navy market class clusters overlap on the right side of the PCA, whereas the red/pink cluster is isolated to the left-hand side. 48 Figure 1.2b: Genomic relationship (proportion of the genome shared) among individuals in the training population (n=303). The Y axis represents the genotyped lines ordered by market class with red/pink, followed by black and navy. The darker the color, the more highly related the lines, with white = no relationship to dark orange (e.g. the diagonal) genomic relationship = 1. 49 Figure 1.3: Comparison of genomic prediction accuracy of 5-fold CV repeated 20 times, for two models rrBLUP and GBLUP, over all training population subsets (entire population, black, navy, black + navy, black predict navy, and navy predict black). 50 Table 1.1: Average genomic prediction accuracies for 5-fold cross validation repeated 20 times across models and subsets. Subset Entire Population Black Navy Black + Navy Model RrBLUP GBLUP RrBLUP GBLUP RrBLUP GBLUP RrBLUP GBLUP Black predict Navy RrBLUP Navy Predict Black RrBLUP GBLUP GBLUP Average Accuracy 0.35 0.36 0.33 0.32 0.37 0.38 0.30 0.30 0.18 0.25 0.12 0.03 51 Table 1.2: Top 20 most frequently called SNPs over 5-fold cross validated GWAS repeated 20 times by chromosome and position. Mean effect is the average effect on phenotype over runs of cross validation. SNP Chromosome Position (Mb) Mean Effect # of Calls Chr02_41657167_A_G sc00687ln167302_103360_G_A_266954597 Chr06_12181540_G_A Chr06_16668700_G_T Chr07_3905254_C_T sc00394ln266395_95537_T_C_204849189 sc00093ln620690_104379_A_G_86490862 sc00146ln499601_116735_C_T_115838631 sc00187ln435150_54957_C_T_135207180 2 4 6 6 7 7 7 8 8 41657167 0.71 13213188 0.52 12181540 -0.61 16668700 -0.61 3905254 0.61 32697386 0.38 33306574 0.35 1525152 0.41 62919922 0.34 13 22 6 9 4 13 5 14 6 sc00007ln1695141_1432256_A_C_13462529 11 4741435 -0.81 44 52 Figure 1.4: GWAS + de novo genomic prediction prediction accuracies over 20 replicates of 5- fold cross validation for 1-7 integrated fixed QTL implemented in rrBLUP. 53 Table 1.3: Average GWAS + de novo genomic prediction prediction accuracies over 20 replicates of 5-fold cross validation for 1-7 integrated fixed QTL implemented in rrBLUP. # of QTL 0 1 2 3 4 5 6 7 Average Prediction Accuracy 0.36 0.39 0.39 0.40 0.37 0.37 0.35 0.22 54 Table 1.4: Ranking of BLUEs for White Mold and Yield. Line N22631 S08418 N21514 N19277 N21519 N21521 N21529 B22855 B22865 B18201 B20590 B20642 N20384 N21520 N19226 N21528 N21511 N22622 N18117 B20547 N17504 N21503 N21531 White Mold Yield (CWT_Acre) Yield Ranking 2.07 2.29 2.62 2.62 2.62 2.95 2.95 3.07 3.07 3.17 3.29 3.29 3.29 3.29 3.29 3.29 3.34 3.4 3.41 3.51 3.56 3.62 3.62 27.72 26.32 25.55 25.48 15.42 19.13 17.22 23.04 19.7 22.52 27.32 26.09 19.35 18.41 16.32 15.58 23.84 28.85 22.52 23.7 24.64 18.79 16.82 55 25 34 47 49 247 179 220 82 162 94 28 38 172 199 230 244 68 16 94 72 57 186 224 Table 1.4 (cont’d) N21534 B22861 N22605 N22637 B16506 B19345 N18103 R17603 B20532 R18402 B20602 N20405 S19307 B20542 N21506 N20317 R17604 N20404 B17220 B22827 B22831 B22840 N22638 B17691 3.62 3.74 3.74 3.74 3.8 3.82 3.9 3.92 3.95 3.95 3.95 3.95 3.95 3.95 3.95 3.95 3.96 4.01 4.06 4.07 4.07 4.07 4.07 4.13 235 19 109 148 137 148 178 2 18 45 50 110 155 217 237 254 13 141 219 8 15 95 250 1 16.18 28.29 22.08 20.61 20.9 20.61 19.16 33.58 28.36 25.59 25.41 22 20.12 17.41 16.05 14.65 29.63 20.79 17.28 30.75 29.15 22.5 15.15 34.12 56 Table 1.4 (cont’d) B16507 B15430 B20591 R16503 B18204 R20629 N21515 S20405 B21707 R20667 B21715 N19253 N21522 N21505 N20401 B20616 N21524 B20627 N17506 B20599 B15451 B22826 B22857 B22834 4.13 4.14 4.18 4.19 4.22 4.29 4.29 4.29 4.29 4.29 4.29 4.29 4.29 4.29 4.29 4.29 4.29 4.29 4.33 4.34 4.34 4.4 4.4 4.4 3 6 111 228 36 22 55 63 69 78 106 144 148 150 158 165 214 259 66 43 92 7 117 123 33.49 31.41 21.96 16.5 26.25 28.07 24.67 24.41 23.79 23.25 22.14 20.73 20.61 20.55 19.93 19.58 17.78 14.12 24.07 25.67 22.63 30.9 21.76 21.52 57 Table 1.4 (cont’d) N22632 N14229 I96417 I09203 N19285 B17536 I11264 B21706 N20352 R17602 R20639 B21720 B21709 N21507 N21509 B20538 B19339 N20391 B21705 B21714 B21724 N21533 N21502 N18105 4.4 4.43 4.48 4.49 4.49 4.54 4.59 4.62 4.62 4.62 4.62 4.62 4.62 4.62 4.62 4.62 4.62 4.62 4.62 4.62 4.62 4.62 4.62 4.62 222 37 269 23 191 126 74 9 10 11 40 81 90 93 93 103 118 119 131 138 174 182 184 188 17.15 26.11 13.03 28.02 18.67 21.33 23.53 30.74 30.49 30.24 25.87 23.07 22.68 22.58 22.58 22.32 21.71 21.68 21.16 20.88 19.33 18.87 18.83 18.75 58 Table 1.4 (cont’d) N20335 N21526 N19223 N19252 B21716 B20536 N19248 B22823 B22825 B22862 N22619 B22838 N22616 B22863 B22835 B22841 N22630 N22610 B22853 B22815 N22621 B22839 N22617 N17505 4.62 4.62 4.62 4.62 4.62 4.68 4.68 4.74 4.74 4.74 4.74 4.74 4.74 4.74 4.74 4.74 4.74 4.74 4.74 4.74 4.74 4.74 4.74 4.82 190 204 209 255 262 26 94 24 27 52 62 80 91 100 108 113 121 122 130 150 159 185 225 134 18.7 18.13 17.94 14.54 13.79 27.45 22.52 27.99 27.4 24.99 24.45 23.1 22.65 22.41 22.11 21.87 21.64 21.54 21.21 20.55 19.91 18.81 16.77 20.95 59 Table 1.4 (cont’d) S18904 B19330 USPT-WM-12 B21708 B04554 B18231 B10244 B21713 B20632 B18504 R20659 R20669 B21702 B20597 B21718 N21525 B19340 B21703 N21504 N20341 B21721 N21518 N21510 N21532 4.82 4.82 4.83 4.84 4.84 4.91 4.93 4.95 4.95 4.95 4.95 4.95 4.95 4.95 4.95 4.95 4.95 4.95 4.95 4.95 4.95 4.95 4.95 4.95 143 193 76 88 94 94 30 4 42 65 67 71 115 134 160 177 187 190 196 215 238 240 271 273 20.77 18.57 23.48 22.8 22.52 22.52 27.26 32.45 25.8 24.09 23.86 23.73 21.81 20.95 19.85 19.17 18.77 18.7 18.47 17.44 15.95 15.9 12.23 11.6 60 Table 1.4 (cont’d) B16501 B17922 B19309 B16504 B22850 B22860 B22820 B22854 B22814 N22624 B22844 B22836 B22866 N22607 B22859 R12844 B15442 N14218 N19246 BC269 N20388 R20683 R20637 R20632 4.96 4.97 5.01 5.04 5.07 5.07 5.07 5.07 5.07 5.07 5.07 5.07 5.07 5.07 5.07 5.11 5.16 5.17 5.18 5.18 5.18 5.29 5.29 5.29 195 94 33 5 32 41 44 51 53 61 84 107 132 156 176 96 12 208 206 207 210 33 54 60 18.52 22.52 26.55 31.99 26.64 25.83 25.66 25.06 24.95 24.56 22.95 22.13 21.12 20.01 19.2 22.49 30.2 17.97 18.07 18 17.92 26.55 24.78 24.61 61 Table 1.4 (cont’d) B21712 R20627 N21517 B20527 N21512 B20549 B20639 B19332 N18130 N20395 B22804 B22876 N22602 B22848 B22828 N22634 B22812 B22830 N22609 B22875 B22873 B22813 N22629 N22633 5.29 5.29 5.29 5.29 5.29 5.29 5.29 5.32 5.32 5.34 5.4 5.4 5.4 5.4 5.4 5.4 5.4 5.4 5.4 5.4 5.4 5.4 5.4 5.4 77 86 192 211 233 256 265 166 245 173 46 48 56 81 83 89 99 101 102 139 151 181 183 234 23.46 22.82 18.59 17.88 16.21 14.46 13.59 19.55 15.53 19.34 25.58 25.5 24.66 23.07 23 22.74 22.42 22.39 22.36 20.86 20.49 18.88 18.84 16.19 62 Table 1.4 (cont’d) B19344 N11283 B18236 N15331 R20633 R20614 B21723 B20623 B21717 R20636 B20579 B19341 N19290 B21722 N21530 N20343 B20617 B21719 B21711 B21710 B22842 N22623 B22837 B22856 5.46 5.48 5.49 5.58 5.62 5.62 5.62 5.62 5.62 5.62 5.62 5.62 5.62 5.62 5.62 5.62 5.62 5.62 5.62 5.68 5.74 5.74 5.74 5.74 242 203 258 31 35 64 71 73 75 98 116 120 140 175 198 243 267 268 272 245 97 105 124 128 15.75 18.17 14.24 26.78 26.29 24.3 23.73 23.65 23.5 22.43 21.79 21.67 20.82 19.22 18.42 15.6 13.33 13.16 12.16 15.53 22.45 22.26 21.46 21.25 63 Table 1.4 (cont’d) B22818 N22608 B22816 B22846 B22821 B22868 N22636 N22620 N22601 B22867 N18122 N22606 N22613 N22603 B22833 B22802 N19239 I81010 R20612 R19502 N21527 N19284 N20336 B19302 5.74 5.74 5.74 5.74 5.74 5.74 5.74 5.74 5.74 5.74 5.74 5.74 5.74 5.74 5.74 5.74 5.82 5.92 5.95 5.95 5.95 5.95 5.95 5.95 133 135 136 142 145 146 147 152 160 184 191 192 194 197 221 232 241 202 79 103 154 200 205 213 21.06 20.92 20.91 20.78 20.69 20.67 20.62 20.47 19.85 18.83 18.67 18.59 18.55 18.46 17.2 16.27 15.82 18.23 23.2 22.32 20.24 18.38 18.08 17.8 64 Table 1.4 (cont’d) B20582 B20629 S18907 B20621 R17605 B22807 B22870 B22874 B22843 B22822 B22806 B22811 B22829 B22801 B22803 B22805 N22627 N22615 N22612 N15341 N16405 R20625 N20346 N19243 5.95 5.95 5.95 5.95 5.96 6.07 6.07 6.07 6.07 6.07 6.07 6.07 6.07 6.07 6.07 6.07 6.07 6.07 6.07 6.15 6.16 6.29 6.29 6.29 226 231 260 264 20 54 70 85 112 125 127 149 161 168 171 201 212 227 263 261 17 87 238 243 16.64 16.3 14.11 13.65 28.26 24.78 23.75 22.88 21.91 21.39 21.3 20.56 19.79 19.45 19.39 18.33 17.87 16.52 13.74 13.82 28.68 22.81 15.95 15.6 65 Table 1.4 (cont’d) B20620 N21523 N19269 I17501 B19346 N16401 B22832 B22847 B22845 N22618 B22819 B22852 N22626 N22614 R98026 R20684 R20624 B21704 R20635 B21701 R20652 N20376 N21516 B22810 6.29 6.29 6.29 6.34 6.35 6.37 6.4 6.4 6.4 6.4 6.4 6.4 6.4 6.4 6.45 6.62 6.62 6.62 6.62 6.62 6.62 6.62 6.62 6.74 252 274 275 162 94 170 114 164 167 218 236 248 251 253 216 21 58 59 104 157 246 257 270 14 15.03 11.07 10.91 19.7 22.52 19.4 21.86 19.61 19.46 17.32 16.15 15.37 15.07 14.95 17.43 28.24 24.63 24.62 22.29 19.97 15.43 14.43 12.75 29.49 66 Table 1.4 (cont’d) N22635 N22639 N18109 R20604 N21535 N21508 B22817 R20653 N21501 R20642 N21513 BC216 I13401 I89011 S20420 N18102 6.74 6.74 6.91 6.95 6.95 6.95 7.07 7.29 7.29 7.29 7.29 7.34 7.46 7.62 7.95 8.41 189 239 94 129 163 266 249 29 153 169 223 180 39 276 229 94 18.71 15.94 22.52 21.24 19.66 13.37 15.27 27.29 20.28 19.41 16.83 19.06 25.91 10.64 16.33 22.52 67 Table 1.5: Line, class, origin, and trial information for the entire training set. Line B04554 B10244 B15430 B15442 B15451 B16501 B16504 B16506 B16507 B17220 B17536 B17691 B17922 B18201 B18204 B18231 B18236 Class Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black B18504 (Adams) Black B19302 B19309 B19330 B19332 B19339 Black Black Black Black Black Origin Trial Historic/2021/2022 Historic/2021/2022 Historic Historic Historic Historic/2021 Historic/2021 Historic Historic Historic Historic Historic Historic Historic Historic/2021 Historic Historic/2021 Historic/2021/2022 2021 2021/2022 Historic/2021 Historic/2021 2021 MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU 68 Table 1.5 (cont’d) B19340 B19341 B19344 B19345 B19346 B20527 B20532 B20536 B20538 B20542 B20547 B20549 B20579 B20582 B20590 B20591 B20597 B20599 B20602 B20616 B20617 B20620 B20621 B20623 Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black 2021 2021 Historic/2021/2022 Historic/2021 Historic 2021 2021 2021/2022 2021 2021 2021/2022 2021 2021 2021 2021 2021/2022 2021 2021/2022 2021 2021 2021 2021 2021 2021 MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU 69 Table 1.5 (cont’d) B20627 B20629 B20632 B20639 B20642 B21701 B21702 B21703 B21704 B21705 B21706 B21707 B21708 B21709 B21710 B21711 B21712 B21713 B21714 B21715 B21716 B21717 B21718 B21719 Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black 2021 2021 2021 2021 2021 2021 2021 2021 2021 2021 2021 2021 2021/2022 2021 2021/2022 2021 2021 2021 2021 2021 2021 2021 2021 2021 MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU 70 Table 1.5 (cont’d) B21720 B21721 B21722 B21723 B21724 B22801 B22802 B22803 B22804 B22805 B22806 B22807 B22810 B22811 B22812 B22813 B22814 B22815 B22816 B22817 B22818 B22819 B22820 B22821 Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black 2021 2021 2021 2021 2021 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU 71 Table 1.5 (cont’d) B22822 B22823 B22825 B22826 B22827 B22828 B22829 B22830 B22831 B22832 B22833 B22834 B22835 B22836 B22837 B22838 B22839 B22840 B22841 B22822 B22823 B22825 B22826 B22827 Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU 72 Table 1.5 (cont’d) B22828 B22829 B22830 B22831 B22832 B22833 B22834 B22835 B22836 B22837 B22838 B22839 B22840 B22841 B22842 B22843 B22844 B22845 B22846 B22847 B22848 B22850 B22852 B22853 Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU 73 Table 1.5 (cont’d) B22854 B22855 B22856 B22857 B22859 B22860 B22861 B22862 B22863 B22865 B22866 B22867 B22868 B22870 B22873 B22874 B22875 B22876 BC216 BC269 Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Pink MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2021/2022 2021/2022 USPT-WM-12 Pinto Miklas, USDA Historic/2021/2022 SR9-5 Merlin Viper Small Red Miklas, USDA Historic/2021/2022 Navy Small Red Provita Provita Historic/2021/2022 Historic/2021 74 Table 1.5 (cont’d) Black Bear Black Provita 2021/2022 Bunsi Navy Historic/2021/2022 Tu, J.C., and W.D. Beversdorf. 1982. Beryl G122 N11283 N14218 N14229 N15331 N15341 N16401 N16405 N17504 N17505 N17506 N18102 N18103 N18105 N18109 N18117 N18122 N18130 N19223 Great Northern Syngenta Historic/2021/2022 Cranberry Landrace Historic/2021/2022 Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Historic/2021/2022 Historic Historic Historic Historic Historic Historic Historic Historic/2021 Historic Historic Historic/2021 2021 Historic Historic Historic/2021 Historic/2021 2021 MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU 75 Table 1.5 (cont’d) N19226 N19239 N19243 N19246 N19248 N19252 N19253 N19269 N19277 N19284 N19285 N19290 N20317 N20335 N20336 N20341 N20343 N20346 N20352 N20376 N20384 N20388 N20391 N20395 Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU 2021 Historic/2021 2021 2021/2022 Historic 2021 2021 2021 2021 2021 Historic/2021 2021 2021 2021 2021 2021 2021 2021 2021 2021 2021 2021/2022 2021 2021/2022 76 Table 1.5 (cont’d) N20401 N20404 N20405 N21501 N21502 N21503 N21504 N21505 N21506 N21507 N21508 N21509 N21510 N21511 N21512 N21513 N21514 N21515 N21516 N21517 N21518 N21519 N21520 N21521 Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy 2021 2021/2022 2021 2021 2021 2021 2021 2021 2021 2021 2021 2021 2021 2021/2022 2021 2021 2021 2021 2021 2021 2021 2021 2021 2021 MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU 77 Table 1.5 (cont’d) N21522 N21523 N21524 N21525 N21526 N21527 N21528 N21529 N21530 N21531 N21532 N21533 N21534 N21535 N22601 N22602 N22603 N22605 N22606 N22607 N22608 N22609 N22610 N22612 Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy 2021 2021 2021 2021 2021 2021 2021 2021 2021 2021 2021 2021 2021 2021 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU 78 Table 1.5 (cont’d) N22613 N22614 N22615 N22616 N22617 N22618 N22619 N22620 N22621 N22622 N22623 N22624 N22626 N22627 N22629 N22630 N22631 N22632 N22633 N22634 N22635 N22636 N22637 N22638 Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 2022 MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU 79 Table 1.5 (cont’d) N22639 R12844 R16503 R17602 R17603 R17604 R17605 R18402 R19502 R20604 R20612 R20614 R20624 R20625 R20627 R20629 R20632 R20633 R20635 R20636 R20637 R20639 R20642 R20652 Navy Red Red Red Red Red Red Red Red Red Red Red Red Red Red Red Red Red Red Red Red Red Red Red 2022 Historic/2021 Historic 2021 Historic Historic/2021 2021 2021 2021 2021 2021 2021 2021 2021 2021 2021 2021 2021 2021 2021 2021 2021 2021 2021 MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU 80 Table 1.5 (cont’d) R20653 R20659 R20667 R20669 R20683 R20684 R98026 S08418 S18904 S18907 S19307 S20405 S20420 Red Red Red Red Red Red Red Pink Pink Pink Pink Pink Pink MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU 2021 2021 2021 2021 2021 2021 Historic/2021 2021 Historic/2021 2021 2021 2021 2021 81 Table 1.6: ANOVA table for white mold resistance in the combined analysis (2021-2022) of the field trial. Variance Standard Error Genotype 0.45* Environment:Genotype 0.24* Error 1.82* 0.10 0.10 0.08 Values marked with an asterisk (*) are significant at the p=0.05 level. 82 Table 1.7: ANOVA table for yield under white mold disease pressure in the combined analysis (2021-2022) of the field trial. Variance Standard Error Genotype 13.51* Environment:Genotype 11.47* Error 13.90* 3.22 2.83 0.73 Values marked with an asterisk (*) are significant at the p=0.05 level. 83 REFERENCES Amadeu RR, Garcia AAF, Munoz PR, Ferrão LFV. AGHmatrix: genetic relationship matrices in R. Bioinformatics. 2023 Jul 1;39(7):btad445. doi: 10.1093/bioinformatics/btad445. PMID: 37471595; PMCID: PMC10371492. Arruda, M. P., Lipka, A. E., Brown, P. J., Krill, A. M., Thurber, C., Brown-Guedira, G., Dong, Y., Foresman, B. J., & Kolb, F. L. (2016). Comparing genomic selection and marker- assisted selection for Fusarium head blight resistance in wheat (Triticum aestivum L.). Molecular Breeding, 36(7), 84. https://doi.org/10.1007/s11032-016-0508-5 Bao, Y., Kurle, J. E., Anderson, G., & Young, N. D. (2015). Association mapping and genomic prediction for resistance to sudden death syndrome in early maturing soybean germplasm. Molecular Breeding, 35(6), 128. https://doi.org/10.1007/s11032-015-0324-3 Bian, Y., & Holland, J. B. (2017). Enhancing genomic prediction with genome-wide association studies in multiparental maize populations. Heredity, 118(6), 585–593. https://doi.org/10.1038/hdy.2017.4 Boland, G. J., & Hall, R. (1994). Index of plant hosts of Sclerotinia sclerotiorum. Canadian Journal of Plant Pathology, 16(2), 93–108. https://doi.org/10.1080/07060669409500766 Bolton, M. D., Thomma, B. P. H. J., & Nelson, B. D. (2006). Sclerotinia sclerotiorum (Lib.) de Bary: Biology and molecular traits of a cosmopolitan pathogen. Molecular Plant Pathology, 7(1), 1–16. https://doi.org/10.1111/j.1364-3703.2005.00316.x Bradbury PJ, Zhang Z, Kroon DE, Casstevens TM, Ramdoss Y, Buckler ES. TASSEL: software for association mapping of complex traits in diverse samples. Bioinformatics. 2007 Oct 1;23(19):2633-5. doi: 10.1093/bioinformatics/btm308. Epub 2007 Jun 22. PMID: 17586829. Butler, D. G., Cullis, B. R., Gilmour, A. R., Gogel, B. J., & Thompson, R. (2023). ASReml estimates variance components under a general linear. 188. de Azevedo Peixoto, L., Moellers, T. C., Zhang, J., Lorenz, A. J., Bhering, L. L., Beavis, W. D., & Singh, A. K. (2017). Leveraging genomic prediction to scan germplasm collection for crop improvement. PLOS ONE, 12(6), e0179191. https://doi.org/10.1371/journal.pone.0179191 De Los Campos, G., Sorensen, D., & Gianola, D. (2015). Genomic Heritability: What Is It? PLOS Genetics, 11(5), e1005048. https://doi.org/10.1371/journal.pgen.1005048 del Río, L. E., Venette, J. R., & Lamey, H. A. (2004). Impact of White Mold Incidence on Dry Bean Yield Under Nonirrigated Conditions. Plant Disease, 88(12), 1352–1356. https://doi.org/10.1094/PDIS.2004.88.12.1352 84 Diaz, L. M., Arredondo, V., Ariza-Suarez, D., Aparicio, J., Buendia, H. F., Cajiao, C., Mosquera, G., Beebe, S. E., Mukankusi, C. M., & Raatz, B. (2021). Genetic Analyses and Genomic Predictions of Root Rot Resistance in Common Bean Across Trials and Populations. Frontiers in Plant Science, 12, 629221. https://doi.org/10.3389/fpls.2021.629221 Đorđević, V., Ćeran, M., Miladinović, J., Balešević-Tubić, S., Petrović, K., Miladinov, Z., & Marinković, J. (2019). Exploring the performance of genomic prediction models for soybean yield using different validation approaches. Molecular Breeding, 39(5), 74. https://doi.org/10.1007/s11032-019-0983-6 Duhnen, A., Gras, A., Teyssèdre, S., Romestant, M., Claustres, B., Daydé, J., & Mangin, B. (2017). Genomic Selection for Yield and Seed Protein Content in Soybean: A Study of Breeding Program Data and Assessment of Prediction Accuracy. Crop Science, 57(3), 1325–1337. https://doi.org/10.2135/cropsci2016.06.0496 Edwards, S. M., Buntjer, J. B., Jackson, R., Bentley, A. R., Lage, J., Byrne, E., Burt, C., Jack, P., Berry, S., Flatman, E., Poupard, B., Smith, S., Hayes, C., Gaynor, R. C., Gorjanc, G., Howell, P., Ober, E., Mackay, I. J., & Hickey, J. M. (2019). The effects of training population design on genomic prediction accuracy in wheat. Theoretical and Applied Genetics. https://doi.org/10.1007/s00122-019-03327-y Endelman, J. B. (2011). Ridge Regression and Other Kernels for Genomic Selection with R Package rrBLUP. The Plant Genome, 4(3), 250–255. https://doi.org/10.3835/plantgenome2011.08.0024 Ender, M., & Kelly, J. D. (2005). Identification of QTL Associated with White Mold Resistance in Common Bean. Crop Science, 45(6), 2482–2490. https://doi.org/10.2135/cropsci2005.0064 Escobar, E., Oladzad, A., Simons, K., Miklas, P., Lee, R. K., Schroder, S., Bandillo, N., Wunsch, M., McClean, P. E., & Osorno, J. M. (2022). New genomic regions associated with white mold resistance in dry bean using a MAGIC population. The Plant Genome, 15(1). https://doi.org/10.1002/tpg2.20190 Fernández-González, J., Akdemir, D., & Isidro Y Sánchez, J. (2023). A comparison of methods for training population optimization in genomic selection. Theoretical and Applied Genetics, 136(3), 30. https://doi.org/10.1007/s00122-023-04265-6 Fuller, P. A., Coyne, D. P., & Steadman, J. R. (1984). Inheritance of Resistance to White Mold Disease in a Diallel Cross of Dry Beans 1. Crop Science, 24(5), 929–933. https://doi.org/10.2135/cropsci1984.0011183X002400050025x Genomestudio Software, www.illumina.com/techniques/microarrays/array-data-analysis- experimental-design/genomestudio.html. Accessed 22 Sept. 2023. 85 Haile, T. A., Walkowiak, S., N’Diaye, A., Clarke, J. M., Hucl, P. J., Cuthbert, R. D., Knox, R. E., & Pozniak, C. J. (2021). Genomic prediction of agronomic traits in wheat using different models and cross-validation designs. Theoretical and Applied Genetics, 134(1), 381–398. https://doi.org/10.1007/s00122-020-03703-z Heffner, E. L., Jannink, J., & Sorrells, M. E. (2011). Genomic Selection Accuracy using Multifamily Prediction Models in a Wheat Breeding Program. The Plant Genome, 4(1), plantgenome2010.12.0029. https://doi.org/10.3835/plantgenome2010.12.0029 Heffner, E. L., Lorenz, A. J., Jannink, J., & Sorrells, M. E. (2010). Plant Breeding with Genomic Selection: Gain per Unit Time and Cost. Crop Science, 50(5), 1681–1690. https://doi.org/10.2135/cropsci2009.11.0662 Hemingway, J., Schnebly, S. R., & Rajcan, I. (2021). Accuracy of genomic prediction for seed oil concentration in high‐oleic soybean populations using a low‐density marker panel. Crop Science, 61(6), 4012–4021. https://doi.org/10.1002/csc2.20607 Hoyos‐Villegas, V., Mkwaila, W., Cregan, P. B., & Kelly, J. D. (2015). Quantitative Trait Loci Analysis of White Mold Avoidance in Pinto Bean. Crop Science, 55(5), 2116–2129. https://doi.org/10.2135/cropsci2015.02.0106 Isidro, J., Jannink, J.-L., Akdemir, D., Poland, J., Heslot, N., & Sorrells, M. E. (2015). Training set optimization under population structure in genomic selection. Theoretical and Applied Genetics, 128(1), 145–158. https://doi.org/10.1007/s00122-014-2418-4 Jannink, J.-L., Lorenz, A. J., & Iwata, H. (2010). Genomic selection in plant breeding: From theory to practice. Briefings in Functional Genomics, 9(2), 166–177. https://doi.org/10.1093/bfgp/elq001 Juliana, P., He, X., Poland, J., Roy, K. K., Malaker, P. K., Mishra, V. K., Chand, R., Shrestha, S., Kumar, U., Roy, C., Gahtyari, N. C., Joshi, A. K., Singh, R. P., & Singh, P. K. (2022). Genomic selection for spot blotch in bread wheat breeding panels, full-sibs and half-sibs and index-based selection for spot blotch, heading and plant height. Theoretical and Applied Genetics, 135(6), 1965–1983. https://doi.org/10.1007/s00122-022-04087-y Juliana, P., Singh, R. P., Singh, P. K., Crossa, J., Rutkoski, J. E., Poland, J. A., Bergstrom, G. C., & Sorrells, M. E. (2017). Comparison of Models and Whole‐Genome Profiling Approaches for Genomic‐Enabled Prediction of Septoria Tritici Blotch, Stagonospora Nodorum Blotch, and Tan Spot Resistance in Wheat. The Plant Genome, 10(2). https://doi.org/10.3835/plantgenome2016.08.0082 Kolkman, J. M., & Kelly, J. D. (2000). An indirect test using oxalate to determine physiological resistance to white mold in common bean. Crop Science, 40(1), 281–285. https://doi.org/10.2135/cropsci2000.401281x 86 Larkin, D. L., Mason, R. E., Moon, D. E., Holder, A. L., Ward, B. P., & Brown-Guedira, G. (2021). Predicting Fusarium Head Blight Resistance for Advanced Trials in a Soft Red Winter Wheat Breeding Program With Genomic Selection. Frontiers in Plant Science, 12, 715314. https://doi.org/10.3389/fpls.2021.715314 Lee, M. (1995). DNA Markers and Plant Breeding Programs. Adv. Agron. doi: 10.1016/S0065- 2113(08)60542-8. Liu, X., Huang, M., Fan, B., Buckler, E. S., & Zhang, Z. (2016). Iterative Usage of Fixed and Random Effect Models for Powerful and Efficient Genome-Wide Association Studies. PLOS Genetics, 12(2), e1005767. https://doi.org/10.1371/journal.pgen.1005767 Lopez-Cruz, M., & de los Campos, G. (2021). Optimal breeding-value prediction using a sparse selection index. Genetics, 218(1), iyab030. https://doi.org/10.1093/genetics/iyab030 Mamidi, S., Miklas, P. N., Trapp, J., Felicetti, E., Grimwood, J., Schmutz, J., Lee, R., & McClean, P. E. (2016). Sequence‐Based Introgression Mapping Identifies Candidate White Mold Tolerance Genes in Common Bean. The Plant Genome, 9(2). https://doi.org/10.3835/plantgenome2015.09.0092 Massman, J. M., Jung, H. G., & Bernardo, R. (2013). Genomewide Selection versus Marker‐ assisted Recurrent Selection to Improve Grain Yield and Stover‐quality Traits for Cellulosic Ethanol in Maize. Crop Science, 53(1), 58–66. https://doi.org/10.2135/cropsci2012.02.0112 Mensack, M. M., Fitzgerald, V. K., Ryan, E. P., Lewis, M. R., Thompson, H. J., & Brick, M. A. (2010). Evaluation of diversity among common beans (Phaseolus vulgaris L.) from two centers of domestication using “omics” technologies. BMC Genomics, 11(1), 686. https://doi.org/10.1186/1471-2164-11-686 Merrick, L. F., Burke, A. B., Chen, X., & Carter, A. H. (2021). Breeding With Major and Minor Genes: Genomic Selection for Quantitative Disease Resistance. Frontiers in Plant Science, 12, 713667. https://doi.org/10.3389/fpls.2021.713667 MBC 2022. Michigan Bean Commission Production Practices Survey. 2022. https://michiganbean.com/https-michiganbean-com-wp-content-uploads-2022- production-practices/ Miklas, P. N. (2007). Marker-Assisted Backcrossing QTL for Partial Resistance to Sclerotinia White Mold in Dry Bean. Crop Science, 47(3), 935–942. https://doi.org/10.2135/cropsci2006.08.0525 Miklas, P. N., Hauf, D. C., Henson, R. A., & Grafton, K. F. (2004). Inheritance of ICA Bunsi‐ Derived Resistance to White Mold in a Navy × Pinto Bean Cross. Crop Science, 44(5), 1584–1588. https://doi.org/10.2135/cropsci2004.1584 87 Miklas, P. N., Johnson, W. C., Delorme, R., & Gepts, P. (2001). QTL Conditioning Physiological Resistance and Avoidance to White Mold in Dry Bean. Crop Science, 41(2), 309–315. https://doi.org/10.2135/cropsci2001.412309x Miklas, P. N., Porter, L. D., Kelly, J. D., & Myers, J. R. (2013). Characterization of white mold disease avoidance in common bean. European Journal of Plant Pathology, 135(3), 525– 543. https://doi.org/10.1007/s10658-012-0153-8 Miorini, T. J. J., Raetano, C. G., & Everhart, S. E. (2017). Control of white mold of dry bean and residual activity of fungicides applied by chemigation. Crop Protection, 94, 192–202. https://doi.org/10.1016/j.cropro.2016.12.023 Mkwaila, W., Terpstra, K. A., Ender, M., & Kelly, J. D. (2011). Identification of QTL for agronomic traits and resistance to white mold in wild and landrace germplasm of common bean: QTL for resistance to white mold in bean germplasm. Plant Breeding, 130(6), 665–672. https://doi.org/10.1111/j.1439-0523.2011.01876.x Osorno, J. M., McClean, P. E., North Dakota State University, USA, Close, T., & University of California (Riverside), USA. (2018). Advanced breeding techniques for grain legumes in the genomics era. In formerly International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), India & S. Sivasankar (Eds.), Burleigh Dodds Series in Agricultural Science (pp. 57–86). Burleigh Dodds Science Publishing. https://doi.org/10.19103/AS.2017.0023.03 Odilbekov, F., Armoniené, R., Koc, A., Svensson, J., & Chawade, A. (2019). GWAS-Assisted Genomic Prediction to Predict Resistance to Septoria Tritici Blotch in Nordic Winter Wheat at Seedling Stage. Frontiers in Genetics, 10, 1224. https://doi.org/10.3389/fgene.2019.01224 Park, S. O., Coyne, D. P., Steadman, J. R., & Skroch, P. W. (2001). Mapping of QTL for Resistance to White Mold Disease in Common Bean. Crop Science, 41(4), 1253–1262. https://doi.org/10.2135/cropsci2001.4141253x Poland, J., & Rutkoski, J. (2016). Advances and Challenges in Genomic Selection for Disease Resistance. Annual Review of Phytopathology, 54(1), 79–98. https://doi.org/10.1146/annurev-phyto-080615-100056 Rice, B., & Lipka, A. E. (2019). Evaluation of RR‐BLUP Genomic Selection Models that Incorporate Peak Genome‐Wide Association Study Signals in Maize and Sorghum. The Plant Genome, 12(1), 180052. https://doi.org/10.3835/plantgenome2018.07.0052 Rutkoski, J., Benson, J., Jia, Y., Brown-Guedira, G., Jannink, J.-L., & Sorrells, M. (2012). Evaluation of Genomic Prediction Methods for Fusarium Head Blight Resistance in Wheat. The Plant Genome, 5(2), 51–61. https://doi.org/10.3835/plantgenome2012.02.0001 88 Sarinelli, J. M., Murphy, J. P., Tyagi, P., Holland, J. B., Johnson, J. W., Mergoum, M., Mason, R. E., Babar, A., Harrison, S., Sutton, R., Griffey, C. A., & Brown-Guedira, G. (2019). Training population selection and use of fixed effects to optimize genomic predictions in a historical USA winter wheat panel. Theoretical and Applied Genetics, 132(4), 1247– 1261. https://doi.org/10.1007/s00122-019-03276-6 Schwartz, H. F., Casciano, D. H., Asenga, J. A., & Wood, D. R. (1987). Field Measurement of White Mold Effects upon Dry Beans with Genetic Resistance or Upright Plant Architecture1. Crop Science, 27(4), cropsci1987.0011183X002700040018x. https://doi.org/10.2135/cropsci1987.0011183X002700040018x Schwartz, H. F., & Singh, S. P. (2013). Breeding Common Bean for Resistance to White Mold: A Review. Crop Science, 53(5), 1832–1844. https://doi.org/10.2135/cropsci2013.02.0081 Shi, A., Gepts, P., Song, Q., Xiong, H., Michaels, T. E., & Chen, S. (2021b). Genome-Wide Association Study and Genomic Prediction for Soybean Cyst Nematode Resistance in USDA Common Bean (Phaseolus vulgaris) Core Collection. Frontiers in Plant Science, 12, 624156. https://doi.org/10.3389/fpls.2021.624156 Singh, S. P., & Schwartz, H. F. (2010). Breeding Common Bean for Resistance to Diseases: A Review. Crop Science, 50(6), 2199–2223. https://doi.org/10.2135/cropsci2009.03.0163 Singh, S. P., Schwartz, H. F., Terán, H., Viteri, D., & Otto, K. (2014). Pyramiding white mould resistance between and within common bean gene pools. Canadian Journal of Plant Science, 94(5), 947–954. https://doi.org/10.4141/cjps2013-321 Spindel, J. E., Begum, H., Akdemir, D., Collard, B., Redoña, E., Jannink, J.-L., & McCouch, S. (2016). Genome-wide prediction models that incorporate de novo GWAS are a powerful new tool for tropical rice improvement. Heredity, 116(4), 395–408. https://doi.org/10.1038/hdy.2015.113 Terán, H., & Singh, S. P. (2009). Gamete selection for improving physiological resistance to white mold in common bean. Euphytica, 167(3), 271–280. https://doi.org/10.1007/s10681-009-9885-8 Tiede, T., & Smith, K. P. (2018). Evaluation and retrospective optimization of genomic selection for yield and disease resistance in spring barley. Molecular Breeding, 38(5), 55. https://doi.org/10.1007/s11032-018-0820-3 Uebersax, M. A., Cichy, K. A., Gomez, F. E., Porch, T. G., Heitholt, J., Osorno, J. M., Kamfwa, K., Snapp, S. S., & Bales, S. (2023). Dry beans ( PHASEOLUS VULGARIS L.) as a vital component of sustainable agriculture and food security—A review. Legume Science, 5(1). https://doi.org/10.1002/leg3.155 USDA Crop Production 2022 Summary 01/12/2023. (2022). Crop Production. 89 Vasconcellos, R. C. C., Oraguzie, O. B., Soler, A., Arkwazee, H., Myers, J. R., Ferreira, J. J., Song, Q., McClean, P., & Miklas, P. N. (2017). Meta-QTL for resistance to white mold in common bean. PLOS ONE, 12(2), e0171685. https://doi.org/10.1371/journal.pone.0171685 Wang, J., & Zhang, Z. (2021). GAPIT Version 3: Boosting Power and Accuracy for Genomic Association and Prediction. Genomics, Proteomics & Bioinformatics, 19(4), 629–640. https://doi.org/10.1016/j.gpb.2021.08.005 Wen, Z., Tan, R., Zhang, S., Collins, P. J., Yuan, J., Du, W., Gu, C., Ou, S., Song, Q., An, Y. C., Boyse, J. F., Chilvers, M. I., & Wang, D. (2018). Integrating GWAS and gene expression data for functional characterization of resistance to white mould in soya bean. Plant Biotechnology Journal, 16(11), 1825–1835. https://doi.org/10.1111/pbi.12918 Zhang, J., Song, Q., Cregan, P. B., & Jiang, G.-L. (2016). Genome-wide association study, genomic prediction and marker-assisted selection for seed weight in soybean (Glycine max). Theoretical and Applied Genetics, 129(1), 117–130. https://doi.org/10.1007/s00122-015-2614-x 90 CHAPTER TWO: SCREENING DRY BEAN BREEDING LINES FOR RESISTANCE TO COMMON MICHIGAN ROOT ROT PATHOGENS ABSTRACT Root rot is a major yield limiting disease of dry bean (Phaseolus vulgaris) production in the US and worldwide. Specifically, disease symptoms conferred by the soil borne fungal pathogens Fusarium oxysporum and Rhizoctonia solani cause up to 84% yield loss in susceptible dry bean cultivars through damage of root biomass, reduced vigor, and plant death. This study evaluated a diverse set of breeding lines and diversity panel lines from 2021 to 2022 for field resistance to root rots conferred by these pathogens. Secondary objectives were to establish correlations between field and greenhouse screens and non-destructive traits correlated to root rot resistance to Fusarium oxysporum for ease of phenotyping. All trials were successful in identifying significant variation for root rot resistance. Significant correlations were found between the field trial and the greenhouse trials (P=0.03, 0.71). Significant trait correlations were also identified between root rating and fresh (P=0.01, -0.67) and dry root (P=0.01, -0.64) and dry shoot (P=0.04, -0.43) weights in the greenhouse. Ultimately multiple lines are recommended as parents for future root rot resistance breeding efforts. INTRODUCTION Dry bean (Phaseolus vulgaris L.) is one of the most important grain legumes for human consumption worldwide, providing an important source of key dietary nutrients such as protein and fiber (Uebersax et al., 2023). Dry beans are a dietary staple in Latin America and sub-Saharan Africa (Leterme & Carmenza Muũoz, 2002; Paparu et al., 2018) and an important agricultural commodity in the US (primarily North Dakota, Minnesota, and Michigan), Latin America, and other regions. However, biotic stress via fungal disease infection is among the main constraints 91 that limit yield and increase annual management cost of dry bean production. Among the most yield limiting fungal borne diseases affecting dry bean production is root rot, which is conferred by a soilborne disease complex of multiple fungal and oomycete pathogens including members of Fusarium sp., Rhizoctonia sp., Pythium sp., Macrophomina sp., and Alternaria sp. (Bilgi et al., 2011; Harveson et al., 2005; Schwartz, 2011; Sendi et al., 2020; Singh & Schwartz, 2010). Root rots are widespread in dry bean production regions and cause significant yield losses in the US, Africa, and Central and South America (Abawi and Pastor Corrales, 1990) For example, root rot is responsible for losses estimated at 221.000 metric tons per year in sub-Saharan Africa (Paparu et al., 2018). Dry beans have two primary gene pools resulting from two independent domestication events in the Andes and Middle-Americas (Gepts, and Debouck, 1991; Mensack et al., 2010). The Andean gene pool primarily consists of large-seeded beans from the kidney (dark red, light red, and white), yellow, and cranberry market classes. The Middle-American gene pool consists of small to medium seeded market classes from the navy, black, small red, pink, pinto and great northern market classes. Previous research has established susceptibility to root rot in both dry bean gene pools, but susceptibility to individual pathogens differs. Large-seeded dry bean cultivars from the Andean gene pool are generally more susceptible to Fusarium species while the small seeded types from the Middle-American gene pool are generally more susceptible to Rhizoctonia species, with Andean types being more susceptible to root rots overall (Conner et al., 2014; Schneider et al., 2001). Andean genotypes in particular, tend to have less robust root systems which is thought to be why they have lower levels of resistance to root rots (Cichy et al., 2007; Román- Avilés & Kelly, 2005; Schneider et al., 2001). 92 Michigan is the second largest dry bean producer in the United States, producing over 400 million pounds of dry edible beans per year and contributing to a farm gate value of over $139 million (USDA Crop Production Summary 2022). It also leads the nation in the black, small red, and navy bean market classes in terms of acreage (USDA Crop Production Summary 2022). Michigan's temperate climate, with warm humid summers, are highly conducive to fungal pathogens. Over the past few years Michigan growers have rated root rot as the second most important disease and it is considered the first most impactful disease in large seeded Andean bean varieties (MBC, 2022). In Michigan, the most common root rot pathogens belong to the Fusarium and Rhizoctonia species as well as multiple oomycete species and species diversity varies by region (Jacobs et al., 2019). Root rots conferred by these fungal pathogens cause up to 84% yield loss in susceptible dry bean cultivars through damage of root biomass, reduced vigor, and whole plant death (Jacobs et al., 2019). Root rot disease symptoms vary depending on the pathogen(s) involved, but are primarily characterized by root lesions, root and foliar biomass loss, and reduced stand. Fusarium and Rhizoctonia infections of dry bean are both primarily characterized by water-soaked, dark brown to rust colored lesions on the root and death of lateral roots (Hall et al. 1991; Hagedorn et al. 1994). Fusarium root rot develops slowly in response to prolonged cool and wet environmental conditions, and generally leads to an overall vigor and yield decline. Entire plant death is rare and usually occurs later in the season in only the most severe infections. Conversely, Rhizoctonia root rot tends to affect plants early in the season primarily through reduced stand resulting from seed rot, seedling blight, and pre/post emergence damping off (Gossen et al., 2016). Rhizoctonia root rot infection is characterized by the formation of deep sunken lesions at the soil level (Conner et 93 al., 2014). Cool wet weather or flooding events can increase infection severity of both pathogens due to additional plant stress (Kumar and Kudada, 2018). Managing root rot can be challenging due to the durability and extended viability of chlamydospores in soil and plant residue and broad host range (Katan, 2017). Current agronomic management practices include fungicidal seed and soil treatments, reduced irrigation, crop rotation, cover crops, seedbed preparation, and other agronomic practices are currently used to combat yield loss from root rots (Abawi and Pastor Corrales, 1990; Gossen et al., 2016; Harveson et al., 2005; Rubiales et al., 2015). However, they are often not sustainable or economically feasible and reduce productivity of dry bean fields. For example, control of root rots through reduced irrigation limits yield per acre. Fungicidal seed and soil treatments are expensive and have adverse effects on the environment. Furthermore, these practices do not entirely prevent root rot infection due to its virulence and ability for thick-walled spores, hyphae, or sclerotia to overwinter in the soil for multiple years (Schwartz, 2011). Biocontrols have emerged as a significant research objective for control of root rot pathogens, but there are currently no commercial biocontrols available that significantly reduce infection (Hassan Dar et al., 1997; Sendi et al., 2020). Therefore, the most sustainable, durable, and cost-effective method to ensure protection against infection would be to develop resistant dry bean cultivars, along with appropriate management practices. Genetic resistance to root rots in dry bean is quantitatively inherited, primarily through root architecture traits such as root weight, root length and root mass (Haus et al., 2020; Kamfwa et al., 2013; Snapp et al., 2003; Wang et al., 2018) and physiological avoidance (Kamfwa et al., 2013; Mukankusi et al., 2011; Román-Avilés et al., 2004; Snapp et al., 2003). The quantitative inheritance of these traits has been demonstrated by the low narrow-sense heritability estimates (26%-44%) reported and many small effect quantitative trait loci (QTL) found for root rot (Román- 94 Avilés et al., 2011; Nakedde et al., 2016; Kamfwa et al., 2018; Wang et al., 2018; Zitnick-Anderson et al., 2020). Previous research has established the importance of high root biomass cultivars with high density of lateral roots, high basal root number, and many adventitious roots for root rot avoidance (Haus et al., 2020; Román-Avilés et al., 2004; Snapp et al., 2003). These traits have been identified as significant targets for improving resistance. Resistance to root rot is primarily found in the small-seeded Middle-American bean germplasm compared to the highly susceptible large-seeded beans of Andean origin and has served as the only source of resistance (Cichy et al., 2007; Mukankusi et al., 2011; Román-Avilés & Kelly, 2005; Schneider et al., 2001) Although few studies to date have evaluated genetic resistance to Fusarium oxysporum or Rhizoctonia solani, multiple studies have established genomic regions associated with resistance to root rots conferred by other Fusarium sp. (Hagerty et al., 2015; Kamfwa et al., 2013; Nakedde et al., 2016; Román-Avilés & Kelly, 2005; Schneider et al., 2001; Wang et al., 2018; Zitnick- Anderson et al., 2020). These studies often identify many different QTL which highlights the highly quantitative inheritance method of resistance. QTL studies of resistance to Fusarium sp. in dry bean have identified genomic regions associated with root traits, plant immune/defense mechanisms, and other disease resistance genes (Hagerty et al., 2015; Nakedde et al., 2016; Wang et al., 2018; Zitnick-Anderson et al., 2020). One study to date has evaluated genetic resistance to Rhizoctonia solani- (Oladzad et al. 2019). In this study, researchers evaluated the Andean and Middle American diversity panels and observed multiple QTL associated with gene models encoding proteins like known disease resistance genes. Other Rhizoctonia solani resistance studies have focused on screening and ranking breeding lines and establishing trait correlations (Adesemoye et al., 2018; Conner et al., 2014; Peña et al., 2013). 95 Progress to introgress root rot resistance has been hindered by the difficulty of pyramiding resistance genes given the quantitative nature of inheritance of root rot, inconsistent screening methods, and screening dependence on the presence of the pathogen under suitable environmental conditions (Hagerty et al., 2015; Nakedde et al., 2016; Wang et al., 2018). Phenotyping for root rot disease in the field often requires laborious and destructive “shovelomics'' techniques (Burridge et al., 2016; Trachsel et al., 2011) that involve digging up the root system of individual plants to evaluate root rot disease severity or manually counting stands multiple times per season to evaluate post-emergence damping off. There have been multiple phenotyping methods proposed for the greenhouse and field evaluation of Fusarium root rots including the liquid inoculum method (Schneider & Kelly, 2000), inoculum layer method (Chaudhary et al., 2006), and nutrient culture (Boomstra et al., 1977). Inoculated grain planted alongside the crop seed or natural infection are common methods for field evaluation (Haus et al., 2020; Pandey et al., 2020; Wang et al., 2018). However, the presence of multiple resistance mechanisms through both physiological resistance and architectural avoidance complicates screening. Greenhouse trials are preferred by researchers because they provide a controlled environment, have higher heritability, ease of screening, and less environmental dependance, but they do not screen for some avoidance mechanisms that can only be observed in the field. Conversely, field screens allow for measurement of both architectural traits and physiological resistance, but are often hampered by low heritability, high coefficient of variation (CV), high error variances, and natural pathogen presence (Guzman, 2016.; Hagerty et al., 2015; Nakedde et al., 2016; Román-Avilés et al., 2004; Román-Avilés & Kelly, 2005). Furthermore, these destructive phenotyping methods require the whole plant to be dug up with a shovel in the field or removed from the pot in the greenhouse to be evaluated for root rot symptoms, which is laborious and prevents additional measurements of yield and other agronomic traits 96 throughout the season. Establishments of aboveground traits related to root rot disease severity would enable high-throughput methods of phenotyping such as UAS imaging to assist breeders in screening for root rot resistant varieties (Guo et al., 2021; Lu et al., 2019; Manganiello et al., 2021; Marzougui et al., 2019). Therefore, the primary objective of this study was to screen advanced breeding lines and a subset of an Andean diversity panel for Fusarium oxysporum and Rhizoctonia solani under field and greenhouse conditions. Specific objectives were to ii) evaluate a set of panels for F. oxysporum and R. solani resistance across multiple years, ii) compare a subset of the Fusarium lines grown in the field and in the greenhouse, and iii) compare the rankings between artificial inoculation and natural infestation in the field, and iv) determine agronomic traits correlated to Fusarium oxysporum root rot resistance that could provide an alternative phenotyping method. MATERIALS AND METHODS Plant Material A set of three panels were used to evaluate Fusarium and Rhizoctonia resistance during the 2021 and 2022 field season (Table 2.1, Table 2.2). Specifically, two panels were used to evaluate F. oxysporum root rot resistance and one for R. solani. The F. oxysporum panels included a set 2021 and 2022 advanced kidney bean breeding lines from the Michigan State University dry bean breeding program (KDB) and a subset of 38 lines from the Andean Diversity Panel (ADP) (Cichy et al., 2015). The KDB trial consisted of 53 advanced breeding lines (AYT) from the large-seeded Andean gene pool (kidney and yellow market classes) of the MSU breeding program and 13 commercial varieties. 29 lines were evaluated in 2021 and 27 lines were evaluated in 2022, with 20 AYT lines conserved in both years. The motivation for this approach was to eliminate inferior breeding lines discarded from the MSU Advanced Yield Trial (AYT) after year one and replace 97 them with newer breeding lines that had performed well in the Preliminary Yield Trials (PYTs). The commercial varieties Coho, Clouseau, Denali, Red Cedar, Red Hawk, and Snowdon were added as performance checks. Clouseau was chosen as it has been widely grown in Michigan, while the others represent recently released MSU developed light red, dark red, or white kidney varieties that have shown superior yield potential despite ambient root rot disease pressure. All lines in the ADP trial were evaluated in both years. The check lines used in the ADP trial included the resistant line VAX 3 (Bilgi et al., 2011; Guzman, 2016), Cabernet (susceptible), Dynasty (moderately resistant), and Talon (moderately resistant). The third panel (MAB) consisted of 65 AYT breeding lines from the small-seeded Middle American gene pool (great northern, pinto, small red, pink, black, and navy market classes) was used to evaluate R. solani root rot resistance in the field during the 2021 and 2022 growing seasons. The 2021 set consisted of 24 AYT lines and 19 commercial lines while the 2022 set consisted of 31 AYT lines and 22 commercial lines. Across the panel 16 breeding lines and 15 commercial lines were conserved in both years. There currently aren’t any established performance checks for Rhizoctonia solani resistance in dry bean. Commercial lines used for comparison are listed in Table 2. Field Experiment Three field trials (ADP, KDB, and MAB) field were planted in a disease nursery in East Lansing, MI using a randomized-complete block design (RCBD) with four replicates. Entries were planted in four-row plots, 160 seeds per replicate (40 per row), 10 ft in length with 30 in row spacing, where two rows contained non inoculated plants, and two rows contained plants inoculated with F. oxysporum isolate F_14-38 or R. solani AG2-2 isolate Rs_14-17 colonized grain. Isolates used for the inoculum were collected from dry bean plants in Michigan. Inoculum 98 consisted of millet (F. oxysporum) or barley (R. solani) kernels colonized with mycelium at a concentration of 4.23 x 106 per gram and air dried in a drying oven. In the F. oxysporum trials (ADP and KDB) the colonized millet seed was planted in the furrow with the dry bean seed at a rate of 34.45 ml/linear meter in 2021 and increased to 42.65 ml/linear meter in 2022 following the protocol described in (Haus et al., 2020; Jacobs et al., 2018) In the R. solani (AM) trial the colonized barley seed was planted in the furrow with the dry bean seed at a rate of 4.27 ml/linear meter in 2021 and 3.28 ml/linear meter in 2022. The KDB trial was also planted at Montcalm Research Farm in the 2021 and 2022 growing season near Entrican in Montcalm County, Michigan which is a location previously established to have substantial root rot disease pressure. Previous literature on root rot disease in Michigan and visual observation has established F. solani fsp. phaseoli present in this site which makes it ideal to compare natural infection conditions with artificial inoculation (Jacobs et al., 2019; Román- Avilés et al., 2004). The experimental design was an alpha-lattice design with three replications using four row plots 6.1 m in length with 50 cm row spacing. Phenotypic Evaluation Plants were collected and rated for disease severity in the ADP and KDB F. oxysporum trials and the KDB natural infection trial. One month after germination (V3-V4 stage), five plants from each line, replicate, and treatment were uprooted with a shovel, washed to remove soil, and evaluated for root rot disease severity using the 1 to 7 scale, developed by Schneider and Kelly where 1 indicates no disease and 7 indicates a non-functional, completely rotted root system (Schneider & Kelly, 2000). Figure 1.1 shows an image of dry bean plants used to calibrate disease scores corresponding to the 1-7 rating scale. In 2022, the five plants from each replication were bulked and weighed (g) after root rot disease severity was recorded. Multiple vigor and stand count 99 measurements were taken throughout the growing season both years. Vigor was measured by visually observing overall plot wise canopy biomass and closure on a scale of 1-9 following the rating system in Van Schoonhoven, 1987. In the MAB R. solani trial, stand count per row for the inoculated and non-inoculated treatments was collected at multiple time points during the growing season. Post-emergence damping off was defined as the difference in stand count between the first and last measurement of the season. In the natural infection trial of the KDB lines, root rot disease severity was evaluated as outlined previously at 6-8 weeks after planting. Standard agronomic traits including plant height, lodging score, days to flower, and maturity were also collected in the field. Plots were trimmed to 4.9m prior to harvest and the center two rows were harvested with a Wintersteiger Classic plot combine (Wintersteiger AG, Austria). Yield data was obtained utilizing a Harvestmaster H2 Classic Graingage (Juniper Systems, Logan, UT) to record plot weight and moisture. Prior to data analysis, seed yield was standardized to 18% moisture. Greenhouse Experiment A trial run of 5 genotypes was used to identify optimal inoculum concentration and harvest date to elucidate differences in resistance to Fusarium oxysporum in the greenhouse and compare rankings for similar genotypes to the field trials. Inoculated grain was prepared using the same procedures as the field trial. The containers used were 354-mL coffee cups with three holes on the bottom for drainage. Different quantities of inoculum (2, 3.5, and 5 grams) were tested to determine the correct ratio of inoculum to best separate resistance among lines in the main trial. Inoculum was mixed thoroughly with vermiculite medium in a plastic bag. Bean seeds were placed on the top and covered with another layer of vermiculite. Non-inoculated replicates followed the 100 same methods, minus the inoculum step. Plants were arranged in a RCBD with 3 replicates. Two harvest dates (3 and 4 weeks) were also tested to determine the best time to evaluate differences in root symptoms. A subset of the 11 most resistant and susceptible lines (n=22) and the resistant check VAX3 obtained from the 2021 KDB and ADP field trials were selected to evaluate root rot disease severity under greenhouse conditions. Based on results of the test trial, a 3 week harvest date and an inoculum concentration of 2 grams per cup was chosen. Production of inoculum and planting methods followed the same procedure as in the test greenhouse trial. For each genotype twelve seeds were planted, one in each coffee cup, using a randomized complete block experimental design with three inoculated and three non-inoculated replicates of 6 plants each. All plants were uprooted, washed with water, and evaluated for root rot disease severity using the 1-7 root rot disease severity scale mentioned previously. Fresh and dry shoot and root weights (g) for each plant were also taken. Data Analysis The statistical analysis for all experiments were conducted in the ASReml-R package in the R programming language (Butler et al., 2023). The normality of residuals was evaluated using quantile plots and residual plots and outliers were removed as necessary. For all experiments a linear mixed model was fit and LSmeans for root rot disease severity (ADP, KDB, GH, and KDB Natural) or post-emergence damping off (MAB) were calculated using the predict function of the ASReml-R package (Butler et al., 2023) utilizing the inoculated treatment only as follows: ADP, KDB, and MAB field trials Yijk = μ + Linei + Envk + Rep(Env)jk + (Line x Env)ik + eijk Eq. 1 101 Where Yijk is the observed phenotype. μ is the overall mean, Linei is the fixed effect of the i-th line, Envk is the fixed effect of the kth environment, Rep(Env)jk is the random effect of the interaction between the the j-th rep within the k-th environment, (Line x Env)ik is the fixed effect of interaction between the i-th line and k-th environment, and eijk is the random residual term. ADP and KDB greenhouse trial Yij = μ + Linei + Repj + eij Eq. 2 Where Yij is the observed phenotype. μ is the overall mean, Linei is the fixed effect of the i-th line, Repj is the random effect of the jth rep, and eij is the random residual term. Natural Infection Trial (KDB) Yijkl = μ + Linei + Envj + iBlock(Rep)kl + Rep(Env)jl + (Line x Env)ij + eijkl Eq. 3 Where Yijkl is the observed phenotype (root rot disease severity score). μ is the overall mean, Linei is the fixed effect of the i-th line, Envj is the fixed effect of the jth environment, iBlock(Rep)kl is the random effect of the interaction between the the k-th incomplete block within the l-th rep, Rep(Env)jl is the random effect of the interaction between the the l-th rep within the j- th environment, (Line x Env)ij is the fixed effect of interaction between the i-th line and j-th environment, and eijkl is the random residual term. To estimate broad sense heritability (H2) for on an entry means basis, variance components were extracted fitting the previous equations with all terms random. Heritability was estimated as follows: 102 𝐻2 = 2 𝜎𝐺 2 𝜎𝐺𝐸 𝑛 + 2 𝜎𝑒 𝑟𝑛 2 + 𝜎𝐺 Eq. 4 Where σ2 G ,σ2 GxE ,and , σ2 e are the genotype, genotype by environment interaction, and error variances, n is the number of environments and r is the number of reps. Pairwise comparisons between lines were calculated using Fisher’s LSD using the assigned LSD value resulting from the allDifferences.data.frame function in the ASRemlPlus package (Brien et al. 2023). The effect of inoculum was validated using a student's t-test of root rot disease severity scores or post- emergence damping off between the inoculated and non-inoculated treatments. Population coefficient of variation (CV) was calculated as ((σ/μ) * 100) = CV%. Correlations Correlations between trials were evaluated to compare phenotyping in the greenhouse vs field as well as artificial vs natural disease pressure using the LSmeans for lines present in both trials. Additionally, correlations between root rot rating (or post emergence damping off) and various agronomic traits were calculated using trait LSmeans. Correlations between trials were performed using the cor.test function in R and correlations between traits were visualized using the corrplot function in the R package corrplot (Wei et al., 2021). RESULTS In this study we evaluated three diverse dry bean panels (ADP, KDB, MAB) in the field for variance in resistance to two prevalent Michigan root rot pathogens. Additionally, a subset of the ADP and KDB trials were evaluated in the greenhouse and the entire KDB trial was also evaluated in natural root rot infection conditions. Significant differences were found between inoculated and non-inoculated plots using a student’s t-test which indicated significant differences 103 in mean value between treatments in all trials and years tested (Table 2.3). Overall, genetic variation was low with the lowest being for the F. oxysporum trials compared to the R. solani trial. This led to a lower heritability (0.12-0.29) in the F. oxysporum trials (KDB, ADP, GH), low to moderate in the KDB natural infection trials (0.11-0.43) and moderate in the MAB R. solani trial (0.49-0.57) with the greenhouse having the lowest heritability overall (Table 2.4). Significant genotypic variation was found for all F. oxysporum trials across all years except the 2022 ADP trial and Greenhouse trial (Table 2.4). There was no significant genotype by year effect across any of the field trials. Coefficient of variation (CV) was low to moderate for all F. oxysporum trials (31.04-49.26%), low in the KDB natural infection trial (23.94-35.32%), and high for the MAB trials (65.11-71.06%) (Table 2.4). Fusarium Oxysporum Field and Greenhouse Screens After outlier removal and model fitting, LSmeans in the field for the ADP trial ranged from 1.8-3.9 in 2021, 2.3-4.6 in 2022, and 2.05-4 in the combined analysis (Figure 2.2a). The five most resistant lines in the combined analysis were ADP99, ADP444, ADP481, Dynasty, and ADP462 with scores ranging from 2.05-2.7 (Table 2.5, Figure 2.3). While the checks, Dynasty, Talon, Cabernet, and VAX3 had disease severity scores of 2.65, 2.87, 3.3, and 3.47 respectively. In 2021, the five most resistant lines were ADP99, ADP444, ADP462, ADP43, and ADP612 with scores ranging from 1.8-2.3 (Table 2.6, Figure 2.4). While the checks, Dynasty, Talon, Cabernet, and VAX3 had disease severity scores of 2.6, 2.7, 3.3, and 3.7 respectively. In 2022, the five most resistant lines were ADP99, ADP111, ADP481, ADP15, and ADP4 with scores ranging from 2.3- 2.65 (Table 2.7, Figure 2.5). While the checks, Dynasty, Talon, Cabernet, and VAX3 had disease severity scores of 2.7, 3.05, 3.3, and 3.25, respectively. 104 The LSmeans in the field for the KDB trial ranged from 2.2-3.3 in 2021, 2.1-4.3 in 2022, and 2.4-4 in the combined analysis (Figure 2.2b). The five most resistant lines in the combined analysis were K20730, Y19817, Beluga, K19832, and Clouseau with scores ranging from 2.39- 2.73 (Table 2.8, Figure 2.6). While the checks, Clouseau, Denali, Snowdon, Coho, and Red Cedar had disease severity scores of 2.7, 2.73, 2.84, 2.75, and 3.35, respectively. In 2021, the five most resistant lines were K19610, K20712, Y19817, K16911, and K18907 with scores ranging from 2.3-2.5 (Table 2.9, Figure 2.7). While the checks, Clouseau, Denali, Snowdon, Coho, and Red Cedar had disease severity scores of 2.73, 2.85, 2.8, 2.75, and 3.35, respectively. In 2022, the five most resistant lines were K20730, K19832, Y19808, Beluga, and Y19817 with scores ranging from 2.09-2.48 (Table 2.10, Figure 2.8). While the checks, Clouseau, Denali, Snowdon, Coho, and Red Cedar had disease severity scores of 2.75, 2.6, 2.89, 3.25, and 3.5, respectively. The LSmeans for the greenhouse trial ranged from 4-6 (Figure 2.2c). The five most resistant lines in the greenhouse were ADP481, K20730, VAX3, ADP462, and K20712 with scores ranging from 4-4.78 (Table 2.11, Figure 2.9). VAX 3 had a disease score of 4.5 and Dynasty had a disease score of 6. Natural Infection The LSmeans in the field for the KDB natural infection trial ranged from 3.45-5.41 in 2021, 2.47-5.58 in 2022, and 3.17-5.06 in the combined analysis (Figure 2.2b). The five most resistant lines in the combined analysis were K19832, K19817, K20745, K20717, and K20730 with scores ranging from 3.17-3.57 (Table 2.12, Figure 2.10). While the checks, Clouseau, Denali, Snowdon, Coho, and Red Cedar had disease severity scores of 4.11, 4.15, 4.26, 4.23, and 5.06 respectively. In 2021, the five most resistant lines were K19832, K20730, K20743, K21904, K16136 with scores ranging from 3.45-3.9 (Table 2.13, Figure 2.11). While the checks, Clouseau, Denali, Snowdon, 105 Coho, and Red Cedar had disease severity scores of 4.44, 4.17, 4.41, 4.17, and 4.53, respectively. In 2022, the five most resistant lines were K19817, K19832, K20745, K20717, and K20730 with scores ranging from 2.47-3.44 (Table 2.14, Figure 2.12). While the checks, Clouseau, Denali, Snowdon, Coho, and Red Cedar had disease severity scores of 3.8, 4.14, 4.12, 4.28, and 5.58, respectively. Rhizoctonia solani Screen The LSmeans in the field for the MAB trial ranged from 2.00-13.50 in 2021, 1.25-16.25 in 2022, and 3.31-12.25 in the combined analysis (Figure 2.2d). The five most resistant lines in the combined analysis were B19344, N20395, N19246, Spectre, and N20404 with values ranging from 3.31-3.69 (Table 2.15, Figure 2.13). In 2021, the five most resistant lines were N19226, N20404, G19611, P19103, and N20395 with values ranging from 2-3.87 (Table 2.16, Figure 2.14). In 2022, the five most resistant lines were Spectre, G21811, Merlin, N19246, and B19344 with values ranging from 1.25-2.5 (Table 2.17, Figure 2.15). Trial and Trait Correlations Most correlations between Fusarium root rot disease scores among genotypes in the field, greenhouse, and natural infection trials were found to be statistically non-significant except for between the 2022 Kidney bean trial and the greenhouse trial with a significant correlation (P =0.03) of 0.71(Table 2.18). Additional measurements were conducted in all trials to establish possible traits correlated to root rot disease severity. In the KDB trials, no significant correlations were found between root rot rating and other traits that were measured (vigor, stand count, bulk weights), but significant correlations were identified among vigor, stand count, and bulk weight measurements (Figure 2.16, Table 2.19). In the KDB trials, significant correlations were found between root rot rating and the first and last vigor measurements in 2021. Significant correlations 106 were also identified between other agronomic traits that were measured (vigor, stand count, bulk weights) Figure 2.17, Table 2.20. In the greenhouse trial significant correlations were found between root rot rating and fresh/dry root and dry shoot measurements (Figure 2.18, Table 2.21). Significant correlations were also found among fresh/dry root and shoot measurements. In the KDB natural infection trial, significant correlations were observed between root rot rating and yield and maturity date (Figure 2.19, Table 2.22). DISCUSSION This diverse study aimed to screen MSU Andean breeding lines and diversity panel lines for field and greenhouse-based resistance to Fusarium oxysporum and screen MSU Middle- American breeding lines for field resistance to Rhizoctonia solani, two pathogens that are particularly detrimental to dry bean production in Michigan. The levels of disease pressure in all experiments were sufficient to examine differences in plant response in inoculated conditions under root rot disease pressure, although low genetic variation limits analysis. Root rot resistance in dry bean is notoriously difficult to screen for due to many environmental factors that contribute to severity of the infection, and the ability of the plant to overcome the pathogen through avoidance or physiological resistance conferred by numerous small effect genetic loci. Although high standard error (SE) limits distinguishing between two similarly resistant genotypes in all trials tested, conclusions can be drawn in distinguishing between the most and least resistant lines, which is ultimately the most important objective when making selections in a plant breeding program. Fusarium oxysporum Study In the ADP trial the commercial checks Dynasty, Cabernet, and Talon behaved as expected based on results from previous literature (Oladzad et al., 2019; Zitnick-Anderson et al., 2020), but VAX3 consistently behaved as susceptible. In the greenhouse, VAX3 behaved as resistant as 107 expected, but Dynasty was the most susceptible line, supporting the observation that there was low correlation between ratings in the field vs greenhouse studies. In the ADP trial multiple lines stood out as possible sources of root rot resistance including ADP444, ADP481, ADP462, and ADP391. In the KDB inoculated and natural infection trials, multiple MSU breeding lines consistently outperformed the commercial lines. K19832, K20717, and K20730 stood out in particular as parents for future root rot resistance breeding efforts as they were top resistant lines in the KDB inoculated and natural infection trials. These lines have been noted previously as vigorous, robust plants with high yield. Their high root rot resistance may contribute to higher stands, vigor, and observed overall yield under ambient disease pressure in Michigan. Root rot rating was highest overall in the greenhouse, which was particularly interesting because during the evaluation we noted multiple “escapes” or inoculated plants that avoided disease infection. This could have also contributed to the lower heritability. In the field, the natural infection condition trial had the highest root rot rating. This is likely due to the effect of the root rot species complex in Montcalm County where these lines were tested, versus the artificial inoculation with one species in the KDB trial. Genetic variation in root rot disease score within trials was relatively small with the ADP trial having the largest variation. Identifying and integrating sources of unadapted resistance such as what is found in the ADP, may be a valid approach for integrating higher levels of resistance in breeding lines. Rhizoctonia solani Study The MAB trial was the first screen of commercial and MSU breeding lines for resistance to post-emergence damping off due to Rhizoctonia solani infection to date. In contrast to the Fusarium oxysporum trials these trials had a higher heritability, likely alluding to the objective phenotyping measurement used (damping off) vs the subjective 1-7 scale or the nature of genetic 108 resistance to the specific species. Significant differences were observed between genotypes. Many MSU breeding lines were highly resistant to post-emergence damping off including B19344, N20395, N19246, N20404, and N19246. B19344 stands out particularly as a promising line to use as a parent in future breeding efforts as it was the most resistant line identified in the combined analysis. Like the resistant lines present in the KDB inoculated and natural trials, this line has been previously identified as high yielding, which may be partially due to its ability to overcome ambient root rot in Michigan. The commercial small red line Viper was consistently the most susceptible line to post-emergence damping off. Field vs Greenhouse Correlations for Fusarium Root Rot Screening for root rots in the field is a challenge due to the potential of other root rot pathogens to confound the study. Furthermore, field trials require large plots of land and in many cases pivot irrigation for inducing root rot. Therefore, developing a greenhouse assay would improve the challenges with field trials and improve the throughput of phenotyping for this trait. The only significant correlation identified was between the KD2022 and GH (0.71) (Table 7). However, there was an almost significant correlation of 0.65 between the combined KDB trial and the greenhouse trial. Increasing inoculum and environmental conditions during the 2022 season could have contributed to this higher correlation. Previous studies identified varying significance between field and greenhouse screens depending on which isolate, and inoculation method was used (Bilgi et al., 2008; Chaudhary et al., 2006; Mukankusi et al., 2010; Nicoli et al., 2012; Schneider & Kelly, 2000) with significant correlations of 0.57-0.92 being found with spore suspension greenhouse methods and natural infection field conditions. Greenhouse screening would be ideal since it takes less time and resources and more lines could be phenotyped in a season, but it is ultimately important that the screening method mimics field conditions in order to 109 make accurate selections. Further exploration to establish correlations between greenhouse and field screening utilizing different methods including the inoculated sorghum seed method is warranted. When comparing results from the greenhouse and field Fusarium oxysporum trials, it is significant to note that although correlations between ratings in the greenhouse and field were not significant, two genotypes (ADP481 and K20730) were identified among the top five most resistant lines in both screening methods. K20730 was also a top five resistant line in the natural infection trial. It is possible that highly resistant lines are identified as resistant no matter the inoculation method used, but further research is needed to establish screening methods with better correlation. Evaluating secondary traits associated with Fusarium oxysporum root rot To score for Fusarium root rot, roots need to be harvested, washed, and scored. This process takes time and is labor intensive. Identifying secondary traits associated with root rot would allow for indirect evaluation of this trait in a non-destructive manner. Several traits were collected for this purpose. Root rot rating was found to be significantly negatively correlated to fresh and dry root weight in the greenhouse, supporting the hypothesis that this disease has significant negative impacts on the plant root biomass (Haus et al., 2020; Nakedde et al., 2016; Román-Avilés et al., 2004; Wang et al., 2018). This is contrary to a previous study by Bilgi et al. 2010 where there was no correlation identified between F. solani f. sp. phaseoli root rot rating and root weight, root- shoot ratio, or yield using the spore suspension method. This observation also supports the finding in the natural infection trial where yield was highly negatively correlated to root rot disease score. In the natural infection trial, standard agronomic traits were measured. Both yield and maturity date were identified as significantly 110 negatively correlated to root rot disease score with correlations of -0.62 and -0.43, respectively. This was encouraging as it suggests that routinely selecting for higher yield and reasonable maturity under natural root rot pressure in breeding trials can be effective at eliminating the most susceptible breeding lines and advancing more tolerant ones. However, there were no significant correlations identified between root rot disease score and stand count or vigor measurements for the ADP and KDB field trials indicating that these may not be reliable measurements for non- destructively measuring root rot resistance. Another method for measuring root rot disease severity that would be beneficial to investigate is UAS based measurement of aboveground biomass. This would be a similar measurement as the ground-truth vigor that was taken in this study, but would be a uniform, high-throughput, unbiased plot wise aerial measurement that would eliminate any scorer error. There is previous research supporting this method for measuring disease traits indicating that it may be beneficial for accurately and non-destructively measuring root rot resistance score (Guo et al., 2021; Marzougui et al., 2019; Moreira et al., 2020). Limitations Effective screening for quantitatively inherited traits is challenging, and root rot disease traits are no exception. Field based root rot screens are preferable to the breeder because they most accurately reflect agronomic conditions, but they are labor and land intensive and confounded by natural pathogen presence, as well as large environmental and error variances. Conversely, greenhouse evaluations provide a controlled environment, cost, and labor effective screening alternative but may not accurately predict resistance in field conditions, which can lead to conflicting results between screening methods. Similar to multiple previous root rot disease studies (Guzman, 2016; Hagerty et al., 2015; Nakedde et al., 2016; Román-Avilés et al., 2004; Román-Avilés & Kelly, 2005), the current study 111 was also hampered by large environmental and error variances, large standard error, and low genetic effect, leading to a low overall heritability and difficulty distinguishing between resistant genotypes for root rot disease rating (Fusarium oxysporum trials) and post emergence damping off (Rhizoctonia solani trials). This is primarily due to low innate heritability and the complex inheritance mechanism of the trait which prevents the separation of genetic and environmental effects and low innate resistance in breeding populations. Our greenhouse trial was also limited by large error variance, low genetic effect, and high CVs. The greenhouse screening method we used was chosen because it was previously shown to reduce environmental variation, decrease CV, and accurately mimic field conditions compared to other inoculation methods (Haus et al., 2020; Nakedde et al., 2016; Sendi et al., 2020; Wang et al., 2018). A consensus has not formed for the best inoculation method for root rot disease studies and various studies utilize different greenhouse screening methods which provide varying results among studies. As stated before, unbiased measurements such as UAS may improve this currently unavoidable aspect of root rot resistance studies as well as identifying the best universal greenhouse and field measurement systems. VAX 3, a line previously established to have moderate resistance to Fusarium brasiliense and Fusarium solani in the greenhouse, behaved moderately resistant in the greenhouse and susceptible in the field, which raises the question of its reliability as a resistant check for Fusarium oxysporum screening in the field. An external factor that could have contributed to variable results in the field is the presence of untested natural pathogens. Natural pathogen presence is virtually unavoidable in the field, especially in a disease nursery, and could have contributed to the observed results. Previous studies have identified Fusarium, Rhizoctonia, and oomycete species as endemic to Michigan growing regions (Jacobs et al., 2019). Another factor that complicates analysis is the presence of avoidance mechanisms and variable environmental conditions in the field that could 112 contribute to resistance or susceptibility that would not be identified under controlled greenhouse conditions. Further research is needed to confirm the validity of this line as a resistant check to screen for Fusarium root rot resistance in the field. In the interim, a better approach would be to use the most resistant lines such as Dynasty as local disease resistant checks for future field studies in Michigan. CONCLUSION In the present study, 98 diverse dry bean lines from the Andean gene pool were screened for Fusarium oxysporum resistance in the field and (a subset) in the greenhouse and 64 lines from the Middle-American gene pool were screened for Rhizoctonia solani resistance in the field. The primary objective of this study was successful. Variance in root rot disease score/damping-off for a diverse set of dry bean lines was identified and the most resistant lines are recommended for future breeding efforts. All trials were successful in identifying significant variation for root rot resistance. Significant correlations were found between the field trial and the greenhouse trials (P=0.03, 0.71). Significant trait correlations were also identified between root rating and fresh (P=0.01, -0.67) and dry root (P=0.01, -0.64) and dry shoot (P=0.04, -0.43) weights in the greenhouse. Ultimately the lines ADP444, ADP481, ADP462, ADP391, K19832, K20717, K20730, B19344, N20395, N19246, N20404, and N19226 are recommended as parents for future root rot resistance breeding efforts. Similar to previous root rot resistance studies, high standard error (SE), error and environmental variance, and low heritability limited interpretation of the results. Ultimately, conclusions can be drawn in distinguishing between the most and least resistant lines, which is the most important objective when making selections in a plant breeding program. Moving forward, objective high-throughput methods such as UAS based phenomics may provide an opportunity for accurate, unbiased, and reproducible screening of root rot resistance if 113 aboveground symptoms can be accurately measured and correlated to root rot rating as was suggested by the correlation between root and shoot biomass in the greenhouse trial. A standardized inoculation method and continued establishment of consistent resistant/susceptible checks under field conditions will also aid future studies. Ultimately, while the current studies presented here will be useful in making selections between the most resistant and most susceptible individuals, further optimization of screening methods is warranted for this highly complex trait. 114 TABLES AND FIGURES Table 2.1: Population description, isolate information, sample number, years tested, and measurements collected for all populations evaluated for root rot resistance in this study. Population Description Isolate Sample Number Years Trait Measurements Andean Diversity Panel (ADP) F. oxysporum 38 2021/2022 Root Rot Disease Score Vigor Stand Count Bulk Fresh Weight (2022) Kidney Breeding Lines (KDB) F. oxysporum 66 2021/2022 Greenhouse trial (GH) F. oxysporum 23 2022 Root Rot Disease Score Vigor Stand Count Bulk Fresh Weight (2022) Root Rot Disease Score Fresh Weight Dry Weight Stand Count Natural Infection F. oxysporum 42 2021/2022 Root Rot Disease Score Flowering Date Maturity Date Yield (CWT Acre) Lodging Score Bulk Fresh Weight (2022) Middle American Breeding Lines (MAB) R. solani 65 2021/2022 Stand Count Vigor 115 Table 2.2: List of lines evaluated in this study including the trials present, market class information, origin, and years tested. Line Trial(s) Market Class / Seed Coat Origin Year(s) ADP001 ADP002 ADP004 ADP015 ADP021 ADP042 ADP ADP ADP ADP ADP ADP ADP043 ADP/GH ADP081 ADP088 ADP091 ADP094 ADP ADP ADP ADP Red Mottled Cranberry Small Red Dark Red Kidney Small Red Purple Yellow Black Light Red Kidney Cranberry Yellow ADP099 ADP/GH Dark Red Kidney ADP111 ADP112 ADP ADP Pink Cranberry Dark Red Kidney ADP186 ADP/GH Red Mottled ADP214 ADP ADP391 ADP/GH ADP392 ADP429 ADP ADP ADP444 ADP/GH ADP462 ADP/GH Yellow Red Mottled Red Mottled Yellow Yellow Yellow ADP474 ADP/GH Cranberry 116 ADP ADP ADP ADP ADP ADP ADP ADP ADP ADP ADP ADP ADP ADP ADP ADP ADP ADP ADP ADP ADP ADP 2021/2022 2021/2022 2021/2022 2021/2022 2021/2022 2021/2022 2021/2022 2021/2022 2021/2022 2021/2022 2021/2022 2021/2022 2021/2022 2021/2022 2021/2022 2021/2022 2021/2022 2021/2022 2021/2022 2021/2022 2021/2022 2021/2022 Table 2.2 (cont’d) ADP481 ADP/GH Light Red Kidney ADP511 ADP513 ADP514 ADP ADP ADP Dark Red Kidney Yellow Light Red Kidney ADP519 ADP/GH White Kidney ADP602 ADP Pink Mottled ADP612 ADP/GH Dark Red Kidney ADP621 ADP626 ADP640 ADP683 ADP684 ADP ADP ADP ADP ADP Purple Speckled Purple Speckled Manteca Yellow Dark Red Kidney ADP ADP ADP ADP ADP ADP ADP ADP ADP ADP ADP ADP 2021/2022 2021/2022 2021/2022 2021/2022 2021/2022 2021/2022 2021/2022 2021/2022 2021/2022 2021/2022 2021/2022 2021/2022 Cabernet ADP Dark Red Kidney Seminis 2021/2022 Dynasty ADP/GH Dark Red Kidney Guelph 2021/2022 Talon VAX3 ADP ADP/GH Dark Red Kidney NDSU 2021/2022 Small Red U of ID 2021/2022 Beluga KDB/Natural (2021) White Kidney MSU 2021/2022 Clouseau KDB/Natural Light Red Kidney Seminis 2021/2022 Coho KDB/Natural Light Red Kidney MSU 2021/2022 Denali KDB/Natural White Kidney MSU 2021/2022 SVS-0863 KDB Yellow Seminis 2021/2022 K16136 KDB/Natural Dark Red Kidney MSU K16640 KDB/GH Light Red Kidney MSU K16911 KDB White Kidney MSU 2021 2021 2021 117 Table 2.2 (cont’d) K17201 KDB Dark Red Kidney MSU K17702 KDB/Natural Light Red Kidney MSU K17703 KDB/Natural Light Red Kidney MSU K17704 KDB/Natural Light Red Kidney MSU K18312 KDB/Natural Dark Red Kidney K18907 K19111 KDB KDB White Kidney Dark Red Kidney K19120 KDB/Natural Dark Red Kidney MSU MSU MSU MSU K19608 KDB/Natural Light Red Kidney MSU 2021 2021 2021 2021 2021 2021 2021 2021 2021 K19610 KDB/GH/Natural Light Red Kidney MSU 2021/2022 K19817 KDB/GH/Natural White Kidney K19830 KDB/Natural White Kidney K19831 KDB/Natural White Kidney K19832 KDB/GH/Natural White Kidney K20210 KDB/Natural Dark Red Kidney K20212 KDB/Natural Dark Red Kidney K20217 KDB/Natural Dark Red Kidney K20221 KDB/Natural Dark Red Kidney K20234 KDB/Natural Dark Red Kidney K20235 KDB/Natural Dark Red Kidney K20239 KDB/Natural Dark Red Kidney MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU 2021/2022 2021/2022 2021/2022 2021/2022 2021 2021/2022 2021/2022 2021/2022 2021 2021 2021/2022 K20712 KDB/GH/Natural Light Red Kidney MSU 2021 K20715 KDB/GH/Natural Light Red Kidney MSU 2021/2022 K20717 KDB/Natural Light Red Kidney MSU 2021/2022 118 Table 2.2 (cont’d) K20720 KDB/Natural Light Red Kidney MSU 2021 K20721 KDB/GH/Natural Dark Red Kidney MSU 2021/2022 K20728 KDB/Natural Light Red Kidney MSU 2021 K20730 KDB/GH/Natural Light Red Kidney MSU 2021/2022 K20732 KDB/GH/Natural Light Red Kidney MSU 2021/2022 K20734 KDB/Natural Light Red Kidney MSU 2021/2022 K20742 KDB/Natural Light Red Kidney MSU 2021/2022 K20743 KDB/Natural Light Red Kidney MSU 2021/2022 K20744 KDB/Natural Light Red Kidney MSU 2021/2022 K20745 KDB/Natural Light Red Kidney MSU 2021/2022 K20749 KDB/Natural Light Red Kidney MSU 2021 Montcalm KDB/Natural (2021) Dark Red Kidney MSU 2021/2022 ND Whitetail KDB/Natural (2021) White Kidney NDSU 2021/2022 Patron KDB Yellow OSU 2021/2022 Red Cedar KDB/Natural Dark Red Kidney Red Hawk KDB/GH/Natural (2021) Dark Red Kidney MSU MSU 2021/2022 2021/2022 Rosie KDB Light Red Kidney NDSU 2021/2022 Snowdon KDB/Natural White Kidney Y17502 Y18703 Y19801 Y19804 Y19808 Y19810 KDB KDB KDB KDB KDB KDB Yellow Yellow Yellow Yellow Yellow Yellow 119 MSU MSU MSU MSU MSU MSU MSU 2021/2022 2022 2022 2022 2022 2022 2022 Table 2.2 (cont’d) Y19815 Y19817 KDB KDB Yellowstone KDB/GH Adams Armada B18504 B19309 B19330 B19332 B19344 B20536 B20547 B20549 B20591 B20597 B20599 B20602 B21708 B21710 B21714 MAB MAB MAB MAB MAB MAB MAB MAB MAB MAB MAB MAB MAB MAB MAB MAB MAB Black Bear MAB Black Beard MAB Black Tails MAB Blizzard MAB MSU MSU MSU MSU 2022 2021/2022 2021/2022 2022 Provita 2021/2022 MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU 2022 2021/2022 2022 2022 2021/2022 2022 2021/2022 2022 2021/2022 2021/2022 2022 2022 2022 2022 2022 Provita 2021/2022 Provita 2021/2022 Provita 2021/2022 Provita 2021/2022 Yellow Yellow Yellow Black Navy Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Black Navy 120 Table 2.2 (cont’d) Caldera MAB Cayenne MAB Charro Coral Eiger G19611 G19613 G21811 MAB MAB MAB MAB MAB MAB HMS Bounty MAB Liberty MAB Medalist MAB Merlin N18103 N19226 N19239 N19246 N19285 N20388 N20395 N20404 N21511 N21525 Nimbus P16901 MAB MAB MAB MAB MAB MAB MAB MAB MAB MAB MAB MAB MAB Provita 2021/2022 MSU MSU MSU MSU MSU MSU MSU 2022 2022 2022 2022 2021 2021/2022 2022 Provita 2021/2022 Provita 2022 Provita 2021/2022 Provita 2021/2022 MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU 2022 2021/2022 2021 2021/2022 2021 2021/2022 2021/2022 2021/2022 2022 2022 Provita 2021 MSU 2021 Small Red Small Red Pinto Pink Great Northern Great Northern Great Northern Great Northern Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Navy Black Pinto 121 Table 2.2 (cont’d) P19103 P19707 P19713 R12844 R20627 R20639 R20652 R20659 R20667 R20669 R20683 Rosetta Ruby S08418 S18904 Spectre Valiant Viper Zenith Zorro MAB MAB MAB MAB MAB MAB MAB MAB MAB MAB MAB MAB MAB MAB MAB MAB MAB MAB MAB MAB Pinto Pinto Pinto Small Red Small Red Small Red Small Red Small Red Small Red Small Red Small Red Pink MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU MSU 2021/2022 2021 2021/2022 2021 2021/2022 2021 2021 2021/2022 2021/2022 2022 2021 2022 Small Red Provita 2021/2022 Pink Pink Black Navy MSU MSU 2021 2021 Provita 2021/2022 Provita 2021/2022 Small Red Provita 2021/2022 Black Black MSU MSU 2021/2022 2021/2022 122 Figure 2.1: Screening scale used to evaluate Fusarium oxysporum inoculated lines for root damage. Lines were evaluated on a 1-7 scale developed by Schneider and Kelly 2001, where 1 indicates no disease and 7 indicates a non functional, completely rotted root system. 123 Table 2.3: T-test of the difference between Least squares means of the inoculated vs non- inoculated treatment to validate inoculum concentration. Population Isolate 2021 2022 KDB Lines- Field F. oxysporum <2.2 x 10^-16 <2.2 x 10^-16 ADP Lines- Field F. oxysporum <2.2 x 10^-16 <2.2 x 10^-16 ADP and KDB Lines- Greenhouse Middle American Lines- Field F. oxysporum - <2.2 x 10^-16 R. solani <2.2 x 10^-16 <2.2 x 10^-16 124 Figure 2.2a: Histogram of Least Squares means in the Andean diversity panel (ADP) Fusarium oxysporum trial (inoculated treatment only) using a 1-7 root rot disease severity score. 125 Figure 2.2b: Histogram of Least Squares means in the kidney breeding line (KDB) Fusarium oxysporum trials and Natural infection trials (inoculated treatment only) using a 1-7 root rot disease severity score. 126 Figure 2.2c: Histogram of Least Squares means in the greenhouse Fusarium oxysporum trial (inoculated treatment only) using a 1-7 root rot disease severity score. 127 Figure 2.2d: Histogram plot of Least Squares means, for the Middle American breeding line trial inoculated with Rhizoctonia solani using the difference in post-emergence damping off in the inoculated treatment. 128 Figure 2.2e: Histograms of Least Squares means in the combined analysis of the Andean diversity panel and Kidney breeding line field trials and the greenhouse trial (2022 only) of the inoculated treatment only using a 1-7 root rot disease severity score. 129 Table 2.4: Table of variance components, heritability, and coefficient of variation for each population, utilizing the inoculated treatment only. Starred values are significant at a 0.05 significance level. Population Year Isolate σ2 G σ2 Y σ2 e H2 CV % σ2 GxY KDB Field Trial (KDB) Combined F. oxysporum 0.08* 0.04 >0.001 1.47* 0.29 42.13 2021 2022 ADP Field Trial (ADP) Combined F. oxysporum 0.09* 0.17* - - - - 0.99* 0.26 34.99 2.08* 0.24 49.26 0.07* 0.01 >0.001 1.80* 0.24 45.59 2021 2022 2022 F. oxysporum 0.11* 0.08 0.10 - - - - - - 1.21* 0.28 39.13 2.10* 0.14 48.19 2.37* 0.12 31.04 Combined N/A 0.06 0.08 0.05 1.13* 0.19 27.69 2021 2022 0.04 0.38* - - - - 0.98* 0.11 23.94 1.50* 0.43 35.32 Combined R. solani 3.31* 2.42 0.00002 1.75* 0.49 67.32 Greenhouse Trial (GH) Natural Infection (KDB) Middle American Field Trial (MAB) 2021 2022 4.95* 6.72* - - - - 0.55 65.11 16.31 * 20.1* 0.57 71.06 130 Figure 2.3: Histogram of Least Squares means by genotype with standard error bars, ADP trial combined analysis. 131 Figure 2.4: Histogram of Least Squares means by genotype with standard error bars, ADP trial 2021 analysis. 132 Figure 2.5: Histogram of Least Squares means by genotype with standard error bars, ADP trial 2022 analysis. 133 Figure 2.6: Histogram of Least Squares means by genotype with standard error bars, KDB trial combined analysis. 134 Figure 2.7: Histogram of Least Squares means by genotype with standard error bars, KDB trial 2021 analysis. 135 Figure 2.8: Histogram of Least Squares means by genotype with standard error bars, KDB trial 2022 analysis. 136 Figure 2.9: Histogram of Least Squares means by genotype with standard error bars, Greenhouse trial. 137 Figure 2.10: Histogram of Least Squares means by genotype with standard error bars, Natural infection trial combined analysis. 138 Figure 2.11: Histogram of Least Squares means by genotype with standard error bars, Natural infection trial 2021 analysis. 139 Figure 2.12: Histogram of Least Squares means by genotype with standard error bars, Natural infection trial 2022 analysis. 140 Figure 2.13: Histogram of Least Squares means by genotype with standard error bars, Middle- American trial combined analysis. 141 Figure 2.14: Histogram of Least Squares means by genotype with standard error bars, Middle- American trial 2021 analysis. 142 Figure 2.15: Histogram of Least Squares means by genotype with standard error bars, Middle- American trial 2022 analysis. 143 Table 2.5: Least squares means, least significant difference, and standard error by line for root rot rating of the ADP Fusarium oxysporum trial combined analysis. Asterisk (*) indicates lines that are not significantly different from the most resistant line. Line ADP99* ADP444* ADP481* Dynasty* ADP462* ADP391* ADP626* ADP392* ADP612* ADP684* ADP214 ADP429 ADP4 ADP511 Talon ADP111 ADP43 ADP514 ADP94 ADP15 ADP513 ADP640 Lsmean 2.05 2.35 2.45 2.65 2.67 2.70 2.70 2.77 2.77 2.77 2.80 2.85 2.87 2.87 2.87 2.90 2.90 2.97 3.00 3.02 3.02 3.05 144 SE 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.31 0.28 Table 2.5 (cont’d) ADP602 ADP42 ADP91 ADP81 ADP683 Cabernet ADP1 ADP21 ADP88 ADP474 ADP2 ADP519 VAX3 ADP112 ADP186 ADP621 LSD 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 - 3.07 3.10 3.20 3.22 3.30 3.30 3.32 3.32 3.35 3.37 3.40 3.42 3.47 3.52 3.60 4.00 0.73 145 Table 2.6: Least squares means, least significant difference, and standard error by line for root rot rating of the ADP Fusarium oxysporum trial 2021 analysis. Asterisk (*) indicates lines that are not significantly different from the most resistant line. Line ADP99* ADP444* ADP462* ADP43* ADP612* ADP391* ADP481* ADP626* Dynasty* ADP429* ADP684* Talon* ADP214 ADP392 ADP511 ADP514 ADP2 ADP91 ADP513 ADP21 ADP4 ADP42 Lsmean 1.80 1.90 2.20 2.30 2.30 2.40 2.40 2.60 2.60 2.70 2.70 2.70 2.80 2.80 2.90 2.90 3.00 3.00 3.09 3.10 3.10 3.10 146 SE 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.44 0.37 0.37 0.37 Table 2.6 (cont’d) ADP94 ADP1 ADP683 ADP81 ADP111 ADP112 ADP602 ADP640 Cabernet ADP15 ADP621 ADP88 ADP186 ADP519 VAX3 ADP474 LSD 3.10 3.20 3.20 3.20 3.30 3.30 3.30 3.30 3.30 3.40 3.40 3.40 3.60 3.60 3.70 3.90 0.97 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 - 147 Table 2.7: Least squares means, least significant difference, and standard error by line for root rot rating of the ADP Fusarium oxysporum trial 2022 analysis. Asterisk (*) indicates lines that are not significantly different from the most resistant line. Line ADP99* ADP111* ADP481* ADP15* ADP4* Dynasty* ADP392* ADP214* ADP444* ADP626* ADP640* ADP474* ADP511* ADP602* ADP684* ADP94* ADP513* ADP391* ADP429* ADP514* Talon* ADP42* Lsmean 2.30 2.50 2.50 2.65 2.65 2.70 2.75 2.80 2.80 2.80 2.80 2.85 2.85 2.85 2.85 2.90 2.95 3.00 3.00 3.05 3.05 3.10 148 SE 0.34 0.34 0.34 0.34 0.34 0.34 0.34 0.34 0.34 0.34 0.34 0.34 0.34 0.34 0.34 0.34 0.34 0.34 0.34 0.34 0.34 0.34 Table 2.7 (cont’d) ADP462* ADP519 ADP81 VAX3 ADP612 ADP88 Cabernet ADP683 ADP91 ADP1 ADP43 ADP21 ADP186 ADP112 ADP2 ADP621 LSD 3.15 3.25 3.25 3.25 3.25 3.30 3.30 3.40 3.40 3.45 3.50 3.55 3.60 3.75 3.80 4.60 0.90 0.34 0.34 0.34 0.34 0.34 0.34 0.34 0.34 0.34 0.34 0.34 0.34 0.34 0.34 0.34 0.34 - 149 Table 2.8: Least squares means, least significant difference, and standard error by line for root rot rating of the Kidney Fusarium oxysporum trial combined analysis. Asterisk (*) indicates lines that are not significantly different from the most resistant line. Line K20730* Y19817* Beluga* K19832* Clouseau* Denali* Snowdon* K19610* K19830* K20717* Yellowstone Coho K20217 K20744 K20221 Rosie K20743 K19831 K20732 K20239 K20734 K20212 Lsmean 2.39 2.47 2.55 2.6 2.73 2.73 2.84 2.85 2.9 2.9 2.98 3 3.05 3.08 3.08 3.08 3.1 3.1 3.15 3.23 3.25 3.25 150 SE 0.22 0.23 0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.22 Table 2.8 (cont’d) K20745 Red Cedar K20742 Montcalm K19817 Red Hawk K20721 K20715 ND Whitetail LSD 3.3 3.35 3.35 3.4 3.58 3.6 3.68 3.95 3.98 0.53 0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.22 - 151 Table 2.9: Least squares means, least significant difference, and standard error by line for root rot rating of the Kidney Fusarium oxysporum trial 2021 analysis. Asterisk (*) indicates lines that are not significantly different from the most resistant line. Line K19610* K20712* Y19817* K16911* K18907* K20728* Rosie* Beluga* Clouseau* K20730* K17702* K17704* K19830* K20234* Coho* K17703* K19608* Snowdon* Denali* K20717* K19120* Montcalm* Lsmean 2.30 2.35 2.45 2.50 2.50 2.55 2.55 2.65 2.70 2.70 2.75 2.75 2.75 2.75 2.75 2.80 2.80 2.80 2.85 2.85 2.85 2.90 152 SE 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 Table 2.9 (cont’d) K19832* K20235* Yellowstone K20744 K19111 K20210 K20720 K20212 K18312 K19831 K17201 K20734 K20745 K20749 K20217 Red Cedar K20743 K20239 K20221 K20742 K16136 ND Whitetail K20732 K20721 2.90 2.90 3.00 3.00 3.00 3.00 3.05 3.05 3.10 3.10 3.15 3.20 3.20 3.20 3.20 3.20 3.25 3.25 3.30 3.35 3.40 3.40 3.45 3.50 153 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 Table 2.9 (cont’d) Red Hawk K16640 K20715 K19817 LSD 3.60 3.80 3.80 3.90 0.62 0.24 0.24 0.24 0.24 - 154 Table 2.10: Least squares means, least significant difference, and standard error by line for root rot rating of the Kidney Fusarium oxysporum trial 2022 analysis. Asterisk (*) indicates lines that are not significantly different from the most resistant line. Line K20730* K19832* Y19808* Beluga* Y19817* Y17502* Y19810* Denali* Y19815* Patron* Clouseau* Y18703* K20221* K20732* Y19804* Snowdon* K20217* K20743* K20717* Yellowstone* K19830 I17506 Lsmean 2.09 2.30 2.45 2.45 2.48 2.50 2.60 2.60 2.65 2.75 2.75 2.80 2.85 2.85 2.85 2.89 2.90 2.95 2.95 2.95 3.05 3.10 155 SE 0.37 0.36 0.36 0.36 0.4 0.36 0.36 0.36 0.36 0.36 0.36 0.36 0.36 0.36 0.36 0.39 0.36 0.36 0.36 0.36 0.36 0.36 0.36 0.36 0.36 0.36 0.36 0.36 0.36 0.36 0.36 0.36 0.36 0.36 0.36 0.36 0.36 0.36 0.36 0.36 - Table 2.10 (cont’d) K19831 K20744 K20239 K19817 Coho K20734 K20742 K20745 K19610 K20212 Red Cedar Rosie Red Hawk Y19801 K20721 Montcalm K20715 ND Whitetail LSD 3.10 3.15 3.20 3.25 3.25 3.30 3.35 3.40 3.40 3.45 3.50 3.60 3.60 3.85 3.85 3.90 4.10 4.55 0.90 156 Table 2.11: Least squares means, least significant difference, and standard error by line for root rot rating of the Greenhouse Fusarium oxysporum trial analysis. Asterisk (*) indicates lines that are not significantly different from the most resistant line. Line ADP481* K20730* VAX3* ADP462* K20712* K20732* ADP474* K16640* ADP186* ADP444* ADP391* ADP519 ADP099 K19832 ADP612 K19610 Red Hawk K20715 Yellowstone ADP043 K19817 K20721 Lsmean 4.00 4.44 4.50 4.61 4.78 4.78 4.89 4.89 4.89 4.94 5.00 5.06 5.11 5.17 5.28 5.39 5.39 5.50 5.67 5.67 5.72 5.72 157 SE 0.39 0.39 0.39 0.39 0.39 0.39 0.39 0.39 0.39 0.39 0.39 0.39 0.39 0.39 0.39 0.39 0.39 0.39 0.39 0.39 0.39 0.39 Table 2.11 (cont’d) DYNASTY LSD 6.00 1.01 0.39 - 158 Table 2.12: Least squares means, least significant difference, and standard error by line for root rot rating of the Natural infection trial combined analysis. Asterisk (*) indicates lines that are not significantly different from the most resistant line. Line K19832* K19817* K20745* K20717* K20730* K20734* K20743 K19830 K19831 K20721 Clouseau K20221 Denali K20715 Coho K19610 Snowdon K20742 K20239 K20217 K20212 K20732 Lsmean 3.17 3.37 3.52 3.55 3.57 3.75 3.87 3.91 4.01 4.06 4.11 4.13 4.15 4.18 4.23 4.25 4.26 4.27 4.33 4.39 4.47 4.54 159 SE 0.21 0.21 0.21 0.21 0.21 0.21 0.21 0.21 0.21 0.22 0.21 0.21 0.21 0.21 0.21 0.21 0.21 0.21 0.21 0.21 0.21 0.21 Table 2.12 (cont’d) K20744 Red Cedar LSD 4.74 5.06 0.59 0.21 0.21 - 160 Table 2.13: Least squares means, least significant difference, and standard error by line for root rot rating of the Natural infection trial 2021 analysis. Asterisk (*) indicates lines that are not significantly different from the most resistant line. Line K19832* K20730* K20743* K21904* K16136* K20720* K17704* K20728* K20734* K17703* K21909* K21902* K21913* K19831* K21901* K20745* K21906* K20717* Beluga* K20235* K20239* I90013* Lsmean 3.45 3.70 3.81 3.88 3.90 3.90 3.92 3.93 3.99 4.01 4.03 4.04 4.05 4.07 4.08 4.08 4.09 4.10 4.14 4.15 4.15 4.15 161 SE 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 Table 2.13 (cont’d) K20721* Denali* K19830* Coho* K20217 K17702 K20744 K21905 K19817 K19610 K20715 Snowdon Clouseau K20210 K21908 K20712 Red Hawk K21912 K21907 Montcalm Red Cedar K20212 K20732 K19120 4.16 4.17 4.17 4.17 4.21 4.23 4.23 4.27 4.27 4.30 4.38 4.41 4.44 4.45 4.46 4.47 4.49 4.50 4.51 4.51 4.53 4.55 4.56 4.57 162 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 Table 2.13 (cont’d) K20749 K19608 K20742 K21910 K20221 ND Whitetail K18312 K21903 K21911 K20234 LSD 4.57 4.59 4.64 4.65 4.65 4.75 4.76 4.82 4.90 5.41 0.75 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 - 163 Table 2.14: Least squares means, least significant difference, and standard error by line for root rot rating of the Natural infection trial 2022 analysis. Asterisk (*) indicates lines that are not significantly different from the most resistant line. Line K19817* K19832* K20745* K20717* K20730* K20734 K20221 K19830 Clouseau K20742 K19831 K20721 K20743 K20715 Snowdon Denali K19610 Coho K20212 K20239 K20732 K20217 Lsmean 2.47 2.89 2.95 3.02 3.44 3.51 3.62 3.66 3.80 3.88 3.94 3.94 3.94 3.99 4.12 4.14 4.20 4.28 4.39 4.51 4.52 4.56 164 SE 0.35 0.35 0.35 0.36 0.35 0.35 0.35 0.35 0.35 0.35 0.35 0.37 0.36 0.35 0.35 0.35 0.35 0.35 0.35 0.35 0.35 0.35 Table 2.14 (cont’d) K20744 Red Cedar LSD 5.25 5.58 0.97 0.35 0.35 - 165 Table 2.15: Least squares means, least significant difference, and standard error by line for root rot rating of the Middle-American Rhizoctonia solani trial combined analysis. Asterisk (*) indicates lines that are not significantly different from the most resistant line. Line B19344* N20395* N19246* Spectre* N20404* N19226* Merlin* B19309* Ruby* P19103* Blizzard N20388 B20591 HMS Bounty Valiant Zorro Armada Nimbus G19613 P19713 R20627 Caldera Lsmean 3.31 3.44 3.50 3.63 3.69 4.38 4.75 5.12 5.75 6.56 7.13 7.19 7.31 7.38 7.50 7.63 8.19 8.19 8.25 8.63 8.69 8.75 166 SE 1.47 1.47 1.47 1.47 1.47 1.47 1.47 1.47 1.47 1.47 1.47 1.47 1.47 1.47 1.47 1.47 1.47 1.47 1.47 1.47 1.47 1.47 Table 2.15 (cont’d) R20667 R20659 Zenith B20547 Medalist Black Bear Black Beard Black Tails Viper LSD 9.44 9.56 9.69 9.75 9.88 10.94 11.31 11.38 12.25 3.55 1.47 1.47 1.47 1.47 1.47 1.47 1.47 1.47 1.47 - 167 Table 2.16: Least squares means, least significant difference, and standard error by line for root rot rating of the Middle-American Rhizoctonia solani trial 2021 analysis. Asterisk (*) indicates lines that are not significantly different from the most resistant line. Line N19226* N20404* G19611* P19103* N20395* B18504* B19344* N20388* N19246* G19613* P19707* Blizzard* N19285* Spectre S08418 B20602 R20627 Armada B19309 B19332 B19330 R20683 Lsmean 2.00 3.12 3.37 3.62 3.87 4.00 4.12 4.37 4.75 5.00 5.12 5.25 5.50 6.00 6.25 6.37 6.37 6.40 6.50 6.75 7.12 7.25 168 SE 1.66 1.66 1.66 1.66 1.66 1.66 1.66 1.66 1.66 1.66 1.66 1.66 1.66 1.66 1.66 1.66 1.66 1.66 1.66 1.66 1.66 1.66 Table 2.16 (cont’d) B20591 Caldera Merlin N19239 S18904 Valiant HMS Bounty B20597 R20652 Ruby R20659 Nimbus B20547 Zorro P16901 B20549 P19713 R12844 R20667 Black Tails Black Bear Black Beard Zenith R20639 7.37 7.50 7.50 7.62 7.75 7.75 8.00 8.12 8.25 8.25 8.37 8.62 8.75 8.75 9.12 9.25 9.25 9.87 9.87 10.25 10.87 11.87 11.87 12.00 169 1.66 1.66 1.66 1.66 1.66 1.66 1.66 1.66 1.66 1.66 1.66 1.66 1.66 1.66 1.66 1.66 1.66 1.66 1.66 1.66 1.66 1.66 1.66 1.66 Table 2.16 (cont’d) Medalist Viper LSD 12.25 13.50 3.96 1.66 1.66 - 170 Table 2.17: Least squares means, least significant difference, and standard error by line for root rot rating of the Middle-American Rhizoctonia solani trial 2022 analysis. Asterisk (*) indicates lines that are not significantly different from the most resistant line. Line Spectre* G21811* Merlin* N19246* B19344* N20395* Ruby* B19309* N18103* N20404* N21511* B20599* Liberty* N21525* Zorro* HMS Bounty* N19226* Charro* B20591* Valiant* Coral Lsmean 1.25 1.75 2.00 2.25 2.5 3.00 3.25 3.75 4.25 4.25 5.25 5.50 5.75 6.25 6.50 6.75 6.75 7.00 7.25 7.25 7.50 171 SE 2.54 2.54 2.54 2.54 2.54 2.54 2.54 2.54 2.54 2.54 2.54 2.54 2.54 2.54 2.54 2.54 2.54 2.54 2.54 2.54 2.54 Table 2.17 (cont’d) Medalist Zenith Nimbus P19713 Eiger Cayenne Rosetta Blizzard R20667 B21714 P19103 Adams Armada Caldera N20388 B20547 Black Beard R20659 B21710 Black Bear R20627 Viper G19613 R20669 7.50 7.50 7.75 8.00 8.25 8.50 8.75 9.00 9.00 9.50 9.50 9.75 10.00 10.00 10.00 10.75 10.75 10.75 11.00 11.00 11.00 11.00 11.50 12.25 172 2.54 2.54 2.54 2.54 2.54 2.54 2.54 2.54 2.54 2.54 2.54 2.54 2.54 2.54 2.54 2.54 2.54 2.54 2.54 2.54 2.54 2.54 2.54 2.54 Table 2.17 (cont’d) Black Tails B21708 B20536 LSD 12.50 14.75 16.25 6.22 2.54 2.54 2.54 - 173 Table 2.18: Correlation of least squares means for root rot rating between trials. Pearson’s Correlation (r) ADP and GH KDB and GH KDB and Natural Combined 2021 2022 0.20 0.04 0.51 0.65* 0.37 0.71** 0.22 0.10 0.24 Values marked with *, **, and *** were significant at the 0.1, 0.05, and 0.01 level, respectively. 174 Figure 2.16: Correlation of root rot rating, stand count, vigor, and weight for the KDB trial. The scale on the right indicates correlation where the values represent the significance of correlation. Table 2.19: Correlation of root rot rating, stand count, vigor, and weight for the KDB trial. Stand Count First Vigor Measurement Vigor Difference Measurement Last Vigor Measurement 2022 Fresh Weight (g) 0.28 -0.27 -0.23 -0.17 -0.15 Correlation with Root Rot Rating Values marked with *, **, and *** were significant at the 0.1, 0.05, and 0.01 level, respectively. 175 Figure 2.17: Correlation of root rot rating, stand count, vigor, and weight for the ADP trial. The scale on the right indicates correlation where the values represent the significance of correlation. Table 2.20: Correlation of root rot rating, stand count, vigor, and weight for the ADP trial. Stand Count First Vigor Measurement Vigor Difference Measurement Last Vigor Measurement 2022 Fresh Weight (g) -0.17 0.3* 0.07 0.25 -0.29* Correlation with Root Rot Rating Values marked with *, **, and *** were significant at the 0.1, 0.05, and 0.01 level, respectively. 176 Figure 2.18: Correlation of root rot rating, root and shoot weights for the GH trial. The scale on the right indicates correlation where the values represent the significance of correlation. Table 2.21: Correlation of root rot rating, root and shoot weight for the GH infection trial. Fresh Root Weight (g) Fresh Shoot Weight (g) Dry Root Weight (g) Dry Shoot Weight (g) -0.67*** -0.27(P=0.22) -0.64*** -0.43** Correlation with Root Rot Rating Values marked with *, **, and *** were significant at the 0.1, 0.05, and 0.01 level, respectively. 177 Figure 2.19: Correlation of root rot rating, yield, flowering date, lodging score, and maturity date for the Natural infection trial. The scale on the right indicates correlation where the values represent the significance of correlation. Table 2.22: Correlation of root rot rating, yield, flowering date, lodging score, and maturity date for the Natural infection trial. Yield (CWT Acre) Flowering Date Lodging Score Maturity Date -0.62*** -0.27 -0.19 -0.43*** Correlation with Root Rot Rating Values marked with *, **, and *** were significant at the 0.1, 0.05, and 0.01 level, respectively. 178 REFERENCES Abawi GS, Pastor-Corrales MA (1990) Root rots of beans in Latin America and Africa: diagnosis, research methodologies, and management strategies. Cali Colombia. CIAT. Adesemoye, A. O., Orrell, T., & Kodati, S. (2018). Effect of Virulence of Root Rot Pathogens and Cultivar Resistance on Disease Occurrence in Dry Beans. Plant Health Progress, 19(3), 237–241. https://doi.org/10.1094/PHP-06-18-0034-RS Bilgi, V. N., Bradley, C. A., Khot, S. D., Grafton, K. F., & Rasmussen, J. B. (2008). Response of Dry Bean Genotypes to Fusarium Root Rot, Caused by Fusarium solani f. Sp. Phaseoli , Under Field and Controlled Conditions. Plant Disease, 92(8), 1197–1200. https://doi.org/10.1094/PDIS-92-8-1197 Bilgi, V. N., Bradley, C. A., Mathew, F. M., Ali, S., & Rasmussen, J. B. (2011). Root Rot of Dry Edible Bean Caused by Fusarium graminearum. Plant Health Progress, 12(1), 14. https://doi.org/10.1094/PHP-2011-0425-01-RS Boomstra, A. G., Bliss, F. A., & Beebe, S. E. (1977). New Sources of Fusarium Root Rot Resistance in Phaseolus vulgaris L.1. Journal of the American Society for Horticultural Science, 102(2), 182-185. Retrieved Sep 24, 2023, from https://doi.org/10.21273/JASHS.102.2.182 Brien, C. J. (2023) asremlPlus: Augments ASReml-R in fitting mixed models and packages in exploring prediction differences. Version 4.4.13. https://cran.r- generally project.org/package=asremlPlus/ or http://chris.brien.name/rpackages/. Burridge, J., Jochua, C. N., Bucksch, A., & Lynch, J. P. (2016). Legume shovelomics: High— Throughput phenotyping of common bean (Phaseolus vulgaris L.) and cowpea (Vigna unguiculata subsp, unguiculata) root architecture in the field. Field Crops Research, 192, 21–32. https://doi.org/10.1016/j.fcr.2016.04.008 Butler, D. G., Cullis, B. R., Gilmour, A. R., Gogel, B. J., & Thompson, R. (2023). ASReml estimates variance components under a general linear. 188. Chaudhary S, Anderson TR, Park SJ, Yu K (2006) Comparison of screening methods or resistance to fusarium root rot in common beans. J Phytopath 154:303–308. https://doi.org/10.1111/j.1439- 0434.2006.01097.x Cichy, K. A., Porch, T. G., Beaver, J. S., Cregan, P., Fourie, D., Glahn, R. P., Grusak, M. A., Kamfwa, K., Katuuramu, D. N., McClean, P., Mndolwa, E., Nchimbi-Msolla, S., Pastor- Corrales, M. A., & Miklas, P. N. (2015). A Phaseolus vulgaris Diversity Panel for Andean Bean Improvement. Crop Science, 55(5), 2149–2160. https://doi.org/10.2135/cropsci2014.09.0653 Cichy, K. A., Snapp, S. S., & Kirk, W. W. (2007). Fusarium root rot incidence and root system 179 architecture in grafted common bean lines. Plant and Soil, 300(1–2), 233–244. https://doi.org/10.1007/s11104-007-9408-0 Conner, R. L., Hou, A., Balasubramanian, P., McLaren, D. L., Henriquez, M. A., Chang, K.-F., & McRae, K. B. (2014). Reaction of dry bean cultivars grown in western Canada to root rot inoculation. Canadian Journal of Plant Science, 94(7), 1219–1230. https://doi.org/10.4141/cjps2013-416 Gepts, P., & Debouck, D. (1991). Origin, domestication, and evolution of the common bean (Phaseolus vulgaris L.). Common beans: research for crop improvement, 7-53. Gossen, B. D., Conner, R. L., Chang, K.-F., Pasche, J. S., McLaren, D. L., Henriquez, M. A., Chatterton, S., & Hwang, S.-F. (2016). Identifying and Managing Root Rot of Pulses on the Northern Great Plains. Plant Disease, 100(10), 1965–1978. https://doi.org/10.1094/PDIS-02-16-0184-FE Guo, W., Carroll, M. E., Singh, A., Swetnam, T. L., Merchant, N., Sarkar, S., Singh, A. K., & Ganapathysubramanian, B. (2021). UAS-Based Plant Phenotyping for Research and Breeding Applications. Plant Phenomics, 2021, 1–21. https://doi.org/10.34133/2021/9840192 Guzman, J. E. V. (2016). A Dissertation Submitted to the Graduate Faculty of the North Dakota State University of Agriculture and Applied Sciences. 86. Hagedorn, D. J. (1994). “Rhizoctonia root rot,” in Compendium of Bean Disease, ed. R. Hall (St Paul, MN: American Phytopathological Society Press), 9–13. Hagerty, C. H., Cuesta‐Marcos, A., Cregan, P. B., Song, Q., McClean, P., Noffsinger, S., & Myers, J. R. (2015). Mapping Fusarium solani and Aphanomyces euteiches Root Rot Resistance and Root Architecture Quantitative Trait Loci in Common Bean. Crop Science, 55(5), 1969–1977. https://doi.org/10.2135/cropsci2014.11.0805 Hall, R. 1991. Fusarium root rot. In: R. Hall, editor, Compendium of bean diseases. Am. Phytopathol. Soc. Press, St. Paul, MN. p. 9–10. Hassan Dar, G., Zargar, M. & Beigh, G. Biocontrol of Fusarium Root Rot in the Common Bean using Symbiotic Glomus mosseae and Rhizobium (Phaseolus leguminosarum . Microb Ecol 34, 74–80 (1997). https://doi.org/10.1007/s002489900036 vulgaris L.) by Harveson, R. M., Smith, J. A., & Stroup, W. W. (2005). Improving Root Health and Yield of Dry Beans in the Nebraska Panhandle with a New Technique for Reducing Soil Compaction. Plant Disease, 89(3), 279–284. https://doi.org/10.1094/PD-89-0279 Haus, M. J., Wang, W., Jacobs, J. L., Peplinski, H., Chilvers, M. I., Buell, C. R., & Cichy, K. (2020). Root Crown Response to Fungal Root Rot in Phaseolus vulgaris Middle American × Andean Lines. Plant Disease, 104(12), 3135–3142. 180 https://doi.org/10.1094/PDIS-05-20-0956-RE Jacobs, J. L., Kelly, J. D., Wright, E. M., Varner, G., & Chilvers, M. I. (2019). Determining the Soilborne Pathogens Associated with Root Rot Disease Complex of Dry Bean in Michigan. Plant Health Progress, 20(2), 122–127. https://doi.org/10.1094/PHP-11-18- 0076-S Jacobs, J. L., Oudman, K., Sang, H., & Chilvers, M. I. (2018). First Report of Fusarium brasiliense Causing Root Rot of Dry Bean in the United States. Plant Disease, 102(10), 2035. https://doi.org/10.1094/PDIS-02-18-0332-PDN Kamfwa, K., Mwala, M., Okori, P., Gibson, P., & Mukankusi, C. (2013). Identification of QTL for Fusarium Root Rot Resistance in Common Bean. Journal of Crop Improvement, 27(4), 406–418. https://doi.org/10.1080/15427528.2013.786775 Katan, J. (2017). Diseases caused by soilborne pathogens: biology, management and challenges. J. Plant Pathol. 99, 305–315. Kumar, M.S., & Kudada, N. (2018). Effect of sowing dates and weather conditions on Rhizoctonia root rot disease incidence and green pod yield of French bean. Journal of Pharmacognosy and Phytochemistry, 7, 893-899. Manganiello, G.; Nicastro, N.; Caputo, M.; Zaccardelli, M.; Cardi, T.; Pane, C. Functional Hyperspectral Imaging by High Related Vegetation Indices to Track the Wide-Spectrum Trichoderma Biocontrol Activity Against Soil-Borne Diseases of Baby-Leaf Vegetables. Front. Plant. Sci. 2021, 12, 630059. Leterme, P., & Carmenza Muũoz, L. (2002). Factors influencing pulse consumption in Latin America. British Journal of Nutrition, 88(S3), 251-254. doi:10.1079/BJN/2002714 Marzougui, A., Ma, Y., Zhang, C., McGee, R. J., Coyne, C. J., Main, D., & Sankaran, S. (2019). Advanced Imaging for Quantitative Evaluation of Aphanomyces Root Rot Resistance in Lentil. Frontiers in Plant Science, 10, 383. https://doi.org/10.3389/fpls.2019.00383 MBC 2022. Michigan Bean Commission Production Practices Survey. 2022. https://michiganbean.com/https-michiganbean-com-wp-content-uploads-2022- production-practices/ Mensack, M. M., Fitzgerald, V. K., Ryan, E. P., Lewis, M. R., Thompson, H. J., & Brick, M. A. (2010). Evaluation of diversity among common beans (Phaseolus vulgaris L.) from two centers of domestication using “omics” technologies. BMC Genomics, 11(1), 686. https://doi.org/10.1186/1471-2164-11-686 Moreira, F. F., Oliveira, H. R., Volenec, J. J., Rainey, K. M., & Brito, L. F. (2020). Integrating High-Throughput Phenotyping and Statistical Genomic Methods to Genetically Improve Longitudinal Traits in Crops. Frontiers in Plant Science, 11, 681. 181 https://doi.org/10.3389/fpls.2020.00681 Mukankusi, C. M., Melis, R. J., Derera, J., Buruchara, R. A., & Mark, D. (2011). A screening technique for resistance to Fusarium root rot of common bean. Afr. J. Plant Sci. Nakedde, T., Ibarra-Perez, F. J., Mukankusi, C., Waines, J. G., & Kelly, J. D. (2016). Mapping of QTL associated with Fusarium root rot resistance and root architecture traits in black beans. Euphytica, 212(1), 51–63. https://doi.org/10.1007/s10681-016-1755-6 Nicoli, A., Zambolim, L., Paula Júnior, T. J., Vieira, R. F., Teixeira, H., & Carneiro, J. E. S. (2012). Resistance of advanced common bean lines to Fusarium root rot. Tropical Plant Pathology, 37(6), 393–398. https://doi.org/10.1590/S1982-56762012000600003 Oladzad, A., Zitnick-Anderson, K., Jain, S., Simons, K., Osorno, J. M., McClean, P. E., & Pasche, J. S. (2019). Genotypes and Genomic Regions Associated With Rhizoctonia solani Resistance in Common Bean. Frontiers in Plant Science, 10, 956. https://doi.org/10.3389/fpls.2019.00956 Pandey, A. K., Burlakoti, R. R., Rathore, A., & Nair, R. M. (2020). Morphological and molecular characterization of Macrophomina phaseolina isolated from three legume crops and evaluation of mungbean genotypes for resistance to dry root rot. Crop Protection, 127, 104962. https://doi.org/10.1016/ j.cropro.2019.104962 Paparu, P., Acur, A., Kato, F., Acam, C., Nakibuule, J., Musoke, S., Nkalubo, S., & Mukankusi, C. (2018). PREVALENCE AND INCIDENCE OF FOUR COMMON BEAN ROOT ROTS IN UGANDA. Experimental Agriculture, 54(6), 888–900. https://doi.org/10.1017/S0014479717000461 Peña, P. A., Steadman, J. R., Eskridge, K. M., & Urrea, C. A. (2013). Identification of sources of resistance to damping-off and early root/hypocotyl damage from Rhizoctonia solani in common bean (Phaseolus vulgaris L.). Crop Protection, 54, 92–99. https://doi.org/10.1016/j.cropro.2013.04.014 Román-Avilés, B., & Kelly, J. D. (2005). Identification of Quantitative Trait Loci Conditioning Resistance to Fusarium Root Rot in Common Bean. Crop Science, 45(5), 1881–1890. https://doi.org/10.2135/cropsci2005.0028 Román-Avilés, B., Snapp, S. S., & Kelly, J. D. (2004). Assessing root traits associated with root rot resistance in common bean. Field Crops Research, 86(2–3), 147–156. https://doi.org/10.1016/j.fcr.2003.08.001 Rubiales, D., Fondevilla, S., Chen, W., Gentzbittel, L., Higgins, T. J. V., Castillejo, M. A., Singh, K. B., & Rispail, N. (2015). Achievements and Challenges in Legume Breeding for Pest and Disease Resistance. Critical Reviews in Plant Sciences, 34(1–3), 195–236. https://doi.org/10.1080/07352689.2014.898445 182 Schneider, K. A., Grafton, K. F., & Kelly, J. D. (2001). QTL Analysis of Resistance to Fusarium Root Rot in Bean. Crop Science, 41(2), 535–542. https://doi.org/10.2135/cropsci2001.412535x Schneider, K. A., & Kelly, J. D. (2000). A Greenhouse Screening Protocol for Fusarium Root Rot in Bean. HortScience, 35(6), 1095–1098. https://doi.org/10.21273/HORTSCI.35.6.1095 Schwartz, H. (2011). Root Rots of Dry Beans. 4. Sendi, Y., Pfeiffer, T., Koch, E., Mhadhbi, H., & Mrabet, M. (2020). Potential of common bean (Phaseolus vulgaris L.) root microbiome in the biocontrol of root rot disease and traits of performance. Journal of Plant Diseases and Protection, 127(4), 453–462. https://doi.org/10.1007/s41348-020-00338-6 Singh, S. P., & Schwartz, H. F. (2010). Breeding Common Bean for Resistance to Diseases: A Review. Crop Science, 50(6), 2199–2223. https://doi.org/10.2135/cropsci2009.03.0163 Snapp, S., Kirk, W., Román-Avilés, B., & Kelly, J. (2003). Root Traits Play a Role in Integrated Management of Fusarium Root Rot in Snap Beans. HortScience, 38(2), 187–191. https://doi.org/10.21273/HORTSCI.38.2.187 Trachsel, S., Kaeppler, S. M., Brown, K. M., & Lynch, J. P. (2011). Shovelomics: High throughput phenotyping of maize (Zea mays L.) root architecture in the field. Plant and Soil, 341(1–2), 75–87. https://doi.org/10.1007/s11104-010-0623-8 Uebersax, M. A., Cichy, K. A., Gomez, F. E., Porch, T. G., Heitholt, J., Osorno, J. M., Kamfwa, K., Snapp, S. S., & Bales, S. (2023). Dry beans ( Phaseolus vulgaris L.) as a vital component of sustainable agriculture and food security—A review. Legume Science, 5(1). https://doi.org/10.1002/leg3.155 USDA Crop Production 2022 Summary 01/12/2023. (2022). Crop Production. Van Schoonhoven, A. (1987). Standard system for the evaluation of bean germplasm. CIAT. Wang, W., Jacobs, J. L., Chilvers, M. I., Mukankusi, C. M., Kelly, J. D., & Cichy, K. A. (2018). QTL Analysis of Fusarium Root Rot Resistance in an Andean × Middle American Common Bean RIL Population. Crop Science, 58(3), 1166–1180. https://doi.org/10.2135/cropsci2017.10.0608 Wei T, Simko V (2021). R package 'corrplot': Visualization of a Correlation Matrix. (Version 0.92), https://github.com/taiyun/corrplot. Zitnick-Anderson, K., Oladzadabbasabadi, A., Jain, S., Modderman, C., Osorno, J. M., McClean, P. E., & Pasche, J. S. (2020a). Sources of Resistance to Fusarium solani and Associated Genomic Regions in Common Bean Diversity Panels. Frontiers in Genetics, 11, 475. 183 https://doi.org/10.3389/fgene.2020.00475 184