COLOR RETENTION AND ANTHOCYANIN CONCENTRATION IN CANNED BLACK BEANS By Othman Mubarak Al Dossary     A THESIS Submitted to  Michigan State University  in partial fulfillment of the requirements  for the degree of Crop and Soil Sciences - Master of Science 2016   ABSTRACT COLOR RETENTION AND ANTHOCYANIN CONCENTRATION IN CANNED BLACK BEANS  By Othman Mubarak Al Dossary Common beans (Phaselous vulgaris L) represents the most widely cultivated of the Phaseolus species. It is a nutritious crop that provides protein, minerals and dietary fiber. Black beans are one of the most popular market classes, with large consumption in the U.S., Mexico, and Central America. Black beans are distinguished from other common beans classes by high anthocyanins in the seed coat which impart the black color. The water soluble anthocyanins make the canning process for black beans challenging. Canning requires several steps, including soaking and boiling. Although the canning process makes beans edible, it also alters the physical and chemical properties of the seeds. In the case of black beans, soaking seeds accelerates anthocyanin loss which lighten the seeds and produce a canned product with poor appearance.  Genetic variability for color retention can be exploited to select and incorporate superior genetics for this trait. The objective of this research was to assess the genotypic and phenotypic diversity for a set of 69 black bean breeding lines and cultivars from the major U.S. public bean breeding programs. Each of the lines was grown in field trials in 2013 and 2014. They were evaluated for agronomic, canning characteristics and anthocyanins profile of raw and canned seed. Color retention as determined by a trained sensory panel on a scale of 1 to 5 was highly variable and ranged from 1.4 to 4.5.  Delphinidin-3-glucoside was identified as the dominant anthocyanin with the highest concentration among black bean genotypes. The anthocyanin malvidin-3-glucoside was found to be retained after canning more than the other two anthocyanins. Genome wide association analysis was conducted to determine genomic regions responsible for color retention and canning quality in black beans that were genotyped with 5398 SNP markers. A region on Pv05 at 39Mb was associated with color retention and was polymorphic candidates for MAS.  iii ACKNOWLEDGEMENTS First of all, I would like to thank my almighty God for honoring me with his grace and mercy, and leading me to what I have achieved. I would like to thank my parents, my wife, and my daughter for their precious patience and infinite support. I will never forget the continued guidance and help from my advisor Dr. Karen Cichy. Many thanks for my committee members Dr. James Kelly and Dr. Kirk Dolan for agreeing to be in my committee. Also, I would like thank my colleagues Dennis Katuramu and Sharon Hooper and all my lab members. Finally, it was a blessing to be a graduate student in one of the best agricultural universities in the world, Michigan State University (MSU).            iv TABLE OF CONTENTS LIST OF TABLES          vi LIST OF FIGURES          x  KEY TO ABBREVIATIONS        xiv  CHAPTER 1            1 INTRODUCTION AND PROBLEM STATEMENT     1 CHAPTER 2           4 LITERATURE REVIEW         4 2.1. Seed Quality in Black Bean        4     2.1.1 Common Bean         4     2.1.2 Black Bean          7     2.1.3 Canning Quality         8     2.1.4 Canning Quality Evaluation       11 2.2. Color Retention and Anthocyanin in Black Bean     13     2.2.1 Impact of Polyphenolics (anthocyanin and tannin) on Black Bean               Color Retention After Canning       13 2.3. Linkage Disequilibrium Mapping       15     2.3.1. Association Mapping        15 CHAPTER 3        17 PHENOTYPIC EVALUATION AND CANNING QUALITY IN BLACK BEAN 17 INTRODUCTION        17 MATERIALS AND METHODS        18 3.1. Plant Materials          18 3.2. Field Description         22 3.3. DNA Extraction          22 3.4. SNP Genotyping          22 3.5. Canning Process          23 3.6. Evaluation Process         23 3.7. Statistical Analysis         25 RESULTS AND DISSCUSSION        27 3.8. Genome-Wide Association Mapping       5  CONCLUSION AND RECOMMENDATION      8 CHAPTER 4        8 PHENOTYPIC EVALUATION OF ANTHOCYANIN IN BLACK BEAN  8 INTRODUCTION        8 MATERIALS AND METHODS        8 4.1. Materials Selection        8 v 4.2. Anthocyanin Extraction and Purification      8 4.3. UV Spectrophotometer Assay         4.4. Anthocyanin Quantification Using Liquid Chromatography Mass Spectrometry (LC-MS)        9 4.5. Statistical Analysis        9 RESULTS AND DISSCUSSION        9 CONCLUSION AND RECOMMENDATION      116 APPENDIX        117 BIBLIOGRAPHY        150          vi LIST OF TABLES Table 1. A list of the 69 black bean genotypes that were used to evaluate color retention of canned black beans.          20  Table 2A. Analysis of variance for yield trait collected on 69 black bean genotypes grown for 2 years (2013 and 2014).         28 Table 2B. Analysis of variance for seed weight trait collected on 69 black bean genotypes grown for 2 years (2013 and 2014).        28 Table 2C. Analysis of variance for appearance trait collected on 69 black bean genotypes grown for 2 years (2013 and 2014).        29 Table 2D. Analysis of variance for color trait collected on 69 black bean genotypes grown for 2 years (2013 and 2014).         29 Table 2E. Analysis of variance for L* value trait collected on 69 black bean genotypes grown for 2 years (2013 and 2014).         30 Table 2F. Analysis of variance for a* value trait collected on 69 black bean genotypes grown for 2 years (2013 and 2014).         30 Table 2G. Analysis of variance for b* value trait collected on 69 black bean genotypes grown for 2 years (2013 and 2014).         31 Table 3. Means and ranges for yield, seed weight, appearance, color, L*, a*, and b* traits for the years 2013 and 2014.          Table 4. Pearson correlation for the seven traits (yield, seed weight, appearance, color, L*, a*, and b*) for year 2013.         4 Table 5. Pearson correlation for the seven traits (yield, seed weight, appearance, color, L*, a*, and b*) for year 2014.         4  Table 6A. Means of the seven traits (yield, seed weight, appearance, color, L*, a*, and b*) of 66 black bean genotypes that obtained from the year 2013.    44 Table 6B. Mean of the seven traits (color, appearance, L*, a*, b*, seed weight, and yield) of 66 black bean genotypes that obtained from the year 2014.    48 Table 7A. Means and standard deviations for all the major and sub branches that obtained from the neighbor joining tree for the seven traits (color, appearance, L*, a*, b*, seed weight, and yield) of 69 black bean genotypes that obtained from the year 2013.   5  vii Table 7B. Means and standard deviations for all the major and sub branches that obtained from the neighbor joining tree for the seven traits (color, appearance, L*, a*, b*, seed weight, and yield) of 69 black bean genotypes that obtained from the year 2014.   5 Table 8A. GWAS significant markers, genome position, p-value, R2, and phenotypic effect associated with the seed yield trait for two-year data (2013 and 2014) and the average of two years.           7  Table 8B. GWAS significant markers, genome position, p-value, R2, and phenotypic effect associated with the seed weight trait for two-year data (2013 and 2014) and the average of two years.          7 Table 8C. GWAS significant markers, genome position, p-value, R2, and phenotypic effect associated with the appearance trait for two-year data (2013 and 2014) and the average of two-years .          7 Table 8D. GWAS significant markers, genome position, p-value, R2, and phenotypic effect associated with the color trait for two-year data (2013 and 2014) and the average of two years.            7 Table 8E. GWAS significant markers, genome position, p-value, R2, and phenotypic effect associated with the L* trait for two-year data (2013 and 2014) and the average of two years.              Table 8F. GWAS significant markers, genome position, p-value, R2, and phenotypic effect associated with the a* trait for two-year data (2013 and 2014) and the average of two years.             8 Table 8G. GWAS significant markers, genome position, p-value, R2, and phenotypic effect associated with the b* trait for two-year data (2013 and 2014) and the average of two years.             8 Table 9. A list of the 12 black bean genotypes evaluated traits for total anthocyanins concentrations and the three specific anthocyanins quantification in raw and canned black beans. 8 Table 10. Delphinidin-3-glucoside calibration curve points made to calculate the total anthocyanins mass in black bean genotype extracts.     9  Table 11. Three major anthocyanins in black beans represent the study standards.  9 Table 12A. Analysis of variance for total anthocyanins (TA) for 12 black bean genotypes with two levels of conditions (raw and canned), two levels of years (2013 and 2014), and two replications.           Table 12B. Analysis of variance for Delphindin-3-glucoside concentration for 12 black bean genotypes with two levels of conditions (raw and canned), two levels of years (2013 and viii 2014), and two replications.         Table 12C. Analysis of variance for petunidin-3-glucoside concentration for 12 black bean genotypes with two levels of conditions (raw and canned), two levels of years (2013 and 2014), and two replications.         Table 12D. Analysis of variance for malvinidin-3-glucoside concentration for 12 black bean genotypes with two levels of conditions (raw and canned), two levels of years (2013 and 2014), and two replications.         Table 13A. Tukey test for the trait of total anthocyanins mg/1g total concentration of 12 raw black bean genotypes in 2013 and 2014, with two replications.     Table 13B. Tukey test for the trait of total anthocyanins mg/1g total concentration of 12 canned black bean genotypes in 2013 and 2014, with two replications.   10 Table 14. Anthocyanins Retention percentage for the total anthocyanin data mg/1g total concentration that obtained from 12 black bean genotypes with two years (2013 and 2014).             10 Table 15A. Tukey test for the trait of SAQ, for the three anthocyanins (delphindin-3-glucoside, petunidin-3-glucoside, and Malvidin-3-glucoside) concentrations for the 12 raw black bean genotypes in 2013with two replication.       10 Table 15B. Tukey test for the trait of SAQ, for the three anthocyanins (delphindin-3-glucoside, petunidin-3-glucoside, and Malvidin-3-glucoside) concentrations for the 12 canned black bean genotypes in 2013 with two replications.      10 Table 15C. Tukey test for the trait of SAQ, for the three anthocyanins (delphindin-3-glucoside, petunidin-3-glucoside, and Malvidin-3-glucoside) concentrations for the 12 raw black bean genotypes in 2014 with two replications.      10 Table 15D. Tukey test for the trait of SAQ, for the three anthocyanins (delphindin-3-glucoside, petunidin-3-glucoside, and Malvidin-3-glucoside) concentrations for the 12 canned black bean genotypes in 2014 with two replications.      10 Table 16. Anthocyanins retention percentage for three anthocyanins concentration mg/1g (delphinidin-3-glucoside, pentuinidin-3-glucoside, and malvinidin-3-glucoside) that obtained from 12 black bean genotypes with two years (2013 and 2014).   1 Table 17A. Pearson Correlation for the eight traits (TAC, Delephinidin-3-glucoside, Pentuendin-3-glucoside, Malvinidin-3-glucoside, color, L*, a*, and b*) of the year 2013.  11  Table 17B. Pearson Correlation for the eight traits (TAC, Delephinidin-3-glucoside, Pentuendin-3-glucoside, Malvinidin-3-glucoside, color, L*, a*, and b*) of the year 2014.  11 ix Table 18. A list of the 69 black bean genotypes that were involved in this study, and each genotype is combined with a picture which has been taken for the canned genotypes in the year of 2014.            11 Table 19A. Means and standard deviations for all the major and sub branches (including genotypes averages) that obtained from the neighbor joining tree for the seven traits (seed yield, seed weight, appearance, color, L*, a*, and b*) of 69 black bean genotypes that obtained from the year 2013.          1  Table 19B. Means and standard deviations for all the major and sub branches (including genotypes averages) that obtained from the neighbor joining tree for the seven traits (seed yield, seed weight, appearance, color, L*, a*, and b*) of 69 black bean genotypes that obtained from the year 2014.          13                   x LIST OF FIGURES Figure 1. A neighbor joining tree of 69 black bean genotypes developed using 2800 SNP         markers.          19  Figure 2A. Color rating chart for black beans.      24 Figure 2B. Appearance rating chart for black beans.      24 Figure 3. Histograms for the data sets grown for two years for all the traits (yield, seed weight,                 appearance, color, L*, a*, and b*) of canned black beans for 2013, and 2014.  3 Figure 4. Pearson correlations for the traits of color, L*, a*, and b* for the 2 years 2013          and 2014.            Figure 5. Manhattan plot of seed yield values averaged across 2013 of 61 Black beans genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 3 principal components to account for population structure. The black line is p=0.01 threshold for significance.    6 Figure 6. Manhattan plot of seed weight values averaged across 2013 of 61 Black beans genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 7 principal components to account for population structure. The black line is p=0.01 threshold for significance.    6 Figure 7. Manhattan plot of appearance rating values averaged across 2013 of 61 Black beans genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 2 principal components to account for population structure. The black line is p=0.01 threshold for significance, and the red line is p=0.001 threshold for significance.       6 Figure 8. Manhattan plot of color ratings averaged across 2013 of 61 Black beans genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix. The black line is p=0.01 threshold for significance. 6 Figure 9. Manhattan plot of the trait of L* values averaged across 2013 of 61 Black beans genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 7 principal components to account for population structure. The black line is p=0.01 threshold for significance.  6 Figure 10. Manhattan plot of the trait of a* values averaged across 2013 of 61 Black beans genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 3 principal components to account for population structure. The black line is p=0.01 threshold for significance.  6 xi Figure 11. Manhattan plot of the trait of b* values averaged across 2013 of 61 Black beans genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 3 principal components to account for population structure. The black line is p=0.01 threshold for significance.  6 Figure 12. Manhattan plots of seed yield values of 2014 and two-year average for 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix. The black line is p=0.01 threshold for significance, and the red line is p=0.001 threshold for significance.   6 Figure 13. Manhattan plots of seed weight values of 2014 and two-year average for 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 7 principal components to account for population structure. The black line is p=0.01 threshold for significance.   Figure 14. Manhattan plots of appearance ratings of 2014 and two-year average for 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 2 principal components to account for population structure. The black line is p=0.01 threshold for significance.  7 Figure 15. Manhattan plots of color ratings of 2014 and two-year average for 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 7 principal components to account for population structure. The black line is p=0.01 threshold for significance.  7 Figure 16. Manhattan plots of the trait of L* of 2014 and two-year average for 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 5 principal components to account for population structure. The black line is p=0.01 threshold for significance, and the red line is p=0.001 threshold for significance.       7 Figure 17. Manhattan plots of the trait of a* of 2014 and two-year average for 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix. The black line is p=0.01 threshold for significance.          7 Figure 18. Manhattan plots of the trait of b* of 2014 and two-year average for 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 3 principal components to account for population structure. The black line is p=0.01 threshold for significance.  7 Figure 19. Manhattan plot of the trait of shiny trait of 61 black bean genotypes with 2800 SNPs (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix. The black line is p=0.01 threshold for significance. The black line is p=0.01 threshold for significance, and the red line is p=0.001 threshold for Significance.             7 xii Figure 20. Calibration curve for delphinidin-3-glucoside standard made to calculate the total anthocyanins mass in black bean genotype extracts.     9 Figure 21. LC-MS chromatograph shows the three anthocyanins peaks have been detected in the studied black bean genotype extracts.       9 Figure 22. Pearson correlations for the trait of total anthocyanin and the two traits of color and L*, a*, and b* color scans for the two years 2013 and 2014.    114 Figure 23. Pearson correlations for the trait of color and malvidin-3-glucoside concentration for the raw and canned genotypes grown for two years 2013 and 2014.   115 Figure 24. QQ plot for the trait of seed yield values averaged across 2013 of 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 3 principal components to account for population structure.         136 Figure 25. QQ plot for the trait of seed weight values averaged across 2013 of 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 7 principal components to account for population structure.         137 Figure 26. QQ plot for the trait of appearance ratings averaged across 2013 of 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 2 principal components to account for population structure.         138 Figure 27. QQ plot for the trait of color ratings averaged across 2013 of 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix.       139 Figure 28. QQ plot for the trait of L* values averaged across 2013 of 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 7 principal components to account for population structure.           140 Figure 29. QQ plot for the trait of a* values averaged across 2013 of 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 3 principal components to account for population structure            141 Figure 30. QQ plot for the trait of seed weight values averaged across 2013 of 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 3 principal components to account for population structure.         142 xiii Figure 31. QQ plot for the trait of seed yield values averaged across 2014 of 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix.     143 Figure 32. QQ plot for the trait of seed weight values averaged across 2014 of 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 7 principal components to account for population structure.         144 Figure 33. QQ plot for the trait of appearance ratings averaged across 2014 of 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 2 principal components to account for population structure.         145 Figure 34. QQ plot for the trait of color ratings averaged across 2014 of 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 7 principal components to account for population structure.           146 Figure 35. QQ plot for the trait of L* values averaged across 2014 of 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 7 principal components to account for population structure.           147 Figure 36. QQ plot for the trait of a* values averaged across 2014 of 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 5 principal components to account for population structure.           148 Figure 37. QQ plot for the trait of b* values averaged across 2014 of 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix.        149        xiv  KEY TO ABBREVIATIONS TAC  Total anthocyanin concentration SAQ  Specific anthocyanin quantification AR  Anthocyanins retention percentage CVD - cardiovascular disease  CBB - Common bacterial blight QTLs - Quantitative trait loci SNP  Single nucleotide polymorphism MA - Middle American gene pool M  Mesoamerica FAO - The Food and Agriculture Organization SSA - Sub-Saharan Africa MAS - Marker assisted selection  RGB - Red, green, blue HSV - Hue, saturation, and value UV  Ultraviolet radiation LD - Linkage disequilibrium GWAS - Genome-wide association mapping SSR - Simple sequence repeats marker  SV - Saginaw Valley MLM - Mixed linear model xv ANOVA - Analysis of variance PCA - Principal Component Analysis LOD - Logarithm (base 10) of odds Q-Q - Quantile- quantile plot LC-MS - Liquid chromatography-mass spectrometry instrument  UPLC - Ultra Performance Liquid Chromatography ESI - Electrospray ionization mass spectrometry CV - Coefficient of Variation  PCA  Principal component analysis           1 CHAPTER 1 INTRODUCTION AND PROBLEM STATEMENT Dry beans (Phaseolus vulgaris L) are the most widely cultivated of the Phaseolus species. They are an important crop in many parts of the world including Africa, Latin America, and North America. Dry beans are a nutritious food, rich in protein, fiber, and some essential minerals such as Ca, Fe, and Zn (Ellis, 2007). Including dry beans in a regular diet may contribute to avoidance and treatment of chronic diseases such as, diabetes, cardiovascular disease (CVD) and cancer, which are leading causes of mortality in the U.S. (Hutchins et al., 2012). There are over 14 commercial dry bean classes including dark red kidney, pinto, cranberry, and black (Singh, 2008). The black bean class is one of the most popular market classes, with large consumption in the U.S., Mexico, Central America, and South America. Like most dry bean types, black beans are considered to be a good source of protein, fiber, folate, and minerals.  The uniqueness of this type of beans lies in their black color, can determined by the high amount of anthocyanin compounds in their seed coat. Black beans contain the highest percentage of anthocyanins among all dry bean types (Dzomba et al., 2013). Anthocyanins offer benefits to human health through their anti-oxidant activity, which can help eliminate free radicals (Díaz et al., 2010; La Cruz et al., 2013). The black color is not only related to anthocyanins, but to other phenolic compounds such as tannins, which are concentrated in the internal layer of the seed coat (Cichy et al., 2014). Higher tannin concentrations in conjunction with low anthocyanin content lead to more brownish seeds (Marles et al., 2010). Genetic variability for anthocyanin concentrations exists among black bean genotypes. Selecting genotypes with higher anthocyanins has the potential to increase the health benefits of the beans. In addition, black beans are usually sold as canned products. Canned bean color retention is an important attribute for processors and 2 consumers. The degree of color retention following canning depenmakeupidentified. The optimal black bean seeds for commercial use not only retain more color, but have the ability to expand rapidly and uniformly during soaking, and remain firm during processing (Kelly et al., 2015).  Currently a small scale canning protocol is the standard method used to identify breeding lines with acceptable canning quality and color retention. This method requires at least 200g of seed per genotype and specialized equipment, therefore, it is not feasible for early generation materials neither it is accessible to all breeders. Molecular markers as a tool are extensively used in plant breeding, and compared to traditional breeding, the use of molecular markers reduces time and effort (Sivolap, 2013).  In beans, molecular markers for numerous disease-resistant genes have been integrated into breeding programs. For instance, common bacterial blight (CBB) is a major yield-reducing factor of common bean. In a previous study, two sequence characterized amplified regions (SCARs) BC420 and SU91 were linked to two CBB resistant quantitative trait loci (QTLs) (Miklas et al., 2006). Through the development of candidate gene marker and candidate gene selection, a recent study determined nine and six polymorphic markers in the BC420- and SU91-QTL loci, respectively (Shi et al., 2012). While QTL studies have been conducted for canning quality and color retention in black beans (Wright and Kelly, 2014; Cichy et al., 2014) molecular markers have yet to be developed for these traits and integrated into breeding programs. Molecular markers for these traits would be valuable to bean breeders.  The appearance and color retention of canned black bean products influences the acceptability of the product by consumers.  During processing the beans stay whole and intact.  Also, consumers 3 prefer that few ingredients are added to the product and would appear on the label. Therefore, the starting seed is very important in determining the end product quality.  While the environmental and handling conditions influence the quality, genetics is also a major determinant of quality. This study aims to identify genomic regions responsible for controlling color retention in 69 black bean breeding lines from the major U.S. bean breeding programs. The objectives of this research are to: 1- Assess genetic variability for color retention in black beans.    2- Assess genetic variability for anthocyanin concentration in black beans. 3- Apply genome wide association analysis (GWAS) to identify genomic region controlling color retention and anthocyanin concentration in black beans. SNP markers were generated from 69 black bean genotypes. Phenotypic data on, canned bean color, appearance, and anthocyanins concentration were collected over two field experiments. Application of this knowledge to breeding effort will contribute to the end use quality of black beans.             4 CHAPTER 2 LITERATURE REVIEW 2.1. Seed Quality in Black Bean 2.1.1 Common Bean Dry bean (Phaseolus vulgaris L) is an important crop in many parts of the world including Africa, Latin America, and North America. Dry bean is considered the third most substantial legume crop , cassava and maize are considered the earliest crops in human history in the Americas (Broughton et al., 2003). Together they were the main staple food in the low to mid-latitudes of the Americas (Broughton et al., 2003). Dry beans are known to have originated in the Americas, based on two discoveries of wild dry beans in Argentina and Guatemala, and some archaeologic remains obtained from the Americas (Mamidi et al., 2011). There are five domesticated species and P. vulgaris is the oldest species. The other four domesticated species are P. lunatus, P. dumosus, P. coccineus, and P. acutifolius (Tuberosa et al., 2014). P. dumosus and P. coccineus are the most closely related to P. vulgaris, and they are considered partially intercrossable. However, the other two domesticated species (P. lunatus, P. acutifolius) are more distantly related (Tuberosa et al., 2014) and only p. acutifolius can be crossed with P. vulgaris using embryo rescue technology.  Dry beans include two major gene pools: The Middle American (MA) gene pool, which extends from Mexico to Central America and Colombia, and the Andean gene pool located in Peru, Chile, Bolivia, and Argentina (Mamidi et al., 2011). These two gene pools have parallel wild and domesticated geographical structures, but the MA possess great diversity than the Andean gene pool (Bitocchi et al., 2012). Gene pool division is based on phaseolin seed protein variation, DNA marker diversity, morphology, isozymes, and mitochondrial DNA differences (Mamidi et al., 5 2011).  Dry beans from the both gene pools are separated into many different market classes that are identified based on growth habit, phonological traits, seed size, shape, color (red, black, white, yellow, and purple), canning and cooking qualities (Singh, 2008). Each gene pool has three races which are characterized based on the variation of plant and seed morphology and adaptation regimes (Singh et al., 1991). The MA gene pool contains the races Mesoamerica (M), Durango, and Jalisco. Black beans are in the (M) race, and are adapted to tropical lowlands (Beebe et al., 2014).  Dry bean is a mainly self-pollinated and annual legume that grows under different environmental conditions such as warm lowlands, dry and humid highlands (Beebe et al., 2014). It is a diploid crop (2n=2x=22) with a small genome ranging between 450-650 Mb and 11 chromosome (Melotto et al., 2005; Schmutz et al., 2014) . Bean plants have a short life cycle (90 days), and a maximum of three generations can be produced annually (Fageria & Santos, 2016). In 2010, world bean production was around 23 million tons, with almost half of it produced in Latin America and Africa (Creamer, 2014). The Food and Agriculture Organization (FAO) reports that low income countries are the source of half of the world bean production. The other half comes from other countries, such as the U.S., where dry bean is economically valuable with 769,000 hectares planted in 2012, and with a farm gate value of $1.5 billion (USDA-NASS, 2015). Dry bean is more economically valuable than all other legumes such as, lentil, pea, chickpea, and cowpea., these data give an insight into the present and possible future importance of this crop (Porch et al., 2013). Globally, dry beans have the highest amount of direct consumption among all legumes (Akibode and Maredia, 2011). It has been estimated that the consumption of beans in Sub-Saharan Africa (SSA) meets more than 50% of dietary protein needs for households  (Beebe et al., 2014). For instance, bean consumption in some regions of Kenya reaches up to 66 kg per 6 person annually, which is much higher than the amount consumed in Latin America which may reach 17 kg per person annually (Broughton et al., 2003).  Although the major protein sources are of animal origin, they are not affordable to many people especially in low-income areas. Dry bean is a major nutritional dietary legume worldwide, due to its high concentration of plant protein, fiber, and complex carbohydrates (Mamidi et al., 2011). It contains numerous essential and health-promoting compounds. It is a rich source of micronutrients (Ca, Cu, Fe, Mg, Mn, and Zn), calories, and vitamins (folate, thiamin, and B group vitamins) (Acosta-Gallegos & Kelly, 2007). Bean consumption is beneficial to human health and may help in the avoidance of many mineral deficiencies and diseases (Zhang et al., 2014). Indeed, experimental, epidemiological and clinical studies have shown that including dry beans in a regular diet may contribute to the prevention and treatment of chronic diseases such as diabetes, CVD, obesity, and cancer, which are the main causes of the death in the U.S. (Hutchins et al., 2012; Bauer et al., 2011; Oseguera-toledo et al., 2011; Thompson et al., 2009).  Dry beans are a source of polyphenols, antioxidants, and many different phytochemicals (Joseph et al., 2014). Analysis of bean phytochemicals indicated the presence of several bioactive compounds including alkaloids, anthocyanin, catechin, flavonoids, phytic acid, quercetin, and tannins (Nyau et al., 2015). Seeds coats of dry beans contain delphinidin, cyanidin, and phenolic acids such as gallic, vanillic, caffeic, coumaric and ferrulic acids. Several studies have revealed that polyphenols, flavonoids, and antioxidants exist in dry beans which are associated with beneficial health effects such as anti-cancer properties (Sancho et al., 2015; Gebrelibanos et al., 2013; Moreno-jiménez et al., 2015). Polyphenolic components of dry beans function as cancer chemopreventive agents, particularly by their antioxidant characteristics (Gebrelibanos et al., 2013).  7 A recent study has shown that different dry bean varieties have remarkably different cancer inhibitory activity associated with their genetic heritage (Zhu et al., 2012). Thompson et al., (2009) have evaluated the anticancer activity for six dry beans (Phaseolus vulgaris L.) market classes. During their experiment, processed dry bean powder were fed to lab rats in order to determine a possible impact of bean on chemically induced mammary carcinogenesis in comparison to the typical diet in a preclinical model for breast cancer. They found that a bean diet was able to diminish the percentage of cancerous rats by 20%, and different bean market classes varied in terms of cancer inhibitory activity (Thompson et al., 2009). Therefore, characterizing the diversity of dry bean cultivars will lead to more understanding to the chemical composition of dry beans that includes bioactive components which are responsible of anticancer activity.  2.1.2 Black Bean The consumption of black bean (Phaseolus vulgaris L) is widespread, and black beans are a  major food in Mexico (Dong et al., 2007). In Brazil, black bean represents 17% of Brazilian bean production with about 490,000 tons produced per year on nearly 370,000 ha ( De Faria et al., 2014). Also, it is the second or third most important bean market class in the U.S. with highest production percentages in Michigan and North Dakota (Kelly et al., 2015). Indeed, Michigan is the top producer of black beans in the United States, contributing 57% of the total production. That is attributed to less risk related to cultivating colored seeds types as black beans, and to the increasing consumption of this type of bean due to a better understanding of its health advantages (Kelly et al., 2015). Furthermore, black bean popularity has increased in the U.S. In fact, the yearly consumption per person in the United States was only 4.5 g in 1970, but it dramatically multiplied 68 times by 2008 to 304g per person (Cichy et al., 2014).  8 The nutritional value of black beans has been known for decades. In recent decades, black bean has been demonstrated to be used in conventional medical practices. For instance, in Japan and Korea, black beans are widely used as a medicinal food since their seed coat is rich in antioxidants activity derived from the anthocyanins. Like most dry bean types, black beans are considered to be a good source of protein, fiber, and folate. Also, they are nutritionally rich in, anthocyanins, and other polyphenols (Jiang et al., 2014). The uniqueness of this type of beans lies in their black color, which is caused by the high level of anthocyanin compounds in their seed coat. Among all dry bean types, black beans contain the highest percentage of anthocyanins including delphinidin 3-glucoside, petunidin 3-glucoside, and malvidin 3-glucoside (Dzomba et al., 2013). Anthocyanins play a role in maintaining human health through their anti-oxidant activity, which helps to capture oxygen free radicals (O2, OH) (Díaz et al. 2010; La Cruz et al. 2013). Those and other reactive oxygen species (H2O2; 1O2) can stimulate lipid peroxidation, which can cause cell death and tissue damage (Lee et al., 2008). Many studies indicate a correlation between arthritis, cirrhosis, emphysema, atherosclerosis, cancer and the oxidative damage catalyzed by the free radicals. Many researchers have found that including more antioxidants in a regular diet may reduce CVD risk. Therefore, consuming food with high anthocyanin content may diminish oxidative damage and contribute to overall health (Lee et al., 2008).  2.1.3 Canning Quality  Canning is an important process to improve the palatability of beans and to reduce flatulence factors (raffinose oligosaccharides) (Xu & Chang, 2009; et al., 2003). During the canning process beans lose many of their anti-nutritional factors (Xu & Chang, 2009). Several studies on canning indicate that food preparation and cooking methods are as important as the food itself to determine nutritional value (Boateng et al., 2008). The canning process includes soaking and 9 cooking the seeds following several variable protocols, and then transferring them to cans. Furthermore, thermal processing (cooking) is a traditional way to preserve food and to ensure its safety and storability (Homayouni & Azizi, 2015). Usually, beans are thermally processed by boiling, pressure boiling, or steaming before consumption (Xu & Chang, 2009). Many thermal processes successfully reduce anti-nutrients such as lectin and trypsin inhibitors (Boateng et al., 2008).   The canning quality of dry beans is affected by the growing environment, canning procedures, and the interaction between the two factors. Moreover, the canning process changes the physical properties and chemical composition of the bean (Xu & Chang, 2009). The impact of the environment was shown by Khanal et al (2015) to be significant for the texture of the beans. Mostly, bean processors prefer seeds that are well hydrated, rapidly expanded, equally cooked, and do not cluster in the bottom of the can during the canning process (Khanal et al., 2015). Other canning quality traits that are susceptible to alteration after processing are for example, color, texture, flavor, visual appearance, and digestibility. Canning quality traits are important to bean consumers and processors, and improving them will contribute to the profitability of canned bean market (Khanal et al., 2015).  Black beans are usually sold as canned products in the U.S. market. Black bean color and appearance are important attributes for both processors and consumers (Cichy et al., 2014). During the canning process, dry bean types that show darker colors such as black, small red, and red kidney are more likely to lose their color. The high amount of anthocyanin in black beans increases their tendency to leach out during industrial canning processes such as washing and soaking, and cause discoloration of canning liquid (Dzomba et al., 2013). Black bean seed coats do not contain only extractable anthocyanins, but also bound condensed tannins, and bound anthocyanins (Kelly 10 et al., 2015). The amounts of each compound vary due to genetic variability and environmental aspects such as growing location (Rocha-Guzmán et al., 2007). These compounds could impact anthocyanin concentration in the seed coat would express darker (blacker) seed color (Mojica et al., 2015). Also, the co-occurrence of the three compounds in the seed coats has an impact on the postharvest appearance, canning quality, and on the dietary influences of the black bean market class (Marles et al., 2010). Indeed, high tannin concentration in conjunction with low anthocyanin content leads to more brownish seeds (Marles et al., 2010). Diaz, Caldas, and Blair (2010) reported that anthocyanins and tannins are primarily inherited together by the seed color genes, whereas anthocyanins concentration can be adjusted by a limited number of genes. Therefore, selecting bean lines with a high anthocyanin: tannin ratio is complex, but would contribute to enhancement of the black color trait.  Black bean canning quality is also influenced by seed coat luster (Cichy et al., 2014). Mature black bean seeds are divided into either opaque (dull) or shiny (glossy). From an industrial point of view, whether raw or canned these two black bean seed types should be treated independently (Cichy et al., 2014)during the canning process faster than the shiny seed type and reduces the processing cost. This difference is due to the different amount of wax layer between the two seed coat luster types (Beninger et al., 2000). Consequently, less wax in the opaque seed type may form some cracks that allow quicker water intake (Bushey et al., 2001; Ma et al., 2004). However, the genetic control of seed coat luster is separate from the genetic control of color retention in canned black beans (Cichy et al., 2014).   11 Color retention of black beans is not only affected by the canning process, but by other factors such as its genetic make-up. There is a significant genetic variability for color retention of black beans following canning (Wright & Kelly, 2011). Cichy and others (2014) in a recent study indicated that color retention trait in black beans they found that quantitative trait loci (QTLs) for color retention and seed anthocyanin concentration following canning in black beans co-localized on chromosome Pv05. They were able to determine a small number of candidate genes for marker assisted selection (MAS) for the two traits on Pv05 and on Pv11 for the color retention trait. In fact, there is significant genetic variability in color retention among black bean genotypes following canning (Wright & Kelly., 2011). A good understanding of the genetic differences in canning quality traits of various black bean market classes will help improve the breeding process.  2.1.4 Canning Quality Evaluation  Black bean canning quality is usually assessed by a well-trained sensory panel (Merwe et al., 2006). Parameters including physical appearance after canning are important (Khanal et al., 2015). The widely used parameters in the evaluation of black beans are color, visual appearance, texture, hydration coefficient, and percent washeddrained solids (Merwe et al., 2006). Indeed, color and appearance are considered essential physiochemical traits of any food, since they mainly correlate with physical, chemical, and sensory indicators of the product quality. Another method of black bean color and appearance evaluation is a machine vision system, which has been considered for automatic assessment of color and appearance in canned black beans. This machine tests a number of color and appearance features such as average, standard deviation, contrast, correlation, energy and homogeneity measurements from red, green, blue (RGB), L*a*b*, and hue, saturation, and value color scales. These images are taken from drained/washed black beans and their brine solution (100 ppm calcium, 5.67 g of salt/454 g of water, 7.09 g of sugar/454 g of water). The 12 quality ratings for color and appearance images are evaluated with sensory scores using multi-statistical models (Mendoza et al., 2014). Texture readings are mostly obtained by specialized equipment (Kramer Shear press), and it is one of the canning quality parameters that measures either hardness or softness (palatability) of seeds after canning (Balasubramanian et al., 2000). In addition, the hydration coefficient is defined as the ratio of dry bean seeds weight prior and after soaking process (De Lange & Labuschagne, 2001). The coefficient is an important parameter in beans marketing, since bean seeds with a high up-take water will increase the canned yield. The percent washeddrained solids parameter means the mass of the rinsed and drained canned bean seeds (Balasubramanian et al., 2000). Low percentage indicates excessive loss of solids during canning process, and a higher tendency of bean seeds to forming clusters in the bottom of the can after processing and during storage (Khanal et al., 2015).              13 2.2. Color Retention and Anthocyanin in Black Bean 2.2.1 Impact of Polyphenolics (anthocyanin and tannin) on Black Bean Color Retention After Canning Black bean color is determined by the existence and concentration of different types of polyphenolic compounds such as flavonol glycosides, anthocyanins, and condensed tannins (proanthocyanidins). The most common groups of flavonoids in beans are proanthocyanidins and anthocyanins, which have been reported to be present only in black and blue-violet beans (Aparicio-Fernandez et al., 2005). Anthocyanins are plant-derived flavonoid compounds that control colors varying from bright pink to red, purple, and deep blue (Wang et al., 2013). They are found in almost all plant parts, mainly flowers and fruits, storage organs, roots, tubers, and stems. They are present in most species in the plant kingdom. Anthocyanins are water-soluble pigments stored in vacuoles. The color expressed by anthocyanins depend on the pH of the vacuole. This pigment is very sensitive to environmental stresses like strong light, ultraviolet (UV) radiation, and high temperature. There are about 500 types of anthocyanins although there are three basic structural categories, pelargonidin, cyanidin, and delphinidin with each being characterized by the number of hydroxyl groups on the B-ring (Gould et al., 2009). Anthocyanins are used broadly as natural food colorants in the food industry as substitutes for artificial colorants (Wu et al., 2006).  Black beans anthocyanins are concentrated in the seed coat (Dzomba et al., 2013). Most anthocyanins that have been extracted from different black bean accessions are delphinidin 3-glucoside, petunidin 3-glucoside, malvidin 3-glucoside, and malvidin 3,5-diglucoside (Zhang et al., 2014). Some anthocyanins have been extracted with lower quantities for instance cyanidin 3-glucoside, cyanidin 3,5-diglucoside, pelargonidin 3-glucoside, and pelargonidin 3,5-diglucoside (Takeoka et al., 1997). Among all these anthocyanins, delphinidin 3-glucoside was found to have 14 high antioxidant activity when it was extracted from the black bean cultivar Yamashirokurosando (Takeoka et al., 1997). Black beans have other phenolic commands such as tannins that also influence color. Tannins are polyphenolic compounds that differ in size, bond type, and concentration among plant species (Díaz et al., 2010). They show different impacts at the ecological and nutritional levels. Tannins in the seed coat play an important role in preventing fungal infection. Nutritionally however, the existence of multiple hydroxyl groups in condensed tannins results in the creation of complexes with proteins, metallic ions like iron, and polysaccharides like starch. The formation of complexes results in reduced bioavailability of the above nutrients (Díaz et al., 2010). Tannins are located in the inner layer of the black bean seed coat, and give canned beans their redbrown color (Marles et al., 2010). Therefore, color retention in black beans after canning is dependent not only on presence of anthocyanins, but also on the presence of tannins (Marles et al., 2010).               15 2.3. Linkage Disequilibrium Mapping  2.3.1. Association Mapping In plant genetics, a goal is to understand the natural phenotypic differences that are caused by alterations in DNA sequence. Linkage mapping has been the wide spread method to achieve this goal by making some empirical crosses to generate a family with known relatedness in order to identify co-segregation of genetic markers and phenotypes within the same family (Myles et al., 2009) In contrast, association mapping (linkage disequilibrium, LD) has been employed to conduct genotype-phenotype correlations in unrelated individuals (Myles et al., 2009). Association mapping is a powerful tool used to detect genes controlling the quantitative variation of the complex traits. Such variation is influenced by several QTLs, the environment effect, and the interfacing between QTLs and the environment (Zhu et al., 2008). Association mapping can be divided into two types, the first is candidate-gene association mapping, which identifies polymorphisms in the elite candidate genes, and becomes responsible of controlling phenotypic differences for the trait of interest. The second is genome-wide association mapping (GWAS), which is capable to scan the whole genome for any possible genetic variation with the goal of discovering any indication of an association for complicated traits (Zhu et al., 2008). The ultimate goal of association mapping is to reveal the complexity of the traits of interest and narrow it down to a gene or a nucleotide. To successfully apply association mapping, knowledge of all genetic characteristics related to germplasm is needed. Adequate numbers of unlinked and selectively impartial background markers are required to accomplish genome-wide coverage and extensively describe the genetic structure of the crop. For example, while 140,000 markers are considered sufficient coverage of the size of Arabidopsis genome 125Mb, two millions markers will be needed to cover the size of the grape genome 475Mb (Kim et al., 2007). 16 and SNPs are the most effective markers in term of assessing population structure (Q) and the relative kinship matrix (K). Also, SNP makers considered to be the best choice for complicated traits, because they have a higher genome density, lower mutation rate, and high-throughput detection platforms (Zhu et al., 2008).                         17 CHAPTER 3 PHENOTYPIC EVALUATION AND CANNING QUALITY IN BLACK BEAN INTRODUCTION Black bean, as most dry beans, is an ideal source of protein, fiber, and folate. Black bean production has expanded in the U.S. within recent years. Annually consumption of black beans per person in the U.S. expanded from 4.5g in 1970 to 304g by 2008 (Cichy et al., 2014). Such a dramatic society diet change toward black beans consumption may be explained on the basis of its health advantages (Kelly et al., 2015). Furthermore, polyphenols and anthocyanins specifically are found in the outer layer of black bean seeds, and it is these compounds which reduce the black color. Black beans are the dry bean type with the highest anthocyanin concentration.  The major anthocyanins in black beans include delphinidin 3-glucoside, petunidin 3-glucoside, and malvidin 3-glucoside (Jiang et al., 2014; Dzomba et al., 2013). Anthocyanins are important chemicals that maintain human health through their anti-oxidant activity, which helps reduce oxygen free radicals (Díaz et al., 2010; La Cruz et al., 2013).  Black bean canning quality may also be associated with anthocyanin levels. This study was conducted on 69 black bean genotypes that had been obtained from U.S. breeding programs. These lines were grown in 2013 and 2014 in replicated field trials at the Saginaw Valley Research Farm and Extension center in Richville, MI. Upon harvest the samples were canned and evaluated for canned bean appearance and color retention. The 69 lines were genotyped with 3975 SNP markers, and genome wide association analysis was conducted for canned bean appearance and color retention related traits.   18 MATERIALS AND METHODS 3.1. Plant Materials A group of 69 different black bean lines were obtained from Michigan State University (MSU), International Center for Tropical Agriculture (CIAT), Colorado State University (CSU), North Dakota State University (NDSU), Universidad de Puerto Rico (UPR), Nebraska (NE), University of Cincinnati (UC), Saskatoon, U.S. Department of Agriculture (USDA-MI, USDA-TARS, and USDA-WA) and includes dull and shiny black bean seed types.                    19 Figure 1. A neighbor joining tree of 69 black bean genotypes developed using 2800 SNP markers.                             A1-1 A1-2 A1 A2 C D B A2-1 A2-2 A2-3 A2-4 20 Table 1. A list of the 69 black bean genotypes that were used to evaluate color retention of canned black beans.  No. Genotype Shiny / Dull Origin Pedigree   1 B11360 Dull MSU  B04644/B05066 2 Condor Dull MSU Phantom / Black Jack 3 Black Rhino Dull CIAT Negro Tacana/G24424(wild bean)  4 B-EL1291d Dull USDA-MI Black Magic/Shiny Crow 5 B-EL01318s Shiny USDA-MI Black Magic/Shiny Crow 6 B-EL01270s Shiny USDA-MI Black Magic/Shiny Crow 7 B11313 Dull MSU B04644//B04349/B05044 8 B10238 Dull MSU Zorro/B05055 9 19365-31 Dull Unknown Unknown 10 B11356 Dull MSU Jaguar/B04644 11 B10213 Dull MSU B04587//Zorro/DPC-1 12 SHINYCROW Shiny CSU Unknown 13 ND071065 Dull NDSU Unknown 14 B-EL01270d Dull USDA-MI Black Magic/Shiny Crow 15 101s-2435 Dull Unknown Unknown 16 TARS-MST1 Dull USDA-TARS Ngro Takana/vax6 17 ND081195 Dull NDSU Unknown 18 B10210 Dull MSU N05324/B04431 19 B11312 Dull MSU B04587//B05070/B05044 20 SB-DT1 Dull USDA-TARS Unknown 21 Domino Dull MSU Nep2/BTS 22 Zorro Dull MSU B00103*/X00822 23 B-EL01260d Dull USDA-MI Black Magic/Shiny Crow 24 ND081343 Dull NDSU Unknown 25 B-EL01268s Shiny USDA-MI Black Magic/Shiny Crow 26 Black Magic Dull USDA-MI Nop-2/black turtle soup 27 BK 11-5 Dull USDA-WA Unknown 28 B11363 Dull MSU B04644/B07554 29 B11329 Dull MSU B04644/B04391 30 ND071089 Dull NDSU Unknown 31 ND071244 Dull NDSU Unknown 32 ND071257 Dull NDSU Unknown 33 DPC-4 Dull UPR Arroyo Lorro Negro/Raven  34 Harblack Opaque Dull USDA-MI Unknown 35 B09119 Dull MSU B04554/X06127 36 Eclipse Dull NDSU ND9902621-2 37 NE14-12-28 Dull NE Unknown 38 B11259 Dull MSU N07009//B04349/B05044 39 T-39 Dull UCDavis SEL-BTS,T-39 40 B-EL0309 Dull USDA-MI Black Magic/Shiny Crow 41 B-EL01294s Shiny USDA-MI Black Magic/Shiny Crow 42 ND081256 Dull NDSU Unknown 43 B11588 Shiny MSU I82054/B07554 44 B11310 Dull MSU B04587//ZORRO/DPC-1 21  45 BK 11-8 Dull USDA-WA Unknown 46 07055-4 Dull CSU DOR 500//NAG 210/UI 906(45735) X ARA 13/UI 906//BelMiDak RR2 47 Zenith (B10244) Dull MSU B04644/B07554 48 B11364 Dull MSU Unknown 49 B11370 Dull MSU B05055/B04265 50 B10243 Dull MSU B04610/N05346 51 BK 11-1 Dull USDA-WA Unknown 52 B09165 Dull MSU B04554/B04587 53 B-EL01335s Shiny USDA-MI Black Magic/Shiny Crow 54 NDFD9302 Dull NDSU Unknown 55 B1133 Dull MSU N08007//B04349/B05044 56 NDF09304 Dull NDSU Unknown 57 OAC Jet Dull Saskatoon OAC Jet 58 OACExpresso Dull Saskatoon Unknown 59 B-EL01291s Shiny USDA-MI Black Magic/Shiny Crow 60 ND081339 Dull NDSU Unknown 61 BK9-2 Dull USDA-WA Unknown 62 B11285 Dull MSU N04152/N05346//N04141/N05317 63 Aifiwuriti Dull UPR Unknown 64 B-EL01308s Shiny USDA-MI Black Magic/Shiny Crow 65 B11311 Dull MSU B04587//Zorro/DPC-1 66 ND081111 Dull NDSU Unknown 67 B11322 Dull MSU B05055/B04644 68 B11309 Dull MSU B04587//Zorro/B05055 69 ICB-3 Dull UPR Unknown             22  3.2. Field Description  The 69 genotypes were grown in 2013 and 2014 at the Saginaw Valley (SV), Valley Research Farm and Extension Center (SV) in Richville, MI (Table 1). Materials were plantedon June 6th in 2014 and 2013, and harvested September 17th in 2014, and September 15th in 2013. The growing season in SV, MI is June-September with an average temperature ranging from 9 to 28°C, the soil type is a Tappan- Londo loam with a 2.9% organic matter and an average pH of 7.9 and the altitude is 242 m. Materials were planted in 4-row plots 6.4 m in length with 0.5-m row spacing. Standard management practices were used including fertilizer and irrigation, and farm machinery was involved for these tasks. Seed was direct harvested with a Hege 140 plot. Moisture- adjusted yield per plot was measured on cleaned seed and used to calculate yield per hectare. A neighbor-joining tree, using the Chord distance matrix, was obtained using the TASSEL software (Bradbury et al., 2007).  3.3. DNA Extraction  DNA was extracted from two-wk-old leaves using the hexadecyltrimethyl ammonium bromide method (Rogers and Bendich, 1985). The DNA concentrations were evaluated by quantifying the optical density at A260 and A280 wavelengths with a Nanodrop Spectrophotometer (ND-8000; NanoDrop Products, Wilmington, DE) (Rogers & Bendich, 1985).  3.4. SNP Genotyping DNA extracts from all 69 black bean genotypes were shipped to the Soybean Genomics and Improvement USDA Laboratory in Beltsville, Maryland, where the DNA extracts were sequenced using an Illumina BARC-Bean 6k-3 beadChip (USDA laboratory, 2014). The whole set of 5398 SNP markers was filtered to remove monomorphic SNPs and those SNPs with a minor allele 23 frequency of 0.02% or less. After filtering, 2800 SNPs remained for GWAS analysis with a mixed linear model (MLM) (Song et al., 2015).   3.5. Canning Process A 100g of black bean seeds were cleaned weighed and placed in mesh bags. They soaked in a cold solution (25°C) of distilled water with calcium chloride for 30min. The seeds were weighted again and transferred to cans. Then, cans were filled with 98.9°C brine solution (1.5% sucrose, 1.2% sodium chloride, 0.03% calcium chloride, and distilled water). The cans were sealed and processed for 45 min at 115.6°C in a steam retort (National Board No. 813, Loveless Manufacturing, Tulsa, OK). Finally, the cans were stored for two weeks at room temperature (25°C) before evaluation. 3.6. Evaluation Process An evaluation process was conducted on the canned black bean seeds to assess the quality of the seeds, and was divided into two methods. The first was a sensory evaluation by a group of 14 trained and expert evaluators who visually rated the color and appearance of the beans by comparing them to quality charts for color (Figure 2A), and appearance (Figure 2B) of canned beans. The first category in color chart is very light brown. The second category is slight dark brown or light gray. The third is average brown black. The forth category is dark brown or medium black. The fifth category is very dark.         24 Figure 2A. Color rating chart for black beans.   In the appearance chart, the first group is unacceptable, with severely split seeds and burst seeds. The second is poor, with seeds badly split but intact. The third is average, showing 60-69 % of seeds intact. The fourth has 70-89 % of seeds moderately intact, and the fifth has an appearance with >90% of seeds intact.  Figure 2B. Appearance rating chart for black beans.    Color was also measured quantitatively to obtain the three color values Lab. In 2013, a colorimeter Hunter Lab scan XE (Hunter Associate Laboratory, Inc., Reston, VA 20190, USA), was used for (Mendoza et al., 2016)   (Mendoza et al., 2016)  25 extracting color parameters from washed and drained canned beans. The data taken from this machine can be set to any CIE units. However, in 2014 a machinery vision system that was made by (Mendoza et al., 2016), was used to obtain the same three color parameters, L*a*b*.  3.7. Statistical Analysis  Analysis of variance (ANOVA) for the seven traits was conducted with the model:  using SAS9.4 statistical software package. The GLM procedure for this paper was generated using SAS software, version 9.4 of the SAS System Copyright © 2016 SAS Institute Inc. SAS and all other SAS Institute Inc.    (Pritchard et al., 2000). Population structure was considered in the model with Principal Component Analysis (PCA) using a correlation matrix in TASSEL 5.2.25 program (Bradbury et al., 2007) and various PCs were accounted. A kinship matrix (K) was also included in the association analysis to account for relatedness. GWAS was conducted using the Mixed Linear Model approach (Pritchard et al., 2000).  The quantile- quantile (QQ) plot was generated from the observed and expected (logarithm (base 10) of odds) LOD scores for each trait. Manhattan plot and QQ plot graphics were developed in TASSEL program (Turner, 2014). The phenotypic data for all the seven traits were averaged across the 2013 and 2014 field seasons prior to GWAS. The traits of seed weight, color, L*, a*, and b* for both years and appearance for 2014 were not normally disrupted and were transformed with the Box-Cox method (Box and Cox, 1964). However, the mentioned traits showed better QQ plots and Manhattan plots with their original data which have been used for GWAS study. The following MLM equation was used: Y (Principle Component Analysis) 26 matrix and both X and p represent fixed effects, K is the relative kinship matrix value, and e is for residual effects.                      27 RESULTS AND DISSCUSSION A set of 69 canned black bean accessions were evaluated for traits related to seed quality included yield, seed weight, appearance, color, L*, a*, and b* over two field seasons. The proc mixed procedure was used for analysis of variance with the model including genotype (69 levels) year (2 levels) and rep (2 levels) and the interaction genotype × year. Analysis of variance (ANOVA), (Table 2A-G) indicated that genotypes and year based on () as level of significance were significantly different. The interactions between genotype and year for all the traits were also significant based on the same level of significance, and that requires a separation for the data analyses by year (Table 2A-G).  28 Table 2A. Analysis of variance for yield trait collected on 69 black bean genotypes grown for 2 years (2013 and 2014).                    Trait  Source of variation Degree of freedom Type III SS    Mean square     F statistic    Pr>F   Yield    Genotype                                  68                          58428662.46              886692.05                        3.01                               <0.0001                                                                                        Year                                             1                          34693849.70              34693849.70                   121.65                           <0.0001                                                                                     Genotype*Year                        68                          29458632.42              433215.18                        1.52                               <0.0202                                                                                      Table 2B. Analysis of variance for seed weight trait collected on 69 black bean genotypes grown for 2 years (2013 and 2014).                    Trait  Source of variation Degree of freedom Type III SS    Mean square     F statistic    Pr>F   Seed Weight   Genotype                                  68                             773.80                   11.38                                 5.38                               <0.0001                                                                                        Year                                            1                              250.61                      250.61                              118.47                            <0.0001                                                                                      Genotype*Year                       68                             270.02                       3.97                                  1.88                                <0.0010                                                                                        29 Table 2C. Analysis of variance for appearance trait collected on 69 black bean genotypes grown for 2 years (2013 and 2014).                    Trait  Source of variation Degree of freedom Type III SS    Mean square     F statistic    Pr>F   Appearance         Genotype                                  68                              69.67                       1.03                                3.77                             <0.0001                                                                                        Year                                             1                               78.70                           78.70                             289.91                         <0.0001                                                                                      Genotype*Year                       68                               26.32                           0.39                               1.43                                0.0417                                                                                    Table 2D. Analysis of variance for color trait collected on 69 black bean genotypes grown for 2 years (2013 and 2014).                    Trait  Source of variation Degree of freedom Type III SS    Mean square     F statistic    Pr>F   Color                     Genotype                                  68                              195.01                        2.86                              19.94                          <0.0001                                                                                       Year                                             1                               63.25                           63.39                           439.87                         <0.0001                                                                                      Genotype*Year                       68                               17.27                           0.25                               1.77                               0.0027                                                                                     30 Table 2E. Analysis of variance for L* value trait collected on 69 black bean genotypes grown for 2 years (2013 and 2014).                    Trait  Source of variation Degree of freedom Type III SS    Mean square     F statistic    Pr>F   L*                  Genotype                                  68                              357.07                       5.25                                3.71                              <0.0001                                                                                        Year                                             1                               433.45                       433.45                             306.28                          <0.0001                                                                                       Genotype*Year                        68                             1689.50                       24.85                              17.56                            <0.0001                                                                                    Table 2F. Analysis of variance for a* value trait collected on 69 black bean genotypes grown for 2 years (2013 and 2014).                    Trait  Source of variation Degree of freedom Type III SS    Mean square     F statistic    Pr>F   a*                  Genotype                                  68                              295.20                    4.34                                 23.28                             <0.0001                                                                                        Year                                             1                              1791.90                   1791.90                           2610.77                          <0.0001                                                                                      Genotype*Year                       68                               40.16                        0.59                                 3.17                                <0.0001                                                                                       31 Table 2G. Analysis of variance for b* value trait collected on 69 black bean genotypes grown for 2 years (2013 and 2014).                    Trait  Source of variation Degree of freedom Type III SS    Mean square     F statistic    Pr>F   b*                  Genotype                                  68                              463.36                    6.81                                 38.70                              <0.0001                                                                                      Year                                             1                              381.24                     381.24                              2165.55                          <0.0001                                                                                      Genotype*Year                        68                              67.10                       0.99                                  5.61                                 <0.0001                                                                                        32 The distribution of the data for all the traits including yield, seed weight, appearance, color, L*, a*, and b* of canned black beans for the 2013, and 2014 harvest seasons are shown in Figure 3. Yield was evaluated as kg per hectare. Yield for both years (2013 and 2014) were normally distributed (Figure 3). In 2013, yield ranged from 828 to 3273 kg ha-1, and the average was 2150 kg ha-1, while in 2014 yield ranged from 1649 to 4648 kg ha-1, and the average was 2867 kg ha-1 which is much larger than 2013 (Table 3). Seed weight was measured using the weight of 100 seeds. The trait of seed weight was not normally distributed for both years (2013 and 2014) (Figure 3). In 2013, seed weight ranged from 14.9 to 22.6g, and the mean seed weights was 19.1 g. In 2014 the seed weights ranged from 17.6 to 27.6 g, and the mean weight was 21.3g (Table 3). Therefore, the seeds used for canning in 2014 were larger than those in 2013.  The trait of appearance was normally distributed in 2013 with rating values ranged from 1.36 to 3.54, and the ratings average was 2.60 (Table 3). In contrast, in 2014, the trait was not normally distributed and was skewed to higher appearance ratings. The ratings ranged from 2.00 to 4.16, and the ratings mean was 3.27, which is higher than 2013. A commercial standard line was drawn on the rating 3.00 to allow us differentiate between acceptable and non-acceptable genotypes canning quality. Most of the genotypes in 2013 were below the commercial standard with low chances of being in the market. However, most of the genotypes in 2014 were acceptable. The lowest -EL01308s, B11313, B-EL01294s) were from the two bean breeding programs MSU and USDA-MI.   Color trait for 2013 as it was determined using the rating system, and rated by a group of trained bean experts showed a binomial distribution, where one group of genotypes appear to have higher color retention than the other (Figure 3). Color rating values ranged from 1.11 to 4.50, and the 33 average of the color ratings was 2.68 in 2013 as shown in Table 4. In comparison with 2014, the distribution of color measured and was more uniform, but was skewed toward better color ratings. This is due to better environmental factors such as weather effects for canned black beans that occurred in 2014. Color rating value in 2014 ranged from 1.72 to 4.72, and the average of color ratings was 3.26 as shown in Table 3. There is a year advantage of 2014 over 2013 in the trait of color as well as the rest of the previous traits. Seed quality traits are linked to environment and there were superior environmental conditions for the canned black bean in 2014. Also, this express the wide genetic variability among black genotypes specially in the trait of color retention and that may do to fewer bean breeding programs that work on color retention, which are MSU and USDA-MI.  Most of the genotypes in 2014 fell below the commercial standard line, where in 2014 a high percentage of the genotype were felled above the commercial standard. The five highest color-rated genotypes in 2013 are Zenith, B-EL01294s, B11364, B11313, and B-EL01308s, while highest color-rated genotypes in 2014 were B-EL01335s, B-EL0309, B11313, B-EL0126, and Zenith respectively (Table 6A-B). All highest color-rated genotypes came from the two breeding programs of MSU and USDA-MI. Also, this elite group of genotypes belong to the two sup-groups A1-2 and A2-2 in the neighbor joining tree (Figure 1). This result reflects the genetic variability of the black bean genotypes that were obtained from different breeders around the country who may have different genetic background for each black bean line. Also, these two sup-groups were known during this study to have the elite color-retaining genotypes.  A colorimeter was also used to detect color parameters from washed and drained canned beans. Three color measurements: L* a* b* were taken on all the canned black bean genotypes. The value L* is on a scale of 0 to 100 and the closer to 0, the darker, while the closer to 100, the lighter. The 34 distributions of L* value in both year (2013 and 2014) were skewed toward higher values. In 2013, L* values ranged from 14.16 to 22.53, and the mean of the values was 17.66, where in 2014 the value ranged from 16.23 to 28.30, and the values mean was 19.29. This data, explains colorimeter detection for darker colors in black bean seeds in the year of 2014. The value of a* is on a scale of -127 to +127 and the minus values are greener while positive values are redder. The a* value for both years were skewed with slightly normalized distribution for the year 2013. The values were ranging from 7.36 to 10.98, and the values average was 8.86 in 2013, where in 2014, the values ranged from 1.45 to 7.79, and the values average was 3.75. The value of b* is on a scale of -127 to +127 and the minus values are bluer while positive values are yellower. The b* values for both years were not normally distributed, and 2014 values were right-skewed toward lower b* values. The b* values of 2013 ranged from 2.95 to 10.29 with an average of 6.37, where in 2014 the values were ranging from 2.47 to 7.30 with an average of 3.94. These results indicate that canned black bean seeds in 2014 were greener and bluer (less red and less yellow on the a* and b* scales respectively. Histograms data for each trait are shown in Figure 3. and Table 3.            35 Table 3. Means and ranges for yield, seed weight, appearance, color, L*, a*, and b* traits for the years 2013 and 2014. Trait Year Mean Std. Deviation Minimum Maximum Yield (kg ha-1) 2013 2150 475 828 3273 Yield (kg ha-1) 2014 2867 607 1649 4648 Seed weight (g) 2013 19.1 1.5 14.9 22.6 Seed weight (g) 2014 21.3 2.3 17.6 27.6 Appearance (Rate 1-5) 2013 2.6 0.5 1.4 3.5 Appearance (Rate 1-5) 2014 3.3 0.6 2.0 4.16 Color (Rate 1-5) 2013 2.7 0.9 1.1 4.5 Color (Rate 1-5) 2014 3.3 0.7 1.7 4.7 L* 2013 17.7 2.1 14.2 22.5 L* 2014 19.3 2.7 16.2 28.3 a* 2013 8.9 0.8 7.4 11.0 a* 2014 3.7 1.3 1.5 7.8 b* 2013 6.4 1.7 3.0 10.3 b* 2014 3.9 1.0 2.5 7.3         36 00.050.10.150.20.250.3010002000300040005000Relative frequencyYield (kg ha-1)Yield 201300.050.10.150.20.2510002000300040005000Relative frequencyYield (kg ha-1)Yield 201400.050.10.150.20.250.314.019.024.029.0Relative frequency100 Seed weight (g)Seed Weight 201300.050.10.150.20.250.30.351015202530Relative frequency100 Seed weight (g)Seed Weight 201400.050.10.150.20.2511.522.533.544.5Relative frequencyAppearance Rate (1-5)Appearance 201300.020.040.060.080.10.120.140.160.1822.533.544.5Relative frequencyAppearance Rate (1-5)Appearance 2014Figure 3. Histograms for the data sets grown for two years for all the traits (yield, seed weight, appearance, color, L*, a*, and b*) of canned black beans for 2013, and 2014.                       37 00.050.10.150.20.2512345Relative frequencyColor Rate (1-5)Color 2013 00.050.10.150.20.2512345Relative frequencyColor Rate (1-5)Color 201400.020.040.060.080.10.120.140.160.180.214192429Relative frequencyL ValuesL *201300.050.10.150.20.250.30.350.41015202530Relative frequencyL ValuesL* 201400.050.10.150.20.250.3789101112Relative frequencya values a* 201300.050.10.150.20.250.3024681012Relative frequencya valuesa* 2014Figure 3                         38 00.050.10.150.20.25246810Relative frequencyb values b* 201300.050.10.150.20.250.3246810Relative frequencyb valuesb* 2014Figure 3         interest (Figure 4). There were several strong correlations detected among the measured traits (Table 4 and Table 5). The trained sensory panel ratings for color 2013 were strongly correlated with L*, a*, and b* values respectively (r=-0.95, p=0.0001, r=-0.79, p=0.0001, and r=0.95, p=0.0001), higher ratings were given for darker lines on the L* scale, and those which tend to be more green and blue on the a* and b* scales respectively. These correlations support findings by Cichy et al (2014) indicating that the trained sensory panel is as efficient and comparable to the Hunter L*a*b* color scans. Color ratings were also strongly correlated with L*, a*, b* values respectively in 2013: (r=-0.95, p=0.0001, r=-0.79, p=0.0001, and r=-0.95, p=0.0001) and 2014: (r=0.81, p=0.0001, r=-0.97, p=0.0001, and r=-0.95, p=0.0001). The L* value in 2014 was positively correlated to color ratings.  It is important to note that two different systems for data measurements were used in 2013 and 2014. A colorimeter (HunterLab) was used to obtain L*a*b* values in 2013, while in 2014 a calibrated digital camera was used. A colorimeter is a color measurement machine that is designed for flat surfaces rather than bean seeds which tend to be uneven and have high light reflection. This reflection may interfere with the sample color and results in biased data. Moreover, this machine has a smaller size sensor where it requires several 39 pictures in order to obtained well representative date for the measured sample. In contrast, machine vision system (calibrated digital camera) has a highly sensitive lens with wider size which is capable of obtaining an image of an entire canned bean sample.  Moderate-weak correlations have been observed between appearance ratings and color traits including color ratings and L*a*b* color parameters in both years (2013 and 2014) (color: r=0.44, p=0.0002, L*: r=-0.44, p=0.0002, a*: r=-0.21, p=0.086, b*: r=-0.43, 0.002) (color: r=0.59, p=0.0001, L*: r=0.47, p=0.0001, a*: r=-0.21, p=0.0001, b*: r=-0.43, p=0.0001). This kind of minor correlation between appearance ratings and color traits including color ratings and L*a*b* color parameters was found because of the bright color of the seed starch that was interfered with lab color parameters detection (Table 4 and Table 5). Canned black beans with poorer appearance tend to split open, therefore the white cotyledons of the seeds will make the overall beans color look whiter than the seed with intact appearance. Also, no correlations were detected between color, L* a* b* values, appearance and the two seed traits for weight and yield. Seed weight was not correlated with yield.  40 Table 4. Pearson correlation for the seven traits (yield, seed weight, appearance, color, L*, a*, and b*) for year 2013.                              Yield            Seed Weight       Appearance          Color                L*              a*                b*   Yield                      .                -0.1966                  0.2940                   0.2175                 -0.1781               0.0132                 -0.1899                                .                 0.1055                  0.0142                   0.0726                   0.1432               0.9144                  0.1181 Seed Weight                               .                        -0.1520                 -0.1455                   0.1082               0.0490                  0.1128                                                       .                         0.2126                   0.2330                   0.3763               0.6895                  0.3562                                                                                                                                                                                                                                                                                            Appearance                                        .                          0.43991                 -0.4359             -0.2080                 -0.4340                                                                                                                                                                                           .                          0.0002                  0.0002               0.0864                  0.0002                                                                            Color                                     .                         -0.9465              -0.7847                -0.9532                                                                                                      .                          <.0001                <.0001                 <.0001                                                              *L                                                           .                     0.8060                 0.9295                                                                                                                       .                      <.0001                 <.0001                                                           *a                                      .                         0.8239                                                                                                                                 .                         <.0001                                                                        *b                                                .                                                                                                                                                           .   1Values with bold have significant p-values       41 Table 5. Pearson correlation for the seven traits (yield, seed weight, appearance, color, L*, a*, and b*) for year 2014.                              Yield            Seed Weight          Appearance  Color                   L*              a*              b*   Yield                      .                   0.0945                    -0.1804                   -0.0815             -0.0582            0.0973                 0.0878                                .                   0.4398                     0.1379                    0.5056               0.6349           0.4263                  0.4732 Seed Weight                               .                             -0.1520                  -0.1519              -0.0977           0.1455                  0.1456                                                       .                              0.2126                    0.2129               0.4244           0.2330                  0.2325                                                                                                                                                                                                                                                                                            Appearance                               .                           0.58601               0.4667          -0.2079                -0.4340                                                                                                                                                                                  .                           <.0001               <.0001           <.0001                  <.0001                                                                            Color                                             .                     0.8122          -0.9647                -0.9455                                                                                                              .                     <.0001           <.0001                 <.0001                                                              L*                                                 .                   -0.7315                -0.6595                                                                                                                       .                    <.0001                 <.0001                                                           a*                                      .                       0.9919                                                                                                                               .                       <.0001                                                                        b*                                              .                                                                                                                                                         .   1Values with bold have significant p-values 42 0510152025300246Lab ValuesColor RatingsCorrelation Analysis 2014Lab05101520250246Lab ValuesColor RatingsCorrelation Analysis 2013 LabFigure 4. Pearson correlations for the traits of color, L*, a*, and b* for the 2 years 2013 and 2014.          In a previous study of Cichy et al., (2014) show that there is no correlation between the trait of color retention and seed luster. However, we decided to include eight shiny seed-coated genotypes since we were not only interested in their seed luster effect, but also in their excellent canning quality. Mean separations were obtained for the seven traits for both years (Table 6A-B). The eight black bean genotypes were involved in means separation analysis to conduct their possible trend through the studied phenotypic traits. Mean separation result for the two years of 2013 and 2014 showed a superiority of the eight genotypes in term of color traits including color retention, L*, a*, and b* and also appearance. However, we have decided to exclude the eight shiny seed-coated genotypes from further analysis because of the role of their gloss in confounding the color measurements comparing to the human evaluators that were effective in rating color.  Three major commercial black bean varieties including Eclipse, Zorro, and Zenith were involved in this study. Eclipse is a major black bean variety that was released in 2004 by NDSU, Zorro which is another major black bean variety that was released in 2008 by MSU, and Zenith the most 43 recent major black bean variety that was released in 2015 by MSU (Kelly et al., 2015). These three verities have different canning properties in term of color, and that is due to genetic and environmental variability. For example, Zenith has a darker color than both of Eclipse and Zorro, which raised the black bean standard color since darker color of black beans is more favorable by consumers. The three genotypes behaved differently for all the seven traits across the two years. The genotype Zenith has an excellent canning quality, and has a high yield potential, where the genotype Eclipse has very poor color retention following canning. The genotype Zorro has a moderate color retention among all the three. Also, the three genotypes have been involved in recent studies that have been conducted by Kelly et al., (2015) and Goffnett et al., (2016).  In 2013, Zenith possessed the highest color rate mean (4.75) and the second highest appearance rate mean (3.25) among all the studied genotypes. Zorro showed moderate rating means for color and appearance (3.00, 2.75) respectively. The variety Eclipse was the lowest among the three with color rating mean (2.50) and appearance rate mean (3.00). In 2014, the variety Zenith was among the highest ten genotype for the color and appearance ratings means (4.00, 4.00). Zorro had a medium color and appearance rate mean (3.50, 3.50), and Eclipse was also the lowest among all the three with color and appearance ratings means (2.75, 2.75). This process supports the results represented by Kelly et al., (2015) and Goffnett et al., (2016) regarding the color differences between the three black bean genotypes. This advantage for the genotype Zenith is also extends to the trait of yield where Zenith had higher yield means over Zorro and Eclipse in both years (2013 and 2014) (2840, 3047 kg ha-1) respectively. The seed quality variation of these three genotypes emphasize a wide genetic variability among the studied population for the seven target traits. One of the reasons why there is a wide genetic variability among black beans is that there are only two programs (USDA-MI and MSU) that work in the improvement of black bean canning quality. 44 Table 6A. Means of the seven traits (yield, seed weight, appearance, color, L*, a*, and b*) of 66 black bean genotypes that obtained from the year 2013. Genotype Yield (kg ha-1) 100 Seed Weight (g) App  (1-5) Color  (1-5) L*  (0-100) a*  (+127-127) b* (+127-127)    Zenith 2840 18.5 3.25 4.75 14.77 7.50 2.95 B-EL01294s 1711 18.6 3.75 4.25 13.98 8.02 3.89 B11364 2164 22.1 3.00 4.25 14.71 7.67 3.56 B11313 2541 17.3 2.75 4.00 14.65 7.59 3.14 B-EL01308s 2457 21.2 3.5 4.00 14.10 7.31 3.74 B-EL01335s 2272 19.1 2.75 4.00 15.42 7.40 4.05 B11309 2178 18.7 1.50 3.75 14.39 8.57 4.55 B-EL0309 1952 18.2 2.00 3.75 14.34 7.84 3.65 B11356 2.085 13.6 2.50 3.50 14.69 8.25 5.51 Condor 2387 17.8 3.25 3.50 15.16 8.32 4.48 B09165 2385 17.3 3.00 3.50 16.55 9.11 5.05 B11360 2335 18.3 2.25 3.50 17.55 8.21 5.63 B-EL01270s 2072 21.3 2.75 3.75 14.69 8.20 4.80 Harblack 2221 18.2 2.00 3.50 16.33 9.05 5.11 B-EL01291s 1681 18.2 2.75 3.50 15.79 7.45 5.21 NE14-12- 2006 20.6 2.75 3.50 14.31 7.43 4.82 DPC-4 1566 18.0 3.00 3.50 15.08 7.87 4.02 SHINYCROW 1858 20.2 3.5 3.50 14.45 7.52 4.23 B-EL01318s 1061 17.4 2.25 3.50 15.97 8.37 5.82 B-EL0126 970 18.4 2.75 3.50 14.16 7.36 3.79 45 Table 6A  B11363 2735 17.0 3.00 3.25 14.70 8.44 4.08 B10238 2593 18.4 3.50 3.25 16.72 8.65 5.18 B11312 2441 19.7 3.50 3.25 15.89 8.56 4.76 B11310 2353 20.7 3.00 3.25 15.89 9.17 5.16 T-39 2194 17.6 2.50 3.25 16.90 8.14 6.01 B09119 2123 18.0 2.25 3.25 17.19 9.52 5.21 B11311 1819 20.3 3.50 3.25 17.35 9.13 4.90 BlackMag 2619 18.7 2.50 3.00 16.56 8.61 6.73 Zorro 2334 18.0 2.75 3.00 16.94 9.14 5.20 OACExpresso 1652 19.1 1.50 3.00 17.81 8.97 7.13 OACJet 2133 19.2 2.25 2.75 16.80 8.21 6.55 B-EL01268s 2135 17.8 2.75 2.75 16.66 8.08 6.30 B11322 2012 13.6 1.75 2.75 15.91 7.99 5.03 ND071065 2005 20.0 1.50 2.50 18.35 8.83 7.30 Domino 1991 20.8 2.00 2.50 16.65 8.96 5.26 Eclipse 1757 17.8 3.00 2.50 18.43 8.94 7.38 TARS-MST 1724 18.6 1.50 2.50 18.36 8.23 5.79 B-EL1291 2699 19.0 2.00 2.25 17.28 8.81 6.66 19365-31 2275 21.4 1.50 2.25 18.05 8.32 6.70 BK11-8 1797 18.8 3.00 2.25 18.27 9.04 7.41 B10213 2313 17.6 2.25 2.00 19.29 9.33 7.14 AifiWuri 2147 19.7 2.50 2.00 16.98 8.41 6.19  46 Table 6A  BK11-5 2108 21.1 1.50 2.00 17.83 8.59 6.98 BK11-1 2079 18.9 2.00 2.00 17.41 8.45 6.71 B11370 2149 18.6 2.75 1.75 18.57 8.59 6.50 ND071244 2134 19.9 1.50 1.75 19.13 9.70 7.75 NDF09302 2123 18.2 2.75 1.75 19.43 8.81 7.66 NDF09304 1833 17.6 2.50 1.75 18.36 8.36 7.01 BlackRhi 1669 22.4 2.00 1.75 17.21 8.59 6.34 ND081256 1541 24.1 2.50 1.75 20.00 10.09 7.83 ND081111 1476 18.0 2.00 1.75 19.46 9.41 8.08 ICB-3 1363 22.4 1.75 1.75 17.99 9.35 7.56 B10243 1357 20.5 3.00 1.75 18.26 8.63 6.48 B10210 3136 19.8 2.75 1.50 19.04 10.51 8.74 BK9-2 2452 19.7 2.25 1.50 20.02 9.76 8.59 B11588 2431 21.6 1.75 1.50 20.20 10.25 8.31 ND071089 2282 19.1 1.75 1.50 19.08 8.98 7.73 B11259 2046 16.4 1.75 1.50 20.75 9.07 7.68 ND081295 1290 19.8 2.50 1.50 20.09 9.33 7.63 07055-4 1166 20.2 1.75 1.50 20.36 9.93 8.64 ND071257 2583 20.6 1.50 1.25 19.93 10.46 8.88 B11285 2212 16.7 2.75 1.00 22.41 10.97 9.41 SB-DT1 2026 18.7 1.50 1.00 19.87 9.72 8.85 ND081343 1684 18.6 2.00 1.00 22.53 10.65 10.29 B11338 1383 16.1 2.50 1.00 20.19 10.32 8.83 47 Table 6A  Mean 2053 19.0 2.44 2.66 17. 8. 6. LSD0.05 1148   2.9 1.23 0.84 1.70 1.03 1.07 Coefficient of Variation 0.21  0.1 0.25 0.37 0.12 0.1 0.2 *Means for all traits have been ordered based on the means of color trait. *Three genotypes were missing across the traits and the two years (101s-2435, B-EL01270d, and ND081339)                       48 Table 6B. Mean of the seven traits (color, appearance, L*, a*, b*, seed weight, and yield) of 66 black bean genotypes that obtained from the year 2014. Genotype Yield (kg ha-1) 100 Seed Weight (g) App  (1-5) Color  (1-5) L*  (0-100) a*  (+127-127) b* (+127-127)    B-EL01335s 2868 22.5 3.5 5.00 24.96 2.07 2.79 B-EL0309 2093 21.5 4.5 5.00 26.74 1.93 2.85 B11313 3503 19.3 3.75 5.00 26.04 1.97 2.58 B-EL0126 2365 20.8 4.00 5.00 28.31 1.45 2.56 Zenith 3047 22.3 4.00 4.75 24.87 1.94 2.59 B-EL01294s 2485 19.1 3.75 4.75 27.31 1.85 2.79 B-EL01270s 2812 26.0 4.25 4.75 26.57 1.54 2.36 B11356 3139 19.2 3.25 4.50 20.97 2.70 3.08 B11363 3861 21.8 4.25 4.50 20.03 2.05 2.53 B11364 3568 21.8 3.50 4.50 21.48 2.63 2.95 B-EL01308s 2460 20.9 4.50 4.50 24.92 1.97 2.76 B11309 3221 18.3 3.25 4.50 21.44 2.30 2.84 B-EL01291s 2649 19.4 3.75 4.50 26.94 1.98 2.82 B11329 3541 18.2 3.75 4.50 21.69 1.86 2.47 B-EL01268s 1345 19.0 2.75 4.25 25.37 1.91 2.71 B-EL01318s 2617 19.1 3.75 4.25 24.77 2.26 2.96 Condor 2624 21.3 3.75 4.25 25.03 2.45 3.08 B11360 3727 20.5 2.50 4.25 20.59 2.60 3.03 NE14-12- 3153 24.2 3.25 4.25 23.59 2.27 3.02 49 Table 6B  SHINYCROW 2499 22.2 4.00 4.25 22.72 2.44 3.02 BK11-5 2713 22.1 3.50 4.00 18.57 3.61 3.69 B11322 2560 18.1 2.75 4.00 19.49 3.15 3.54 B-EL1291 2592 19.3 3.75 4.00 18.96 2.77 3.19 DPC-4 1209 20.5 3.75 4.00 19.80 3.11 3.44 BlackMagic 3425 19.4 3.50 4.00 22.81 2.56 3.28 OACExpresso 2658 20.0 3.50 3.75 18.84 3.90 3.99 B10243 2567 21.5 3.25 3.75 18.47 3.41 3.54 TARS-MST 2719 21.8 2.00 3.75 19.26 3.69 3.95 B-EL0127 3555 21.5 3.50 3.75 20.59 3.15 3.53 harblack 2910 18.9 3.50 3.50 18.40 3.53 3.58 T-39 3062 20.0 3.50 3.50 19.95 2.96 3.27 B09165 2950 22.6 4.25 3.50 17.83 3.47 3.63 B11370 2625 18.0 3.25 3.50 16.33 3.55 3.83 BlackRhi 3856 24.1 3.00 3.50 17.70 3.58 3.71 B09119 2792 20.5 4.25 3.50 18.43 3.46 3.76 NDF09304 2601 19.8 3.75 3.50 18.95 3.58 3.72 B10238 2409 18.8 4.25 3.50 17.64 3.75 3.81 B11311 2476 18.1 3.75 3.50 18.35 3.13 3.48 OACJet 3403 20.7 4.00 3.50 18.42 3.71 4.04 B11310 2799 22.2 4.00 3.50 19.99 3.05 3.38 Zorro 2324 21.1 3.50 3.25 18.00 3.73 3.86 B10213 2430 19.2 4.00 3.25 17.42 3.97 4.09 50 Table 6B  BK11-1 2219 21.0 3.75 3.25 19.29 3.57 3.75 19365-31 1416 23.8 3.50 3.25 17.57 3.86 3.95 Eclipse 2773 19.7 2.75 3.25 16.43 4.30 4.26 B11588 2891 24.8 3.00 3.00 18.09 4.59 4.58 NDF09302 2619 18.5 3.25 3.00 18.51 4.59 4.66 Domino 1844 20.2 3.75 3.00 17.13 4.05 4.20 BK11-8 1828 22.0 4.00 3.00 17.77 3.97 3.97 ICB-3 1347 21.3 2.50 3.00 16.87 4.56 4.53 AifiWuri 1750 19.6 2.75 3.00 17.96 3.80 4.08 ND081295 2583 25.5 2.25 3.00 17.65 4.47 4.50 ND071257 3534 21.8 3.50 3.00 16.72 4.68 4.44 ND071065 1884 21.3 3.00 3.00 16.23 4.67 4.48 ND081256 2380 19.8 4.25 3.00 17.04 4.15 4.10 B11259 3345 19.1 3.75 2.75 17.25 5.17 5.07 ND071244 3092 21.3 3.25 2.75 17.18 4.55 4.49 ND081111 3009 22.3 3.00 2.75 18.72 3.80 4.12 B10210 4716 27.4 4.50 2.75 16.72 4.91 4.74 SB-DT1 2814 26.3 2.50 2.75 17.20 5.59 5.41 ND081339 3117 26.4 3.25 2.75 18.53 5.43 5.44 07055-4 1499 23.3 3.25 2.75 18.34 5.15 5.12 B11338 3329 17.3 2.75 2.25 18.27 7.12 6.62 BK9-2 2777 20.0 3.00 2.25 16.39 5.56 5.22 B11285 3430 19.2 3.75 2.00 18.16 6.74 6.37 51 Table 6B  ND081343 2789 20.4 3.25 1.75 19.65 7.79 7.30 Mean 2745 21.0 3.50 3.61 20.03 3.54 3.81 LSD0.05 983 2.9 0.82 0.67 2.91 0.63 0.52 Coefficient of Variation 0.24 0.1 0.16 0.22 .16 0.37 0.26  *Means for all traits have been ordered based on the means of color trait. *Three genotypes were missing across the traits and the two years (101s-2435, B-EL01270d, and ND081339)  A neighbor joining tree developed with the data revealed that this study population consists of four major branches A, B, C, and D. The branch A was split into six sub-branches A1-A, A1-2, A2-1, A2-2, A2-3, and A2-4 (Figure 1). All major branches have several genotypes except for the branch B. Analysis of variance was conducted for the neighbor joining tree groups for the two years (2013 and 2014). The means of all the genotypes for the seven traits were averaged for each major and sub branches, and the standard deviation were also obtained.  The sub-branch A2-2 was found to have higher means for the traits of appearance, a*, and b* for both years (2.73, 3.78, 7.85, 2.11, 4.73, and 2.86) respectively (Table 7A-B). The sub-branch A2-2 has all the shiny seed-coated genotypes tested. This result suggests the superiority of shiny seed-coated genotypes over dull seed-coated genotypes, and that is due not only to their seed luster, but to their excellent canning quality that have been chosen based on. All the other traits varied between the two years. For the trait of yield, the sub-branch A1-2 had the higher yielding mean in the year of 2013, while the mean of the major branch C was the highest in 2014 respectively (2323, 3121 kg ha-1). The sub-branch A2-1 had the highest mean of seed weight in 2013 (20.14) with a standard deviation of 1.27, while the major branch D had the highest seed weight mean in 2014 (23.40) with a standard deviation of 2.15. The two traits of color retention and L* have been dominated by the sub-group A2-2 with the highest mean values (3.61, 15.17) and standard 52 deviation (0.58 and 1.08) respectively. The same two traits were varied in the year of 2014. The sub-group A1-2 had the highest color retention mean (3.68) with a standard deviation of 0.53. This group also had the second highest color retention mean in the year of 2013 (3.06) with standard deviation of 0.96. In 2014, the highest mean of the trait of L* was or the major branch C (17.68) with standard deviation of 0.54 (Table 7 A-B).  53 Table 7A. Means and standard deviations for all the major and sub branches that obtained from the neighbor joining tree for the seven traits (color, appearance, L*, a*, b*, seed weight, and yield) of 69 black bean genotypes that obtained from the year 2013.         NJ Group Average/St. Deviation Yield (kg ha-1) Seed Weight (g) Appearance (1-5) Color (1-5) L* (0-100) a* (-127+127) b* (-127+127) A1-1 Avg 2081 18.4 2.45 3.00 16.87 8.39 5.56  Std.dev 267 0.4 0.37 0.69 1.17 0.40 0.94 A1-2 Avg 2323 18.5 2.64 3.06 16.50 8.77 5.38  Std.dev 440 2.4 0.64 0.96 2.05 0.89 1.80 A2-1 Avg 1969 20.1 1.86 2.11 17.96 8.86 7.03  Std.dev 294 1.3 0.51 0.62 1.32 0.73 1.05 A2-2 Avg 1894 19.2 2.73 3.61 15.17 7.85 4.73  Std.dev 480 1.3 0.44 0.58 1.08 0.47 1.04 A2-3 Avg 2050 19.0 2.46 2.38 18.11 8.96 6.93  Std.dev 441 1.1 0.30 0.67 1.51 0.51 1.09 A2-4 Avg 2079 18.3 2.33 1.67 18.96 8.72 7.47  Std.dev 185 0.6 0.42 0.12 0.45 0.26 0.32 B Avg 1684 18.6 2.00 1.00 22.53 10.65 10.29 C  1795 18.3 2.38 1.31 20.84 10.11 8.44  Std.dev 343 3.4 0.38 0.32 0.95 0.68 0.71 D Avg 1417 19.2 2.50 1.75 18.86 9.02 7.28  Std.dev 59 1.3 0.50 0.00 0.60 0.39 0.80 54 Table 7B. Means and standard deviations for all the major and sub branches that obtained from the neighbor joining tree for the seven traits (color, appearance, L*, a*, b*, seed weight, and yield) of 69 black bean genotypes that obtained from the year 2014.  NJ Group Average/St. Deviation Yield (kg ha-1) Seed Weight (g) Appearance (1-5) Color (1-5) L* (0-100) a* (-127+127) b* (-127+127) A1-1 Avg 2775.22 19.69 3.40 3.40 18.71 3.30 3.58  Std.dev 871.05 1.07 0.51 0.51 1.45 0.40 0.35 A1-2 Avg 2984.58 20.95 3.68 3.68 19.91 3.26 3.64  Std.dev 664.08 2.39 0.53 0.53 2.74 1.03 0.98 A2-1 Avg 2499.21 22.15 3.14 3.14 18.20 4.05 4.10  Std.dev 750.23 1.74 0.53 0.53 1.75 0.67 0.53 A2-2 Avg 2565.53 20.94 3.78 4.28 24.84 2.11 2.86  Std.dev 434.90 2.17 0.24 0.38 2.69 0.46 0.29 A2-3 Avg 2744.48 20.78 3.13 3.13 18.39 3.98 4.12  Std.dev 483.41 2.13 0.52 0.52 2.31 0.99 0.68 A2-4 Avg 2610.10 19.10 3.50 3.50 18.73 4.09 4.19  Std.dev 8.9 0.65 0.25 0.25 0.22 0.505 0.47 B Avg 2789.70 20.40 3.25 3.25 19.65 7.79 7.30 C  3121.35 18.83 3.63 3.63 17.68 5.80 5.54  Std.dev 429.34 0.95 0.54 0.54 0.54 1.20 1.02 D Avg 2898.10 23.40 3.17 3.17 18.57 4.21 4.37  Std.dev 238.09 2.15 0.12 0.12 0.11 0.87 0.80 55 3.8. Genome-Wide Association Mapping GWAS was conducted for the seven traits (yield, seed weight, canned appearance, color, L*, a*, and b*) of 61 black bean genotypes that were grown for two years 2013 and 2014. Eight genotypes with shiny seed coat were excluded from the original 69 genotypes, because of their luster that interfered with the color of the dull genotypes, which are the dominant group in the study. Including the shiny seed-coated genotypes may confound the final result by influencing L*, a*, and b* scores.  For all traits, there were significant SNP markers detected above LOD 2 cutoff. Even though the size of the studied population was small, it was strong enough to detect some significant SNPs above LOD 6 on Pv07 where the region of ASP gene, which is responsible of seed luster in black bean (Figure 19).  For the trait of seed yield, significant SNP marker trait associations were found on Pv04 and Pv06, the significant SNPs mapped to two locations with different effects on yield. (Figure 5 and Figure 12). The three associations on Pv04 for 2013, 2014, and the two-year average (where the seed yield values from the two years were averaged) were the most significant and also explained the most phenotypic variation with R2 values of 0.26 for all the three associations (Table. 8A). The significant SNPs from 2013 was detected on the physical position ranged from 2.27 to 3.36Mb and had an R2 of 17%. The SNP markers from 2014 were identified on the physical positions ranged (between 5.63 to 5.71Mb) and carried 26% of phenotypic variation. The significant SNPs from the two-year average were detected on the physical position ranged from 2.04 to 2.36Mb and had an R2 ranged between 22-23%. The two association on Pv06 for 2013 and the two-year average explained from 17% to 23% of the phenotypic variation. The SNP markers from 2013 were identified on the physical positions ranged (between 20.43 to 20.57Mb) and carried phenotypic variation ranged from 17-20%. The SNP markers from the two-year average 56 were identified on the physical position of 20.63Mb and carried 23% of phenotypic variation. We could not find any previous studies matching what was found in this study result for the trait of seed yield. The significant marker trait associations were found on Pv06 and Pv10 for the trait of seed weight (Figure 6 and Figure 13). The two associations on Pv06 for 2013 and 2014 were significant and explained 15% and 17% phenotypic variation. The significant SNPs for the two years (2013 and 2014) were identified on the physical positions of 24.82 and 20.81 Mb respectively, and mapped to one region with different effects on seed weight. This result agreed with what was discovered in the study of Cichy et al., where they detected some QTLs for the trait of seed weight on Pv06 for two years of 2000 and 2005. The QTL that was found in 2000 for the trait of seed weight explained phenotypic variation ranged from 10-18%, where the one in 2005 had R2 ranged from 21-27%. One significant SNP marker for 2013 was found associated with seed weight on Pv10 with a physical position of 40.83Mb and explain 17% from the phenotypic variation (Table 8B).    Significant marker trait associations were found on Pv04, Pv07, and Pv10 for the trait of appearance (Figure 7 and Figure 14). The significant associations on Pv04 for 2013 were identified on the physical regions ranged between 41.56 and 42.31Mb, and explained 17-21% of the phenotypic variation. The significant SNPs on Pv07 for the two-year average were detected on the physical locations ranged between 4.23 to 10.56Mb, and explained phenotypic variation ranged between 17-20%. The significant associations on Pv07 for 2013 were identified on the physical regions ranged between 3.92 and 4.19Mb, and had R2 ranged between 18-23%. The significant SNP on Pv07 was identified on the physical location of 36.47Mb, and had R2 of 18% (Table 8C). The significant SNP on Pv10 for 2014 was identified on the physical location of 8.25Mb, and had  57 R2 of 17%. We were not able to find any previous studies matching what was found in this study result for the trait of appearance. Significant marker trait associations were found on Pv04, Pv05, Pv06, and Pv09 for the trait of color (Figure 8 and Figure 15). The overall associations were significant and explained 13-23% of the phenotypic variation. The significant markers on Pv04 for 2013 were detected on the physical region ranged between 2.28 and 2.46Mb, and explained phenotypic variation ranged between 13-15%. The significant SNP on Pv04 for the two-year average was detected on the physical region of 10.56Mb, and explained phenotypic variation of 18%. The significant SNPs on Pv05 for 2013 were detected on the physical region ranged between 39.47 and 39.51Mb, and explained phenotypic variation ranged between 39.47 and 39.51Mb, and explained phenotypic variation of 13%. The significant SNP on Pv05 for 2014 was detected on the physical region of 39.47Mb and explained phenotypic variation of 20%. The significant SNP on Pv05 for the two-year average was detected on the physical region of 39.47Mb, and explained phenotypic variation of 22%. In a study conducted by Wright and Kelly, (2011), two QTLs on Pv05 was also found associated with the trait of color with R2 of 0.10 and 0.13, and carried different effect on the trait. On Pv06 for 2013, significant SNPs were identified on the physical position ranged between 0.03 and 18.33Mb, and explained phenotypic variation ranged between 13-17%. On Pv09 for 2013, significant SNP was also identified on the physical position of 32.56Mb, and explained phenotypic variation of 13%. On the same chromosome for 2014, a significant SNP was found on the physical region of 35.32 and explained phenotypic region of 22%. The significant SNP on Pv09 for the two-year average was detected on the physical region of 35.32Mb, and explained phenotypic region of 23% (Table 8D). In a study by Cichy et al., (2014), QTLs on Pv04 and Pv05 were identified for the trait of color retention which agrees with our finding in this study.  58 Significant marker trait associations were found on Pv03, Pv04, Pv05, Pv06, and Pv09 for the trait of L* (Figure 9 and Figure 16). The significant SNPs on Pv03 for 2013 were identified on the physical region ranged between 0.89 and 50.12Mb, and had R2 ranged between 14-17%. The significant SNP on the same chromosome for 2014 was identified on the physical region of 45.59Mb with R2 of 19%. The significant SNP markers on Pv04 for 2013 were identified on the physical region ranged between 2.27 to 4.4Mb, with R2 ranged between 15-20%. The significant SNP markers that were found on Pv05 and were identified on the physical region ranged between 4.81 and 39.47Mb, and had R2 ranged between 14-16%. On Pv06 for 2013, significant SNPs were identified on the physical region ranged between 18.33 and 22.22Mb, and had R2 ranged 12-20%. The significant SNP on Pv09 for 2014 was identified on the physical region of 32.95MB, and had R2 of 13%. On the same chromosome for the two-year average significant SNP markers were identified on the physical region ranged between 20.58 and 25.25Mb, and had R2 ranged between 14-16% (Figure 18E). In a study by Cichy et al., (2014), a QTL was also detected on Pv04 and Pv05 associated with L* trait and had R2 of 3.3 and 9.8 respectively (Table 8E). Significant SNPs associated with a* were found on Pv02, Pv05 and Pv08 (Figure 10 and Figure 17). On Pv02 for 2013, the significant SNPs mapped to on the physical region ranged between 32.86and 44.91Mb with phenotypic variation ranged between 15-18%. The significant SNP on Pv05 for 2013 were identified on the physical positions of 39.18Mb, and carried 14% of phenotypic variation. The significant SNP on Pv05 for 2014 were identified on the physical positions of 39.47Mb, and carried 23% of phenotypic variation. The significant SNP on Pv05 for the two-year average were identified on the physical positions of 39.47Mb, and carried 25% of phenotypic variation. On Pv08 for 2013, the significant markers were identified on the physical 59 region ranged between 57.94 and 59.01Mb, and had R2 ranged 14-15% (Table 8F). We could not find any previous studies matching what was found in this study result for the trait of a*. For the trait of b*, significant marker trait associations were found on Pv04, Pv05, Pv06, and Pv07 (Figure 11 and Figure 18). The overall associations were significant and explained 16-29% of the phenotypic variation. The significant markers on Pv04 for 2013 were detected on the physical region ranged between 2.30 and 2.47Mb, and had phenotypic variation ranged between 16-20%. The significant SNPs on the same chromosome for 2014 mapped to the physical location which ranged between 10.56 and 19.99Mb, and had phenotypic variation of 18%. On the same chromosome for the two-year average, significant SNPs were detected in the physical region ranged between 10.56 and 10.99Mb, with phenotypic variation of 17%. On Pv05 for 2013, significant SNPs were detected in the physical region ranged between 39.18 and 39.20Mb, with phenotypic variation of 19%. On the same chromosome for 2014, significant SNPs detected and mapped to the physical region ranged between 39.47 and 39.48Mb, with phenotypic variation of 21%, For the two-year average, significant SNPs were detected in the physical region ranged between 39.47 and 39.51Mb, with R2 of 29%. The significant SNP markers on Pv06 for 2013 were detected and mapped to the physical region ranged between 18.33 and 18.51Mb, with R2 of 16%. On the same chromosome for the two-year average, significant SNP markers were detected on the physical location ranged between 18.50 and 18.50Mb. The significant SNPs on Pv07 for 2014 were detected on the physical location of 38.89Mb, and had phenotypic variation of 19%. On the same chromosome for the two-year average significant SNP markers were detected on the physical location ranged between 36.47 and 38.89Mb, with phenotypic variation ranged between 18-19% (Table 8G). In a study by Cichy et al., (2014), their finding was agreed with our result in this study where they identified a QTL on Pv05 that was associated with b* trait  60 The significant SNP markers that were found associated with color retention, L*, a*, and b* shared same physical position (39Mb) on the Pv05. This result assured what was found by the study of Cichy et al., (2014), where they identified some significant SNP markers on Pv04 and Pv05 for the traits of color retention, L*, b*, and anthocyanin. However, by comparing the two results, the physical regions that were identified in this study and associated with color traits (color retention, L*, a*, and b*) were novel locations. The traits of color retention and L* were also co-localized on Pv04, Pv05, Pv6, and Pv09 and shared very similar regions. Through GWAS, we have also identified regions on Pv04 with physical locations ranged from (4.43 to 10.56Mb) that were related to the traits of color retention, L* and appearance. These findings indicate an important role for the identified SNP markers that are associated with the traits of color and appearance in black beans. This research serves as an important foundation for further studies to understand the genetic control of color retention after canning in black beans.                  61 Figure 5. Manhattan plot of seed yield values averaged across 2013 of 61 Black beans genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 3 principal components to account for population structure. The black line is p=0.01 threshold for significance.                        62 Figure 6. Manhattan plot of seed weight values averaged across 2013 of 61 Black beans genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 7 principal components to account for population structure. The black line is p=0.01 threshold for significance.                 63 Figure 7. Manhattan plot of appearance rating values averaged across 2013 of 61 Black beans genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 2 principal components to account for population structure. The black line is p=0.01 threshold for significance, and the red line is p=0.001 threshold for significance.               64 Figure 8. Manhattan plot of color ratings averaged across 2013 of 61 Black beans genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix. The black line is p=0.01 threshold for significance.                65 Figure 9. Manhattan plot of the trait of L* values averaged across 2013 of 61 Black beans genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 7 principal components to account for population structure. The black line is p=0.01 threshold for significance.                66 Figure 10. Manhattan plot of the trait of a* values averaged across 2013 of 61 Black beans genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 3 principal components to account for population structure. The black line is p=0.01 threshold for significance.                67 Figure 11. Manhattan plot of the trait of b* values averaged across 2013 of 61 Black beans genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 3 principal components to account for population structure. The black line is p=0.01 threshold for significance.                68 Figure 12. Manhattan plots of seed yield values of 2014 and two-year average for 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix. The black line is p=0.01 threshold for significance, and the red line is p=0.001 threshold for significance.     69 Figure 13. Manhattan plots of seed weight values of 2014 and two-year average for 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 7 principal components to account for population structure. The black line is p=0.01 threshold for significance.     70 Figure 14. Manhattan plots of appearance ratings of 2014 and two-year average for 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 2 principal components to account for population structure. The black line is p=0.01 threshold for significance.     71 Figure 15. Manhattan plots of color ratings of 2014 and two-year average for 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 7 principal components to account for population structure. The black line is p=0.01 threshold for significance.     72 Figure 16. Manhattan plots of the trait of L* of 2014 and two-year average for 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 5 principal components to account for population structure. The black line is p=0.01 threshold for significance, and the red line is p=0.001 threshold for significance.    73 Figure 17. Manhattan plots of the trait of a* of 2014 and two-year average for 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix. The black line is p=0.01 threshold for significance.     74 Figure 18. Manhattan plots of the trait of b* of 2014 and two-year average for 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 3 principal components to account for population structure. The black line is p=0.01 threshold for significance.     75 Figure 19. Manhattan plot of the trait of shiny trait of 61 black bean genotypes with 2800 SNPs (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix. The black line is p=0.01 threshold for significance. The black line is p=0.01 threshold for significance, and the red line is p=0.001 threshold for significance.                     76 Table 8A. GWAS significant markers, genome position, p-value, R2, and phenotypic effect associated with the seed yield trait for two-year data (2013 and 2014) and the average of two years. Yield Year SNP CHR1 Position (Mb) P-Value Add Effect R2% 2014 ss715648226 Pv04 5.63 8.60E-04* -2.48E+02 26 2014 ss715648222 Pv04 5.70 8.60E-04* 248.45886 26 2014 ss715648223 Pv04 5.71 8.60E-04* 248.45886 26 2014 ss715648228 Pv04 5.66 9.03E-04* 247.6965 26 Average ss715647257 Pv06 20.63 9.89E-04* -1.98E+02 23 Average ss715650012 Pv04 2.04 0.00116 -2.60E+02 23 Average ss715646240 Pv04 2.27 0.00141 145.27562 22 Average ss715646234 Pv04 2.36 0.00141 -1.45E+02 22 2013 ss715648612 Pv06 20.95 0.00475 185.14331 20 2013 ss715645763 Pv06 26.57 0.00616 392.06839 18 2013 ss715647253 Pv06 20.43 0.00736 199.61102 18 2013 ss715646240 Pv04 2.27 0.00817 158.80195 17 2013 ss715646234 Pv04 2.36 0.00817 -1.59E+02 17 2013 ss715646240 Pv04 2.27 0.00817 158.80195 17 2013 ss715647257 Pv06 20.63 0.00848 -2.07E+02 17 2013 ss715649182 Pv04 43.34 0.00906 -8.32E+01 17 1CHR: chromosome.  *SNPs with 3 LOD cutoff.              77 Table 8B. GWAS significant markers, genome position, p-value, R2, and phenotypic effect associated with the seed weight trait for two-year data (2013 and 2014) and the average of two years. Seed Weight Year SNP CHR1 Position (Mb) P-Value Add Effect R2% 2013 ss715645537 Pv010 40.83 0.00497 -4.36E-01 17 2014 ss715646411 Pv06 24.82 0.00666 1.97785 17 2013 ss715649416 Pv06 20.81 0.00844 0.60545 15 1CHR: chromosome.                   78 Table 8C. GWAS significant markers, genome position, p-value, R2, and phenotypic effect associated with the appearance trait for two-year data (2013 and 2014) and the average of two-years. Appearance Year SNP CHR1 Position (Mb) P-Value Add Effect R2% 2013 ss715646463 Pv07 4.19 0.00104 0.2241 23 2013 ss715646475 Pv07 4.03 0.00111 0.22083 23 2013 ss715646473 Pv07 4.05 0.00161 -2.19E-01 21 2013 ss715649259 Pv04 42.31 0.00198 0.29723 21 2013 ss715651054 Pv04 41.63 0.00209 0.23837 20 2013 ss715649261 Pv04 42.28 0.0024 0.3073 20 2013 ss715648973 Pv04 41.56 0.00257 0.24427 20 2013 ss715649276 Pv07 3.92 0.00364 -1.96E-01 18 Average ss715648120 Pv04 4.23 0.00376 1.14E-01 20 2013 ss715646464 Pv07 4.14 0.00425 -1.94E-01 18 2013 ss715648138 Pv04 42.08 0.00435 -3.10E-01 18 2013 ss715649205 Pv04 41.78 0.00468 -3.02E-01 17 Average ss715647800 Pv07 36.47 0.00807 0.49437 18 Average ss715650075 Pv04 10.56 0.00854 -3.90E-01 17 2014 ss715649365 Pv10 8.25 0.00995 -0.52951 17 1CHR: chromosome.                 79 Table 8D. GWAS significant markers, genome position, p-value, R2, and phenotypic effect associated with the color trait for two-year data (2013 and 2014) and the average of two years. Color Year SNP CHR1 Position (Mb) P-Value Add Effect R2% 2013 ss715646829 Pv06 18.33 0.00194 0.64152 17 Average  ss715646072 Pv09 35.32 0.00234 -0.35366 23 Average  ss715645369 Pv05 39.47 0.00287 0.35163 22 2014 ss715646072 Pv09 35.32 0.00293 -0.36126 22 2013 ss715646236 Pv04 2.30 0.00455 -0.40351 15 2013 ss715646235 Pv04 2.34 0.00455 -0.40351 15 2014 ss715645369 Pv05 39.47 0.00459 0.34479 20 2013 ss715646238 Pv04 2.28 0.00623 -0.39433 14 2013 ss715645369 Pv05 39.47 0.00737 0.37543 13 2013 ss715645370 Pv05 39.48 0.00737 0.37543 13 2013 ss715645371 Pv05 39.49 0.00737 -0.37543 13 2013 ss715645372 Pv05 39.49 0.00737 -0.37543 13 2013 ss715645375 Pv05 39.51 0.00737 -0.37543 13 2013 ss715645376 Pv05 39.51 0.00737 -0.37543 13 2013 ss715646230 Pv04 2.40 0.00881 -0.41829 13 2013 ss715646229 Pv04 2.43 0.00881 0.41829 13 2013 ss715646227 Pv04 2.46 0.00881 -0.41829 13 2013 ss715645617 Pv09 32.56 0.00883 -0.50188 13 2013 ss715640556 Pv04 4.43 0.00884 -0.23133 13 Average  ss715650075 Pv04 10.56 0.00896 0.27443 18 2013 ss715650985 Pv06 0.03 0.00972 0.41838 13 2013 ss715648431 Pv06 0.18 0.00972 0.41838 13 2013 ss715648429 Pv06 0.20 0.00972 -0.41838 13 2013 ss715648496 Pv06 3.38 0.00972 -0.41838 13 2013 ss715648495 Pv06 3.39 0.00972 -0.41838 13 1CHR: chromosome.         80 Table 8E. GWAS significant markers, genome position, p-value, R2, and phenotypic effect associated with the L* trait for two-year data (2013 and 2014) and the average of two years. L* Year SNP CHR1 Position (Mb) P-Value Add Effect R2%  2014 ss715639244 Pv03 45.59 0.00125 4.09077 19  2013 ss715646829 Pv06 18.33 0.00142 -1.5125 20  2013 ss715650222 Pv06 18.50 0.00142 -1.51E+00 20  2013 ss715641091 Pv06 18.51 0.00142 -1.51E+00 20  2013 ss715646227 Pv04 2.46 0.00162 1.21984 20  2013 ss715646236 Pv04 2.30 0.0017 1.01947 20  2013 ss715646235 Pv04 2.34 0.0017 1.01947 20  2013 ss715646238 Pv04 2.28 0.00322 0.9469 17  2013 ss715646230 Pv04 2.40 0.00378 1.35858 17  2013 ss715646229 Pv04 2.43 0.00378 -1.36E+00 17  Average ss715646059 Pv09 25.25 0.00386 0.53827 16  2013 ss715650435 Pv03 0.89 0.00397 -0.85332 17  2013 ss715645369 Pv05 39.47 0.00444 -0.91271 16  2013 ss715646240 Pv04 2.27 0.00477 -0.79655 16  2013 ss715646234 Pv04 2.36 0.00477 0.79655 16  2013 ss715650285 Pv06 22.22 0.00501 NaN 12  2013 ss715645575 Pv03 50.12 0.0058 -2.0857 15  2013 ss715640336 Pv05 8.09 0.0058 2.08575 15  2013 ss715640556 Pv04 4.43 0.00752 0.50605 15  2013 ss715639244 Pv03 45.59 0.00773 -1.6219 14  Average ss715648756 Pv09 20.58 0.00798 0.52461 14  2014 ss715645638 Pv09 32.95 0.0087 -2.70E+00 13  2013 ss715650411 Pv05 4.81 0.00873 0.63716 14 1CHR: chromosome.          81 Table 8F. GWAS significant markers, genome position, p-value, R2, and phenotypic effect associated with the a* trait for two-year data (2013 and 2014) and the average of two years. a* Year SNP CHR1 Position (Mb) P-Value Add Effect R2%  Average ss715645369 Pv05 39.47 0.0013 -0.49187 25  2014 ss715645369 Pv05 39.47 0.0022 -0.56645 23  2013 ss715647408 Pv02 39.66 0.00269 -3.89E-01 18  2013 ss715647243 Pv02 44.72 0.00358 0.40797 17  2013 ss715647244 Pv02 44.74 0.00358 0.40797 17  2013 ss715647242 Pv02 44.68 0.00409 0.40824 17  2013 ss715645959 Pv02 44.80 0.0046 0.39699 16  2013 ss715645968 Pv02 44.80 0.0046 -3.97E-01 16  2013 ss715645958 Pv02 44.81 0.0046 -3.97E-01 16  2013 ss715645970 Pv02 44.89 0.0046 -3.97E-01 16  2013 ss715645955 Pv02 44.91 0.0046 0.39699 16  2013 ss715645963 Pv02 44.84 0.0053 0.3969 16  2013 ss715645956 Pv02 44.92 0.0053 0.3969 16  2013 ss715647099 Pv02 32.86 0.00709 0.37005 15  2013 ss715646101 Pv08 58.01 0.00749 -3.07E-01 15  2013 ss715646094 Pv08 57.94 0.00869 0.30925 14  2013 ss715646095 Pv08 57.94 0.00869 -3.09E-01 14  2013 ss715645338 Pv05 39.18 0.00982 -4.13E-01 14 1CHR: chromosome.            82 Table 8G. GWAS significant markers, genome position, p-value, R2, and phenotypic effect associated with the b* trait for two-year data (2013 and 2014) and the average of two years. b* Year SNP CHR1 Position (BP) P-Value Add Effect R2%  Average ss715645370 Pv05 39.48 5.53E-04* -6.57E-01 29  Average ss715645371 Pv05 39.49 5.53E-04* 0.6574 29  Average ss715645372 Pv05 39.49 5.53E-04* 0.6574 29  Average ss715645375 Pv05 39.51 5.53E-04* 0.6574 29  Average ss715645376 Pv05 39.51 5.53E-04* 0.6574 29  Average ss715645369 Pv05 39.47 5.53E-04* -6.57E-01 29  2013 ss715646227 Pv04 2.46 0.00178 0.83891 20  2013 ss715645338 Pv05 39.18 0.002 -9.60E-01 19  2013 ss715645340 Pv05 39.20 0.002 0.95978 19  Average ss715650222 Pv06 18.50 0.00258 -6.77E-01 22  Average ss715641091 Pv06 18.51 0.00258 -6.77E-01 22  2013 ss715646230 Pv04 2.40 0.00323 0.87974 18  2013 ss715646230 Pv04 2.40 0.00323 0.87974 18  2013 ss715646229 Pv04 2.43 0.00323 -8.80E-01 18  2014 ss715645369 Pv05 39.47 0.00323 -3.80E-01 21  2014 ss715645370 Pv05 39.48 0.00323 -3.80E-01 21  2013 ss715646236 Pv04 2.30 0.00448 0.68126 16  2013 ss715646236 Pv04 2.30 0.00448 0.68126 16  2013 ss715646235 Pv04 2.34 0.00448 0.68126 16  Average ss715649164 Pv07 38.89 0.00502 -4.39E-01 19  2013 ss715646829 Pv06 18.33 0.00569 -9.82E-01 16  2013 ss715646829 Pv06 18.33 0.00569 -9.82E-01 16  2013 ss715650222 Pv06 18.50 0.00569 -9.82E-01 16  2013 ss715641091 Pv06 18.51 0.00569 -9.82E-01 16  2014 ss715649164 Pv07 38.89 0.0057 -3.18E-01 19  2014 ss715646943 Pv04 10.84 0.00747 0.30394 18  2014 ss715646944 Pv04 10.99 0.00747 0.30394 18  2014 ss715650075 Pv04 10.56 0.00747 -3.04E-01 18  Average ss715647800 Pv07 36.47 0.00807 0.49437 18  Average ss715646943 Pv04 10.84 0.00854 0.38985 17  Average ss715646944 Pv04 10.99 0.00854 0.38985 17  Average ss715650075 Pv04 10.56 0.00854 -3.90E-01 17 1CHR: chromosome. *SNPs with 3 LOD cutoff.      83 CONCLUSION AND RECOMMENDATION The appearance and color retention of canned black bean products affects consumer decisions to accept or reject black bean products.  In this study, 69 black bean genotypes from the major U.S. bean breeding programs were assessed for their genetic variability for seven quality traits (seed yield, seed weight, canned appearance, color, L*, a*, and b*) over two field experiments grown in 2013 and 2014. A year advantage of 2014 over 2013 was observed in the trait of color retention as well as the rest of the previous traits (yield, seed weight, appearance, L*, a*, and b*). Seed quality traits are linked to environment and there were superior environmental conditions for the canned black bean in 2014. Also, this express the wide genetic variability among black genotypes specially in the trait of color retention and that may do to fewer bean breeding programs that work on color retention, which are MSU and USDA-MI. Strong correlation analysis results were found between color ratings and the traits of L*, a*, b* colors scans in both years 2013 and 2014. This assures that the trained sensory panel was efficient and comparable to the colorimeter that was used to obtain L* a* b* color measurements.  The result of neighbor joining tree revealed that the sub-branch A2-2, which has all the shiny seed-coated genotypes, has higher means for the traits of appearance, a*, and b* for both years (2013 and 2014). That suggests shiny seed-coated genotypes to have superior color retention over opaque seeds, and that is due not only to their seed luster, but to their excellent canning quality that have been chosen based on. All the other traits were varied between the two years. For the trait of yield, the sub-branch A1-2 had the higher yielding mean in the year of 2013, while the mean of the major branch C was the highest in 2014. The two traits of color retention and L* have been dominated by the sub-group A2-2 with the highest mean values (3.61, 15.17) and standard deviation (0.58 and 1.08) respectively. The same two traits were varied in the year of 2014.  84 GWAS was conducted for the seven traits (seed yield, seed weight, appearance, color, L*, a*, and b*) of 61 black bean genotypes that were obtained from the two years 2013 and 2014. Even though the size of the studied population was small, it was strong enough to detect some significant SNPs above LOD 6 on Pv07 where the region of ASP gene, which is responsible of seed luster in black bean. The significant marker trait associations were found on Pv06 and Pv10 for the trait of seed weight (Figure 6 and Figure 13). The two associations on Pv06 for 2013 and 2014 were significant and explained 15% and 17% phenotypic variation. The significant SNPs for the two years (2013 and 2014) were identified on the physical positions of 24.82 and 20.81 Mb respectively, and mapped to one region with different effects on seed weight. This result agreed with what was discovered in the study of Cichy et al., where they detected some QTLs for the trait of seed weight on Pv06 for two years of 2000 and 2005.  The significant SNP markers that were found associated with color retention, L*, a*, and b* shared  same physical position (39Mb) on the Pv05. This result assured what was found by the study of Cichy et al., (2014), where they identified some significant SNP markers on Pv04 and Pv05 for  the traits of color retention, L*, b*, and anthocyanin. However, by comparing the two results, the  physical regions that were identified in this study and associated with color traits (color retention,  L*, a*, and b*) were novel locations. The traits of color retention and L* were also co-localized  on Pv04, Pv05, Pv6, and Pv09 and shared very similar regions. Through GWAS, we have also  identified regions on Pv04 with physical locations ranged from (4.43 to 10.56Mb) that were related  to the traits of color retention, L* and appearance. These findings indicate an important role for  the identified SNP markers that are associated with the traits of color and appearance in black  beans. This research serves as an important foundation for further studies to understand the genetic  control of color retention after canning in black beans. 85 CHAPTER 4 PHENOTYPIC EVALUATION OF ANTHOCYANIN IN BLACK BEAN INTRODUCTION Anthocyanins are water-soluble pigments, and are triggered environmental stresses such as bright light, UV radiation, and high temperature. There are about 500 types of anthocyanins in nature (Gould et al., 2009). Black bean anthocyanins are concentrated in the seed coat (Dzomba et al., 2013). The most common anthocyanins found in black beans are delphinidin 3-glucoside, petunidin 3-glucoside, malvidin 3-glucoside, and malvidin 3,5-diglucoside (Zhang et al., 2014). Some anthocyanin compounds have been extracted in lower quantities for example cyanidin 3-glucoside, cyanidin 3,5-diglucoside, pelargonidin 3-glucoside, and pelargonidin 3,5-diglucoside (Takeoka et al., 1997). Black bean seed coats also contain extractable anthocyanins and bound anthocyanins (Kelly et al., 2015). Anthocyanins are bound either by oxidation or by cross-linking, and the concentration of the bound anthocyanins in some black bean accessions are higher than the extractable ones  (Marles et al., 2010). The amounts of each compound vary due to genetic variability and environmental aspects such as the growing location (Rocha-Guzmán et al., 2007). These compounds could impact black bean canning quality since they are responsible for seed coat color. Anthocyanins also play a role in maintaining human health through their anti-oxidant activity, which help to reduce oxygen free radicals (Díaz et al., 2010; La Cruz et al., 2013). In this study, concentrations of three anthocyanins (delphinidin-3-glucoside, petunidin-3-glucoside, and malvidin-3-glucoside) were quantified using liquid chromatography-mass spectrometry (LC-MS). Total anthocyanins were also quantified using an UV assay. Both approaches were carried out on 12 black bean genotypes where raw and canned samples grown for two years (2013 and 2014). 86 The 12 genotypes were chosen to represent a wide range of the anthocyanin concentration based on color retention of the 69 genotypes described in the third chapter.                   87 MATERIALS AND METHODS 4.1. Materials Selection A group of 12 genotypes was previously selected based on the canning process evaluation result that obtained earlier in this study (Table 9). The 12 genotypes represent a range of variability in term of color and appearance traits. This group included 96 samples grown for two years (2013 and 2014), two conditions (raw and canned), and two replications, and one duplicate for each replication per sample was involved.                     88 Table 9. A list of the 12 black bean genotypes evaluated traits for total anthocyanins concentrations and the three specific anthocyanins quantification in raw and canned black beans. No. Genotype Shiny / Dull Origin Pedigree   1 B-EL1291d Dull USDA-MI Black Magic/Shiny Crow 2 B11313 Dull MSU B04644//B04349/B05044 3 B11356 Dull MSU Jaguar/B04644 4 ND081343 Dull NDSU Unknown 5 Eclipse Dull NDSU ND9902621-2 6 Zenith (B10244) Dull MSU B04644/B07554 7 B11370 Dull MSU B05055/B04265 8 B10243 Dull MSU B04610/N05346 9 B11338 Dull MSU N08007//B04349/B05044 10 NDF09304 Dull NDSU Unknown 11 B11285 Dull MSU N04152/N05346//N04141/N05317 12 OACExpresso Dull Saskatoon Unknown  4.2. Anthocyanin Extraction and Purification The seed varieties were separately freeze-dried using Virtis genesis 12el freeze dryer. Dried seeds from each variety of bean was then separately ground in a PX-MFC 90 D mill to yield a 30-40 mesh powder. One gram each of the powdered bean was transferred to 50mL flasks. Total anthocyanins in the bean powder was then extracted by adding a solvent system containing 80 methanol: 20 HPLC H2O: 0.3HCl (20mL) to each flask, shaken for 4h on a Gyrotory to 50 mL tubes and centrifuged (Centrifuge Sorvall RT7 Benchtop Centrifuge) for 10min (270rpm, -10°C). The combined supernatants were then concentrated to 5mL under vacuum using a rotary 89 evaporator at 35°C (Buchi). The pellets were discorded. The resulting samples were analyzed immediately for total anthocyanin content (Chandra et al., 1992; Bowen-Forbes et al., 2010). 4.3. UV Spectrophotometer Assay The total anthocyanin was measured using a spectrophotometry (Genesys 10S UV-Vis) for a UV assay. The total anthocyanin was absorbed at anthocyanins maximum wavelengths of 520nm. The calibration curve was made using delphinidin-3-glucoside as a standard (purchased from Cerilliant, TX, USA) that represents the major anthocyanin in black beans (Table 11) (Figure 20) (Takeoka et al., 1997). Several dilution factors have been made based on the preliminary spectrophotometer absorbents for all the samples, and used to obtain a moderate absorbent range of (0.01-1.00) (Table 10). Three different amounts of the extracted stock based on the preliminary samples absorbance were taken to a 15ml tube (300µ/g, 200µ/g, and 100µ/g), and three different amount of the extraction buffer were added to obtain a total volume of 3000µ for each sample (2700µ/g, 2800µ/g, and 2900µ/g). Two duplicates, 1ml each were transferred from the 3000µ-total volume to two 1.5-Cuvettes and were taken for absorbance measurements. Moreover, some sample extracts, especially the raw types, were over the absorbent range (higher than 1.00), and accordingly three dilution factors were obtained (as described) for those over-rang samples 10, 15, and 30. The spectrophotometer was blanked by using distilled water filled in a 1.5-Cuvette and measured at the same wavelength level (Chandra et al., 1992; Bowen-Forbes et al., 2010).        90 Figure 20. Calibration curve for delphinidin-3-glucoside standard made to calculate the total anthocyanins mass in black bean genotype extracts.                y = 0.1201x + 0.0008R² = 0.999600.20.40.60.811.20123456789Absorbance (Au)Delphinidin-3-glucoside Conc (µ/g) 91 Table 10. Delphinidin-3-glucoside calibration curve points made to calculate the total anthocyanins mass in black bean genotype extracts.                             Delphinidin-3-glucoside conc µg/g Absorbent  0 0 0.126 0.016 0.189 0.023 0.2625 0.034 0.42 0.051 0.525 0.07 1.05 0.13 2.1 0.24 4.2 0.508 6.3 0.761 7.35 0.867 7.98 0.976 8.358 1.001 92 Table 11. Three major anthocyanins in black beans represent the study standards. Anthocyanin Structure Delphinidin-3-glucoside  Petunidin-3-glucoside  Malvidin-3-glucoside  (Takeoka et al., 1997) 4.4. Anthocyanin Quantification Using Liquid ChromatographyMass Spectrometry (LC-MS) High-performance liquid chromatography combined with mass detection (HPLCMS) is the most commonly used technique for analyzing anthocyanins. HPLC provides powerful and rapid separations of anthocyanins, and MS makes a quick and sensitive detection. HPLCMS is an important frequently-used instrument for anthocyanin characterization, offering mass spectrum of intact molecular and fragment ions (Acevedo et al., 2012). Anthocyanin extracts for the same chosen genotypes group were analyzed by LC-MS with an Acquity Ultra Performance Liquid Chromatography (UPLC) coupled to a Xevo G2QT of spectrometer (Waters Corp).  For LC analysis, 5µL samples of extract was injected and passed through a HSS C18 1.8 mm2.1 x 100mm column (Waters) at 0.4mL/min. The mobile phase contained 0.1% formic acid in water (solvent A) and 0.1% formic acid in acetonitrile (solvent B).  Anthocyanins were eluted using linear gradients of 97.6 to 80% A in 2.5min, 80 to 60% A in 0.5min, 60 to 48% A in 2min, 48 to 5% A 93 in 0.5min, 5 to 97.6% A in 1min, and a 0.5min hold at 97.6% A.  Electrospray ionization mass spectrometry (ESI) was made in positive VIII ionization mode with a scan speed of 5/s in the mass range from 50 to 1200Da. Lock-mass correction was used, with a fluidic system to deliver LockSpray reference which is leucine enkephalin as the external lock-mass standard. LC-MS data was analyzed using the MassLynx software (Rockenbach et al., 2012). 4.5. Statistical Analysis  Analysis of variance (ANOVA) for the two traits have been obtained using SAS9.4 statistical software package. The GLM procedure was used for analysis of variance (ANOVA) for TAC trait with the model:  ditions j + (Genotypes * Conditions) ij + Block (Year) k + e ij  including genotype (12 levels) condition (2 level) year (2 levels) and rep (2 levels) and genotype*condition. Data means were separated using POST-significance.           94 RESULTS AND DISSCUSSION A set of 12 black bean genotypes was obtained from the study population based on the result of the trained sensory panel ratings evaluation for canned color and the L* a* b* color parameters. This sub-population represents the range of color ratings including, moderate, and low. The sub-population had been evaluated for total anthocyanin concentration (TAC) and for three specific anthocyanins quantification (SAQ) (delphinidin-3-glucoside, petunidin-3-glucoside, and malvidin-3-glucoside) using spectrophotometer and LC-MS respectively over two field seasons. The result of ANOVA in (Table 12A) indicated that the variation among genotypes was significant  0.05) as a level of significance as well as the difference between the two conditions (raw and canned) based on the same level. This result suggests that canned black beans have considerable total anthocyanins concentrations comparing to the three anthocyanins quantification in canned black beans. The interactions between genotype and condition for total anthocyanin trait was also not significant. The same statistical model was used to obtain ANOVA for the trait of SAQ. The result was found to be significant for genotypes and conditions for all the three anthocyanins used for SAQ trait based on the same level of significance. The interactions between genotypes and conditions were significant for all the quantified anthocyanins except for delphinidin-3-glucoside where it was not significant (Table 12 B-D).  There was genetic variability for TAC in raw black beans in 2013 and 2014. Two genotypes, B-EL1291d and Eclipse (Table 13A) had significantly higher and lowest levels respectively among the other genotypes in both years 2013 and 2014 as follow; B-EL1291d (1.70, 2.024mg/g), Eclipse (0.49, 0.47mg/g) (Table 8C). The TAC in canned black beans was also genetically variable for the both years (2013 and 2014) (Table 13B). Genotype B-EL1291d was significantly higher than the other genotypes in both years, where the genotype ND081343 was the lowest as follow; B-95 EL1291d (0.118, 0.169mg/g), Eclipse (0.051, 0.047mg/g) (Table 14B). B-EL1291d was the highest in TAC across all conditions and years. This finding suggests the genotype B-EL1291d has more anthocyanins than the other genotypes in both raw and canned processed products. The genotype B-EL1291d is a black bean accession with dull seeds luster that was obtained from USDA-MI, and was developed as a result of the cross BlackMagic/ShinyCrow (these two genotypes are included in the studied population for seed quality evaluation). Anthocyanin retention after canning is an important attribute in black beans for the canning industry. Anthocyanin retention percentage for the 12 black bean genotypes were calculated by subtracting the total anthocyanins for a canned genotype from the raw one, and presenting data on a percentage (Table 14). The 12 genotypes varied in the ability to retain anthocyanins. The genotype with the highest total anthocyanin retention percentage in TAC experiment was Eclipse (15.92, 27.10) for both years (2013 and 2014) respectively. Even though the genotype Eclipse had the highest anthocyanin retention value (Table 14), it had a very low total anthocyanin concentration. In 2013, Eclipse had the lowest TAC among all the other genotypes, and it was among the lower TAC-genotypes in 2014. Therefore, higher anthocyanins retention does not necessary mean higher anthocyanins concentration. Two genotypes (B11313 and ND081343) had very similar anthocyanin retention (AR) across the two years, whereas the rest of the genotypes varied for AR in TAC experiment.  Three anthocyanins delphinidin-3-glucoside, petunidin-3-glucoside, and malvidin-3-glucoside were found to have the highest concentration in black beans (Dzomba et al., 2013). In this experiment, the concentrations of these three anthocyanins were quantified in raw and canned beans grown for two years (2013 and 2014). The three anthocyanins standards were used in the LC-MS quantification to identify them in the black bean extracts. The LC-MS chromatograph of 96 black bean extracts detected the three major anthocyanins in black beans with retention times of 3.73, 3.97, and 4.20 respectively (Figure 21). The electrospray (ES) mass spectrum of the three anthocyanins showed the molecular mass ions at m/z 465.11, m/z 479.12, and m/z 493.14 respectively. Figure 21. LC-MS chromatograph shows the three anthocyanins peaks have been detected in the studied black bean genotype extracts.  97 Table 12A. Analysis of variance for total anthocyanins (TA) for 12 black bean genotypes with two levels of conditions (raw and canned), two levels of years (2013 and 2014), and two replications.                    Trait  Source of variation Degree of freedom Type III SS    Mean square     F statistic    Pr>F   TA                        Genotype                                  11                              1.50                       0.14                                4.06                              0.0001                                                                                        Condition               1    0.05           0.05        1.34                0.0001      Year                                          1                              11.70                          11.70                             348.77                            0.2504                                                                              Genotype*Condition                       11                              0.50                              0.11                                3.39                              0.0008                                                                                    Table 12B. Analysis of variance for Delphindin-3-glucoside concentration for 12 black bean genotypes with two levels of conditions (raw and canned), two levels of years (2013 and 2014), and two replications.                    Trait  Source of variation Degree of freedom Type III SS    Mean square     F statistic    Pr>F   Del-3-gluc             Genotype                                  11                              0.38                       0.03                                1.14                               0.034                                                                                        Condition               1   24.68          24.68      956.76              <0.0001      Year                                         1                               0.05                              0.05                                 1.70                               0.20                                                                              Genotype*Condition                       11                              0.36                              0.03                                1.08                                0.39                                                                                     98 Table 12C. Analysis of variance for petunidin-3-glucoside concentration for 12 black bean genotypes with two levels of conditions (raw and canned), two levels of years (2013 and 2014), and two replications.                    Trait  Source of variation Degree of freedom Type III SS    Mean square     F statistic    Pr>F   Pent-3-gluc           Genotype                               11                             0.47                       0.04                              2.52                               0.0096                                                                                        Condition               1   8.63           8.63     508.54            <0.0001      Year                                         1                              0.001                            0.001                            0.03                               0.001                                                                              Genotype*Condition                       11                             0.43                              0.04                              2.31                               0.02                                                                                    Table 12D. Analysis of variance for malvinidin-3-glucoside concentration for 12 black bean genotypes with two levels of conditions (raw and canned), two levels of years (2013 and 2014), and two replications.                    Trait  Source of variation Degree of freedom Type III SS    Mean square     F statistic    Pr>F   Malv-3-gluc           Genotype                              11                             0.89                       0.08                              7.29                               0.0001                                                                                        Condition               1                4.38           4.38     393.86            <0.0001      Year                                         1                              0.01                               0.01                              0.78                               0.38                                                                              Genotype*Condition                      11                              0.79                              0.72                              6.47                                0.001                                                                                   99 Table 13A. Tukey test for the trait of total anthocyanins mg/1g total concentration of 12 raw black bean genotypes in 2013 and 2014, with two replications. Genotype Year  Mean mg/g   Genotype Year  Mean   B-EL1291d 2013 1.697 A  B-EL1291d 2014 2.0235 A  B11338 2013 1.3505 A B B11313 2014 1.4845 A B B11356 2013 1.3125 A B OACExpre 2014 1.179 A B NDF09304 2013 1.3115 A B B11356 2014 1.177 A B B11285 2013 1.299 A B Zenith 2014 1.139 A B B11370 2013 1.2555 A B B11370 2014 1.0995 A B Zenith 2013 1.236 A B B11285 2014 1.0935 A B OAC Expresso 2013 1.229 A B B11338 2014 0.724  B B10243 2013 0.9545 A B B10243 2014 0.7075  B B11313 2013 0.9045 A B NDF09304 2014 0.649  B ND081343 2013 0.657 A B ND081343 2014 0.639  B Eclipse 2013 0.493  B Eclipse 2014 0.4725  B a  Means followed with the same letter are not significantly different.          100 Table 13B. Tukey test for the trait of total anthocyanins mg/1g total concentration of 12 canned black bean genotypes in 2013 and 2014, with two replications. Genotype Year  Mean mg/g    Genotype Year  Mean      B-EL1291d 2013 0.118 A   B-EL1291d 2014 0.169 A     Zenith 2013 0.097 A B  B11313 2014 0.157 A B    B11313 2013 0.096 A B  B11370 2014 0.134 A B C   B11370 2013 0.089 A B C Zenith 2014 0.130 A B C D  B11338 2013 0.089 A B C Eclipse 2014 0.128 A B C D  B11356 2013 0.083 A B C OACExpresso 2014 0.117  B C D  OACExpresso 2013 0.082 A B C B11356 2014 0.103   C D  Eclipse 2013 0.079 A B C NDF09304 2014 0.103   C D  NDF09304 2013 0.076 A B C B11285 2014 0.101   C D  B11285 2013 0.070  B C B11338 2014 0.095   C D  B10243 2013 0.068  B C B10243 2014 0.086    D E ND081343 2013 0.051   C ND081343 2014 0.047     E a  Means followed by the same letter are not significantly different.      101 Table 14. Anthocyanins Retention percentage for the total anthocyanin data mg/1g total concentration that obtained from 12 black bean genotypes with two years (2013 and 2014).                    Genotype  Year  Anthocyanins retention percentage% (AR) Year Anthocyanins retention percentage% (AR)  Eclipse   2013    15.92    2014   27.10 B11313   2013    10.61    2014   10.54 Zenith   2013    7.81    2014   11.37 ND081343  2013    7.76    2014   7.28 B11370   2013    7.10    2014   12.19 B10243   2013    7.07    2014   12.16 B-EL1291d  2013    6.92    2014   8.35 OACExpresso  2013    6.63    2014   9.88 B11338   2013    6.59    2014   13.05 B11356   2013    6.29    2014   8.75 NDF09304  2013    5.80    2014   15.87 B11285   2013    5.39    2014   9.24 102 The mean differences were calculated for each anthocyanin. There were no significant differences among delphinidin-3-glucoside means for either conditions nor years (Table 15A-D). Such a result confirms that delphinidin-3-glucoside is the dominant anthocyanin with the highest concentration in black beans. Means separation for petunidin-3-glucoside in the raw black bean genotypes showed the genotype B11356 to be significantly higher than the rest, while the genotype ND081343 was the lowest (Table 15A-D). The means for petunidin-3-glucoside quantification showed no significant differences in the canned genotypes in 2013, and 2014 for raw and canned beans processed products. The overall concentrations of petunidin-3-glucoside was significantly lower than delphinidin-3-glucoside concentrations among the studied sup-population. Genotype B11356 possessed the highest concentration among all the other genotypes for malvidin-3-glucoside, whereas genotype ND081343 had the lowest for both years and both conditions except for the canned genotypes in 2013 (Table 15A-D). Means separation for malvidin-3-glucoside for the canned genotype in 2013 showed genotype B-EL1291d to be significantly higher than all the other genotypes, while the genotype NDF09304 was the lowest. In 2014, no significant differences were found among raw and canned beans. This result assures the two raw genotypes B11356 and B11370 exhibited superiority for SAQ for all anthocyanins in both years (2013 and 2014). Delphinidin-3-glucoside had the highest concentration among the raw black bean genotypes across the two years, while similar concentrations of delphinidin-3-glucoside, petunidin-3-glucoside, and malvidin-3-glucoside were detected in the canned genotypes. This result suggests the delphinidin-3-glucoside dominance over petunidin-3-glucoside and malvidin-3-glucoside in black beans, and agreed with previous finding of Marles et al., (2010).    103 Table 15A. Tukey test for the trait of SAQ, for the three anthocyanins (delphindin-3-glucoside, petunidin-3-glucoside, and Malvidin-3-glucoside) concentrations for the 12 raw black bean genotypes in 2013with two replication. a  Means followed with the same letter are not significantly different. bMean D, P, and M are the mean values for the anthocyanin delphindin-3-glucoside, petunidin-3-glucoside, and Malvidin-3-glucoside respectively.         Genotype Mean D mg/g  Genotype Mean P mg/g   Genotype Mean M mg/g    B11356 1.395 A B11356 0.950 A  B11356 0.945 A   NDF09304 1.390 A Zenith 0.905 A B Zenith 0.790 A B  B11338 1.310 A B-EL1291d 0.745 A B B11313 0.575 A B C B11285 1.290 A B11285 0.690 A B B-EL1291d 0.480 A B C Zenith 1.225 A NDF09304 0.640 A B B11285 0.395 A B C B10243 1.210 A B11313 0.625 A B NDF09304 0.380 A B C OACExpresso 1.150 A OACExpresso 0.590 A B OACExpresso 0.355 A B C B-EL1291d 1.110 A B11338 0.550 A B B11370 0.320  B C B11370 1.085 A B10243 0.540 A B Eclipse 0.310  B C B11313 1.060 A B11370 0.525 A B B10243 0.280  B C Eclipse 0.995 A Eclipse 0.520 A B B11338 0.235  B C ND081343 0.705 A ND081343 0.205  B ND081343 0.090   C 104 Table 15B. Tukey test for the trait of SAQ, for the three anthocyanins (delphindin-3-glucoside, petunidin-3-glucoside, and Malvidin-3-glucoside) concentrations for the 12 canned black bean genotypes in 2013 with two replications. a  Means followed with the same letter are not significantly different. bMean D, P, and M are the mean values for the anthocyanin delphindin-3-glucoside, petunidin-3-glucoside, and Malvidin-3-glucoside respectively.          Genotype Mean D mg/g  Genotype Mean P mg/g  Genotype Mean M mg/g   OACExpresso 0.030 A B-EL1291d 0.025 A B-EL1291d 0.025 A  B-EL1291d 0.025 A Eclipse 0.015 A B11356 0.015 A B Eclipse 0.020 A OACExpresso 0.015 A Zenith 0.015 A B B11338 0.020 A Zenith 0.012 A OACExpresso 0.01 A B B11370 0.020 A B11338 0.010 A B11313 0.01 A B B11313 0.015 A B11356 0.010 A Eclipse 0.007 A B Zenith 0.015 A B11370 0.010 A B11370 0.0065 A B ND081343 0.015 A B11313 0.008 A B11285 0.0065 A B B11285 0.015 A B11285 0.007 A B10243 0.0065 A B NDF09304 0.010 A B10243 0.007 A ND081343 0.0055 A B B11356 0.010 A ND081343 0.006 A B11338 0.0035 A B B10243 0.007 A NDF09304 0.003 A NDF09304 0.001  B 105 Table 15C. Tukey test for the trait of SAQ, for the three anthocyanins (delphindin-3-glucoside, petunidin-3-glucoside, and Malvidin-3-glucoside) concentrations for the 12 raw black bean genotypes in 2014 with two replications.                         a  Means followed with the same letter are not significantly different. bMean D, P, and M are the mean values for the anthocyanin delphindin-3-glucoside, petunidin-3-glucoside, and Malvidin-3-glucoside respectively.  Genotype Mean D mg/g  Genotype Mean P mg/g  Genotype Mean M mg/g  B11370 1.405 A B11370 0.815 A B11356 0.780 A NDF09304 1.170 A B-EL1291d 0.775 A B11370 0.635 A B-EL1291d 1.120 A B11356 0.700 A Zenith 0.550 A OACExpresso 1.110 A OACExpresso 0.655 A B11313 0.545 A B11356 1.095 A Zenith 0.620 A B-EL1291d 0.540 A B11338 1.085 A NDF09304 0.610 A B11285 0.450 A B11285 1.065 A Eclipse 0.605 A OACExpresso 0.435 A ND081343 0.985 A B11285 0.590 A Eclipse 0.425 A Eclipse 0.965 A B11313 0.565 A NDF09304 0.375 A B10243 0.930 A B11338 0.505 A B10243 0.315 A B11313 0.915 A B10243 0.475 A B11338 0.245 A Zenith 0.910 A ND081343 0.355 A ND081343 0.170 A 106 Table 15D. Tukey test for the trait of SAQ, for the three anthocyanins (delphindin-3-glucoside, petunidin-3-glucoside, and Malvidin-3-glucoside) concentrations for the 12 canned black bean genotypes in 2014 with two replications.                           a  Means followed with the same letter are not significantly different. bMean D, P, and M are the mean values for the anthocyanin delphindin-3-glucoside, petunidin-3-glucoside, and Malvidin-3-glucoside respectively.       Genotype Mean D mg/g  Genotype Mean P mg/g  Genotype Mean M mg/g  B10243 0.035 A OACExpresso 0.025 A B11356 0.040 A B11285 0.030 A B10243 0.025 A Zenith 0.030 A OACExpresso 0.030 A Zenith 0.025 A B-EL1291d 0.025 A B11356 0.030 A B11356 0.025 A B10243 0.025 A NDF09304 0.030 A Eclipse 0.025 A Eclipse 0.025 A B11338 0.025 A B-EL1291d 0.020 A B11313 0.025 A Zenith 0.025 A B11338 0.020 A B11285 0.020 A Eclipse 0.025 A B11285 0.020 A OACExpresso 0.020 A B-EL1291d 0.020 A NDF09304 0.020 A NDF09304 0.020 A B11313 0.015 A B11313 0.015 A B11338 0.015 A ND081343 0.010 A B11370 0.010 A B11370 0.010 A B11370 0.010 A ND081343 0.0035 A ND081343 0.002 A 107 traits. The first group is color and L*a*b* values from the previous chapter, and the second is TAC and the three SAQ of the three anthocyanins (delphinidin-3-glucoside, petunidin-3-glucoside, and malvidin-3-glucoside). Two moderate correlations were found between color rating and the two traits TAC and SAQ of malvinidin-3-glucoside for both years 2013 and 2014 (r=-0.58, p=0.0475, r=0.58, p=0.0496, r=0.63 p=0.0287, and r=0.69, p=0.0133) respectively (Table 17A-B). Such correlations suggest an important role of total anthocyanins and the SAQ of malvinidin-3-glucoside in the color concentration of the canned black beans. Also, there were some moderate correlations between the traits of L*a*b* values and the two traits of TAC and the SAQ of malvinidin-3-glucoside. However, the correlations between the above traits were varied across the two years. In 2013, moderate correlations were detected between TAC and the L* a* b* values (r=-0.62, p=0.0313, r=-0.53, p=0.0800, and r=-0.60 p=0.0394) respectively (Figure 22). The correlations between TAC and the a* b* values traits were stronger in 2014 (r=-0.71, p=0.0100 and r=-0.70, p=0.0115) respectively. However, the trait of L* was not correlated to TAC and that is due to the usage of two different machinery systems of obtaining the L*a*b* color values (differences explained earlier in chapter 3). In 2013, the trait of SAQ of malvinidin-3-glucoside was found to be strongly correlated to L* values trait and weakly correlated to a* and b* respectively (r=-0.51, p=0.0935, r=-0.35, p=0.0.2690, and r=-0.42, p=0.1755) (Figure 23). The correlations between the trait of SAQ of malvinidin-3-glucoside and the traits of a* and b* were found to be strong in 2014 (r=-0.69, p=0.0128 and r=-0.70, p=0.0111) respectively. The trait of L* was weakly correlated to SAQ of malvinidin-3-glucoside (r=0.41, p\0.1885). This also assures the impact of TAC and the trait of SAQ of malvinidin-3-glucoside specifically on the color retained by the evaluated canned black bean genotypes. 108 In a comparison between TAC and the trait of SAQ. In 2013, TAC was strongly correlated to SAQ of pentunidin-3-glucoside and malvinidin-3-glucoside, and weakly correlated to delphinidin-3-glucoside respectively (r=0.72, p=0.0089, r=0.70, p=0.0106, and r=0.43, and p=0.1617). However, in 2014 no correlations were observed between TAC and SAQ of delphinidin-3-glucoside and pentunidin-3-glucoside, while SAQ of malvinidin-3-glucoside was found to be moderately correlated to TAC. There were strong correlations between SAQ of delphinidin-3-glucoside and SAQ pentunidin-3-glucoside across the two years 2013 and 2014 (r=0.71, p=0.0094, r=0.86, p=0.0003) respectively. The trait of SAQ of pentunidin-3-glucoside was to be strongly correlated to SAQ of malvinidin-3-glucoside across the two years 2013 and 2014 (r=0.81, p=0.0016, r=0.80, p=0.0020) respectively. The clear differences observed of the trait of the three SAQ across the two years may suggest a location effect on the concentration of different anthocyanins that involved in black bean color synthesis. This result agrees with the study of Connor et al., (2002c), where significant variances were observed in anthocyanins concentration in blueberries between the same cultivars and different cultivars grown in the same location. Anthocyanins retention percentage for SAQ experiment showed the superiority of the genotype B10243 across all anthocyanins in 2014 (Table 16). Three different genotypes (OAC Expresso, B-EL1291, and B11356) were capable of retaining their anthocyanins during canning more than the other genotypes in 2013 for all three anthocyanin respectively (delphinidin-3-glucoside, petunidin-3-glucoside, and malvidin-3-glucoside). The anthocyanin malvidin-3-glucoside was found to be retained more than the other two anthocyanins among the genotypes tested. In a comparison between TAC and SAQ anthocyanins retention percentages, TAC has been observed to have higher color retention percentages than SAQ. This finding agreed with several studies results that 109 other anthocyanins are also involved in the black beans color formation (Zhang et al., 2014; Akond et al., 2011; Takeoka et al., 1997). It was expected that Zenith would have higher anthocyanins than Eclipse based on color differences (Kelly et al., 2015; Goffnett et al., 2016). This result was proven through the trained rating panel data of the seven traits, although TAC and SAQ data for the two genotypes were interesting. In both experiments (TAC and SAQ) anthocyanins concentrations were higher for Zenith than Eclipse in raw condition, and in some cases the difference was significant. For example, in TAC experiment anthocyanin concentration for Zenith (1.236mg/1g) was significantly higher than Eclipse (0.493mg/1g). However, both genotypes tend to have very similar anthocyanins concentrations in the canned product across all experiments in both years. This finding suggests factors other than anthocyanin content are important for dark black color in Zenith as compared to Eclipse. Anthocyanins degrade thermally during canning process which affects color quality, and may also affect nutritional attributes (Nakitto et al., 2015). Such a stress causes the loss of glycosyl moieties of anthocyanins by hydrolysis of the glycoside bond (Rodrigues and Narciso, 2012). This result in the development of different polyphenolic degradation products based on the stress degree and nature of the heating process (Patras et al., 2010). Nakitto et al., (2015) reported that chemical structure and some organic acids possess a serious impact on anthocyanin degradation mechanism. Markakis, Livingstone, and Fillers (1957) stated that anthocyanins may decompose because of heating into a chalcone structure, and they could even be transformed into a coumarin glucoside derivative with the B-ring being removed. Adams (1973) also mentioned that the aglycon-sugar bond is more unstable. The degradation of anthocyanins leads to color change and ultimately lead to brown products, especially in the presence of oxygen. As in other polyphenols, anthocyanins could be enzymatically degraded in the presence of PPO. 110 Lower temperature degree may act as an inactivating factor for PPO enzyme which may lead to more anthocyanins retention (Rodrigues and Narciso, 2012) 111 Table 16. Anthocyanins retention percentage for three anthocyanins concentration mg/1g (delphinidin-3-glucoside, pentuinidin-3-glucoside, and malvinidin-3-glucoside) that obtained from 12 black bean genotypes with two years (2013 and 2014).                                 Del-3-gluc 2013     Pent-3-gluc 2013    Malv-3-gluc 2013          Del-3-gluc 2014   Pent-3-gluc 2014  Malv-3-gluc 2014    Genotype      AR%               AR%  AR%   AR%  AR%  AR%     OACExpresso     2.61   2.54  1.83   3.76  5.26  7.94 B-EL1291d     2.25   3.36  2.08   2.82  3.39  4.44 ND081343     2.13   2.68  1.11   2.75  4.03  5.46 Eclipse      2.01   2.89  2.10   2.74  3.57  5.13 B11370      1.84   1.91  2.03   2.70  3.82  4.60 B11338      1.53   1.82  1.49   2.59  4.13  5.88 B11313      1.42   1.20  2.61   2.56  3.28  5.33 Zenith      1.23   1.33  1.90   2.30  3.96  6.12 B11285      1.16   1.02  2.53   1.79  5.58  4.63 NDF09304     0.72   0.47  1.84   1.64  2.66  4.59 B11356      0.72   1.05  2.65   1.02  0.99  1.18 B10243      0.58   1.20  1.96   0.71  1.23  1.58     aAR% is the percent of anthocyanins retention. b Mean Del-3-gluc, Pent-3-gluc, and Malv-3-gluc are the three anthocyanins delphindin-3-glucoside, petunidin-3-glucoside, and Malvidin-3-glucoside respectively.    112 Table 17A. Pearson Correlation for the eight traits (TAC, Delephinidin-3-glucoside, Pentuendin-3-glucoside, Malvinidin-3-glucoside, color, L*, a*, and b*) of the year 2013.                               TAC            Del-3-gluc           Pent-3-gluc    Malv-3-gluc        Color                L*              a*                b*   TAC                        .                 0.4312                 0.7151               0.7039                      0.5812             -0.6205            -0.5245                 -0.5995                                .                 0.1617                 0.0089               0.0106     0.0475              0.0313             0.0800                   0.0394 Del-3-gluc                                    .                       0.7121               0.3173                    -0.1013              0.0761             0.1647                   0.1236                                                       .                        0.0094              0.3150       0.7541              0.8141             0.6089                   0.7020                                                                                                                                                                                                                                                                                           Pent-3-gluc                                        .                   0.8049         0.2469             -0.2718            -0.1270                 -0.1576                                                                                                                                                                                           .                   0.0016         0.4392              0.3928              0.6941                  0.6247                                                                            Malv-3-gluc                .   0.5767           -0.5057         -0.3471          -0.4188                  .   0.0496             0.0935           0.2690           0.1755 Color                                            .                   -0.9661             -0.8937                -0.9480                                                                                                             .                    <.0001               <.0001                 <.0001                                                              *L                                                             .                       0.9324                 0.9407                                                                                                                         .                      <.0001                 <.0001                                                           *a                                        .                        0.9367                                                                                                                                   .                       <.0001                                                                        *b                                                  .                                                                                                                                                             .                                                                                   113 Table 17B. Pearson Correlation for the eight traits (TAC, Delephinidin-3-glucoside, Pentuendin-3-glucoside, Malvinidin-3-glucoside, color, L*, a*, and b*) of the year 2014.                              TAC            Del-3-gluc           Pent-3-gluc    Malv-3-gluc        Color                L*              a*                b*   TAC                        .               -0.1466                 0.2580               0.4197                      0.6283             0.2715            -0.7081                 -0.6985                                .                 0.6494                0.4182               0.1744     0.0287             0.3933             0.0100                  0.0115 Del-3-gluc                                    .                      0.8634               0.5864                     0.1192            -0.1118            -0.1293                 -0.1671                                                       .                       0.0003              0.0451       0.7121             0.7295             0.6888                   0.6037                                                                                                                                                                                                                                                                                           Pent-3-gluc                                        .                  0.7960         0.4091             0.0568            -0.4367                 -0.4678                                                                                                                                                                                           .                  0.0020         0.1867             0.8609             0.1558                   0.1251                                                                            Malv-3-gluc                .   0.6885            0.4076       -0.6910         -0.7009                  .   0.0133            0.1885         0.0128          0.0111 Color                                            .                    0.6639            -0.9665                -0.9596                                                                                                             .                    0.0186             <.0001                 <.0001                                                              *L                                                             .                    -0.4770                 -0.4473                                                                                                                         .                      0.1169                 0.1449                                                           *a                                        .                       0.9985                                                                                                                                   .                      <.0001                                                                        *b                                                  .                                                                                                                                                             .                114 01020300.000.050.100.150.20Lab Scan ValuesTotal Anthocyanin Content mg/gCorrelation Analysis 2014L14a14b14Linear (a14)0.000.020.040.060.080.100.120.140246Total Anthocyanin Content mg/gColor Ratings 1-5Correlation Analysis Canned 201301020300.000.050.100.15Lab Scan ValuesTotal Anthocyanin Content mg/gCorrelation AnalysisCanned2013L 13a13b13Linear (a13)0.000.050.100.150.20024Total Anthocyanin Content mg/gColor Ratings 1-5Correlation Analysis 2014Figure 22. Pearson correlations for the trait of total anthocyanin and the two traits of color and L*, a*, and b* color scans for the two years 2013 and 2014.                                   115 00.20.40.60.810246Malvidin Concmg/gColor Ratings 1-5Correlation Analysis Raw 20130.000.010.020.030246Malvidin Concmg/gColor Ratings 1-5Correlation Analysis Canned 201300.20.40.60.810246Malvidin Concmg/gColor Ratings 1-5Correlation Analysis Raw 20140.0000.0100.0200.0300.0400.0500246Malvidin Concmg/gColor Ratings 1-5Correlation Analysis Canned 2014Figure 23. Pearson correlations for the trait of color and malvidin-3-glucoside concentration for the raw and canned genotypes grown for two years 2013 and 2014.                               116 CONCLUSION AND RECOMMENDATION In this study, concentrations of three anthocyanins (delphinidin-3-glucoside, petunidin-3-glucoside, and malvidin-3-glucoside) were quantified using liquid chromatography-mass spectrometry. Total anthocyanins were also quantified using a UV assay. Both approaches were carried out for 12 genotypes of raw and canned black beans in two years (2013 and 2014). The result of TAC was greater than the SAQ in canned black beans. TAC experiment suggested that the genotype B-EL1291d as the highest in TAC across all the conditions and years. There were two genotypes (B11313 and ND081343) had constant AR across the two years, where the rest of the genotypes were varied for AR in TAC experiment. In the experiment of the three SAQ, the superiority of delphinidin-3-glucoside was conformed as the dominant anthocyanin with the highest concentration in black beans. Two moderate correlations were found between color rating and the two traits TAC and SAQ of malvinidin-3-glucoside for both years 2013 and 2014 (r=-0.58, p=0.0475, r=0.58, p=0.0496, r=0.63 p=0.0287, and r=0.69, p=0.0133) respectively. Such correlations suggest an important role of total anthocyanins and the SAQ of malvinidin-3-glucoside specifically in the color concentration of the canned black beans. The anthocyanin malvidin-3-glucoside was found to be retained more than the other two anthocyanins among all the sup-population genotypes. In both experiments (TAC and SAQ) anthocyanins concentrations were higher for Zenith than Eclipse in raw condition, and in some cases the difference was significant. However, both genotypes tend to have very similar anthocyanins concentrations in canned condition across all experiments including the two years. This finding suggests different factors other than anthocyanin content are important for dark black color in Zenith as compared to canned color of Eclipse.  117 APPENDIX 118 APPENDIX Table 18. A list of the 69 black bean genotypes that were involved in this study, and each genotype is combined with a picture which has been taken for the canned genotypes in the year of 2014. No. Genotype  1 B11360  2 Condor  3 Black Rhino  4 B-EL1291d  5 B-EL01318s            119 Table 18  6 B-EL01270s  7 B11313  8 B10238  9 19365-31  10 B11356  11 B10213      120 Table 18  12 SHINYCROW  13 ND071065  14 B-EL01270d  15 101s-2435  16 TARS-MST1  17 ND081195 - 18 B10210    121 Table 18  19 B11312  20 SB-DT1  21 Domino  22 Zorro  23 B-EL01260d  24 ND081343     122 Table 18  25 B-EL01268s  26 Black Magic  27 BK 11-5  28 B11363  29 B11329  30 ND071089      123 Table 18  31 ND071244  32 ND071257  33 DPC-4  34 Harblack Opaque  35 B09119  36 Eclipse      124 Table 18  37 NE14-12-28  38 B11259  39 T-39  40 B-EL0309  41 B-EL01294s  42 ND081256     125 Table 18  43 B11588  44 B11310  45 BK 11-8  46 07055-4  47 B10244 (Zenith)  48 B11364       126 Table 18  49 B11370  50 B10243  51 BK 11-1  52 B09165  53 B-EL01335s  54 NDFD9302        127 Table 18  55 B11338  56 NDF09304  57 OAC Jet  58 OACExpresso  59 B-EL01291s  60 ND081339       128 Table 18  61 BK9-2  62 B11285  63 Aifiwuriti  64 B-EL01308s  65 B11311  66 ND081111    129 Table 18  67 B11322  68 B11309  69 ICB-3                 130 Table 19A. Means and standard deviations for all the major and sub branches (including genotypes averages) that obtained from the neighbor joining tree for the seven traits (seed yield, seed weight, appearance, color, L*, a*, and b*) of 69 black bean genotypes that obtained from the year 2013.  NJ Group Genotypes 2013 Yield Seed Weight Appearance Color L* a* b*  A1-1 B11360 2335 18.3 2.25 3.50 17.55 8.21 5.63  DPC-4 1566 18.0 3.00 3.50 15.08 7.87 4.02  B11370 2149 18.6 2.75 1.75 18.57 8.59 6.50  HarblackOpaque 2221 18.2 2.00 3.50 16.33 9.05 5.11  OAC Jet 2133 19.2 2.25 2.75 16.80 8.21 6.55  Average 2081 18.4 2.45 3.00 16.87 8.39 5.56  Standard Deviation 267 0.4 0.37 0.69 1.17 0.40 0.94 A1-2 Condor 2387 17.8 3.25 3.50 15.16 8.32 4.48  B11311 1819 20.3 3.50 3.25 17.35 9.13 4.90  B10210 3136 19.8 2.75 1.50 19.04 10.51 8.74  B11313 2541 17.3 2.75 4.00 14.65 7.59 3.14  B11356 2.085 13.6 2.50 3.50 14.69 8.25 5.51  Zenith 2840 18.5 3.25 4.75 14.77 7.50 2.95  B11363 2735 17.0 3.00 3.25 14.70 8.44 4.08  B11312 2441 19.7 3.50 3.25 15.89 8.56 4.76  B11588 2431 21.6 1.75 1.50 20.20 10.25 8.31  B11310 2353 20.7 3.00 3.25 15.89 9.17 5.16  Zorro 2334 18.0 2.75 3.00 16.94 9.14 5.20  B10213 2313 17.6 2.25 2.00 19.29 9.33 7.14  B11309 2178 18.7 1.50 3.75 14.39 8.57 4.55  B11364 2164 22.1 3.00 4.25 14.71 7.67 3.56  B11322 2012 13.6 1.75 2.75 15.91 7.99 5.03  07055-4 1166 20.2 1.75 1.50 20.36 9.93 8.64  Average 2323 18.5 2.64 3.06 16.50 8.77 5.38  Standard Deviation 440 2.4 0.64 0.96 2.05 0.89 1.80 A2-1 ND071257 2583 20.6 1.50 1.25 19.93 10.46 8.88  19365-31 2275 21.4 1.50 2.25 18.05 8.32 6.70  AifiWuriti 2147 19.7 2.50 2.00 16.98 8.41 6.19  ND071244 2134 19.9 1.50 1.75 19.13 9.70 7.75  BK11-5 2108 21.1 1.50 2.00 17.83 8.59 6.98  BK11-1 2079 18.9 2.00 2.00 17.41 8.45 6.71  SB-DT1 2026 18.7 1.50 1 19.87 9.72 8.85  NE14-12- 2006 20.6 2.75 3.50 14.31 7.43 4.82  ND071065 2005 20.0 1.50 2.50 18.35 8.83 7.30  BK11-8 1797 18.8 3.00 2.25 18.27 9.04 7.41  TARS-MST 1724 18.6 1.50 2.50 18.36 8.23 5.79  Black Rhino 1669 22.4 2.00 1.75 17.21 8.59 6.34                  131 Table 19A   OAC Expresso 1652 19.1 1.50 3.00 17.81 8.97 7.13  ICB-3 1363 22.4 1.75 1.75 17.99 9.35 7.56  Average 1969 20.1 1.86 2.11 17.96 8.86 7.03  Standard Deviation 294 1.3 0.51 0.62 1.32 0.73 1.05 A2-2 B-EL01268s 2084 17.8 2.61 3.21 16.66 8.08 6.25  B-EL01270s 1992 20.6 3.11 3.68 14.69 8.20 4.80  B-EL01291s 1797 18.5 2.93 3.46 15.79 7.45 5.21  B-EL01294s 1747 18.8 3.07 4.46 13.98 8.02 3.89  B-EL01308s 2451 20.3 3.46 4.36 14.10 7.31 3.74  B-EL01318s 1179 17.8 2.46 3.21 15.97 8.37 5.82  B-EL01335s 2225 19.7 2.61 4.04 15.42 7.40 4.05  SHINYCROW 1734 22.0 3.04 3.82 14.45 7.52 4.23  B-EL1291d 2699 19.0 2.00 2.25 17.28 8.81 6.66  B-EL0309 1952 18.2 2.00 3.75 14.34 7.84 3.65  B-EL01260d 970 18.4 2.75 3.50 14.16 7.36 3.79  Average 1894 19.2 2.73 3.61 15.17 7.85 4.73  Standard Deviation 480 1.3 0.44 0.58 1.08 0.47 1.04 A2-3 Black Magic 2619 18.7 2.50 3.00 16.56 8.61 6.73  BK9-2 2452 19.7 2.25 1.50 20.02 9.76 8.59  T-39 2194 17.6 2.50 3.25 16.90 8.14 6.01  Domino 1991 20.8 2.00 2.50 16.65 8.96 5.26  Eclipse 1757 17.8 3.00 2.50 18.43 8.94 7.38  ND081295 1290 19.8 2.50 1.50 20.09 9.33 7.63  Average 2050 19.0 2.46 2.38 18.11 8.96 6.93  Standard Deviation 441 1.1 0.30 0.67 1.51 0.51 1.09 A2-4 ND071089 2282 19.1 1.75 1.50 19.08 8.98 7.73  NDF09302 2123 18.2 2.75 1.75 19.43 8.81 7.66  NDF09304 1833 17.6 2.50 1.75 18.36 8.36 7.01  Average 2079 18.3 2.33 1.67 18.96 8.72 7.47  Standard Deviation 185 0.6 0.42 0.12 0.45 0.26 0.32 B ND081343 1684 18.6 2.00 1.00 22.53 10.65 10.29 C B11285 2212 16.7 2.75 1.00 22.41 10.97 9.41  B11259 2046 16.4 1.75 1.50 20.75 9.07 7.68  ND081256 1541 24.1 2.50 1.75 20.00 10.09 7.83  B11338 1383 16.1 2.50 1.00 20.19 10.32 8.83  Average 1795 18.3 2.38 1.31 20.84 10.11 8.44  Standard Deviation 343 3.4 0.38 0.32 0.95 0.68 0.71 D ND081111 1476 18.0 2.00 1.75 19.46 9.41 8.08 132 Table 19A         B10243 1357 20.5 3.00 1.75 18.26 8.63 6.48  Average 1417 19.2 2.50 1.75 18.86 9.02 7.28  Standard Deviation 59 1.3 0.50 0.00 0.60 0.39 0.80                          133 Table 19B. Means and standard deviations for all the major and sub branches (including genotypes averages) that obtained from the neighbor joining tree for the seven traits (seed yield, seed weight, appearance, color, L*, a*, and b*) of 69 black bean genotypes that obtained from the year 2014.  NJ Group Genotypes 2014 Yield Seed Weight Appearance Color L* a* b*  A1-1 B11360 3727 20.5 2.50 2.50 20.59 2.60 3.03  OACJet 3403 20.7 4.00 4.00 18.42 3.71 4.04  Harblack 2910 18.9 3.50 3.50 18.40 3.53 3.58  B11370 2625 18.0 3.25 3.25 16.33 3.55 3.83  DPC-4 1209 20.5 3.75 3.75 19.80 3.11 3.44  Average 2775 19.7 3.40 3.40 18.71 3.30 3.58  Standard Deviation 871 1.1 0.51 0.51 1.45 0.40 0.35 A1-2 B10210 4716 27.4 4.50 4.50 16.72 4.91 4.74  B11363 3861 21.8 4.25 4.25 20.03 2.05 2.53  B11364 3568 21.8 3.50 3.50 21.48 2.63 2.95  B11329 3541 18.2 3.75 3.75 21.69 1.86 2.47  B11313 3503 19.3 3.75 3.75 26.04 1.97 2.58  101s-243 3331 21.3 2.50 2.50 16.29 4.98 4.83  B11309 3221 18.3 3.25 3.25 21.44 2.30 2.84  B11356 3139 19.2 3.25 3.25 20.97 2.70 3.08  Zenith 3047 22.3 4.00 4.00 24.87 1.94 2.59  B09165 2950 22.6 4.25 4.25 17.83 3.47 3.63  B11588 2891 24.9 3.00 3.00 18.09 4.59 6.37  B11310 2799 22.2 4.00 4.00 19.99 3.10 3.38  B09119 2792 20.5 4.25 4.25 18.43 3.46 3.76  Condor 2624 21.3 3.75 3.75 25.03 2.45 3.08  B11322 2560 18.1 2.75 2.75 19.49 3.15 3.54  B11311 2476 18.1 3.75 3.75 18.35 3.13 3.48  B10213 2430 19.2 4.00 4.00 17.42 3.97 4.09  B10238 2409 18.8 4.25 4.25 17.64 3.75 3.81  Zorro 2324 21.1 3.50 3.50 18.00 3.73 3.86  07055-4 1499 23.3 3.25 3.25 18.34 5.15 5.12  Average 2984 21.0 3.68 3.68 19.91 3.26 3.64  Standard Deviation 664 2.4 0.53 0.53 2.74 1.03 0.98 A2-1 BlackRhi 3856 24.1 3.00 3.00 17.70 3.58 3.71  ND071257 3534 21.8 3.50 3.50 16.72 4.68 4.44  NE14-12- 3153 24.2 3.25 3.25 23.59 2.72 3.02  ND071244 3092 21.3 3.25 3.25 17.18 4.55 4.49  SB-DT1 2814 26.3 2.50 2.50 17.20 5.59 5.41  TARS-MST 2719 21.8 2.00 2.00 19.26 3.69 3.95  BK11-5 2713 22.1 3.50 3.50 18.57 3.61 3.69  OACExpre 2658 20.0 3.50 3.50 18.84 3.90 3.99  BK11-1 2219 21.0 3.75 3.75 19.29 3.57 3.75  ND071065 1884 21.3 3.00 3.00 16.23 4.67 4.48 134 Table 19B         BK11-8 1828 22.0 4.00 4.00 17.77 3.97 3.97  AifiWuri 1750 19.6 2.75 2.75 17.96 3.80 4.08  19365-31 1416 23.8 3.50 3.50 17.57 3.86 3.95  ICB-3 1347 21.3 2.50 2.50 16.87 4.56 4.53  Average 2499 22.2 3.14 3.14 18.20 4.05 4.10  Standard Deviation 750 1.7 0.53 0.53 1.75 0.67 0.53 A2-2 B-EL01260d 2420 20.7 4.06 4.72 28.30 1.44 2.55  B-EL01268s 1839 18.7 3.72 4.22 25.37 1.91 2.71  B-EL01270d 3636 22.7 3.34 3.44 20.59 3.15 3.53  B-EL01270s 2874 26.2 3.94 4.66 26.56 1.54 2.36  B-EL01291s 2686 19.2 3.69 4.31 26.94 1.98 2.82  B-EL01294s 2582 19.3 3.94 4.72 27.31 1.85 2.78  B-EL01308s 2489 20.9 3.94 4.28 24.92 1.97 2.76  B-EL01318s 2430 18.8 4.06 4.06 24.77 2.26 2.96  B-EL01335s 2858 22.7 3.34 4.53 24.95 2.06 2.79  B-EL0309 2027 21.4 3.91 4.59 26.73 1.93 2.85  B-EL1291d 2623 18.6 3.63 3.75 18.96 2.76 3.19  SHINYCROW 2316 22.1 3.81 4.13 22.72 2.44 3.01  Average 2565 20.9 3.78 4.28 24.84 2.11 2.86  Standard Deviation 434 2.2 0.24 0.38 2.69 0.46 0.29 A2-3 BlackMag 3425 19.4 3.50 3.50 22.81 2.56 3.28  T-39 3062 20.0 3.50 3.50 19.95 2.96 3.27  BK9-2 2777 20.0 3.00 3.00 16.39 5.56 5.22  Eclipse 2773 19.7 2.75 2.75 16.43 4.30 4.26  ND081295 2583 25.5 2.25 2.25 17.65 4.47 4.50  Domino 1844 20.2 3.75 3.75 17.13 4.05 4.20  Average 2744 20.8 3.13 3.13 18.39 3.98 4.12  Standard Deviation 483 2.1 0.52 0.52 2.31 0.99 0.68 A2-4 NDF09302 2619 18.5 3.25 3.25 18.51 4.59 4.66  NDF09304 2601 19.8 3.75 3.75 18.95 3.58 3.72  Average 2610 19.1 3.50 3.50 18.73 4.09 4.19  Standard Deviation 8 0.7 0.25 0.25 0.22 0.505 0.47 B ND081343 2789 20.4 3.25 3.25 19.65 7.79 7.30 C B11285 3430 19.2 3.75 3.75 18.16 6.74 6.37  B11259 3345 19.1 3.75 3.75 17.25 5.17 5.07  B11338 3329 17.3 2.75 2.75 18.27 7.12 6.62  ND081256 2380 19.8 4.25 4.25 17.04 4.15 4.10  Average 3121 18.8 3.63 3.63 17.68 5.80 5.54  Standard Deviation 429 1.0 0.54 0.54 0.54 1.20 1.02 D ND081339 3117 26.4 3.25 3.25 18.53 5.43 5.44 135 Table 19B   ND081111 3009 22.3 3.00 3.00 18.72 3.80 4.12  B10243 2567 21.5 3.25 3.25 18.47 3.41 3.54  Average 2898 23.4 3.17 3.17 18.57 4.21 4.37  Standard Deviation 238 2.2 0.12 0.12 0.11 0.87 0.80                         136 Figure 24. QQ plot for the trait of seed yield values averaged across 2013 of 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 3 principal components to account for population structure.                137 Figure 25. QQ plot for the trait of seed weight values averaged across 2013 of 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 7 principal components to account for population structure.               138 Figure 26. QQ plot for the trait of appearance ratings averaged across 2013 of 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 2 principal components to account for population structure.              139 Figure 27. QQ plot for the trait of color ratings averaged across 2013 of 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix.               140 Figure 28. QQ plot for the trait of L* values averaged across 2013 of 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 7 principal components to account for population structure.              141 Figure 29. QQ plot for the trait of a* values averaged across 2013 of 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 3 principal components to account for population structure.              142 Figure 30. QQ plot for the trait of seed weight values averaged across 2013 of 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 3 principal components to account for population structure.              143 Figure 31. QQ plot for the trait of seed yield values averaged across 2014 of 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix.                144 Figure 32. QQ plot for the trait of seed weight values averaged across 2014 of 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 7 principal components to account for population structure.              145 Figure 33. QQ plot for the trait of appearance ratings averaged across 2014 of 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 2 principal components to account for population structure.              146 Figure 34. QQ plot for the trait of color ratings averaged across 2014 of 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 7 principal components to account for population structure.              147 Figure 35. QQ plot for the trait of L* values averaged across 2014 of 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 7 principal components to account for population structure.              148 Figure 36. QQ plot for the trait of a* values averaged across 2014 of 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix and 5 principal components to account for population structure.              149 Figure 37. QQ plot for the trait of b* values averaged across 2014 of 61 black bean genotypes with 2800 SNP (>0.02 minor allele frequency) and mapping conducted with multiple linear model (MLM) using a kinship matrix.    150 BIBLIOGRAPHY151 BIBLIOGRAPHY Adams, J. B. (1973). Thermal degradation of anthocyanin with partiular reference on 3 glucosides of º C. Journal of the Science of Food and Agriculture, 24, 747-762. Acevedo, A., Cruz, D., Hilbert, G., Rivière, C., Mengin, V., Ollat, N., Richard, T. (2012). Analytica Chimica Acta Anthocyanin identification and composition of wild Vitis spp . accessions by using LC  MS and LC  NMR. Analytica Chimica Acta, 732, 145152.  Acosta-Gallegos, J., Kelly, J., & Gepts, P. (2007). Prebreeding in common bean and use of genetic diversity from wild germplasm. Crop Science, 47(Supplement_3), S44S59. Akibode, S. and Maredia, M. (2011). Global and regional trends in production, trade and consumption of food legume crops. [online] East Lansing: Michigan State University, p.42. Available at: http://impact.cgiar.org/sites/default/files/images/Legumetrendsv2.pdf [Accessed 23 Mar. 2016]. Akond, M., Laila, K., Janelle, B., Khwaja, H. (2011). Anthocyanin, Total Polyphenols and Antioxidant Activity of Common Bean. Academic Journals Inc, 5, 385394. Aparicio-Fernandez, X., Yousef, G. G., Loarca-Pina, G., De Mejia, E., & Lila, M. A. (2005). Characterization of Polyphenolics in the Seed Coat of Black Jamapa Bean (Phaseolus vulgaris L.). Agric Food Chem Balasubramanian, P., Slinkard, A., Tyler, R., & Vandenberg, A. (2000). A modified laboratory canning protocol for quality evaluation of dry bean ( Phaseolus vulgaris L). Sci Food Agric, 738, 732738. Bauer, U.E., Briss, P. A., Goodman, R. A., Bowman, B.A. (2011). The Health of Americans 1 Prevention of chronic disease in the 21st century: elimination of the leading preventable causes of premature death and disability in the USA. The Lancet, 384(9937), 4552.  Beebe A, S. E., Rao A, I. M., Jyostna Devi, M. B., & Polania, J. A. (2014). Common beans, biodiversity, and multiple stresses: challenges of drought resistance in tropical soils. Crop & Pasture Science, 65, 667675. Beebe, S., Skroch, P. W., Tohme, J., Duque, M. C., Pedraza, F., & Nienhuis, J. (2014). Structure of Genetic Diversity among Common Bean Landraces of Middle American Origin Based on Correspondence Analysis of RAPD. Crop Sci, 40, 264273. Beninger, C. W., Hosfield, G. L., Bassett, M. J., & Owens, S. (2000). Chemical and morphological expression of the B and Asp seedcoat genes in Phaseolus vulgaris. Journal of the American Society for Horticultural Science, 125(1), 5258. 152  Bewley, J. D., Black, M., & Halmer, P. (2006). The Encyclopedia of Seeds: Science, Technology and Uses. CABI. Cromwell.  Bitocchi, E., Bellucci, E., Giardini, A., Rau, D., Rodriguez, M., Biagetti, E., Papa, R. (2012). Molecular analysis of the parallel domestication of the common bean (Phaseolus vulgaris L.) in Mesoamerica and the Andes. Pnas, 109(14), 788796. Boateng, J., Verghese, M., Walker, L. T., & Ogutu, S. (2008). Effect of processing on antioxidant contents in selected dry beans (Phaseolus spp. L.). Food Science and Technology, 41, 15411547. Bowen-forbes, C. S., Zhang, Y., & Nair, M. G. (2010). Journal of Food Composition and Analysis Anthocyanin content , antioxidant , anti-inflammatory and anticancer properties of blackberry and raspberry fruits. Food Composition and Analysis, 23, 554560.  Box GE, Cox DR (1964). An analysis of transformations. J R Stat Soc Series B (Methodological), 26, 211252. Bradbury, PJ., Zhang, Z., Kroon DE., Casstevens, TM., Ramdoss, Y., Buck-ler, ES. (2007). TASSEL: software for association mapping of complex traits in diverse samples. Bioinformatics, 23, 26332635. Broughton, W. J., Hern, G., Blair, M., Beebe, S., Gepts, P., & Vanderleyden, J. (2003). Beans (Phaseolus spp.)  model food legumes. Plant and Soil, 252, 55128. Bushey SM, Hosfield GL, Beninger CW. (2000). Water uptake and its relationship to pigment leaching in black beans (Phaseolus vulgaris L.). Bean Improv Coop, 43, 104105. Chandra, A., Nair, M. G., & Iezzoni, A. (1992). Evaluation and Characterization of the Anthocyanin Pigments in Tart Cherries (Prunus Cerasus L.). Agrlc. Food Chem, 40, 967969. Cichy, K., Fernandez, A., Kilian, A., Kelly, J., Galeano, C., Shaw, S., Troxtell, E. (2014). QTL analysis of canning quality and color retention in black beans (Phaseolus vulgaris L.). Molecular Breeding, 33(1), 139154.  Cichy, K. A., G. V. Caldas, S. S. Snapp, and M. W. Blair. (2009). QTL Analysis of Seed Iron, Zinc, and Phosphorus Levels in an Andean Bean Population All rights reserved. Crop Sci, 49, 1742-1750. Creamer, B. (2014). Major constraints and trends for common bean production and establishing priorities for future research. Agronomía Colombiana, 32(3), 423431.  Connor, A. M., Luby, J. J., Tong, C. B. S., Finn, C. E., & Hancock, J. F. (2002). Genotypic and Environmental Variation in Antioxidant Activity , Total Phenolic Content , and 153 Anthocyanin Content among Blueberry Cultivars. J Amer Soc Hort Sci, 127(1), 8997. De Faria, L. C., Melo, P. G. S., Pereira, H. S., Wendland, A., Borges, S. F., Filho, I. A. P.. . Melo, L. C. (2014). Genetic progress during 22 years of black bean improvement. Euphytica, 199(3), 261272. De Lange, A., and M. Labuschagne. (2001). Multivariate assess- ment of canning quality, chemical characteristics and yield of small white canning beans (Phaseolus vulgaris L) in South Africa. J. Sci. Food Agric, 81, 3035. Díaz, A. M., Caldas, G. V., & Blair, M. W. (2010). Concentrations of condensed tannins and anthocyanins in common bean seed coats. Food Research International, 43(2), 595601.  Dong, M., He, X., & Liu, R. H. (2007). Phytochemicals of black bean seed coats: Isolation, structure elucidation, and their antiproliferative and antioxidative activities. Journal of Agricultural and Food Chemistry, 55(15), 60446051. Dzomba, P., Togarepi, E., & Mupa, M. (2013). Anthocyanin content and antioxidant activities of common bean species ( Phaseolus vulgaris L.) grown in Mashonaland Central Zimbabwe. Afr. J. Agric. Res, 8(25), 33303333.  Ellis, R. H. (2007). The encyclopaedia of seeds: science, technology and uses. Wallingford, UK: CABI. Fageria, N. K., & Santos, A. B. (2008). Yield Physiology of Dry Bean, Journal of plant nutrition, 31, 983100. Gebrelibanos, M., Tesfaye, D., Raghavendra, Y., & Sintayeyu, B. (2013). Nutritional and health implications of legumes. International Journal of Pharmaceutical Sciences and Research IJPSR, 4(4), 12691279. (Phaseolus vulgaris L.) in different weed control methods. Journal of Elementology, 20(2), 293303. Goffnett, A., Sprague, C., Mendoza, F., Cichy, C. (2016). Preharvest Herbicide Treatments Affect Black Bean Desiccation,  Yield, and Canned Bean Color. Crop Science Society of America, 56,19621969. Gould, K., Davies, K. M., & Winefield, C. (2009). Anthocyanins: Biosynthesis, functions, and applications. New York, NY: Springer. preservation of date. Journal of Food Science and Technology, 52(April), 18721880. Hutchins, A. M., Winham, D. M., & Thompson, S. V. (2012). Phaseolus beans: impact on 154 glycaemic response and chronic disease risk in human subjects. British Journal of Nutrition, 108(S1), S52S65. Jiang, L., Wang, J., Li, Y., Wang, Z., Liang, J., Wang, R., Zhang, M. (2014). Effects of ultrasound on the structure and physical properties of black bean protein isolates. FRIN, 62, 595601.  Joseph, O., Phelomene, M., Helene, N., Valens, H., Patrick, O. M., Thavarajah, D., & Thavarajah, P. (2014). Phenolic Compound Profiles of Two Common Beans Consumed by Rwandans. American Journal of Plant Sciences, 5, 29432947. Black Bean. Journal of Plant Registrations, 9(1), 1520. Khanal, R., Burt, A. J., Woodrow, L., Balasubramanian, P., & Navabi, A. (2015). Genotypic Association of Parameters Commonly Used to Predict Canning Quality of Dry Bean. CROP SCIENCE, 54, 25642573. Kim, S., Plagnol, V., Hu, T. T., Toomajian, C., Clark, R. M., Ossowski, S., Nordborg, M. (2007). Recombination and linkage disequilibrium in Arabidopsis thaliana. Nature Genetics, 39(9), 11511155.   content of dry and processed beans. Food Chemistry, 80(2), 231235. La Cruz, A. A. De, Hilbert, G., Mengin, V., Rivière, C., Ollat, N., Vitrac, C., Richard, T. (2013). Anthocyanin phytochemical profiles and anti-oxidant activities of vitis candicans and vitis doaniana. Phytochemical Analysis, 24(5), 446452.  Lee, I. H., Hung, Y. H., & Chou, C. C. (2008). Solid-state fermentation with fungi to enhance the antioxidative activity, total phenolic and anthocyanin contents of black bean. International Journal of Food Microbiology, 121(2), 150156.  Ma, F., Cholewa, E., Mohamed, T., Peterson, C. A., & Gijzen, M. (2004). Cracks in the palisade cuticle of soybean seed coats correlate with their permeability to water. Annals of Botany, 94(2), 213228.  Mamidi, S., Rossi, M., Annam, D., Moghaddam, S., Lee, R., Papa, R., and McClean, P. (2011). Investigation of the domestication of common bean (Phaseolus vulgaris L.) using multilocus sequence data. Journal compilation, 38, 953967. Markakis, P., Livingstone, G. E., & Fillers, G. R. (1957). Quantitativeaspects of strawberry pigment degradation. Food Research, 22, 117130. Marles, M. A. S., Balasubramanian, P., & Bett, K. E. (2010). Differential Accumulation of Polyphenolics in Black Bean Genotypes Grown in Four Environments. J. Agric. Food 155 Chem, 58, 70017006.  Melotto, M., Monteiro-Vitorello, C. B., Bruschi, A. G., & Camargo, L. E. A. (2005). Comparative bioinformatic analysis of genes expressed in common bean (Phaseolus vulgaris L.) seed. Genome Jun, 48(3), 562570. Mendoza, F. A., Cichy, K., Lu, R., & Kelly, J. D. (2014). Evaluation of Canning Quality Traits in Black Beans ( Phaseolus vulgaris L.) by Visible / Near-Infrared Spectroscopy. Food Bioprocess Technol, 7, 26662678. Mendoza, F. A., Kelly, J. D., & Cichy, K. (2016). Automated prediction of sensory scores for color and appearance in canned black beans (Phaseolus vulgaris L.) using machine vision. International Journal of Food Properties, 2912, 15322386. Merwe, D. Van Der, Osthoff, G., & Pretorius, A. J. (2006). Comparison of the canning quality of small white beans ( Phaseolus vulgaris L.) canned in tomato sauce by a small-scale and an industrial method. Journal of the Science of Food and Agriculture, 1056(August 2005), 10461056. Miklas, PN., Kelly, JD., Beebe, SE., Blair, MW. (2006a). Common bean breeding for resistance against biotic and abiotic stresses: from classical to MAS breeding. Euphytica 147, 105131. Miklas, PN., Hu, J., Grunwald, NJ., Larsen, KM. (2006b). Potential application of TRAP (targeted region amplified poly-483 morphism) markers for mapping and tagging disease resistance traits in common bean. Crop Sci, 46, 910916. Mojica, L., Meyer, A., Berhow, M. A., & de Mejía, E. G. (2015). Bean cultivars (Phaseolus --glucosidase while diverse phenolic composition and concentration. Food Research International, 69, 3848. Moreno-jiménez, M. R., Cervantes-cardoza, V., Gallegos-infante, J. A., González-laredo, R. F., Estrella, I., García-gasca, T. D. J., Rocha-guzmán, N. E. (2015). Phenolic composition changes of processed common beans: their antioxidant and anti-inflammatory effects in intestinal cancer cells. FRIN, 76, 7985. Myles, S., Peiffer, J., Brown, P. J., Ersoz, E. S., Zhang, Z., Costich, D. E., & Buckler, E. S. (2009). Association mapping: critical considerations shift from genotyping to experimental design. The Plant Cell, 21(8), 21942202.  Nakitto, A. M., Muyonga, J. H., & Nakimbugwe, D. (2015). Effects of combined traditional processing methods on the nutritional quality of beans. Food Science & Nutrition, 131, 233241.  Nyau, V., Prakash, S., Rodrigues, J., & Farrant, J. (2015). HPLC-PDA-ESI-MS Identification of 156 Polyphenolic Phytochemicals in Different Market Classes of Common Beans (Phaseolus vulgaris L.). IJBCRR, 8(4), 111.  Oseguera-toledo, M. E., Gonzalez, E., Mejia, D., Dia, V. P., & Amaya-llano, S. L. (2011). Common bean (Phaseolus vulgaris L.) hydrolysates inhibit inflammation in LPS-induced macrophages through suppression of NF- j B pathways. Food Chemistry, 127(3), 11751185. Trust, D. (2013). Use of Wild Relatives and Closely Related Species to Adapt Common Bean to Climate Change. Agronomy (Basel), 3, 433461. Rocha-Guzmán, N. E., Herzog, A., González-Laredo, R. F., Ibarra-Pérez, F. J., Zambrano-Galván, G., & Gallegos-Infante, J. A. (2007). Antioxidant and antimutagenic activity of phenolic compounds in three different colour groups of common bean cultivars (Phaseolus vulgaris L.). Food Chemistry, 103(2), 521527. Rockenbach, I., Jungfer, E., Ritter, C., Santiago-schübel, B., Thiele, B., Fett, R., & Galensa, R. (2012). Characterization of fl avan-3-ols in seeds of grape pomace by CE , HPLC-DAD-MS and LC-ESI-FTICR-MS. FRIN, 48(2), 848855. Rodrigues, S., & Fernandes, F. A. N. (2012). Contemporary Food Engineering: Advances in Fruit Processing Technologies(1). London, US: CRC Press. Patras, A., Brunton, N. P., O'Donnell, C., & Tiwari, B. K. (2010). Effect of thermal processing on anthocyanin stability in foods; mechanisms and kinetics of degradation. Trends in Food Science & Technology, 21(1), 311. Pritchard, JK., Stephens, M., Donnelly, P. (2000). Inference of population structure using multilocus genotype data. Genetics 155, 945959. Trust, D. (2013). Use of Wild Relatives and Closely Related Species to Adapt Common Bean to Climate Change, Agronomy, 3(2), 433461. Sancho, R. A. S., Pavan, V., & Pastore, G. M. (2015). Effect of in vitro digestion on bioactive compounds and antioxidant activity of common bean seed coats. FRIN, 76, 7478. Schmutz, J., McClean, PE., Mamidi, S., Wu, GA., Cannon, SB., Grimwood, J., Jenkins, J., Shu, S., Song, Q., Chavarro, C., Torres-Torres, M., Gef- froy, V., Moghaddam, SM., Gao, D., Abernathy, B., Barry, K., Blair, M., Brick, MA., Chovatia, M., Gepts, P., Goodstein, DM., Gonzales, M., Hellsten, U., Hyten, DL., Jia, G., Kelly, JD., Kudrna, D., Lee, R., Richard, MMS., Miklas, PN., Osorno, JM., Rodrigues, J., Thareau, V., Urrea, CA., Wang, M., Yu, Y., Zhang, M., Wing, RA., Cregan, PB., Rokhsar, DS., Jackson, SA. (2014). A reference genome for com- mon bean and genome-wide analysis of dual domestications. Nat Genet, 46, 707713. 157 Shi, C., Yu, K., Xie, W., & Deidre, P. N. M. (2012). Development of candidate gene markers associated to common bacterial blight resistance in common bean. Theoretical and Applied Genetics, 125(7), 15251537. Singh, G. (2008). Seed Quality. Delhi, IND: Gene-Tech Books. Sivolap, Y. M. (2013). Molecular markers and plant breeding. Cytology and Genetics, 47(3), 188195. Takeoka, G. R., Dao, L. T., Full, G. H., Wong, R. Y., Harden, L. A., Edwards, R. H., Berrios, J. (1997). Characterization of black bean (Phaseolus vulgarisL.) anthocyanins. Journal of Agricultural and Food Chemistry, 45(9), 33953400. Thompson, M. D., Brick, M. A., McGinley, J. N., & Thompson, H. J. (2009). Chemical Composition and Mammary Cancer Inhibitory Activity of Dry Bean. Crop Science, 49, 179186. Tuberosa, R., Graner, A., & Frison, E. (2014). Genomics of plant genetic resources : Volume 2. crop productivity, food security and nutritional quality. Dordrecht, Netherlands: Springer. Turner SD (2014) qqman: an R package for visualizing GWAS results using Q-Q and manhattan plots. bioRxiv (2014): 005165. USDA Laboratory. (2014). Soybean Genomics and Improvement www.ars.usda.gov/main/site_main.htm?modecode=80-42-05-70 (accessed 5 April. 2016).  USDA-NASS (National Agricultural Statistics Service) (2015) 2014 State agriculture overview. http://nass.usda.gov/Quick_Stats/Ag_Overview/stateOverview.php?state=MICHIGAN. Accessed April 8, 2015. Wang, X., Hansen, C., & Allen, K. (2013). Identification of Anthocyanins Isolated from Black Bean Canning Wastewater by Macroporous Resin Using Optimized Conditions. Food and Nutrition Sciences, 4, 174181. Wright, E. M., & Kelly, J. D. (2011). Mapping QTL for seed yield and canning quality following processing of black bean (Phaseolus vulgaris L.). Euphytica, 179(3), 471484.  Wu, X., Beecher, G. R., Holden, J. M., Haytowitz, D. B., Gebhardt, S. E., & Prior, R. L. (2006). Concentrations of Anthocyanins in Common Foods in the United States and Estimation of Normal Consumption. J.Agric.FoodChem Xu, B., Huang, Y., & Watson, M. D. (2001). Cotton Color Distributions in the CIE L*a*b* System. Textile Rex J, 71(11), 10101015. Zhang, C., Monk, J. M., Lu, J. T., Zarepoor, L., Wu, W., Liu, R., Power, K. a. (2014). Cooked navy and black bean diets improve biomarkers of colon health and reduce inflammation 158 during colitis. The British Journal of Nutrition, 111, 15491563. Zhu, Z., Jiang, W., & Thompson, H. J. (2012). Edible dry bean consumption (Phaseolus vulgaris L.) modulates cardiovascular risk factors and diet-induced obesity in rats and mice. British Journal of Nutrition, 108, S66S73.