EVALUATION OF ROOT TRAITS ASSOCIATED WITH DROUGHT TOLERANCE IN DRY BEAN (Phaseolus vulgaris L.) By Amy Lydia Lasley A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Plant Breeding, Genetics and Biotechnology - Crop and Soil Sciences – Master of Science 2013 ABSTRACT EVALUATION OF ROOT TRAITS ASSOCIATED WITH DROUGHT TOLERANCE IN DRY BEAN (Phaseolus vulgaris L.) By Amy Lydia Lasley Drought is one of the most important abiotic plant stresses that negatively effects yield and quality of crops. Dry bean (Phaseolus vulgaris L.) is an important crop in Michigan and decreased rainfall is affecting productivity of the bean crop which is largely raised under rainfed conditions. Ninety-six dry bean genotypes were grown under drought stress in the field in Michigan in 2011 and 2012. Plants were excavated at flowering to investigate the number of basal roots, basal root angle, taproot diameter, and overall root score. At harvest, biomass was measured and harvest index and yield were calculated. Several root characteristics were identified as contributing to drought stress tolerance and specific genotypes were identified as possessing superior drought tolerant root traits including high basal root numbers in most pinto bean genotypes. The same genotypes were grown in growth pouches and analyzed using WinRhizo root analysis software. Data for basal root whorl number, root length, surface area, average diameter, root volume and root tips were recorded. Basal root whorl number and root length were identified as significant traits and pink beans had the largest number of basal root whorls. The same genotypes were subjected to drought stress under limited root growth in small pots in the growth chamber and rated for wilting, stem greenness and unifoliate senescence after 21 days of stress. Watering was resumed and ratings for survival were taken after a 14 day recovery period. Wilting was identified as an important trait for survival and several genotypes expressed high survival rates. Superior genotypes were identified for both root and shoot traits that could be combined in future breeding activities to enhance the drought tolerance of dry beans. ACKNOWLEDGEMENTS I want to express my gratitude to Dr. James Kelly, my major advisor, for his kindness, encouragement and knowledge throughout my master’s education at Michigan State University. Special thanks to Dr. Wayne Loescher and Dr. Karen Cichy for serving on my committee and providing valuable insight and recommendations throughout my research. Many thanks to the whole dry bean breeding and genetics lab: Evan Wright, Norm Blakely, Halima Awale, Shitaye Moges, Gerardine Mukeshimana, Valerio Hoyos Villegas, Jim Heilig, Wezi Mkwaila, Kelvin Kamfwa, and Beth Brisco for the support, assistance and enjoyable times while at Michigan State. A special thanks to Evan and Norm for facilitating my field experiments as well as Gerardine and Valerio for their unending patience with my many questions. I would also like to express my gratitude to my friends who became like family during my stay in East Lansing. Thank you for the support and memories. Finally, I wish to thank my family, especially my parents, Jim and Shari Lasley, for their love, support and ever encouragement throughout my education. iii TABLE OF CONTENTS LIST OF TABLES ......................................................................................................................... vi LIST OF FIGURES ..................................................................................................................... viii Literature Review............................................................................................................................ 1 World and Local Dry Bean Production .......................................................................................... 1 Drought and Dry Beans .................................................................................................................. 1 Dry Bean Races, Seed Size and Drought Tolerance ....................................................................... 3 Yield ................................................................................................................................................ 5 Selection Indices ............................................................................................................................. 7 Studying Shoot Traits for Drought Tolerance ................................................................................ 8 Studying Root Architecture for Drought Tolerance ..................................................................... 10 Taproot .......................................................................................................................................... 11 Basal root whorls .......................................................................................................................... 12 Angle of Basal Roots .................................................................................................................... 12 Root characteristics ....................................................................................................................... 13 Studying the Root System in the Field ......................................................................................... 15 Studying the Root System in the Laboratory ................................................................................ 15 Objectives ..................................................................................................................................... 16 REFERENCES ............................................................................................................................. 17 Chapter 1. Identification of root traits associated with drought tolerance in 96 dry bean genotypes ................................................................................................................................. 23 Introduction ................................................................................................................................... 23 Materials ....................................................................................................................................... 27 Methods......................................................................................................................................... 46 Statistical Analysis ........................................................................................................................ 50 Results and Discussion ................................................................................................................. 53 Genotype× Environment Interaction............................................................................................. 53 Basal Root Number ....................................................................................................................... 53 Basal Root Angle .......................................................................................................................... 54 Taproot Diameter .......................................................................................................................... 55 Overall Root Score ........................................................................................................................ 55 Yield .............................................................................................................................................. 56 Harvest Index ................................................................................................................................ 57 Comparisons of Root Traits with Yield ........................................................................................ 59 Conclusions ................................................................................................................................... 60 iv REFERENCES ............................................................................................................................. 62 Chapter 2. Evaluation of 96 dry bean genotypes grown in pouches for seedling root characteristics related to drought tolerance .............................................................................. 66 Introduction ................................................................................................................................... 66 Materials ....................................................................................................................................... 69 Methods......................................................................................................................................... 77 Statistical Analysis ........................................................................................................................ 79 Results and Discussion ................................................................................................................. 80 Basal Root Whorls ........................................................................................................................ 80 Total Root Length ......................................................................................................................... 81 Total Surface Area ........................................................................................................................ 83 Total Root Volume ....................................................................................................................... 84 Average Root Diameter ................................................................................................................ 84 Number of Tips ............................................................................................................................. 85 Comparison of Pouch Root Traits to Field Root Traits, Harvest Index and Yield ....................... 85 Conclusions ................................................................................................................................... 87 REFERENCES ............................................................................................................................. 89 Chapter 3. Evaluation of 95 dry bean genotypes for reaction to drought in a 5-week growth chamber experiment ................................................................................................................. 93 Introduction ................................................................................................................................... 93 Objectives ..................................................................................................................................... 95 Materials ....................................................................................................................................... 95 Methods......................................................................................................................................... 96 Statistical Analysis ........................................................................................................................ 98 Results and Discussion ................................................................................................................. 99 Wilting .......................................................................................................................................... 99 Unifoliate Senescence ................................................................................................................. 104 Stem Greenness ........................................................................................................................... 105 Survival ....................................................................................................................................... 106 Genotypic Responses .................................................................................................................. 107 Correlation Analysis ................................................................................................................... 108 Breeding Strategies to Enhance Drought Resistance in Dry Bean ............................................. 109 Conclusions ................................................................................................................................. 111 REFERENCES ........................................................................................................................... 113 v LIST OF TABLES Table 1. 1 Genotypes of the Mesoamerican race listed by commercial seed types with combined taproot diameter data grown in two locations in Michigan over two seasons 2011 and 2012. .................................................................................................................................. 28 Table 1. 2 Genotypes of the Durango race listed by commercial seed types with combined taproot diameter data grown in two locations in Michigan over two seasons 2011 and 2012. 29 Table 1. 3 Genotypes of the Jalisco race listed by commercial seed types with combined taproot diameter data grown in two locations in Michigan over two seasons 2011 and 2012. 31 Table 1. 4 Genotypes of the Mesoamerican race listed by commercial seed type with data collected for different root traits and agronomic traits grown in two locations in Michigan in 2011...................................................................................................................................... 32 Table 1. 5 Genotypes of the Durango race listed by commercial seed type with data collected for different root traits and agronomic traits grown in two locations in Michigan in 2011. .... 34 Table 1. 6 Genotypes of the Jalisco race listed by commercial seed type with data collected for different root traits and agronomic traits grown in two locations in Michigan in 2011. .... 37 Table 1. 7 Genotypes of the Mesoamerican race listed by commercial seed type with data collected for different root traits and agronomic traits grown in two locations in Michigan in 2012...................................................................................................................................... 39 Table 1. 8 Genotypes of the Durango race listed by commercial seed type with data collected for different root traits and agronomic traits grown in two locations in Michigan in 2012. .... 41 Table 1. 9 Genotypes of the Jalisco race listed by commercial seed type with data collected for different root traits and agronomic traits grown in two locations in Michigan in 2012. .... 44 Table 1. 10 Analysis of variance (degrees of freedom, mean squares, and F values) for root data collected on 96 dry bean genotypes grown over two field seasons in Michigan, 2011 and 2012. .................................................................................................................................. 51 Table 1. 11 Analysis of variance (degrees of freedom, mean squares, and F values) for yield and harvest index for 96 dry bean genotypes grown in two locations in Michigan over two field seasons, 2011 and 2012. .................................................................................................. 52 vi Table 2. 2 Average values for six root traits grown in the pouches and analyzed using WinRhizo for 26 Mesoamerican race seed types. .................................................................... 70 Table 2. 3 Average values for six root traits grown in the pouches and analyzed using WinRhizo for 50 Durango race seed types. ............................................................................. 72 Table 2. 4 Average values for six root traits grown in six pouches and analyzed using WinRhizo for 19 Jalisco race seed types.................................................................................. 75 Table 2. 1 High, low, mean, F test and LSD(0.05) values for six root parameters of 96 dry bean genotypes plus tepary bean tested in the growth pouch and analyzed using WinRhizo software. ................................................................................................................................... 79 Table 3. 1 Average scores for wilting, stem greenness, unifoliate senescence after 21 days without water, and survival after 14 day recovery period of 28 Mesoamerican genotypes grown in the growth chamber. ............................................................................................... 100 Table 3. 2 Average scores for wilting, stem greenness, unifoliate senescence after 21 days without water, and survival after 14 day recovery period 50 Durango genotypes grown in the growth chamber. ............................................................................................................... 101 Table 3. 3 Average scores for wilting, stem greenness, unifoliate senescence after 21 days without water, and survival after 14 day recovery period for 18 Jalisco genotypes grown in the growth chamber. ............................................................................................................... 103 vii LIST OF FIGURES Figure 1.1 Yearly rainfall totals for the Saginaw Valley Research and Extension Center from 1979 to 2012. Trends show an overall decreasing amount of rainfall. Prior to 1997, yearly rainfall averaged above 750 mm, and since then has decreased. .................................................. 24 Figure 1.2 Example of protractor apparatus used in the measurement of basal root angle of the 96 bean genotypes grown in the field in Michigan in 2011 and 2012. For interpretation of the references to color in this and all other figures, the reader is referred to the electronic version of this thesis....................................................................................................................................... 49 Figure 2. 1 Comparison of root scans of tepary (left) and dry bean (right) grown for two weeks in growth pouches. ............................................................................................................................ 83 Figure 3. 1 Wilting scores in dry bean after 14 days without watering. 0 is no sign of wilting. 5 is completely wilted .......................................................................................................................... 98 viii Literature Review World and Local Dry Bean Production Dry bean (Phaseolus vulgaris) is the most important grain legume crop grown for direct human consumption (Broughton et al., 2003). Production within the United States accounts for 6% of world dry bean production. In 2010, approximately 462,000 hectares were planted in the United States, over half of which is produced in North Dakota and Michigan combined (USDAERS, 2011). Michigan is the second leading producer of dry beans in the United States with more than 95,000 hectares harvested in 2010 and was valued at over $120 million. Black and navy beans are the two most important classes produced in Michigan, but other classes grown include cranberry, kidney, small red and pinto beans. Production acreage lies primarily in the Saginaw Valley and Bay Thumb regions of central and eastern Michigan (USDA-ERS, 2011). Drought and Dry Beans A major abiotic constraint of dry bean production is drought. Sixty percent of dry beans grown worldwide are produced under drought conditions (Singh, 1995), which affects yield, quality, and market value of the crop (Urrea et al., 2009). Major dry bean production areas in the U.S., like the Midwest and Upper Midwest, experience periods of unpredictable drought each season (Urrea et al., 2009). Although drought is less common in Michigan than in many production areas worldwide, it has become an increasingly frequent problem. Typically Michigan is not considered an area that suffers from a lack of water due to the surrounding Great Lakes. However, as climate begins to change the world’s meteorological landscape, Michigan weather patterns are also changing. Concerning trends in yearly rainfall totals for the Saginaw Valley Research Farm, located in the heart of Michigan bean production are emerging. Between 1979 and 1994, annual rainfall totals of at least 750 mm were common. From 1995 to present, 1 the annual rainfall has only exceeded 750 mm in two years. In fact, only 471mm of rainfall was recorded in 2010. These are concerning trends especially since ninety-five percent of Michigan bean production is rainfed (Kelly and Cichy, 2012). So in a drought year like 2012 bean yields were drastically reduced under prevailing drought conditions. Drought can be defined meteorologically as a period of prolonged and abnormal moisture deficiency (Palmer, 1965) or insufficient moisture necessary for a plant to grow normally and complete its life cycle (Graham and Ranalli, 1997). Agriculture accounts for 70% of the fresh water withdrawals worldwide, more than any other use category on the planet (FAO, 2007). As population increases, more of the world fresh water resources will be needed for direct use by humans, reducing the amount that is available for crop production. Drought stress can be classified into three types, early, terminal or intermittent drought (Ludlow and Muchow, 1990). Terminal drought is defined as optimum rainfall until some point in the growing season when rainfall greatly diminishes and does not resume for the remainder of the season usually occurring during the reproductive period of the crop cycle. Terminal drought often occurs in regions with clearly defined rainy and dry seasons like the tropics. Terminal drought is endemic to the intermountain growing regions in the United States (Teran and Singh, 2002) and most of Central America (Frahm et al., 2004). Intermittent drought occurs when periods of rain are limited and erratic during the growing season and thus cause periods of stress (Schneider et al., 1997). Intermittent drought is common in the semiarid highlands of the subtropics and the Great Plains (Ludlow and Muchow, 1990; Urrea et al., 2009). The majority of drought studies use terminal drought as it is much easier to stimulate than intermittent drought. Regardless of the type of drought, the effects are devastating in much of the developing world 2 where dry bean are a major subsistence crop and are a major source of protein (Broughton et al., 2003). Dry bean is an excellent crop for breeding for drought tolerance because it has an average growing season of less than 100 days and thus has much lower water requirements than many other crops (White, 1992). Breeding for drought tolerance can be difficult as the effects are complex and plant response is highly variable and based on many interacting factors (RamirezVallejo and Kelly, 1998). However, developing drought resistant crop varieties is a slow and very difficult process (Subbarao et al., 1995), but is particularly important for dry beans, because 60% of dry bean production worldwide is raised under some level of drought stress (White and Singh, 1991). Genetic diversity for drought tolerance is found in wild, cultivated and related bean species and can be utilized when breeding for drought tolerance (Beebe et al., 2008). Dry Bean Races, Seed Size and Drought Tolerance Dry beans exhibit an impressive amount of diversity in seed size and growth habit. Dry beans were domesticated in two separate gene pools, Andean and Middle American (Gepts, 1996;Gepts and Bliss, 1985). Middle American bean races are considered more drought tolerant as their domestication origin was in the arid and semi-arid regions of Mexico and Central America. The Andean gene pool was domesticated in more humid regions of the Andean mountains and rainforests. The two gene pools are further separated into three races in each gene pool based on agronomic and adaptation traits (Singh et al., 1991). Historically, important drought tolerance sources have been identified in the Durango and Jalisco races in the Middle American gene pool. These races were bred and grown in the arid highlands of Mexico as was the Mesoamerican race, bred in Central America under terminal droughts (Singh et al., 1991; Singh, 2007; Teran and Singh, 2002). These three races of the Middle American gene pool 3 represent important genetic variation that could be exploited in breeding for drought tolerance in beans. Durango race beans were cultivated and bred in the highlands of Mexico for centuries and are known to possess drought tolerance (Singh, 1989). Members of this race include pinto and great northern (GN) seed types. Race Jalisco is also known to possess drought tolerance (Hayes and Singh, 2007; Singh, 2007). These genotypes were also developed in the semi-arid highlands of Mexico, but in a more Southern location to Durango race where it is less humid and cooler (Teran and Singh, 2002), Jalisco race seed types include modern small reds (formerly known as red Mexican beans) and pink beans. Mesoamerican races include small seeded black and tan beans that are more tropical in origin and better adapted to lowland regions (Singh et al., 1991). Specific bean genotypes have been identified as being drought tolerant. These include, BAT 477 a historically recognized source of drought tolerance developed by CIAT (International Center for Tropical Agriculture), Cali, Colombia (White and Castillo, 1989). The drought tolerance of BAT477 has been investigated in comparison to other less drought tolerant genotypes (Sponchiado et al. ,1989). Grown under drought conditions, BAT 477 produced significantly deeper roots than the less drought tolerant lines. More recently Asfaw and Blair (2012) used BAT 477 to study quantitative trait loci (QTL) associated with rooting patterns in drought stress versus non-stress conditions. They found that BAT 477 had significantly better root characteristics than the drought susceptible genotype. Especially under water stressed conditions, BAT 477 had greater rooting depth, thick root length, root volume, average root diameter, root length distribution with depth and root biomass distribution with depth than the susceptible parent, DOR 364, of the population. 4 Matterhorn, a GN seed type that belongs to the Durango race was bred in Michigan under rain fed conditions to combine traits from both the Durango and Mesoamerican races (Kelly et al., 1999). Matterhorn has been shown in multiple studies to exhibit drought tolerance when specifically comparing drought stress performance to non-stress performance (Muñoz-Perea et al., 2006; Singh, 2007; Urrea et al., 2009). Muñoz-Perea, et al. (2006) found that Matterhorn outperformed the majority of genotypes regardless of race or origin. Matterhorn had greater biomass yield, seed yield, harvest index, and seed weight under drought stress. Urrea et al. (2009) also identified Matterhorn as having a lower percent yield reduction under stress conditions as compared to non-stress conditions. Tepary bean (P. acutifolius A. Gray) is also recognized as possessing important drought tolerance traits. Native to the southwestern United States and northwestern Mexico tepary bean is a desert plant known for its drought tolerance (Thomas et al., 1983). Drought tolerance in tepary is attributed to its deep rooting capabilities and sensitive stomatal control. Tepary beans have been observed to root approximately one meter deeper than common beans when relying on soil water alone (Thomas et al., 1983). Tepary beans represent an important resource for drought tolerance that could be introgressed into common bean (Pratt, 1983). Yield Yield is the main focus of most breeding programs, and is the key economical trait most affected by drought stress. Many studies have been designed to measure seed yield under various levels of water stress and identify mechanisms used by individual bean cultivars to adapt to drought stress (Schneider et al., 1997). Regardless of which characteristic is being directly investigated in a given study, results always include seed yield as it is the most important economic trait of dry bean (Acosta-Gallegos and Adams, 1991). Drought tolerance is a 5 quantitatively inherited trait and thus yield is a cumulative measure of the combined effect of the many genes influencing the traits that contribute to yield (Blair et al., 2012). Hence yield is usually chosen as the best measure of genotype performance as it is cumulative of all growth and physiological processes throughout the season. Teran and Singh (2002) confirmed that the most effective way of identifying drought resistant genotypes is based on seed yield. They identified cultivars that yielded well in both drought stressed and well watered environments. Many studies have compared yield under stress and non-stress environments (Acosta-Gallegos and Shibata, 1989; Abebe et al., 1998; Porch et al., 2009; Ramirez-Vallejo and Kelly, 1998; Rosielle and Hamblin, 1981; Schneider et al., 1997; Teran and Singh, 2002; Urrea et al., 2009). Abebe et al. (1998) showed differential genotypic response among 20 different dry bean lines grown under drought stress and non-stress. Yield varied among lines within water regimes which confirmed that yield is the most common trait for measuring genotypic drought tolerance in dry bean. Yield components that significantly contribute to yield and differ between genotypes include pod number, seeds per pod, seed number per plant, seed size, days to flowering and days to maturity (Blair et al., 2012; Ramirez-Vallejo and Kelly, 1998). Drought stress during flowering greatly contributes to flower abortion as the number of flowers aborted increases, the greater the yield loss. One common factor in drought tolerance research is determining whether crops will survive extreme drought scenarios. Although this concept is interesting, it is not entirely valid if these traits do not show any yield advantage under more typical productive conditions (Davies et al., 2011). 6 Selection Indices Plant breeders are constantly searching for selection techniques that are fast and reliable. As a result, numerous selection indices have been developed and are used as simple indicators of a cultivar’s reaction to drought. Considerable research has focused on verifying the accuracy of these drought selection indices. Indices are not species-specific, and therefore many investigations have been conducted to determine their application in dry bean. Abebe et al. (1998) compared selection indices for dry bean lines grown under drought. The indices were calculated when beans were grown under drought stress and non-stress conditions and included arithmetic mean, geometric mean, drought response index, drought susceptibility index, response to drought, and percent yield reduction. Abebe et al. (1998) found that arithmetic mean and geometric mean were the best indicators of high yielding lines under both stressed and non-stressed conditions and that drought response index is a good indicator in stressed environments. The underlying concept between comparing plant responses under stress to non-stress studies is the comparison of arithmetic mean and geometric means. Geometric mean identifies genotypes that are high yielding under water and better accounts for the reduction in yield under stress than using simple arithmetic yields. Abebe et al. (1998) concluded that the other indices were not highly correlated in either environment and could not be recommended as a basis for selection. Schneider et al. (1997) suggested using a simple breeding strategy where lines were first selected for high geometric mean, and then further selected those with the highest yields grown under stress. Another popular selection index is harvest index. Harvest index is the proportion of the whole plant mass that is partitioned to the seed (Sadok and Sinclair, 2011). The index is believed to be a good indicator of how effectively the plant is able to partition resources between 7 vegetative and reproductive growth. Lines with low harvest index are often thought to be genetically closer to wild viney ancestors. Wild types produce more vegetative growth including stems and leaves to compete in their natural forest habitat where sunlight is limiting (Debouck, 1999). These wild relatives are adapted to prolonged periods of drought by slowing growth and then resuming growth when more favorable conditions return. As a result of adaption, modern cultivars often exhibit the same trait and mature later than expected. This trait is referred to by breeders as a stay-green characteristic (Evans, 1993). When selecting for yield under drought stress, breeders are inadvertently selecting against these characteristics and for higher partitioning as measured by harvest index (Beebe et al., 2008). Among the different partitioning influencing drought indices compared by Ramirez-Vallejo and Kelly (1998), harvest index was found to be the most reliable. For these reasons, harvest index is considered a good measure of drought resistance or susceptibility among genotypes. Indices can play an important role in identifying overall drought tolerance of genotypes grown under drought stress. Studying Shoot Traits for Drought Tolerance Indices are an important tool used by breeders to quickly assess the drought tolerance of different genotypes. By studying specific portions of plant morphology and physiology, scientists can gain greater insight into specific mechanisms that are important in drought response. Shoots play a very important role in overall plant productivity. Shoots are the portion of the plant that are directly interacting with environmental factors throughout the growing season and thus show the first symptoms of drought stress. Beans have diverse shoot types or growth habits that have been classified into plant types I to IV. Type I growth habit is determinate, highly branched and exhibit upright growth (Singh, 1982). Type II plants are indeterminate and upright; type III plants are indeterminate, weak stemmed and prostrate and non-climbing; and type IV plants are indeterminate, weak stemmed with a strong ability to climb 8 (Singh, 1982). Singh et al. (1991) further classified races of common bean into races based on characteristics such as leaf shape and size, leaf pubescence, length of internodes, number of node to flower, flower and pod types, days to maturity, seed size and shape and geographic distribution. Determinacy is an important shoot characteristic. Determinate types stop growing when the plant starts flowering. Indeterminate types grow throughout the season and continue to produce leaves and flowers until sink strength ceases vegetative growth. If determinate plants are stressed during the season, flowers and pods abort resulting in significant yield reductions. In contrast, indeterminate growth allows the plant to continue producing flowers and pods, once more favorable growing conditions have returned and recovery results in less yield loss (Kelly et al., 1987). One important shoot trait that may be associated with drought tolerance is slow wilting. The trait has been identified and is being utilized in breeding crops such as maize (Zea mays L.), sorghum (Sorghum bicolor L.) and soybean (Glycine max L.) for drought tolerance when grown under drought stress (Sadok and Sinclair, 2011). The trait has yet to be identified in dry bean. Sadok and Sinclair (2011) found that genetic variation for the trait exists in soybean, so there is an expectation that the same trait exists in dry bean. This trait relies on the interactions between hydraulic conductance in the leaves, the xylem and guard cells (Sadok and Sinclair, 2011). Plants with a greater transpiration rates yield more because they are utilizing more water. When transpiration stops, growth stops as photosynthesis is halted. Slow-wilting genotypes have an overall lower transpiration rate and hence a lower yield potential when compared with normal genotypes under ideal conditions. Genotypes without the trait have a higher yield potential when conditions are ideal because they are able to transpire and grow at a faster rate (Sadok and 9 Sinclair, 2011). Interestingly, these slow wilting genotypes are quite valuable in geographical regions where drought is a consistent occurrence each year because they are able to slow the rate at which water is lost and survive during intense drought periods. When genotypes not possessing the trait have completely stopped transpiring and photosynthesizing under stress, slow wilting genotypes continue to grow and thus have an advantage. Researchers have not determined what mechanisms underlie the slow wilting trait (Ries et al., 2012). White and Castillo (1992) investigated the importance of bean shoot diversity by grafting different shoot genotypes to contrasting root genotypes and subjecting the grafted plants to drought. The different root and shoot genotypes had known drought tolerance or susceptibility and combinations of grafting were made between both types. Root genotype had the greatest effect on performance under stress, whereas shoot genotypes produced smaller but still important effects (White and Castillo, 1992). Studying Root Architecture for Drought Tolerance Root architecture is an important component of plant growth and drought tolerance in dry beans. The existing root architecture of bean genotypes needs to be investigated before those traits that influence tolerance can be studied. Dry beans have three root types: taproot, basal roots and adventitious roots. The taproot is the first root to emerge from the germinating seed. The taproot has few branches, is the deepest extending root and has the largest diameter of all roots that the bean plant produces (Gregory, 2006a). Basal roots emerge secondary to the taproot from the base of the bean hypocotyl and establish the architecture of the entire root system (Bonser et al., 1996). Dry beans have basal roots arranged in whorls around the hypocotyl and the number of basal root whorls varies by genotype but typically ranges from 1 to 4 whorls (Lynch, 2012). Basal roots extend laterally in relation to the taproot at varying degrees. 10 Adventitious roots emerge as the last set of roots from the below-ground base of the plant stem. The adventitious roots grow parallel to the soil horizon (Burridge, 2012). The exact spatial distribution of these root types is dependent on genotype and significant phenotypic variation is known to exist. White and Castillo (1989) believe that root architecture is of primary importance in determining and developing drought tolerant bean lines as opposed to shoot characteristics. Taproot Kelly (1998) offers his perspective as a plant breeder on bean roots in relation to their above ground growth habit. For example, type II indeterminate growth habit as demonstrated in genotypes from the Mesoamerican race are characterized by a very upright growth structure with little branching. Kelly (1998) suggests that the above ground structures strongly resemble the below ground structures. He cites evidence from his many years pulling plants by hand that these type II plants have a strong, deep rooting taproot with limited branching that can sustain a plant during periods of drought by mining the soil profile to greater depths for available moisture. He describes type III indeterminate growth habits often found in the Durango race as producing more lateral roots rather than vertical roots (Kelly, 1998). The plants branch profusely and opportunistically and so he believes the roots behave in a similar fashion. These genotypes exploit the upper profiles of the soil with more profuse roots and are highly adaptive and able to adjust to shallow soil conditions and sporadic rainfall (Kelly, 1998). Overall, it is very important to exploit both root system types as they are important for plant survival under terminal and intermittent drought. Taproot length is an important trait as those genotypes with deeper taproots are able to capture water reserves deep in the soil profile. Few studies have been conducted to determine root length in the field as excavation of complete root systems is difficult. Sponchiado et al. 11 (1989) showed that under drought conditions, genotypes that were known to be drought tolerant rooted 0.4 m deeper than less drought tolerant genotypes. Those genotypes with deeper roots had threefold the yield of drought susceptible lines and also left lower soil water content at the end of the season (Sponchiado et al., 1989). Rooting depth is an important component in determining overall drought tolerance in dry bean. Basal root whorls Basal root whorls are a known component of bean root architecture, but relatively little is known about their role in overall drought tolerance. Dry bean genotypes differ in the total number of root whorls they produce (Lynch, 2012), but no relationship between number of basal root whorls and drought tolerance has been reported. Angle of Basal Roots Basal roots are an important part of the bean root system critical for nutrient and water acquisition. Bonser et al. (1996) studied basal root angle relative to phosphorus presence in the soil profile. They hypothesized that a shallower basal root angle in the soil is desirable for phosphorus acquisition. Several dry bean genotypes differing in growth habit, gene pool origin and performance were chosen for study in low phosphorus conditions. Some of these genotypes chosen were also known to be drought tolerant. Using growth pouches and various phosphorus treatments, root growth angles were measured daily. Basal root angle growth varied with phosphorus treatment. When phosphorus is not limiting in the soil profile, roots would grow deeper and thus acquire greater amounts of water, since genotypes that exhibit a shallow branching pattern may be at a disadvantage for water acquisition (Bonser et al., 1996). Nutrient availability appears to be a confounding factor for water acquisition by basal roots deeper in the soil profile. 12 Root tradeoffs between water and phosphorus acquisition were also investigated by Ho et al. (2005). Using field and greenhouse techniques, the effects of water and phosphorus stress were tested on bean root architecture. The premise for the study was that phosphorus is fixed in the upper soil layers where water is quickly depleted under terminal drought conditions so survival of bean plants under drought is dependent on deep rooting traits. The authors used recombinant inbred lines from a cross between genotypes of the Mesoamerican race with type II growth habit known to be drought tolerant developed by Frahm et al. ( 2004). The lines used had contrasting basal rooting depth. Greater root density was found in the shallow soil layers when phosphorus was limiting whereas in water limiting conditions, greater root density was found at greater depths. Their hypothesis that shallow rooting genotypes would perform better in phosphorus stressed conditions and that deeper rooting genotypes would perform better under drought was supported. Those lines with greater basal root angle had greater water acquisition under terminal drought stress and optimal nutrient conditions (Ho et al., 2005). Root characteristics Scientists believe that a strong, deep tap root is the key to drought tolerance because the root is able to reach water reserves deeper in the soil profile (Gregory, 2006b). Some genotypes have a greater mass of roots in the upper soil layers for nutrient acquisition and fewer roots deep in the soil profile for water acquisition. The most useful scenario is one in which bean plants have a balanced root architecture with a substantial taproot to sustain plant through intermittent drought and sufficient adventitious roots for nutrient acquisition (Asfaw and Blair, 2012). This scenario is particularly important in areas where rainfall is intermittent making it difficult to establish reliable and resilient crop productivity (Davies et al., 2011). 13 A major limiting factor in studying root architecture is the great heterogeneity that exists within the soil profile and throughout various types and classes of soils (Asfaw and Blair, 2012). Soil type can vary across a single field; replications must be made to account for such variation. Across Michigan, soil types vary so greatly that data valid in one location may not be applicable elsewhere. For these reasons, few studies have been conducted on root architecture and scientists are constantly searching for mechanisms to model typical root growth outside of the field. Gregory (2006b) discusses root thickness as a possible means of inferring drought tolerance in a genotype. The theory is that thick roots can penetrate hardened soil more easily and reach soil moisture reserves deep in the soil profile. Since the taproot in beans grows vertically in the soil profile, an investigation into the relationship between taproot diameter and drought tolerance would be appropriate. The actual diameter of those roots and the angle of basal roots appear to be important components of drought tolerance exhibited in bean roots. Phenotyping root characteristics is a very important component to discovering molecular markers linked to root traits and conducting genetic studies to develop drought tolerant dry beans. The first QTL study for root characteristics was performed by Beebe et al. (2006) and investigated root architectural traits that were correlated with phosphorus acquisition. Several QTLs that were identified as being involved in phosphorus acquisition were correlated to basal root development. QTLs were also identified for basal and taproot development that were independent of phosphorus acquisition QTLs. In a later study, QTL were identified based on expression of genes for deep roots, root length and root diameter traits supporting the importance of reliable phenotypic data in molecular breeding for drought tolerance (Asfaw and Blair, 2012). 14 Studying the Root System in the Field Excavating roots for phenotyping can be arduous and time consuming and in some cases can also involve use of expensive machinery. Shovelomics is a technique first used in maize to visually phenotype roots (Trachsel et al., 2011). The goal was to identify root architectural traits important to plant productivity under edaphic stress. The study used standard shovels to remove the crown area of the maize plant for study and recorded 10 traits. The study was designed to handle large numbers of genotypes and thus was considered a high throughput technique for gaining phenotypic knowledge of the maize root system. The research team at Pennsylvania State University used a similar technique to study bean roots. They produced a video showing how they excavated the roots and their simple methodology for phenotyping roots (Burridge, 2012). The system they describe appears to be a simple and efficient way to study major root traits in a field setting. Studying the Root System in the Laboratory Studies have been conducted in the laboratory or greenhouse in more controlled settings to attempt to model the field and study the relationship between roots and drought in greater detail (Himmelbauer et al., 2004; Jones and Ljung, 2012; Lobet et al., 2011; Lodeiro et al., 2000; Lynch and Vanbeem, 1993; Sassi Aydi et al., 2008). Many areas of bean production in temperate zones have only one growing season per year and thus field space is limited. Thus the decision as to the number of genotypes and how they can be studied in the field must be made prudently. Studies conducted during the off-season are effective tools in modeling field conditions since they are cheaper and time saving. Many systems exist to study roots under more controlled environmental conditions. Those include hydroponics, growth pouches (McMichael et al., 1985), slant tubes, the use of sand systems for easy extraction of roots, and various other containers to study young seedling 15 plants (Lynch and Vanbeem, 1993). More methods for imaging, processing and data collection of roots have been developed including computer software like WinRHIZO, ROOTEDGE and SmartRoot (Himmelbauer et al., 2004; Lobet et al., 2011). These programs assess parameters like root length, surface area, diameter, tips and branching (Himmelbauer et al., 2004). A more advanced technique uses a transparent gellan gum system for plant growth, a camera collecting 2-dimensional photos and RootReader3D software to create three-dimensional platforms for seedling root architecture study (Clark et al., 2011). Software is valuable in the study of roots as computers can record thousands of data points over time that can be assembled into a model root system using appropriate software programs. Thus the use of computer-based systems has and will have great potential to revolutionize root architecture phenotyping. It is important, however, that the data collected in the laboratory setting can be associated with field-based traits. Objectives The overall objectives of this study were to examine the role of specific root traits as the basis for drought tolerance in dry beans. The second objective was to identify seedling root and shoot traits that were associated with enhanced performance under drought stress in the field. The final objective was to explore germplasm from the Middle American gene pool for diversity of root characteristics and performance under drought stress in the field. 16 REFERENCES 17 REFERENCES Abebe A., Brick M.A., Kirkby R.A. (1998) Comparison of selection indices to identify productive dry bean lines under diverse environmental conditions. Field Crops Research 58:15-23. Acosta-Gallegos J.A., Adams M.W. (1991) Plant traits and yield stability of dry bean ( Phaseolus vulgaris) cultivars under drought stress. The Journal of Agricultural Science 117:213219. Acosta-Gallegos J.A., Shibata J.K. (1989) Effect of water-stress on growth and yield of indeterminate dry-bean (Phaseolus-vulgaris) cultivars. Field Crops Research 20:81-93. Asfaw A., Blair M.W. (2012) Quantitative trait loci for rooting pattern traits of common beans grown under drought stress versus non-stress conditions. Molecular Breeding 30:681695. Beebe S.E., Rao I.M., Cajiao C., Grajales M. (2008) Selection for drought resistance in common bean also improves yield in phosphorus limited and favorable environments. Crop Science 48:582-592. Beebe S.E., Rojas-Pierce M., Yan X., Blair M.W., Pedraza F., Muñoz F., Tohme J., Lynch J.P. (2006) Quantitative trait loci for root architecture traits correlated with phosphorus acquisition in common bean. Crop Science 46:413-423. Blair M.W., Galeano C.H., Tovar E., Torres M.C.M., Castrillon A.V., Beebe S.E., Rao I.M. (2012) Development of a Mesoamerican intra-genepool genetic map for quantitative trait loci detection in a drought tolerant x susceptible common bean (Phaseolus vulgaris L.) cross. Molecular Breeding 29:71-88. Bonser A.M., Lynch J., Snapp S. (1996) Effect of phosphorus deficiency on growth angle of basal roots in Phaseolus vulgaris. New Phytologist 132:281-288. Broughton W.J., Hernandez G., Blair M., Beebe S., Gepts P., Vanderleyden J. (2003) Beans (Phaseolus spp.) - model food legumes. Plant and Soil 252:55-128. Burridge J. (2012) High throughput phenotyping of root system architecture, http://plantscience.psu.edu/research/labs/roots/methods/field-methods/field-crop-rootcrown-sampling-and-analysis. Clark R.T., MacCurdy R.B., Jung J.K., Shaff J.E., McCouch S.R., Aneshansley D.J., Kochian L.V. (2011) Three-dimensional root phenotyping with a novel imaging and software platform. Plant Physiology 156:455-465. 18 Davies W.J., Zhang J., Yang J., Dodd I.C. (2011) Novel crop science to improve yield and resource use efficiency in water-limited agriculture. The Journal of Agricultural Science 149:123-131. Debouk D.G. (1999) Diversity in Phaseolus species in relation to the common bean, in: S. P. Singh (Ed.), Common bean improvement in the twenty-first century., Kluwer Academic Publishers, Dordrecht, Netherlands. Evans, L.T. (1993) Crop evolution, adaptation and yield. Cambridge University Press, Cambridge, U.K. FAO. (2007) Water at a glance, http://www.fao.org/nr/water/art/2007/glance/facts3.html. Accessed Oct. 2011. Frahm M.A., Rosas J.C., Mayek-Perez N., Lopez-Salinas E., Acosta-Gallegos J.A., Kelly J.D. (2004) Breeding beans for resistance to terminal drought in the lowland tropics. Euphytica 136:223-232. Gepts P. (1996) Origin and evolution of cultivated Phaseolus species. In: Pickersgill B, Lock J (eds) Advances in Legume Systematics, Part 8. Legumes of economic importance. Kew Botanical Garden, Kew, UK, pp 65–74. Gepts P., Bliss F.A. (1985) F1 hybrid weakness in the common bean: differential geographic origin suggests two gene pools in cultivated bean germplasm. Journal of Heredity 76:447-450. Graham P.H., Ranalli P. (1997) Common bean (Phaseolus vulgaris L.). Field Crops Research 53:131-146. Gregory P.J. (2006b) Genetic control of root system properties. In:Plant Roots: Growth, Activity, and Interaction with Soils. Blackwell Publishing Ltd, Oxford, UK. pp. 253-285. Gregory P.J. (2006a) Roots and the architecture of root systems, In:Plant Roots: Growth, Activity, and Interaction with Soils. Blackwell Publishing Ltd, Oxford, UK. pp. 18-44. Hayes R., Singh S.P. (2007) Response of cultivars of race Durango to continual dry bean versus rotational production systems. Agronomy Journal 99:1458-1462. Himmelbauer M.L., Loiskandl W., Kastanek F. (2004) Estimating length, average diameter and surface area of roots using two different Image analyses systems. Plant and Soil 260:111120. Ho M.D., Rosas J.C., Brown K.M., Lynch J.P. (2005) Root architectural tradeoffs for water and phosphorus acquisition. Functional Plant Biology 32:737-748. 19 Jones B., Ljung K. (2012) Subterranean space exploration: the development of root system architecture. Current Opinion in Plant Biology 15:97-102. Kelly J.D. (1998) Bean Roots - A plant breeder’s perspective. Annual Report of the Bean Improvement Cooperative 41:214-215. Kelly J.D., Cichy K.A. (2012) Dry bean breeding and production technologies, in: M. Siddiq, M.A. Uebersax (Ed.), Dry Beans and Pulses: Production, Processing, and Nutrition, Wiley Blackwell Publishing Co. Oxford, U.K. pp. 23-54. Kelly J.D., Adams M.W., Varner, G.V. (1987) Yield stability of determinate and indeterminate dry bean cultivars. Theoretical Applied Genetics 74:516-521. Kelly J.D., Hosfield G.L., Varner G.V., Uebersax M.A., Taylor J. (1999) Registration of 'Matterhorn' great northern bean. Crop Science 39:589-590. Lobet G., Pages L., Draye X. (2011) A novel image-analysis toolbox enabling quantitative analysis of root system architecture. Plant Physiology 157:29-39. Lodeiro A.R., Gonzalez P., Hernandez A., Balague L.J., Favelukes G. (2000) Comparison of drought tolerance in nitrogen-fixing and inorganic nitrogen-grown common beans. Plant Science 154:31-41. Ludlow M.M., Muchow R.C. (1990) A critical-evaluation of traits for improving crop yields in water-limited environments. Advances in Agronomy 43:107-153. Lynch J.P. ( 2012) Genotype screening for number of whorls, http://plantscience.psu.edu/research/labs/roots/methods/methods-info/screening-of-rootwhorls/genotype-screening-for-number-of-root-whorls. Lynch J., Vanbeem J.J. (1993) Growth and architecture of seedling roots of common bean genotypes. Crop Science 33:1253-1257. McMichael B.L., Burke J.J., Berlin J.D., Hatfield J.L., Quisenberry J.E. (1985) Root vascular bundle arrangements among cotton strains and cultivars. Environmental and Experimental Botany 25:23-30. Muñoz-Perea C.G., Terán H., Allen R.G., Wright J.L., Westermann D.T., Singh S.P. (2006) Selection for drought resistance in dry bean landraces and cultivars. Crop Science 46:2111-2120. Palmer W.C. (1965) Meteorological drought. Office of Climatology, U.S. Weather Bureau, Washington, D.C. 20 Porch T.G., Ramirez V.H., Santana D., Harmsen E.W. (2009) Evaluation of common bean for drought tolerance in Juana Diaz, Puerto Rico. Journal of Agronomy and Crop Science 195:328-334. Pratt R.C. (1983) Gene transfer between tepary and common beans. Desert Plants 5:57-63. Ramirez-Vallejo P., Kelly J.D. (1998) Traits related to drought resistance in common bean, Euphytica. pp. 127-136. Ries L.L., Purcell L.C., Carter T.E., Edwards J.T., King C.A. (2012) Physiological traits contributing to differential canopy wilting in soybean under drought. Crop Science 52:272-281. Rosielle A.A., Hamblin J. (1981) Theoretical aspects of selection for yield in stress and nonstress environments. Crop Science 21:943-946. Sadok W., Sinclair T.R. (2011) Crops yield increase under water-limited conditions: review of recent physiological advances for soybean genetic improvement, Advances in Agronomy, 113: 313-337. Sassi Aydi S., Aydi S., Gonzalez E., Abdelly C. (2008) Osmotic stress affects water relations, growth, and nitrogen fixation in Phaseolus vulgaris plants. Acta Physiologiae Plantarum 30:441-449. Schneider K.A., Rosales-Serna R., Ibarra-Perez F., Cazares-Enriquez B., Acosta-Gallegos J.A., Ramirez-Vallejo P., Wassimi N., Kelly J.D. (1997) Improving common bean performance under drought stress. Crop Science 37:43-50. Singh S.P. (1989) Patterns of variation in cultivated common bean (Phaseolus vulgaris, Fabaceae). Economic Botany 43:39-57. Singh S.P. (1982) A key for indentification of different growth habits of Phaseolus vulgaris L. Annual Report of the Bean Improvement Cooperative 25:92-95. Singh S.P. (1995) Selection for water-stress tolerance in interracial populations of common bean. Crop Science 35:118-124. Singh S.P. (2007) Drought resistance in the race Durango dry bean landraces and cultivars. Agronomy Journal 99:1219-1225. Singh S.P., Gepts P., Debouck D. (1991) Races of common bean (Phaseolus vulgaris, Fabaceae). Economic Botany 45:379-396. Singh S.P., Terán H., Gutiérrez J.A. (2001) Registration of SEA 5 and SEA 13 drought tolerant dry bean germplasm .Crop Science 41:276-277. 21 Sponchiado B.N., White J.W., Castillo J.A., Jones P.G. (1989) Root growth of four common bean cultivars in relation to drought tolerance in environments with contrasting soil types. Experimental Agriculture 25:249-257. Subbarao G.V., C. Johansen, A.E. Slinkard, R.C. Nageswara, R.N.P. Saxena, and Y.S. Chauhan (1995) Strategies for improving drought resistance in grain legumes. Critical reviews in plant sciences 14 (6):469-523. Teran H., Singh S.P. (2002) Comparison of sources and lines selected for drought resistance in common bean. Crop Science 42:64-70. Thomas C.V., Manshardt R.M., Waines J.G. (1983) Teparies as a source of useful traits for improving common beans. Desert Plants 5:43-48. Trachsel S., Kaeppler S.M., Brown K.M., Lynch J.P. (2011) Shovelomics: high throughput phenotyping of maize (Zea mays L.) root architecture in the field. Plant and Soil 341:7587. Urrea C.A., Yonts C.D., Lyon D.J., Koehler A.E. (2009) Selection for drought tolerance in dry bean derived from the Mesoamerican gene pool in western Nebraska. Crop Science 49:2005-2010. USDA-ERS. (2011) U.S. dry bean area planted and harvested, http://usda.mannlib.cornell.edu/MannUsda/viewDocumentInfo.do?documentID=1394. Accessed Nov. 2012. White J.W., Castillo J.A. (1989) Relative effect of root and shoot genotypes on yield of common bean under drought stress. Crop Science 29:360-362. White J.W., Castillo J.A. (1992) Evaluation of diverse shoot genotypes on selected root genotypes of common bean under soil-water deficits. Crop Science 32:762-765. White J.W., Singh, S.P. (1991) Breeding for adaptation to drought. In: A. van Schoonhoven and O. Voysest (ed.) Common beans: Research for crop improvement. CAB International, Wallingford, UK & CIAT, Cali, Colombia pp. 501-560. 22 Chapter 1. Identification of root traits associated with drought tolerance in 96 dry bean genotypes Introduction Dry beans (Phaseolus vulgaris L.) are an important agricultural commodity in Michigan which is the second leading producer of dry beans in the United States (USDA-ERS, 2012). Bean production in Michigan is limited by certain abiotic and biotic constraints. Drought is one abiotic stress that is becoming an increasingly important problem and a major constraint to increasing bean production in Michigan. Typically the state is not considered as an area that lacks water due to the surrounding Great Lakes. But as climate change begins to influence the world’s meteorological landscape, Michigan is being affected. Yearly rainfall totals for the Saginaw Valley Research Farm, located in the heart of Michigan bean production, shows concerning trends. Over the fifteen year period from 1979 to 1994, precipitation averaged at least 750 mm, whereas from 1995 to present, rainfall exceeded 750 mm in only two years. That same 18 year period experienced an average rainfall of 663 mm, a difference of 133 mm between these two time periods. In certain years moderate drought occurred. For example, only 472 mm of rainfall was recorded in 2010 (Figure 1.1). These are concerning trends especially since ninety-five percent of Michigan bean production is grown under rainfed conditions (Kelly and Cichy, 2012). In 2010, bean yields were reduced by 50% as the crop relied on seasonal rainfall to provide adequate water. Thus the need to recognize and understand drought as well as to breed crops with drought tolerance is becoming increasingly important in Michigan. 23 Yearly Rainfall Totals (mm) 1200 1000 800 600 400 200 0 Year Figure 1.1 Yearly rainfall totals for the Saginaw Valley Research and Extension Center from 1979 to 2012. Trends show an overall decreasing amount of rainfall. Prior to 1997, yearly rainfall averaged above 750 mm, and since then has decreased. Breeding for drought tolerance is largely dependent on the environment under which the plants are grown (Ramirez-Vallejo and Kelly, 1998). Historically, important sources of bean germplasm possessing drought tolerance have been found in the Middle American bean races Durango and Jalisco which were domesticated in the arid highlands of Mexico (Singh et al., 1991). In addition genotypes from race Mesoamerica, bred in Central America were also reported as possessing drought tolerance (Singh, 2007; Teran and Singh, 2002). These three races represent important genetic variation that could be exploited in breeding for drought tolerance in beans. 24 Roots play an important role in overall drought tolerance. Some dry bean genotypes avoid drought by rooting at greater depths and thus reaching and extracting soil moisture found in these regions of the soil profile (Asfaw and Blair, 2012). Lynch and Van Beem (1993) showed that early seedling establishment is dependent on taproot diameter whereas the mature root system depends on basal root development. A study of root systems throughout the growing season is vital to elucidate mechanisms that condition drought tolerance as drought can effect plant growth and development anytime during the season. Studies have been conducted to identify root architecture differences in beans (Ho et al., 2005; Robertson et al., 1985), but knowing which root traits are important to overall drought tolerance is critical when breeding for root systems that contribute to drought tolerance. Excavating roots for phenotyping can be arduous and time consuming and in some cases can also involve the use of expensive equipment. Shovelomics (Trachsel et al., 2011) is a simple technique first used in maize (Zea mays L.) to visually phenotype roots. Few field studies have been conducted on roots although shoots have previously been visually phenotyped for specific traits. The goal was to identify root architectural traits important to plant productivity under edaphic stress. The study used standard shovels to remove the crown area of the maize plant for study and 10 traits were recorded. The study was designed to handle large numbers of genotypes and thus considered a high throughput technique for gaining phenotypic knowledge of the maize root system. The same group at Pennsylvania State University used a similar technique to study bean roots. They produced a video showing how they excavated the roots and a relatively simple methodology for phenotyping different root traits (Burridge, 2012). Many studies have been conducted comparing bean yield under stress and non-stress environments (Acosta-Gallegos and Shibata, 1989; Porch et al., 2009; Ramirez-Vallejo and 25 Kelly, 1998; Rosielle and Hamblin, 1981; Schneider et al., 1997; Teran and Singh, 2002; Urrea et al., 2009). Yield under stress is the main focus of these studies as yield is the most economically important trait and is the main factor in farmer’s choice of which variety to grow. In addition, geometric mean yield is also measured to combine yield data from stress and nonstress trials. High geometric mean yields ensure that breeders are developing high yielding lines with drought tolerance not low yielding lines that only express stress tolerance. Hall (1993) defines drought tolerance as the ability of a genotype to yield more than other genotypes that were subjected to the same drought stress. Yield is the cumulative result of all the plant processes and thus is an effective tool in measuring a genotype’s performance. Thus comparing genotypic yield is an important component of drought tolerance research (Schneider et al., 1997). Other factors such as harvest index can influence yield. Harvest index is the proportion of the whole plant mass that is partitioned to the seed (Sadok and Sinclair, 2011). Thus partitioning becomes an important component in drought tolerance. Drought stress is known to reduce biomass accumulation thus seed yield and seed weight when fewer resources are available for all plant growth processes (Muñoz-Perea et al., 2006). Some genotypes may produce large amounts of biomass that are not efficiently partitioned into seed. Thus those genotypes with high harvest indices are more desirable as they effectively partition nutrients to the seed. The objective of this study was to compare 96 dry bean genotypes grown under rainfed conditions for two field seasons in Michigan and collect data on: basal root number, basal root angle, taproot diameter, overall root score, yield and harvest index. Root characteristics identified may explain genotypic performance under drought stress. Harvest index was measured to assist in identifying partitioning characteristics that may infer drought tolerance 26 between bean genotypes. Bean genotypes were identified that had superior root, performance and harvest index traits that could be used to advance breeding for drought tolerance. Materials Ninety-six dry bean genotypes were selected based on prior breeder knowledge of their reaction to drought. The genotypes were assembled as part of the Common Bean Coordinated Agricultural Project (BeanCAP) and have been used in a number of field studies across the country to gain greater insights into drought tolerance in dry bean. The 96 genotypes were previously selected by a group of breeders in the United States and Puerto Rico for more extensive testing as part of the BeanCAP research projects being conducted in different production areas. The selections represent past and present dry bean varieties that belong to the Middle American gene pool (Gepts and Bliss, 1985). In the second year, four genotypes needed to be substituted as insufficient amounts of seed were available to complete a second year of study due to lack of adaptation in 2011. So each year 92 common genotypes were tested for a total of 96 genotypes. Those genotypes that were grown in 2011 and not in 2012 were Sawtooth (GN), San Juan (pinto), GN#1Sel27 (GN) and Fisher (pinto). Those replacement genotypes grown in 2012 were SER 16 (small red), SER 48 (small red), Rosetta (pink) and Eldorado (pinto). A complete list of the genotypes used in this study is shown in Tables 1.1-1.9. 27 Table 1. 1 Genotypes of the Mesoamerican race listed by commercial seed types with combined taproot diameter data grown in two locations in Michigan over two seasons 2011 and 2012. Genotype Seed Type Eclipse Raven A-55 Shania Domino F04-2801-4-1-2 Zorro Aifi Wuriti Midnight Jaguar I9365-31 T-39 Shiny Crow PR 0443-151 115M Navigator Medalist C-20 Mayflower Avalanche Seafarer Schooner A285 Verano SEA 10 BAT 477 Mean LSD0.05 CV% black black black black black black black black black black black black black black black navy navy navy navy navy navy navy carioca small white tan tan Taproot Diameter mm 3.3 3.2 3.1 2.9 2.8 2.7 2.6 2.6 2.4 2.3 2.3 2.2 2.1 2.1 1.8 3.3 3.2 3.0 2.9 2.8 2.6 2.1 2.3 2.1 2.2 2.2 2.6 0.7 23.0 28 Table 1. 2 Genotypes of the Durango race listed by commercial seed types with combined taproot diameter data grown in two locations in Michigan over two seasons 2011 and 2012. Genotype Seed Type Matterhorn GN9-4 Orion Weihing GN9-1 Coyne Gemini Beryl R US-1140 UI-59 Marquis UI-425 BelNeb-RR-1 GN#1Sel27 Eldorado ABCP-8 PT7-2 Medicine Hat Kimberly PT9-17 Chase USPT-CBB-1 Buster Maverick Croissant ND-307 Santa Fe La Paz Kodiak GN GN GN GN GN GN GN GN GN GN GN GN GN GN pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto Taproot Diameter mm 2.7 2.6 2.4 2.3 2.2 2.0 2.0 2.0 1.9 1.9 1.9 1.7 1.5 1.4 3.0 2.7 2.7 2.7 2.7 2.6 2.5 2.5 2.5 2.5 2.4 2.4 2.4 2.4 2.3 29 Table 1.2 (cont’d) NE2-09-3 UI-114 Poncho Stampede Sierra Sawtooth Lariat Quincy Common Pinto NW-590 Buckskin Topaz Othello Bill Z Montrose Nodak TARS-VCI-4B NW-63 USPT-CBB-5 Fisher San Juan Mean LSD0.05 CV% pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto 2.3 2.3 2.3 2.2 2.2 2.1 2.1 2.1 2.0 2.0 2.0 2.0 2.0 1.9 1.9 1.9 1.8 1.8 1.8 1.6 1.5 2.2 0.7 23.0 30 Table 1. 3 Genotypes of the Jalisco race listed by commercial seed types with combined taproot diameter data grown in two locations in Michigan over two seasons 2011 and 2012. Genotype Seed Type Sedona Harold PR 0340-3-3-1 CDC Crocus UI-537 Roza Viva Pink Floyd ROG 312 Victor Gloria Yolano Rosetta DOR 364 SER16 Merlot CENTA Pupil USRM-20 F07-449-9-3 SER48 CRM UI-239 IBC 301-204 IJR Mean LSD0.05 CV% pink pink pink pink pink pink pink pink pink pink pink pink small red small red small red small red small red small red small red small red small red small red small red red mottled Taproot Diameter mm 2.6 2.5 2.4 2.4 2.3 2.1 2.1 2.0 2.0 1.9 1.9 1.7 3.0 2.6 2.5 2.4 2.3 2.3 2.2 2.2 2.1 2.1 2.0 1.5 2.2 0.7 23.0 31 Table 1. 4 Genotypes of the Mesoamerican race listed by commercial seed type with data collected for different root traits and agronomic traits grown in two locations in Michigan in 2011. Genotype 115M A-55 Aifi Wuriti Domino Eclipse F04-2801-4-1-2 I9365-31 Jaguar Midnight PR 0443-151 Raven Shiny Crow T-39 Zorro Seed Type Basal Root Average Angle black black black black black black black black black black black black black black count 7.3 6.5 5.9 5.9 8.6 6.4 5.4 9.4 8.2 6.7 5.8 6.5 11.4 6.8 degrees 39.2 36.0 35.5 21.5 40.5 33.3 44.5 29.0 33.8 56.5 37.0 43.0 45.5 34.1 Overall Score Yield Harvest Index 2.9 4.3 4.6 1.8 3.0 3.6 3.0 4.3 5.4 3.5 5.7 4.7 4.0 4.2 kg ha-1 3077 3753 4306 4175 4265 3791 3582 3619 3915 2761 3602 2938 3034 4318 % 11.9 31.1 39.3 40.1 39.5 41.7 30.6 42.8 38.5 29.6 37.2 30.6 37.3 43.9 32 Table 1.4 (cont’d) Genotype Avalanche C-20 Mayflower Medalist Navigator Schooner Seafarer Verano BAT 477 SEA 10 Mean LSD0.05 CV% Seed Type navy navy navy navy navy navy navy small white tan tan Basal Root Average Angle count 7.3 6.7 7.6 7.2 8.8 5.7 6.8 4.8 9.4 8.9 7.3 degrees 43.5 34.6 31.5 30.9 36.1 38.0 53.0 31.7 31.0 33.0 37.2 4.1 26.0 22.0 26.0 Overall Score Yield -1 Harvest Index 4.7 3.5 2.8 1.5 3.9 2.7 4.8 2.6 3.3 3.3 3.7 kg ha 3319 3015 3644 3062 4140 3950 3632 3829 3116 3333 3607 % 34.8 25.7 29.7 34.8 38.5 36.4 36.1 38.8 33.5 42.9 36.5 2.4 31.0 1079 16.0 12.3 18.0 33 Table 1. 5 Genotypes of the Durango race listed by commercial seed type with data collected for different root traits and agronomic traits grown in two locations in Michigan in 2011. Genotype BelNeb-RR-1 Beryl R CDC Crocus Coyne Gemini GN#1Sel27 GN9-1 GN9-4 Marquis Matterhorn Orion Sawtooth UI-425 UI-59 US-1140 Weihing Seed Type GN GN GN GN GN GN GN GN GN GN GN GN GN GN GN GN Basal Root count 4.5 6.8 8.2 4.0 8.2 6.1 7.7 10.7 8.7 10.3 10.1 9.6 8.1 8.0 5.0 6.2 Average Angle degrees 33.0 33.0 54.5 49.0 43.0 39.5 41.5 35.5 52.0 30.2 49.5 49.5 30.2 45.0 31.7 35.5 34 Overall Score Yield -1 2.9 3.5 3.4 4.8 2.7 5.0 4.6 4.2 4.2 6.2 3.8 6.1 3.2 4.8 3.0 3.2 kg ha 3241 3146 2743 2940 3859 632 2994 4232 3446 4768 4318 581 2482 1984 4057 2920 Harvest Index % 38.0 37.7 39.8 33.2 49.2 2.7 31.6 35.0 34.7 46.8 37.0 3.5 28.5 26.8 52.1 33.2 Table 1.5 (cont’d) Genotype Croissant ABCP-8 Bill Z Buckskin Buster Chase Common Pinto Fisher Kimberly Kodiak La Paz Lariat Maverick Medicine Hat Montrose ND-307 NE2-09-3 Nodak NW-590 Seed Type pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto Basal Root count 13.5 8.0 8.8 9.9 11.8 8.7 6.4 6.0 6.0 7.1 7.6 8.4 10.5 11.0 8.4 8.3 9.7 7.3 7.3 Average Angle degrees 52.5 49.5 50.0 48.0 41.5 54.0 34.0 41.5 11.0 48.0 44.5 52.5 57.5 56.2 48.0 44.5 43.7 39.0 40.5 35 Overall Score Yield -1 4.5 3.9 4.3 3.1 2.9 4.8 2.6 4.2 2.5 3.9 6.9 4.2 2.9 5.0 3.4 5.0 3.7 2.6 3.3 kg ha 3146 2962 3552 2588 4010 2969 4022 2073 3328 3128 4428 4536 3277 4006 3407 3647 3199 3688 2710 Harvest Index % 42.0 29.5 30.1 32.0 52.0 36.9 35.9 11.9 40.6 44.4 46.7 43.3 50.6 47.3 35.5 32.9 43.1 49.4 24.2 Table 1.5 (cont’d) Genotype Seed Type NW-63 pinto Othello pinto Poncho pinto PT7-2 pinto PT9-17 pinto Quincy pinto San Juan pinto Santa Fe pinto Shania pinto Sierra pinto Stampede pinto TARS-VCI-4B pinto Topaz pinto UI-114 pinto USPT-CBB-1 pinto USPT-CBB-5 pinto Mean LSD0.05 CV% Abbreviations: GN= great northern Basal Root Average Angle count 8.3 6.1 10.5 9.2 8.0 10.1 8.0 9.2 8.9 14.4 9.3 5.8 5.7 9.0 8.7 6.5 8.6 degrees 36.8 50.0 46.8 36.8 46.5 54.5 43.1 39.2 49.7 48.0 54.0 61.0 37.5 54.5 52.5 43.8 46.7 4.1 26.0 22.0 26.0 36 Overall Score Yield -1 Harvest Index 3.3 4.7 4.2 6.1 4.9 4.1 2.7 4.1 3.0 4.4 3.8 2.2 2.2 2.5 4.5 3.9 3.8 kg ha 3038 3445 4224 4281 3894 4042 1491 4729 4147 3837 3720 1499 3153 2994 3240 3366 3443 % 21.4 37.2 32.1 37.8 38.8 38.2 6.8 43.7 34.3 35.8 41.3 6.3 43.3 35.6 23.4 46.0 34.3 2.4 31.0 1079 16.0 12.3 18.0 Table 1. 6 Genotypes of the Jalisco race listed by commercial seed type with data collected for different root traits and agronomic traits grown in two locations in Michigan in 2011. Genotype Gloria Harold Pink Floyd ROG 312 Roza Sedona UI-537 Victor Viva Yolano Seed Type Basal Root Average Angle pink pink pink pink pink pink pink pink pink pink count 6.1 8.6 8.1 9.1 6.3 10.9 6.1 7.7 7.6 8.1 Overall Score degrees 44.0 44.0 37.2 49.0 39.5 42.0 45.0 43.5 40.7 41.5 Yield -1 2.4 5.6 2.0 3.5 3.6 3.2 4.5 3.9 3.2 3.1 37 kg ha 3444 1863 3599 3488 3652 3615 3422 2212 1397 3408 Harvest Index % 24.8 8.8 39.7 38.3 32.6 37.8 36.6 18.2 4.8 37.1 Table 1.6 (cont’d) Genotype CENTA Pupil CRM DOR 364 F07-449-9-3 IBC 301-204 Merlot PR 0340-3-3-1 UI-239 USRM-20 Verano IJR Mean LSD0.05 CV% Seed Type small red small red small red small red small red small red small red small red small red small white red mottled Basal Root Average Angle count 7.6 8.3 6.1 8.7 6.3 8.8 5.0 8.6 5.4 4.8 6.3 7.4 degrees 35.5 41.3 39.5 37.5 49.0 40.5 50.5 44.0 48.8 31.7 40.8 42.2 4.1 26.0 Overall Score 22.0 26.0 Yield -1 Harvest Index 2.8 2.9 3.4 3.6 4.1 6.1 4.1 3.6 3.4 2.6 2.2 3.5 % 45.0 32.8 38.5 34.9 48.1 40.7 41.2 30.6 40.4 38.8 31.7 33.4 2.4 31.0 38 kg ha 3987 2768 3395 4234 4381 4802 3521 3165 3163 3829 3168 3358 1079 16.0 12.3 18.0 Table 1. 7 Genotypes of the Mesoamerican race listed by commercial seed type with data collected for different root traits and agronomic traits grown in two locations in Michigan in 2012. Genotype PR 0443-151 Domino 115M Jaguar Zorro Aifi Wuriti Eclipse Midnight I9365-31 Raven Shiny Crow F04-2801-4-1-2 A-55 T-39 Seed Type Basal Root Average Angle black black black black black black black black black black black black black black count 5.8 5.5 5.1 4.9 4.3 4.2 4.1 4.1 4.1 4.0 3.9 3.7 3.7 3.6 degrees 31.6 23.6 26.9 31.5 28.1 28.2 29.5 29.3 23.2 20.8 37.3 34.6 25.4 25.4 Overall Score Yield -1 4.8 4.5 2.8 3.5 3.5 4.5 3.8 3.9 2.5 3.2 3.1 3.2 3.6 2.9 39 kg ha 2415 3462 3987 2800 3426 2938 3403 2973 3961 3061 2654 2733 3231 2949 Harvest Index % 29.6 43.4 40.9 44.7 44.1 43.5 43.8 37.3 37.7 42.1 43.4 45.5 27.1 47.6 Table 1.7 (cont’d) Genotype Avalanche Schooner Navigator Seafarer Medalist C-20 Mayflower A285 Verano SEA 10 BAT 477 Mean LSD0.05 CV% Seed Type navy navy navy navy navy navy navy carioca small white tan tan Basal Root Average Angle count 4.7 4.6 4.2 4.1 3.7 3.7 3.2 4.4 3.2 4.8 3.3 4.0 degrees 27.7 30.3 25.8 31.1 23.9 42.6 32.8 38.4 28.4 23.6 34.2 35.4 2.0 29.0 2.0 28.0 Overall Score Yield -1 Harvest Index 4.0 3.9 3.5 4.3 4.1 3.3 2.7 4.1 2.8 3.0 2.7 3.5 % 41.8 37.3 40.5 39.2 38.7 35.4 41.2 42.9 40.8 42.6 43.1 40.9 2.0 29.0 40 kg ha 3119 2881 2666 2879 3636 3599 2960 3413 1880 2790 2921 3040 739 15.0 6 9.0 Table 1. 8 Genotypes of the Durango race listed by commercial seed type with data collected for different root traits and agronomic traits grown in two locations in Michigan in 2012. Marquis GN9-4 UI-425 Matterhorn US-1140 GN9-1 Orion Coyne Weihing CDC Crocus Gemini BelNeb-RR-1 Beryl R UI-59 Seed Type GN GN GN GN GN GN GN GN GN GN GN GN GN GN Basal Root Average Angle count Genotype degrees 5.5 5.0 4.8 4.7 4.3 4.3 4.2 4.1 4.1 4.1 4.1 4.1 3.7 3.5 31.1 32.0 35.2 33.1 27.6 35.2 26.2 32.9 29.6 28.2 29.1 36.5 35.1 37.9 Overall Score Yield kg ha 3.1 5.2 3.3 4.7 4.0 2.8 2.6 4.5 3.5 2.6 2.5 2.1 3.0 2.5 Abbreviations: GN= great northern 41 -1 3123 3433 2294 3510 3015 3637 3341 2794 2829 2607 2818 3010 3110 2864 Harvest Index % 41.1 45.5 29.8 46.5 47.5 38.1 44.6 37.7 38.1 45.5 41.3 46.7 44.0 44.3 Table 1.8 (cont’d) Genotype La Paz UI-114 Chase Maverick NE2-09-3 Medicine Hat Common Pinto Othello Quincy Sedona Poncho PT9-17 NW-590 Sierra Nodak Santa Fe Eldorado Stampede USPT-CBB-1 PT7-2 Croissant Buckskin Seed Type Basal Root Average Angle pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto count 5.8 5.8 5.0 4.9 4.9 4.8 4.7 4.6 4.6 4.6 4.5 4.5 4.5 4.4 4.3 4.3 4.2 4.2 4.2 4.2 4.1 4.1 degrees 27.7 34.5 35.1 34.7 29.7 31.6 31.7 38.1 32.1 36.7 25.4 33.4 33.2 35.7 35.2 34.0 30.0 28.5 28.2 25.1 41.8 29.3 Overall Score Yield -1 3.7 3.6 4.0 3.4 4.8 3.1 4.5 4.4 3.5 3.7 3.5 3.3 3.3 3.7 3.4 2.8 3.9 3.9 3.7 3.5 3.7 3.0 42 kg ha 3594 2015 3077 2953 2706 3607 3332 3390 3530 3349 2051 3033 3781 2980 3063 3126 3733 2917 3392 4415 3240 2855 Harvest Index % 42.6 39.3 44.5 43.2 40.7 50.1 43.6 48.1 43.8 38.9 39.9 45.5 34.7 31.5 44.2 43.3 35.6 43.8 41.7 49.8 43.5 43.4 Table 1.8 (cont’d) Genotype USPT-CBB-5 Kodiak Buster NW-63 Kimberly ND-307 TARS-VCI-4B Lariat Shania Bill Z ABCP-8 Topaz Montrose Mean LSD0.05 CV% Seed Type Basal Root Average Angle pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto count 4.0 4.0 3.9 3.9 3.8 3.7 3.7 3.6 3.6 3.5 3.4 3.4 2.9 3.9 degrees 32.2 25.5 32.0 31.2 36.9 31.0 35.2 29.8 34.6 33.4 33.6 30.1 37.1 32.3 2.0 29.0 2.0 28.0 Overall Score Yield -1 Harvest Index 3.1 2.9 3.9 3.4 3.4 4.3 3.0 3.1 3.4 2.9 3.5 3.2 2.3 3.4 kg ha 2561 3662 3489 3112 2341 3608 2373 3704 3552 2625 2563 2705 2859 3131 % 47.0 48.6 46.0 34.0 45.2 42.6 25.8 38.8 37.2 43.2 37.8 45.5 51.6 41.8 2.0 29.0 739 15.0 6.0 9.0 43 Table 1. 9 Genotypes of the Jalisco race listed by commercial seed type with data collected for different root traits and agronomic traits grown in two locations in Michigan in 2012. Genotype Harold Rosetta UI-537 ROG 312 Roza Gloria Victor Viva Yolano Pink Floyd Seed Type Basal Root Average Angle pink pink pink pink pink pink pink pink pink pink count 5.3 5.2 5.1 5.1 4.7 4.3 4.1 3.7 3.4 2.2 degrees 31.9 202.9 32.8 26.7 30.3 191.3 34.2 25.1 30.6 28.0 Overall Score Yield 3.3 5.1 5.5 2.7 3.7 3.7 3.6 3.4 3.4 3.7 kg ha 3095 2936 3039 3332 2179 1812 3125 2936 2716 3442 -1 44 Harvest Index % 22.8 47.6 48.1 45.0 23.3 37.5 30.4 20.1 44.7 48.9 Table 1.9 (cont’d) Genotype Seed Type Basal Root Average Angle count 4.9 4.8 4.3 4.3 4.3 4.2 4.1 3.5 3.5 3.4 2.8 4.1 4.2 degrees 29.2 25.9 31.5 34.6 29.3 26.6 34.1 31.9 24.2 22.3 32.4 32.2 44.9 SER48 small red USRM-20 small red Merlot small red DOR 364 small red SER16 small red CRM small red CENTA Pupil small red UI-239 small red F07-449-9-3 small red IBC 301-204 small red PR 0340-3-3-1 small red IJR red mottled Mean LSD0.05 2.0 2.0 29.0 28.0 CV% Abbreviations: IJR=Indeterminate Jamaica Red Overall Score Yield 2.2 3.0 4.0 3.7 3.0 3.9 3.3 2.8 3.8 2.5 2.1 2.9 3.4 kg ha 2153 3464 3792 3390 2837 2915 2558 2764 3218 2307 4407 2682 2959 % 38.9 46.3 37.0 45.8 45.9 37.2 47.6 39.5 43.6 45.9 46.7 42.5 40.2 2.0 29.0 739 15.0 6 9.0 -1 45 Harvest Index Methods The 2011 field drought study was planted at the Michigan State University Montcalm Research Farm in Entrican, MI on June 15. The site was selected to induce drought as the soil type is a coarse textured McBride sandy loam. The 96 genotypes were planted in four-row plots at a rate of 13 seeds per meter in an incomplete randomized block design. The center two rows were planted to the 96 experimental genotypes with the two outside rows planted with a single red bean cultivar, Merlot. Two replications of the 96 genotypes were planted. Weeds and insects were controlled throughout the season following recommended practices provided by the Michigan State University weed science group (Sprague, 2012). In 2012, the study was planted at the Saginaw Valley Research and Extension Center located near Richville, MI on June 12, where the soil type is a Tappan-Londo loam. The genotypes were planted in 4.9 m rows in four-row plots in an incomplete randomized block design. The center rows were planted to the experimental genotype with the outside rows planted to a single black bean cultivar of Jaguar. There were three replications of the 96 genotypes in 2012. In 2012, there was insufficient seed for some genotypes grown in 2011 so substitutions were made. Prior to planting broadcast fertilizer was incorporated as recommended. Weeds and insects were controlled throughout the season following recommended practices for beans (Sprague, 2012). No supplemental water was applied to plots in 2011 or 2012. In 2011, the plot received about 202 mm of rainfall from planting to harvest (June to September) which is below the thirty year average of 432 mm in Entrican, MI. In 2012, the plots received about 241 mm of rainfall from planting to harvest which is below the thirty year average of 310 mm in Richville, MI. The two 46 locations received considerably different amounts of rain during each growing season. These differences in thirty year averages can be attributed to their location in the state. Entrican is located in central Michigan whereas Richville is located in eastern Michigan close to the Thumb region. When 50% of the plants were flowering, five interior plants from each plot were excavated so as to keep roots as intact as possible following the Shovelomics protocol (Burridge, 2012). Roots were excavated on July 27, 2011 and August 15, 2012. In 2011, the plots in Entrican, MI had received 72 mm of rainfall from planting to excavation. In 2012, the plots in Richville, MI had received 187 mm of rainfall from planting to excavation. Shoot portion was cut from root growth at soil level. Roots were placed in plastic bags by plot and refrigerated until analysis could be completed. In 2011, data was collected on: basal root angles, basal root branching in a three centimeter segment, number of basal roots, number of adventitious roots, adventitious root branching in a three centimeter segment, tap root diameter, tap root branching in a three centimeter segment and an overall root score. In 2012, the number of plot replications of the 96 genotypes was increased from two to three. In order to speed up processing, data was recorded on total basal root number, basal root angle, taproot diameter and an overall root score. Only data on root categories from 2012 were reported because root traits collected in 2011 that were not recorded in 2012 were found to be highly variable and not useful. Basal root number is simply a count of the number of basal roots on an individual plant. Basal root angle was measured using a protractor on a plastic tray. Measurements were taken by arranging the approximate soil line on the stem at zero degrees (Figure 1.2). Angles were then determined for each basal root from that zero line. Ordinarily the taproot fell close to the 90 degree mark on the protractor. For ease of analysis, following data collection, basal root angles 47 were averaged for each plant and then for each genotype. Taproot diameter was recorded approximately 3 cm below the basal root area of the root system to standardize the point where data were collected. Overall root score consisted of a scale ranging from zero to ten. A score from zero to three was an overall low root system that looked weak and not well developed. A score from four to seven was considered an middle root system and seven to ten was an high overall root system consisting of many strong roots, a thick taproot and deeply angled basal roots. Overall root score was designed to be a quick and easy way to visually rank the roots that were analyzed for genotypic comparisons. 48 Figure 1.2 Example of protractor apparatus used in the measurement of basal root angle of the 96 bean genotypes grown in the field in Michigan in 2011 and 2012. For interpretation of the references to color in this and all other figures, the reader is referred to the electronic version of this thesis. 49 Plots were harvested on October 6, 2011 and on September 28, 2012. At harvest, plots were end trimmed to uniform lengths of 4.6 m and 4.3 m in 2011 and 2012 respectively. Plants were pulled using a Pickett one step bean rod-puller and plants were stacked. Total biomass was recorded and the plants were threshed using a Wintersteiger combine to obtain seed yield. Harvest index was calculated using adjusted seed weight (18% moisture) to account for various moisture levels at harvest divided by total biomass recorded in the field. Statistical Analysis Data was analyzed using SAS 9.3 (SAS Institute Inc., Cary, NC, 2010). Analysis of variance was conducted using PROC MIXED with random statements being row, column and replication. Significant differences were found for the genotype × environment interaction for many of the traits measured and thus the years were analyzed separately (Table 1.1-2). PROC GLM was used to determine the least significant differences for each variable at a level of alpha equals 0.05. 50 Table 1. 10 Analysis of variance (degrees of freedom, mean squares, and F values) for root data collected on 96 dry bean genotypes grown over two field seasons in Michigan, 2011 and 2012. Grand Mean LSD(0.05) CV Source Genotype Year Genotype×Year Grand Mean LSD(0.05) CV Source Genotype Year Genotype×Year Grand Mean LSD(0.05) CV Source Genotype Year Genotype*Year Grand Mean LSD(0.05) CV Source Genotype Year Genotype×Year df 99 1 91 Basal Root Number 5.7 2.0 27.1 MS 499.75 1501.43 402.34 F-Test 2.13**** 633.39**** 1.87**** df 99 1 91 Basal Root Angle 35.4 12.2 26.5 MS 10891.71 13707.6 9232.66 F-Test 1.25 ns 155.86**** 1.15 ns df 99 1 91 Taproot Diameter (mm) 2.3 0.69 23.0 MS 45.88 54.62 16.7 F-Test 1.67*** 196.31**** 0.66 ns df 99 1 91 Overall Score 3.6 1.4 29.5 MS 215.11 8.6 133.54 F-Test 1.94**** 7.66** 1.31* 51 Table 1. 11 Analysis of variance (degrees of freedom, mean squares, and F values) for yield and harvest index for 96 dry bean genotypes grown in two locations in Michigan over two field seasons, 2011 and 2012. Yield Grand Mean LSD(0.05) CV Source Genotype Year Genotype×Year Grand Mean LSD(0.05) CV Source Genotype Year Genotype×Year df 99 1 91 kg ha-1 3191 633 15.3 MS 143540114.8 19801525 58736667.1 F-Test 6.11**** 83.4**** 2.72**** df 99 1 91 Harvest Index % 38.7 6.1 12.0 MS 27564.32 3156.27 4669.06 F-Test 12.90**** 146.25**** 2.38**** 52 Results and Discussion Genotype× Environment Interaction As the two years of this study were performed in different locations, 2011 Montcalm, MI and 2012 Richville, MI, it was imperative to investigate the genotype × environment interaction (Table 1.1-2). Significant genotype × environment interactions were identified for basal root number, score, yield and harvest index. The analysis of these traits was then separated by year. Two traits did not have significant genotype × environment interactions; basal root angle and taproot diameter. Basal root angle also did not have significant differences between genotypes and thus was deemed an unreliable trait and not analyzed any further. Taproot diameter did have significant differences between genotypes and so their analysis was combined and was discussed as such. Basal Root Number In 2011, significant differences in basal root number were detected between genotypes (p = 0.01). The mean number of basal roots per genotype in 2011 was 7.89. LSD (0.05) between genotypes was 4.09 (Tables 1.6-8). Genotypes with the greatest numbers of basal root were Sierra, Croissant, Buster, T-39, Medicine Hat, Sedona, GN9-4, Maverick, Poncho and Matterhorn. These genotypes were not statistically different for basal root number. Genotypes with the least number of basal roots were Coyne, BelNeb-RR-1, Verano, PR-0340-3-3-1, US1140, USRM-20, I9365-31, Topaz, Schooner and TARS-VCI-4B. Forty-nine different genotypes were not statistically different for lowest number of basal roots. 53 In 2011, many of the genotypes with the greatest number of basal roots where from the Durango class with pinto or great northern seed types. Those genotypes with the lowest number of basal roots represented all seed classes. In 2012, significant differences in basal root number were not detected for genotype (p= 0.78). The mean number of basal roots per genotype was 4.21 and the LSD was 1.97 (Tables 1.9-11). Those with the greatest number of basal roots in 2012 were La Paz, UI-114, PR-0443151, Marquis, Domino, Harold, Rosetta, 115M, UI-537 and ROG 312. Those with the lowest numbers of basal roots were Pink Floyd, PR-0340-3-3-1, Montrose, Mayflower, Verano, BAT 477, Topaz, ABCP-8, Yolano and IBC-301-204. It is interesting to observe significant basal root number differences in 2011, but not in 2012. The growing environment was different between seasons, and there are clearly many factors contributing to the basal root number characteristic. Those factors are not limited to, but appear to include soil type and seasonal rainfall. Basal Root Angle The overall ANOVA of basal root angle showed that there was no significant differences for the genotype by environment interaction (p=0.11) and so data for basal root angle was analyzed as a compiled data set without a separate analysis for year (Tables 1.6-11). The ANOVA for the combined analysis showed no significant differences for genotype (p=0.06). Average basal root angle ranged from 20° to 50° with a mean of 35°. We expected these values to be highly dynamic and provide an indirect measure of rooting depth but this simply was not true for the genotypes in the years and locations of our study. 54 Taproot Diameter ANOVA analysis of taproot diameter detected no significant differences for taproot diameter for the genotype × environment interaction (p=0.98). However, the ANOVA detected significant differences for genotype (p=0.0004). This suggests that taproot diameter is a highly consistent trait across environments. The analysis of taproot diameter was combined and environment used as a random statement nested in replication (Tables 1.3-5). Average taproot diameter for the genotypes ranged from 1.4 mm to 3.5 mm with a mean of 2.3 mm and there was highly significantly differences between genotypes (p<0.0001). Interestingly Rosetta, a new pink bean variety from Michigan State University (Kelly et al., 2012) had the largest average taproot diameter at 3.5 mm. Thirteen genotypes did not differ significantly from Rosetta for average taproot diameter. This group consists primarily of upright type II navy and black bean genotypes with a few great northerns, and one pinto and one small red genotype. When examining taproot diameter by race, Durango and Jalisco had identical average values at 2.2, whereas genotypes of the Mesoamerican race had a mean of 2.5 mm. This is consistent with prior observations of race Mesoamerican genotypes, as they are known to have a dominant taproot associated with the upright type II indeterminate growth habit (Kelly, 1998). Overall Root Score In 2011, significant differences for overall root score were detected (p=0.02). Overall root score ranged from 1.5 to 6.9 with a mean of 3.79 and the LSD between genotypes was 2.39 (Tables 1.6-8). Genotypes with the highest scores were La Paz, Matterhorn, Merlot, PT 7-2, Sawtooth, Raven, Harold, Midnight, GN#1Sel127 and Medicine Hat. Those with the lowest average scores were Medalist, Domino, Pink Floyd, Indeterminate Jamaica Red, TARS-VCI-4B, Topaz, Gloria, UI-114, Kimberly and Nodak. 55 Significant differences were detected between genotypes in 2012 for overall root score (p=0.02). Overall root score ranged from 2.0 to 5.5 with a mean of 3.5 and the LSD between genotypes was 1.97 (Tables 1.9-11). Those genotypes with the highest overall root score were UI-537, GN9-4, Rosetta, NE2-09-3, PR-0443-151, Matterhorn, Coyne, Aifi-Wuriti, Common Pinto and Domino. Those genotypes with the lowest scores were BelNeb-rr-1, PR-0340-3-3-1, SER48, Montrose, IBC-301-204, Gemini, I9365-31, UI-59, CDC-Crocus and Orion. It is encouraging to see that several genotypes repeat in 2011 and 2012 as being not significantly different than the best performing genotype. Those genotypes include: Matterhorn, Othello, Seafarer, Avalanche, Merlot, Midnight and La Paz. All of these genotypes were included in the best performing genotypes for taproot diameter and basal root number in at least one of the years. These data suggest that this rating method identifies genotypes with superior qualities for either root trait and this method of rating genotypes could be used by a breeder from to quickly evaluate genotypes in the field on an annual basis. Yield In 2011, significant differences were detected between genotypes for yield (p<0.0001). Yield ranged from 581 to 4802 kg ha-1 with a mean of 3391 kg ha-1. LSD between genotypes for yield in 2011 was 1079 kg ha-1 (Tables 1.6-8). Genotypes with the highest yield in 2011 were Merlot, Matterhorn, A285, Santa Fe, Lariat, La Paz, IBC-301-204, Orion, Zorro, Aifi Wuriti and PT7-2. Those with the lowest yield were Sawtooth, GN#1Sel27, UI-425, Viva, San Juan, TARS-VCI-4B, Harold, UI-259, Fisher, and Victor. In 2012, significant differences were detected between genotypes for yield (p<0.0001). Yield ranged from 1812 to 4415 kg ha-1 with a mean of 3057 kg ha-1. LSD between genotypes 56 for yield in 2012 was 739 kg ha-1 (Tables 1.9-11). Genotypes with the highest yield were PT7-2, PR-0320-3-3-1, 115M, I9365-31, Merlot, NW-590, Eldorado, Lariat, Kodiak and GN9-1. Genotypes with the lowest yields were Gloria, Verano, UI-114, Poncho, SER48, Roza, UI-425, IBC-301-204, Kimberly and TARS-VCI-4B. A comparison of the highest performing genotypes for yield over the two years showed that only three genotypes repeat in both years; PT7-2, Merlot (Hosfield et al., 2004) and Lariat (Osorno et al., 2010). The differences in yield performance over the two locations must be attributed to the differences in environments. These three genotypes however maintained consistent performance over the two years. PT 7-2 also performed among the best genotypes for basal root number in 2012, overall taproot diameter, overall score in 2011 and harvest index in 2012. Merlot performed equivalent to the best genotypes for overall score in 2011 and basal root number in 2012. Lariat, however, did not perform as well as any of the genotypes for root traits measured or harvest index in either year. Lariat must have another mechanism contributing to its yield which may be worth investigating. Harvest Index In 2011, significant differences were detected between genotypes for harvest index (p<0.0001). Harvest index ranged from 2.7 to 52% with a mean of 34.7%. LSD between genotypes for harvest index in 2011 was 12.3% (Tables 1.6-8). Genotypes with the highest harvest indexes were US-1140, Buster, Maverick, Nodak, Gemini, IBC-301-204, Medicine Hat, Matterhorn, La Paz and USPT-CBB-5. Genotypes with the lowest harvest indexes were GN#1Sel27, Sawtooth, Viva, TARS-VCI-4B, San Juan, Harold, Fisher, 115M, Victor, and NW63. 57 Genotypes that exhibited photoperiod sensitivity did not perform well in Michigan due to lack of adaptation. Those genotypes did not mature and were green at harvest so a large part of their biological weight was retained in the vegetative parts resulting in low harvest index. Two genotypes that failed to mature were San Juan and Fisher (Fisher et al., 1995) pinto bean cultivars that were bred for severe drought conditions in southwest Colorado and perform quite well under drought stress in that region. In Michigan, however, these genotypes are photoperiod sensitive and produced extensive vegetative growth and did not sense the day length or temperature trigger to produce seed and thus performed poorly due to poor partitioning. This serves as an example that in breeding for drought tolerance local adaptation is more critical than certain drought tolerance traits. Due to low yield and lack of adaption, these genotypes were not included in the study in 2012. In 2012, significant differences were detected between genotypes (p<0.0001). Harvest index ranged from 20.6 to 51.6% with a mean of 41.3% and the LSD between genotypes in 2012 was 6.0% (Tables 1.9-11). Genotypes with the highest harvest indexes were Montrose, Medicine Hat, PT7-2, Pink Floyd, Kodiak, UI-537, Othello, CENTA Pupil, T-39 and Rosetta. Genotypes with the lowest harvest indexes were Viva, Harold, Roza, TARS-VCI-4B, A-55, PR-443-151, UI-425, Victor, Sierra and NW-63. Genotypes that were not significantly different from the best performing genotype in both years were Medicine Hat, Kodiak (Kelly et al., 1999a), CENTA Pupil, US-1140, Matterhorn, USRM-20, Buster and IBC 301-204. These genotypes represent important sources of partitioning under drought stress and should be used in further development of drought tolerant dry beans. 58 Replacement genotypes were used in 2012 for those genotypes not producing enough quality seed due to photoperiod sensitivity issues mentioned earlier. One of those replacement genotypes, Rosetta, performed very well in 2012. Rosetta is a new pink bean cultivar developed at Michigan State University (Kelly et al., 2012) that had high harvest index and could be used as an important source for efficient partitioning. Comparisons of Root Traits with Yield Interestingly many of those genotypes that ranked highest for basal roots ranked lowest for average taproot diameter. Those genotypes with the greatest taproot diameter were mainly from the Mesoamerican race whereas those with greater number of basal roots were Durango seed types. When comparing those observations with yield and harvest index, some of the best performing genotypes belonged to the Durango class including: Matterhorn, Medicine Hat and PT7-2. Interestingly, those genotypes were also the only three non-Mesoamerican genotypes that performed equivalent to the best genotype for taproot diameter. Likewise, Rosetta of the Jalisco race was included as one of the best performing for taproot diameter in the combined analysis and for basal root number in 2012. These genotypes are superior sources for root traits as well as for yield and harvest index and could be used for breeding purposes. These genotypes represent prior introgression between Durango and Mesoamerican races where the goal was to combine improved upright plant architecture from race Mesoamerican with seed traits of the race Durango pinto and great northern genotypes (Kelly, 2000; Kelly et al., 1999b). It is not surprising that some of the distinct root traits of these races were also combined but not evaluated in prior studies. Integrating taproot diameter and basal root number would be important to improving drought tolerance of dry beans. As this study demonstrated, it seems that combining these two traits may be difficult as most genotypes possessing one of these two traits 59 did not possess the other. Since Matterhorn and Medicine Hat have both qualities, it seems possible to combine the traits in a single genotype. The best strategy to combine several traits like taproot diameter, basal root number and yield, may be to follow the strategy suggested by Singh (1995). In the study, he used high yielding double-cross lines to study drought tolerance. The lines were also derived from interracial crosses of Durango and Mesoamerican seed types and inter gene pool crosses between Mesoamerica and Nueva Granada. The study tested performance of these high yielding lines under drought stress and compared them to non-stress performance. The study concluded that using four parents provided enough variability at the F2 stage to use single seed descent and bulk breeding methods to produce drought tolerant dry bean lines. This strategy of combining an interracial four-way cross with SSD and bulk breeding methods was an effective strategy in breeding future drought tolerant dry beans such as SEA 10. Another breeding method used successfully by bean breeders to maximize recombination between races of the same gene pool was recurrent selection (Kelly and Adams, 1987). Phenotypic recurrent selection was used to introduce plant architectural traits form the Mesoamerican race into race Durango where the trait was previously absent. Conclusions Root traits, yield and harvest index were identified as being highly significant between the 96 genotypes studied. These included number of basal roots, taproot diameter, overall root score, yield and harvest index. Yield performance was the most important characteristic when identifying drought tolerance but certain root traits were also essential for drought tolerance. The cultivars, Matterhorn and Medicine Hat, were identified as superior genotype as they ranked high for all root traits as well as harvest index and yield in at least one year of the study. In general, 60 genotypes from the Mesoamerican race possessed the best overall drought tolerance, as determined by yield performance under stress, among the range of the 96 genotypes evaluated. 61 REFERENCES 62 REFERENCES Acosta-Gallegos J.A., Shibata J.K. (1989) Effect of water-stress on growth and yield of indeterminate dry-bean (Phaseolus-vulgaris) cultivars. Field Crops Research 20:81-93. Asfaw A., Blair M.W. (2012) Quantitative trait loci for rooting pattern traits of common beans grown under drought stress versus non-stress conditions. Molecular Breeding 30:681695. Bonser A.M., Lynch J., Snapp S. (1996) Effect of phosphorus deficiency on growth angle of basal roots in Phaseolus vulgaris. New Phytologist 132:281-288. Burridge J. (2012) High throughput phenotyping of root system architecture, http://plantscience.psu.edu/research/labs/roots/methods/field-methods/field-crop-rootcrown-sampling-and-analysis. Fisher A.G., Brick M.A., Wood D.R., Stack M., Schwartz H.F., Ogg J.B., Pearson C.H., Shanahan J.F., Ballarin M. (1995) Registration of Fisher pinto bean.Crop Science 35:1511-1511. Gepts P., Bliss F.A. (1985) F1 hybrid weakness in the common bean: Differential geographic origin suggets two gene pools in cultivated bean germplasm. Journal of Heredity 76:447450. Ho M.D., Rosas J.C., Brown K.M., Lynch J.P. (2005) Root architectural tradeoffs for water and phosphorus acquisition. Functional Plant Biology 32:737-748. Hosfield G.L., Varner G.V., Uebersax M.A., Kelly J.D. (2004) Registration of 'Merlot' small red bean. Crop Science 44:351-352. Kelly J.D. (1998) Bean roots - a plant breeder’s perspective. Annual Report Bean Improvement Cooperative 41:214-215. Kelly J.D. (2000) Remaking bean plant architeture for efficient production. Advances in Agronomy 71:109-143. Kelly J.D., Adams, M.W. (1987) Phenotypic recurrent selection in ideotype breeding of pinto beans. Euphytica 36:69-80. Kelly J.D., Cichy K.A. (2012) Dry bean breeding and production technologies, in: M. Siddiq, M.A. Uebersax (Ed.), Dry Beans and Pulses: Production, Processing, and Nutrition, Wiley Blackwell Publishing Co. Oxford, U.K. pp. 23-54. 63 Kelly J.D., Hosfield G.L., Varner G.V., Uebersax M.A., Taylor J. (1999a) Registration of 'Kodiak' pinto bean. Crop Science 39:292-293. Kelly J.D., Hosfield G.L., Varner G.V., Uebersax M.A., Taylor J. (1999b) Registration of 'Matterhorn' great northern bean. Crop Science 39:589-590. Kelly J.D., Varner G.V., Cichy K.A., Wright E.M. (2012) Registration of 'Rosetta' pink bean. Journal of Plant Registrations 6:229-232. Muñoz-Perea C.G., Terán H., Allen R.G., Wright J.L., Westermann D.T., Singh S.P. (2006) Selection for drought resistance in dry bean landraces and cultivars. Crop Science. 46:2111-2120. Osorno J.M., Grafton K.F., Rojas-Cifuentes G.A., Gelin R., Wal A.J.V. (2010) Registration of 'Lariat' and 'Stampede' Pinto Beans. Journal of Plant Registrations 4:5-11. Porch T.G., Ramirez V.H., Santana D., Harmsen E.W. (2009) Evaluation of common bean for drought tolerance in Juana Diaz, Puerto Rico. Journal of Agronomy and Crop Science 195:328-334. Ramirez-Vallejo P., Kelly J.D. (1998) Traits related to drought resistance in common bean, Euphytica. 99:127-136. Robertson B.M., Hall A.E., Foster K.W. (1985) A field technique for screening for genotypic differences in root-growth. Crop Science 25:1084-1090. Rosielle A.A., Hamblin J. (1981) Theoretical aspects of selection for yield in stress and nonstress environments. Crop Science 21:943-946. Sadok W., Sinclair T.R. (2011) Crops yield increase under water-limited conditions: review of recent physiological advances for soybean genetic improvement, in: D. L. Sparks (Ed.), Advances in Agronomy, Vol 113. pp. 313-337. Schneider K.A., Rosales-Serna R., Ibarra-Perez F., Cazares-Enriquez B., Acosta-Gallegos J.A., Ramirez-Vallejo P., Wassimi N., Kelly J.D. (1997) Improving common bean performance under drought stress. Crop Science 37:43-50. Singh S.P. (1995) Selection for water-stress tolerance in interracial populations of common bean. Crop Science 35:118-124. Singh S.P. (2007) Drought resistance in the race Durango dry bean landraces and cultivars. Agronomy Journal 99:1219-1225. Singh S.P., Gepts P., Debouck D. (1991) Races of common bean (Phaseolus vulgaris, Fabaceae). Economic Botany 45:379-396. 64 Singh S.P., Teran H., Lema M., Dennis M.F., Hayes R., Robinson C. (2008) Development of large-seeded high-quality, high-yielding great northern dry bean 'Hungerford' and 'Sawtooth'. Journal of Plant Registrations 2:174-179. Sprague C.L. (2012) Weed control guide for field crops, Extension Bulletin E-434, Michigan State Univeristy, East Lansing, http://www.msuweeds.com/publications/weed-controlguide/. Teran H., Singh S.P. (2002) Comparison of sources and lines selected for drought resistance in common bean. Crop Science 42:64-70. Trachsel S., Kaeppler S.M., Brown K.M., Lynch J.P. (2011) Shovelomics: high throughput phenotyping of maize (Zea mays L.) root architecture in the field. Plant and Soil 341:7587. Urrea C.A., Yonts C.D., Lyon D.J., Koehler A.E. (2009) Selection for drought tolerance in dry bean derived from the Mesoamerican gene pool in western Nebraska. Crop Science 49:2005-2010. USDA-ERS. (2012) Dry Beans, http://ers.usda.gov/topics/crops/vegetables-pulses/drybeans.aspx#major. Accessed December 201 65 Chapter 2. Evaluation of 96 dry bean genotypes grown in pouches for seedling root characteristics related to drought tolerance Introduction Common bean (Phaseolus vulgaris L.) is the most important food legume grown for direct human consumption worldwide (Broughton et al., 2003). Beans are an important source of protein in many developing nations. Bean production is also an important part of Michigan’s agricultural landscape and economy. Michigan is the second largest producer of dry beans in the United States producing fourteen percent of U.S. production. Michigan produces primarily black and navy beans, but grows smaller quantities of various other classes of dry beans including small red and cranberry beans (USDA-ERS, 2012). Biotic and abiotic stresses pose a major threat to production of dry beans worldwide. The most limiting factor in dry bean production worldwide is drought stress. In fact, ninety-five percent of dry beans produced in Michigan depend on seasonal rain fall to complete their life cycle. As the world climate changes so has the climate in Michigan. Rainfall measurements dating back to 1979 from the Saginaw Valley Research Farm showed that yearly rainfall is decreasing. Prior to 1997, yearly rainfall averaged above 750 mm. From 1998 to present yearly rainfall has been below that average with the exception of two years (SVREC, 2012). Drought drastically affects both yield and quality of dry bean (Urrea et al., 2009). Increasing scarcity of fresh water resources makes the development of dry beans that are tolerant to drought an important objective especially in Michigan if future trends of decreasing yearly rainfall continue. Elucidating the precise factors that confer drought tolerance is very difficult due to the quantitative and inconsistent nature of the traits that contribute to plant productivity. One such trait is the root morphology. Roots are the physical feature of the plant that facilitates water uptake, which makes them an important trait to investigate as one component to improve drought 66 tolerance. Field studies have been conducted to determine which root traits contribute to drought tolerance in common bean (Ho et al., 2005; Robertson et al., 1985). Other studies have been conducted in more controlled systems like the laboratory and greenhouse to study in greater detail the relationship between roots and drought stress (Himmelbauer et al., 2004; Jones and Ljung, 2012; Lobet et al., 2011; Lodeiro et al., 2000; Lynch and Vanbeem, 1993; Sassi Aydi et al., 2008). Laboratory experiments provide the opportunity to eliminate certain environmental factors that confound genotypic response with the environment. One concern is that laboratory and greenhouse conditions do not represent conditions in the field and results are difficult to relate to actual field conditions. Field work is an important complement to greenhouse studies as field conditions document real scenarios confronting the crop. However, greenhouse studies are often utilized because collecting phenotypic data on bean roots in the field is arduous and time consuming. Many areas of bean production in temperate zones have only a single annual growing season with limited field space. Thus the decision on which traits to be studied in the field must be made prudently as the numbers of genotypes increase. Studies that can be conducted during the off-season are effective tools as they are time saving and resource efficient. Many systems exist to study roots under more controlled environmental conditions. Those include hydroponics, growth pouches (McMichael et al., 1985), slant tubes, the use of sand systems for easy extraction of roots, and different container designs to study young seedling plants (Lynch and Vanbeem, 1993). New methods for imaging, processing and data collection of roots have been developed such as computer software programs: WinRHIZO, ROOTEDGE and SmartRoot (Himmelbauer et al., 2004; Lobet et al., 2011). A more advanced technique uses a transparent gellan gum system for plant growth, a camera collecting 2-dimensional photos and 67 RootReader3D software to create three-dimensional platforms for seedling root architecture study (Clark et al., 2011). Software programs have an advantage in the study of roots as computers can record thousands of data points over time that can be assembled into constructing a model root system. Thus the use of computer-based systems has and will have great potential to revolutionize root architecture phenotyping. It is important, however, that the data collected in the laboratory setting can be associated with field-based traits that regulate plant growth and productivity under different environmental conditions. Several groups are using data obtained by these programs to make selections and identify QTL associated with these traits in common bean (Asfaw and Blair, 2012; Beebe et al., 2006; Butare et al., 2011). Beebe et al. (2006) used total root length, total root surface area and average diameter obtained from root analysis software to identify QTLs for root architecture correlated with phosphorus acquisition. Likewise, Butare et al. (2011) used mean root diameter, number of root tips and specific root length from a root analysis program to identify sources of resistance for aluminum toxicity and soil drying stresses in dry bean. These are practical applications for using root analysis software for specific selection in dry bean breeding. Different methods of obtaining data on root studies are available, but many factors must be considered when selecting a suitable method. Although different systems offer specific advantages, the availability of financial resources or even knowledge to use the system may prevent their use. The methods used in this study, namely the pouch method for seedling root growth and analysis using WinRhizo software, were chosen for their relatively low cost and availability of resources. The objectives of this study were to compare differences in root morphology in a large group of bean genotypes that represent different seed classes using the pouch method. The pouch method for seedling root growth and the WinRhizo software (Reagent Instruments Inc., 68 2012, Canada) were chosen to analyze root morphology in seedling stages. Findings of this study were compared to historical knowledge of the genotypes response to drought and to field performance data collected during the 2011 and 2012 field seasons in Michigan. Materials Ninety-six dry bean genotypes were selected based on prior breeder knowledge of their reaction to drought. The genotypes were assembled as part of the Common Bean Coordinated Agricultural Project (BeanCAP) and have been used in a number of field studies across the country to gain greater insights into drought tolerance of common bean. The 96 genotypes were previously selected by a group of breeders in the United States and Puerto Rico for more extensive testing as part of the Beancap research projects being conducted in different production areas. The selections represent past and present dry bean varieties that belong to the Middle American gene pool (Gepts and Bliss, 1985). There were fourteen black seeded, sixteen great northern, eight navy, ten pink, thirty-five pinto, eight small red, one red-mottled, one carioca and three tan genotypes. Tepary bean was also added to this study due to its recognized drought tolerance and the limited information on its seedling rooting characteristics. A complete list of the genotypes used in this study is shown in Tables 2.2-4. 69 Table 2. 1 Average values for six root traits grown in the pouches and analyzed using WinRhizo for 26 Mesoamerican race seed types. Genotype 115M A-55 Aifi Wuriti Domino Eclipse F04-2801-4-1-2 I9365-31 Jaguar Midnight PR 0443-151 Raven Shania Shiny Crow T-39 Zorro Mean (black) Seed Type black black black black black black black black black black black black black black black Length Whorl cm count 412 1.60 558 1.58 720 1.60 602 1.50 646 2.00 535 1.40 552 1.58 516 1.90 821 1.97 458 2.00 508 1.00 538 1.30 292 1.80 402 1.60 545 1.40 540 1.61 Avg Diameter mm 0.32 0.32 0.30 0.33 0.30 0.33 0.32 0.34 0.33 0.32 0.31 0.31 0.36 0.31 0.29 0.32 Surface Area cm2 42 54 68 62 61 55 123 57 86 46 50 53 31 39 49 58 70 Root Volume cm3 0.34 0.42 0.51 0.51 0.46 0.45 0.50 0.57 0.72 0.37 0.39 0.42 0.27 0.31 0.35 0.44 Tips count 1422 1588 1639 1224 1558 1318 1401 1570 1859 1407 1543 1266 934 1597 1795 1475 Table 2.1 (cont’d). Genotype Medalist Navigator Schooner Seafarer Verano Avalanche C-20 Mayflower BAT 477 BAT93 SEA 10 A285 tepary Mean (navy) Grand Mean LSD 0.05 CV% Seed Type navy navy navy navy navy navy navy navy tan tan tan carioca tepary Length Whorl cm count 353 2.67 304 2.33 297 1.50 332 1.25 370 1.67 481 1.30 407 1.38 426 1.50 459 2.88 496 2.20 645 1.90 655 1.55 188 1.67 435 1.84 494 1.72 388 33 0.98 25 Surface Area cm2 34 34 30 35 42 45 41 42 44 45 69 67 20 44 52 Avg Diameter mm 0.32 0.38 0.33 0.39 0.36 0.30 0.32 0.34 0.30 0.28 0.34 0.33 0.37 0.33 0.32 Root Volume cm3 0.27 0.31 0.25 0.30 0.37 0.34 0.34 0.33 0.33 0.32 0.59 0.54 0.18 0.36 0.40 Tips count 1379 855 749 658 705 1474 1110 1292 1802 1710 1320 1550 416 1217 1360 ns ns ns ns ns ns ns ns 71 Table 2. 2 Average values for six root traits grown in the pouches and analyzed using WinRhizo for 50 Durango race seed types. BelNeb-RR-1 Beryl R CDC Crocus Coyne Gemini GN#1Sel27 GN9-1 GN9-4 Marquis Matterhorn Orion Sawtooth UI-425 UI-59 US-1140 Weihing Mean (GN) Seed Type GN GN GN GN GN GN GN GN GN GN GN GN GN GN GN GN Length Whorl cm Genotype Surface Area count cm2 mm cm3 count 842 832 573 586 1047 628 525 834 587 899 723 425 558 670 1109 476 707 1.46 1.70 1.90 1.20 1.90 2.36 2.60 1.70 1.30 2.03 1.15 2.10 2.33 2.10 1.70 1.10 1.79 78 85 64 65 110 64 63 94 62 97 71 43 53 68 108 51 74 0.31 0.33 0.35 0.37 0.33 0.33 0.39 0.36 0.34 0.34 0.31 0.40 0.31 0.33 0.32 0.38 0.34 0.58 0.70 0.57 0.58 0.93 0.53 0.60 0.84 0.52 0.84 0.56 0.36 0.40 0.56 0.84 0.44 0.62 2083 2118 1693 1088 2065 1749 1217 1432 1722 2288 1819 1001 2176 1969 1928 1137 1718 72 Avg Diameter Root Volume Tips Table 2.2 (cont’d). ABCP-8 Bill Z Buckskin Buster Chase Common Pinto Croissant Fisher Kimberly Kodiak La Paz Lariat Maverick Medicine Hat Montrose ND-307 NE2-09-3 Nodak NW-590 NW-63 pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto Length Whorl Surface Area cm Genotype Seed Type count cm2 mm cm3 count 670 778 705 689 571 602 687 429 809 501 488 504 541 860 326 799 648 807 462 542 1.70 1.68 2.38 2.38 1.48 2.10 2.63 1.80 1.50 1.33 1.40 1.75 1.92 2.30 2.00 1.30 1.70 2.00 1.50 1.63 70 75 78 72 65 70 66 45 82 65 53 66 64 84 34 83 73 83 52 65 0.33 0.31 0.36 0.33 0.40 0.40 0.31 0.35 0.34 0.49 0.36 0.43 0.46 0.31 0.42 0.33 0.39 0.37 0.42 0.38 0.59 0.58 0.69 0.60 0.60 0.66 0.51 0.38 0.67 0.68 0.46 0.71 0.61 0.66 0.30 0.69 0.67 0.68 0.47 0.62 2036 2431 1424 1469 940 1442 2838 1866 2315 841 2034 1329 789 2747 802 2629 1600 1456 981 904 73 Avg Diameter Root Volume Tips Table 2.2 (cont’d). Othello Poncho PT7-2 PT9-17 Quincy San Juan Santa Fe Sierra Stampede TARS-VCI-4B Topaz UI-114 USPT-CBB-1 USPT-CBB-5 Mean (pinto) Grand Mean LSD 0.05 CV% pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto Length Whorl Surface Area cm Genotype Seed Type count cm2 mm cm3 count 779 593 542 889 557 487 893 712 589 828 700 845 688 753 655 672 1.60 1.17 2.03 2.23 1.60 2.00 1.30 2.50 1.40 1.80 2.00 1.88 1.25 1.70 1.79 1.79 86 64 68 97 69 47 102 77 63 92 88 95 74 76 72 72 0.36 0.34 0.41 0.35 0.40 0.32 0.36 0.40 0.35 0.36 0.44 0.36 0.36 0.32 0.37 0.36 0.76 0.54 0.68 0.87 0.69 0.36 0.93 0.68 0.54 0.83 0.89 0.85 0.63 0.62 0.64 0.63 1923 1841 1055 2513 1518 1442 2078 1700 2615 1489 1233 1543 1356 2760 1704 1708 388 33 0.98 25 ns ns ns ns ns ns ns ns 74 Avg Diameter Root Volume Tips Table 2. 3 Average values for six root traits grown in six pouches and analyzed using WinRhizo for 19 Jalisco race seed types. Genotype Gloria Harold Pink Floyd ROG 312 Roza Sedona UI-537 Victor Viva Yolano Mean (pink) Seed Type pink pink pink pink pink pink pink pink pink pink Length cm 425 447 554 344 397 421 348 496 328 516 427 Whorl count 1.66 2.23 1.60 2.80 2.00 3.00 2.00 2.13 2.75 2.33 2.25 75 Surface Area cm2 45 45 56 36 46 43 36 55 33 51 44.65 Avg Diameter mm 0.35 0.34 0.33 0.34 0.44 0.35 0.42 0.37 0.37 0.35 0.37 Root Volume cm3 0.39 0.37 0.46 0.30 0.43 0.34 0.33 0.49 0.27 0.42 0.38 Tips count 1288 1658 1910 1526 1049 1350 1751 1208 1193 2253 1518.60 Table 2.3 (cont’d). Genotype Seed Type CENTA Pupil Common Red Mexican DOR 364 F07-449-9-3 IBC 301-204 Merlot UI-239 USRM-20 Indeterminate Jamaica Red Mean (small red) Grand Mean LSD 0.05 CV% small red small red small red small red small red small red small red small red red mottled Length cm 534 490 630 471 510 519 452 710 702 558 489 Whorl count 1.93 2.04 1.70 1.70 1.90 2.50 1.67 1.78 2.65 1.98 2.12 Surface Area cm2 48 59 64 51 56 57 54 79 70 60 52 Avg Diameter mm 0.29 0.41 0.33 0.36 0.36 0.37 0.38 0.35 0.33 0.35 0.36 388 33 0.98 25 ns ns ns ns 76 Root Volume cm3 0.35 0.56 0.53 0.45 0.50 0.52 0.52 0.70 0.56 0.52 0.45 Tips count 2335 760 1390 1196 1247 1580 1371 1987 2113 1553 1535 ns ns ns ns Methods The genotypes used in this study were grown using the pouch method and analyzed using WinRhizo root analysis software (Reagent Instruments Inc., 2012, Canada) from March to October 2012. A plastic pouch method using plastic sheet protectors, paper towel as a wick and an office paper binder were used to grow bean seedling roots. Rigorous seed cleaning measures were performed to avoid fungal problems during germination. Briefly, ten seeds of each genotype were selected based on a visually clean appearance and were first soaked in a three percent Clorox bleach and water solution for three minutes. The bleach solution was then discarded and seeds were rinsed with Millipore water and then soaked in 70 percent ethanol for three minutes after which the ethanol solution was discarded. Seeds were then rinsed three times with Millipore water. Excess water was removed. Seeds were then coated with 200 ml of the ampicillin antibiotic solution (80µg/ml-1) to control bacterial outbreaks. Using forceps, seeds were inserted into sterile water agar in a petri dish with the hypocotyls down. Petri dishes containing surface sterilized seeds were labeled with the genotype and date and then placed in an incubator at 29°C for approximately 72 hours. Paper towels (Scotts Extreme Shop Towels) were cut to fit into a clear plastic sheet protector approximately 20 × 28 cm in size. The top edge was folded down about 2.5cm and then folded back up to create a trough for seed placement. The prepared paper towels to be used as pouch wicks were wrapped in aluminum foil and autoclaved for 25 minutes at 25 psi. After about 72 hours in the incubator, the radicles on the seeds were about 2.5 to 3 cm in length and ready to be placed in the pouch. In a sterile hood, five suitable, clean seeds from the 10 that were germinated were selected. The root research group at Pennsylvania State University led by Jonathon Lynch have identified the basal root whorl number trait as a potentially important trait in common bean 77 drought tolerance. According to Lynch (2012), this trait is highly variable between dry bean genotypes. During germination, basal roots emerge after the taproot from the bean hypocotyl (Bonser et al., 1996). Basal roots emerge in groups and form a ring around the taproot in common bean. Each ring of basal roots is considered a whorl. The number of root whorls of each genotype was recorded before beans were placed in the pouches. At this point, each selected seed was designated a letter code, A-E, to differentiate between seeds of the same genotype. An autoclaved paper towel was placed in a sheet protector with the trough at the top. Each pouch was then labeled with the date, identification number and its designation code. Using ethanol sterilized scissors; a small slit was cut into the trough to allow root passage. Carefully, the selected seedling was positioned in the pouch with roots oriented down toward the bottom of the pouch. The paper towel was then wetted with 30 mL of half-strength Hoagland’s solution. The entire pouch was then placed in a binder. Approximately ten pouches were stored in a single 8 cm binder and stored in a growth chamber. The growth chamber was set on 16-hr day and 8-hr night at 26°C. Pouches were watered every other day with 10 mL of half-strength Hoagland’s solution (Hoagland and Arnon, 1938) for the first week and 15-20 mL for the second week. After 2 weeks, plants were removed from the paper towel, submerged in water on an acrylic tray and scanned at 400 dpi and a 16-bit gray scale using an Epson 1700 scanner. Root images were edited using photo editing software to remove the edges of the tray and any background noise that would be incorrectly counted as roots when analyzed. Root images were then analyzed using WinRhizo (Regent Instruments, Inc., 2012) in a single batch analysis. WinRhizo software recorded data for root length, surface area, root volume, average root diameter and number of root tips. Length is the sum of the lengths of all roots in the sample. Surface area is the total area of all surfaces of the roots in the sample. Root volume 78 is the area that the roots occupy. Average root diameter is the diameter of every root in the sample averaged. Root tips are the number of root tips in the sample. Statistical Analysis Data was analyzed using SAS 9.3 (SAS Institute Inc., Cary, NC, 2010). Analysis of variance was conducted using PROC MIXED. Mean comparisons were performed using least significant difference (LSD) with a significance level at α=0.05 (Table 2.1). In order account for the effect of genotypic seed size on root traits due to differences in vigor, germination and growth, 100-seed weight was introduced as a covariate into the model. Pearson correlation analysis was conducted using PROC CORR to determine associations between root traits and yield data from the field for the genotypes grown in the 2011 and 2012 seasons. Pearson correlations were performed and were either insignificant or were highly correlated but were found to be intuitive and thus did not contribute further insight to the analysis. Table 2. 4 High, low, mean, F test and LSD(0.05) values for six root parameters of 96 dry bean genotypes plus tepary bean tested in the growth pouch and analyzed using WinRhizo software. Length Whorl Surface Area Avg Diameter 2 Root Volume 3 cm cm cm count mm 1109 3.00 123 0.49 0.93 High 188 1.00 20 0.28 0.18 Low 581 1.84 62 0.35 0.53 Mean 0.080 0.014 0.119 0.268 0.269 F testgenotype 388 0.98 ns ns ns LSD (0.05) ns= LSD(0.05) not listed for traits that were not significantly different. 79 Tips count 2838 416 1564 0.508 ns Results and Discussion Basal Root Whorls The number of whorls for each individual germinated seed was collected and averages for each genotype were compiled (Tables 2.2-4). Whorl number for each seedling was counted and although this was thought to be a constant number for each genotype, variation was noted. Whorl numbers ranged from one to three with the mean value for all genotypes at 1.8 (Table 2.1). Significant genotypic differences were found for average whorl number (p= 0.0141). Twenty-seven genotypes did not differ significantly in whorl number from Sedona and BAT 477, each with three whorls. Sedona (Kelly et al., 2006) is a pink bean variety released by Michigan State University and BAT 477 is a recognized source of drought tolerance from CIAT (Sponchiado et al., 1989). Sedona was developed under rainfed conditions in Michigan and may possess drought tolerance recorded in other pink bean genotypes (Singh, 2007). BAT 477 developed at CIAT in Cali, Colombia has an indeterminate prostrate type III growth habit (Singh et al., 2001) and a root system that is known to be highly adaptable to drought conditions (Sponchiado et al., 1989). The genotype with the lowest basal root whorl number was Raven with a single root whorl. Raven is a black bean variety from Michigan State University released in 1994 and is not recognized as being drought tolerant (Kelly et al., 1994). Other genotypes that had greater than 2.5 basal root whorls were Medalist (navy), Croissant (pinto), Sierra (pinto), GN9-1 (GN), Indeterminate Jamaica Red (red mottled), ROG 312 (pink) and Viva (pink) (Table 2.2, 2.3, 2.4). Many different seed types studied are included in this group showing that variation existed for the trait. Possessing variation in all the seed classes means that this trait could possibly be introgressed into different seed classes. Notably, the pink seed type represents an important source of high whorl number with an average whorl number of 2.25. No other seed class had an average whorl number greater than 2. Pink beans have been identified as a source 80 of drought tolerance (Hayes and Singh, 2007; Singh, 2007). Viva (pink) specifically mentioned above as having a high basal root whorl number, has a history of possessing drought tolerance in the field (Singh, 2007). Pink beans appear to be an important source of high basal root whorl numbers. This trait is a relatively new trait to the bean breeding community that is easy to measure and thus it is of interest to find significant differences in this group of genotypes. Total Root Length Average root length in the pouches ranged from 188 cm to 1109 cm with a mean value of 582 cm across the 96 genotypes. Average root length was significantly different (p=0.0795) between genotypes. Nineteen genotypes with the greatest lengths were not significantly different and included great northern cultivars US-1140, Gemini, Matterhorn and pinto varieties Santa Fe and PT9-17 (Table 2.3). Forty-seven genotypes with the lowest average root length were not significantly different from one another and the five genotypes with the lowest average root lengths included tepary, Shiny Crow (black), Schooner (navy), Navigator (navy) and Montrose (pinto) (Table 2.2, 2.3, 2.4). Genotypes with the greatest average root length were either great northern or pinto seed types which belong to the Durango race of beans. Durango race originates from the semi-arid highlands of Mexico where precipitation is erratic and scarce (Schneider et al., 1997). The genotypes with the highest average root length in the pouches were bred in programs across the United States in both irrigated and non-irrigated conditions. This may suggest that although these genotypes were not necessary selected under drought pressure, they still retain the longer root length trait that was an important adaptive trait in areas with sporadic rainfall. Establishing deeper roots early in their growing season may aid the genotype in accessing adequate water for 81 the plant throughout the growing season as the plant advances from vegetative to reproductive growth. Genotypes with the shortest average root length were navy, black, one pinto and the tepary bean. The navy and the black seed types are of the Mesoamerican race and are small seeded which may have contributed to lower root length in this short term experiment. Tepary bean, native to the southwest is generally recognized as being highly drought tolerant (Thomas et al., 1983). In this study, tepary possessed a root system with a predominant taproot and very little branching and this made it to rank among the lowest of all genotypes in root length yet it is considered to be able to reach deep soil moisture reserves. Since root length is measured over the entire root system, tepary did not produce as many roots as common bean genotypes in the short time period of the experiment (Figure 2.1). 82 Figure 2. 1 Comparison of root scans of tepary (left) and dry bean (right) grown for two weeks in growth pouches. There is clearly a link here between seed size and performance for root length at this stage of seedling development. The Durango race beans have greater seed size and also had greater root length whereas the smaller seeded Mesoamerican race beans had lower root length. Therefore, we cannot be justified in making comparisons across seed sizes. Total Surface Area 2 2 Root surface area ranged from 20 to 123 cm with a mean value of 62 cm (Table 2.1-4). No significant differences were found between genotypes (p=0.1185). The top eighteen genotypes were not significantly different and as expected, many of those with the greatest 83 average length also had the greatest average surface area. Genotypes with the greatest surface area were I9365-31(black), Gemini (GN), Santa Fe (pinto), US-1140 (GN), Matterhorn (GN) and PT9-17(pinto). Those with the least amount of average surface area were Schooner (navy), Shiny Crow (black), Viva (pink), Navigator (navy) and tepary bean. Total Root Volume 3 3 Root volume values ranged from 0.18 to 0.93 cm and averaged 0.53 cm . Root volume was not significantly different between genotypes (p=0.2692). Forty-three genotypes were not significantly different for the top entries (Table 2.1-4). Genotypes with the greatest root volume were Santa Fe (pinto), Gemini (GN), Topaz (pinto), PT9-17 (pinto) and UI-114 (pinto). These are the same genotypes that had the greatest root length and surface area with the addition of Topaz and UI-114. Topaz is a pinto bean developed by a private company under irrigation in the western United States. UI-114 is an old pinto bean variety from the University of Idaho also bred under irrigation. Those with the least root volume were Schooner (navy), Shiny Crow (black), Viva (pink), Medalist (navy), Seafarer (navy), Montrose (pinto) and tepary bean similar to the genotypes with least surface area. Average Root Diameter Average root diameter ranged from 0.28 to 0.49 mm with an average of 0.28 mm. Average root diameter was not significantly different (p=0.2678) among genotypes (Table 2.14). The top fourteen genotypes for average root diameter were not significantly different and included Kodiak (pinto), Maverick (pinto), Roza (pink), Topaz (pinto) and Lariat (pinto). Those genotypes with the lowest average root diameter were all small seeded genotypes: BAT 93 (tan), CENTA Pupil (small red), Zorro (black), BAT 477 (tan) and Eclipse (black). 84 Number of Tips Average number of root tips ranged from 416 to 2838 and averaged 1564 (Table 2.1). Average number of root tips was not significantly different between genotypes (p=0.5076). The trait appears to be opportunistic and highly dynamic; adjusting to environmental conditions and is unlikely to be a useful or consistent root trait to use in selection for drought tolerance. Overall only whorl number was found to be highly significantly different among the 96 genotypes. Root length differences were just outside the significance range but this trait could be valuable for selection. Surface area, total root volume, average diameter and total number of root tips were not significantly different across genotypes. These traits, specifically length and surface area are highly associated with seed and cell size. Genotypes with larger seed size grow faster and have larger cell size than those with smaller seed size. Larger seed types contain more reserves and are able to produce more plant biomass early. After only two weeks in the pouches, these traits were confounded by larger cell size of individual genotypes and thus comparisons should really only be made between genotypes of similar seed size. Two weeks is a short amount of time to detect significant trait differences compared to the total time that a plant needs to complete its life cycle. Although, screening genotypes for greater drought tolerance at a juvenile stage of development would be very useful, variability among traits identified must be confirmed with results from genotypes grown in the field. Comparison of Pouch Root Traits to Field Root Traits, Harvest Index and Yield An important aspect of this study was to compare root traits of genotypes grown in the pouches to root traits measured in the field and compare with postharvest data harvest index and yield. A major objective was to identify specific traits of roots grown in the pouches that could to be utilized by breeders to develop drought tolerant dry bean cultivars. 85 Matterhorn (GN) identified in Chapter 1 as having superior field performance (Kelly et al., 1999) also performed well in the pouches for overall root length and basal root whorl number. In the field, Matterhorn was not significantly different from the best genotypes for basal root number and yield in 2011, taproot diameter in the combined analysis, and overall score and harvest index in both years. As discussed earlier, Matterhorn has a history of performing well under drought stress (Singh, 2007; Urrea et al., 2009) and appears to possess useful drought tolerance traits. Another genotype that has exhibited outstanding traits in both the pouch and the field was the pinto cultivar, Medicine Hat. Medicine Hat did not differ significantly from the best performing genotypes for length, basal root whorl number, number of basal roots and harvest index in both years and yield in 2011. Medicine Hat was bred by a private company in the western United States under irrigated conditions and resulted from a cross of Buster (pinto) and a pinto bean breeding line from MSU. Medicine Hat does appear to possess drought tolerance when compared to other genotypes and could be an important parent in breeding for drought tolerance. Great Northern seed-type, GN9-4 developed by the USDA-ARS breeding program in Prosser, WA also performed well and did not differ significantly from the best genotypes for root length in the pouch, basal root number in both years, taproot diameter in the combined analysis, overall score in 2012 and yield in 2011. Due to its performance in the pouch and field, GN9-4 appears to possess drought tolerance and could be a valuable source of genes for increased drought tolerance. 86 Merlot (Hosfield et al., 2004), small-red seeded genotype bred at Michigan State University performed well in both the pouch and field. Merlot belongs to the Jalisco race that also originated in the semi-arid highlands of Mexico. Merlot did not differ significantly from the best genotypes for whorl number in the pouch, overall score and yield in both years. Merlot is an especially important genotype as it represents genetic variation not previously identified in the small-red seed type (Hayes and Singh, 2007). BAT477 (tan) and Sedona (pink) were identified as possessing superior basal root whorl numbers, but neither of these genotypes had superior yield or harvest index performance in the field. In the field, Mesoamerican types especially black beans had the largest taproot diameter. BAT477, Sedona and the pink genotypes represent important sources of high basal root whorl numbers and could be combined with high yielding lines lacking in these traits. Among black, navy and pink bean seed types, no navy beans performed above average in specific root traits in pouches or the field. Navy beans represent an important class for improvement of drought stress tolerance. Conclusions In this study, two root characteristics were identified as being significantly different between the 96 diverse genotypes from the Middle American gene pool. Those characteristics were whorl number and overall root length. Number of root tips, average diameter, root volume and surface area did not differ significantly among genotypes. Although we were not able to correlate yield data to root characteristics in the pouch, we did identify genotypes Matterhorn, Medicine Hat, GN9-4 and Merlot that produced more extensive root systems in the pouches and also exhibited superior performance in the field for root characteristics, harvest index and yield. It is apparent that significant phenotypic and thus genotypic variation exists for various root traits 87 measured in the growth pouches but some traits are confounded by seed size. At two weeks of growth, only a small portion of the total root system is being measured and compared among genotypes. This seedling assay could be an important tool to survey large numbers of potential genotypes of similar seed size or market class for use in breeding. It enables breeders to investigate many more genotypes than can be grown in the field. However, combined with the field root study in Chapter 1, a more complete analysis of seasonal root growth is obtained. Analyzing the root system throughout the season can help determine what traits or mechanisms are important and functional in drought tolerance of dry bean. 88 REFERENCES 89 REFERENCES Asfaw A., Blair M.W. (2012) Quantitative trait loci for rooting pattern traits of common beans grown under drought stress versus non-stress conditions. Molecular Breeding 30:681695. Beebe S.E., Rojas-Pierce M., Yan X., Blair M.W., Pedraza F., Muñoz F., Tohme J., Lynch J.P. (2006) Quantitative trait loci for root architecture traits correlated with phosphorus acquisition in common bean. Crop Science. 46:413-423. Bonser A.M., Lynch J., Snapp S. (1996) Effect of phosphorus deficiency on growth angle of basal roots in Phaseolus vulgaris. New Phytologist 132:281-288. Broughton W.J., Hernandez G., Blair M., Beebe S., Gepts P., Vanderleyden J. (2003) Beans (Phaseolus spp.) - model food legumes. Plant and Soil 252:55-128. Butare L., Rao I., Lepoivre P., Polania J., Cajiao C., Cuasquer J., Beebe S. (2011) New genetic sources of resistance in the genus Phaseolus to individual and combined aluminium toxicity and progressive soil drying stresses. Euphytica 181:385-404. Clark R.T., MacCurdy R.B., Jung J.K., Shaff J.E., McCouch S.R., Aneshansley D.J., Kochian L.V. (2011) Three-dimensional root phenotyping with a novel imaging and software platform. Plant Physiology 156:455-465. Gepts P., Bliss F.A. (1985) F1 hybrid weakness in the common bean: Differential geographic origin suggets two gene pools in cultivated bean germplasm. Journal of Heredity 76:447450. Hayes R., Singh S.P. (2007) Response of cultivars of race Durango to continual dry bean versus rotational production systems. Agronomy Journal 99:1458-1462. Himmelbauer M.L., Loiskandl W., Kastanek F. (2004) Estimating length, average diameter and surface area of roots using two different Image analyses systems. Plant and Soil 260:111120. Ho M.D., Rosas J.C., Brown K.M., Lynch J.P. (2005) Root architectural tradeoffs for water and phosphorus acquisition. Functional Plant Biology 32:737-748. Hosfield G.L., Varner G.V., Uebersax M.A., Kelly J.D. (2004) Registration of 'Merlot' small red bean. Crop Science. 44:351-352. Jones B., Ljung K. (2012) Subterranean space exploration: the development of root system architecture. Current Opinion in Plant Biology 15:97-102. 90 Kelly J.D., Hosfield G.L., Varner G.V., Uebersax M.A., Haley S.D., Taylor J. (1994) Registration of 'Raven' black bean. Crop Science 34:1406-1407. Kelly J.D., Hosfield G.L., Varner G.V., Uebersax M.A., Taylor J. (1999) Registration of 'Matterhorn' great northern bean. Crop Science. 39:589-590. Kelly J.D., Varner G.V., Hosfield G.L., Uebersax M.A., Taylor J. (2006) Registration of ‘Sedona’ Pink Bean. Crop Science. 46:2707-2708. Lobet G., Pages L., Draye X. (2011) A novel image-analysis toolbox enabling quantitative analysis of root system architecture. Plant Physiology 157:29-39. Lodeiro A.R., Gonzalez P., Hernandez A., Balague L.J., Favelukes G. (2000) Comparison of drought tolerance in nitrogen-fixing and inorganic nitrogen-grown common beans. Plant Science 154:31-41. Lynch J.P. ( 2012) Genotype screening for number of whorls, http://plantscience.psu.edu/research/labs/roots/methods/methods-info/screening-of-rootwhorls/genotype-screening-for-number-of-root-whorls. Accessed October 2012. Lynch J., Vanbeem J.J. (1993) Growth and architecture of seedling roots of common bean genotypes. Crop Science. 33:1253-1257. McMichael B.L., Burke J.J., Berlin J.D., Hatfield J.L., Quisenberry J.E. (1985) Root vascular bundle arrangements among cotton strains and cultivars. Environmental and Experimental Botany 25:23-30. Robertson B.M., Hall A.E., Foster K.W. (1985) A field technique for screening for genotypic differences in root-growth. Crop Science. 25:1084-1090. Sassi Aydi S., Aydi S., Gonzalez E., Abdelly C. (2008) Osmotic stress affects water relations, growth, and nitrogen fixation in Phaseolus vulgaris plants. Acta Physiologiae Plantarum 30:441-449. Schneider K.A., Rosales-Serna R., Ibarra-Perez F., Cazares-Enriquez B., Acosta-Gallegos J.A., Ramirez-Vallejo P., Wassimi N., Kelly J.D. (1997) Improving common bean performance under drought stress. Crop Science. 37:43-50. Singh S.P. (2007) Drought resistance in the race durango dry bean landraces and cultivars. Agronomy. Journal. 99:1219-1225. Singh S.P., Terán H., Gutiérrez J.A. (2001) Registration of SEA 5 and SEA 13 drought tolerant dry bean germplasm. Crop Science. 41:276-277. 91 Sponchiado B.N., White J.W., Castillo J.A., Jones P.G. (1989) Root growth of four common bean cultivars in relation to drought tolerance in environments with contrasting soil types. Experimental Agriculture 25:249-257. SVREC. (2012) Saginaw Valley Research and Extension Center, http://agbioresearch.msu.edu/saginawvalley/index.html. Accessed November 2012. Thomas C.V., Manshardt R.M., Waines J.G. (1983) Teparies as a source of useful traits for improving common beans. Desert Plants 5:43-48. Urrea C.A., Yonts C.D., Lyon D.J., Koehler A.E. (2009) Selection for drought tolerance in dry bean derived from the Mesoamerican gene pool in western Nebraska. Crop Science. 49:2005-2010. USDA-ERS. (2012) Dry Beans, http://ers.usda.gov/topics/crops/vegetables-pulses/drybeans.aspx#major. Accessed December 2012. 92 Chapter 3. Evaluation of 95 dry bean genotypes for reaction to drought in a 5-week growth chamber experiment Introduction Dry bean (Phaseolus vulgaris L.) is the most important food legume grown worldwide (Broughton et al., 2003) and as such drought is the most important abiotic stress that significantly effects its production (Ramirez-Vallejo and Kelly, 1998). As a major production constraint, drought can have devastating effects on yield especially in the developing world where dry bean is primarily grown under rainfed conditions. The reaction of different dry bean genotypes to drought could have major ramifications on how devastating drought events are on the crop productivity and quality and on food supplies in localized production areas. Breeding for drought tolerance is challenging as the effects are complex and plant response is highly variable and based on many interacting factors (Ramirez-Vallejo and Kelly, 1998). For that reason screening for drought tolerance in the growth chamber or greenhouse and using that data to infer genotypic performance in the field could be beneficial. When attempting to develop drought tolerance in any crop several approaches can be employed. Many studies focus on whole plant response to drought using yield as the main indicator of a genotype’s overall drought tolerance (Acosta-Gallegos and Shibata, 1989; Porch et al., 2009; RamirezVallejo and Kelly, 1998; Rosielle and Hamblin, 1981; Schneider et al., 1997; Teran and Singh, 2002; Urrea et al., 2009). Selecting for yield under stress is a valid approach as yield is a cumulative trait that incorporates all mechanisms of tolerance and avoidance in the final seed yield. Shoot traits represent a major source of variation that could be used in breeding for drought tolerance in beans. Beans have four distinct growth habits that are classified into four 93 shoot types I to IV each with its own properties (Singh, 1982). These growth habits can be determinate or indeterminate and differ in spatial layout of branching and climbing ability (Kelly, 1998; Singh, 1982). Watanabe et al. (1997) found that shoots play an important role in drought tolerance in cowpeas when grown in root-limiting environments. White and Castillo (1992) grafted different bean shoot genotypes onto different root genotypes and tested their performance under drought stress. They found that the effect of shoot genotype was small in comparison to root genotype but the shoot genotype still played an important role in overall drought tolerance. Water use efficiency as determined by above ground structures of a crop also plays a critical role in determining drought tolerance (Davies et al., 2011). Dry bean is known to alter the orientation of its leaves in response to stress and ultimately will wilt its leaves in order to tolerate drought stress. These mechanisms are important first responses to stress as plants close stomata and restrict photosynthesis to conserve water. Recent investigations in soybean have uncovered variation in plant wilting response to drought in relation to the plant’s transpiration rate under water limited conditions (Ries et al., 2012; Sadok and Sinclair, 2011). Ries et al. (2012) found that higher soil moisture content after a drought period was only associated with soybean genotypes that were slow to wilt, but they concluded that there were many factors affecting water use efficiency across genotypes. Sadok and Sinclair (2011) identified an important maximal transpiration trait in soybean (Glycine max L.), and mention similar findings in corn (Zea mays L.) (Ray et al., 2002), sorghum (Sorghum bicolor L.) (Bunce, 2003) and in recent crop varieties that have been released as slow wilting under moisture stress. This trait seems to be very important specifically for regions that experience predictable drought. The majority of studies are either conducted in the field and investigate the tolerance of the plant over 94 the entire season or have used indirect screenings based on traits linked to molecular markers conducted in the laboratory. Objectives This study was conducted to determine the relative drought response of different bean genotypes grown under root limiting conditions for a short 5-week period in the growth chamber during the off-season. Materials Ninety-five dry bean genotypes were selected based on prior knowledge of reaction to drought. The study was part of the Common Bean Coordinated Agricultural Project (BeanCAP) to gain greater insights in drought tolerance in common beans. The 96 genotypes were previously selected by a group of breeders in the United States and Puerto Rico for more extensive testing as part of the Beancap research projects being conducted in different production areas. The selections represent old and contemporary dry bean varieties (BeanCAP, 2010). The study was limited to bean genotypes from the Middle American gene pool (Gepts and Bliss, 1985). The experiment included fifteen black, sixteen great northern (GN), eight navy, ten pink, thirty-three pinto, nine small red, two tan and one carioca genotypes (Table 3.1-3). Jaguar, B98311 and Fuji served as checks as their reaction to drought conditions in Michigan in the greenhouse is known. The black genotype, B98311 is considered tolerant to drought due to its deep rooting behavior (Frahm et al., 2004). Fuji (Kelly et al., 2009), a determinate, small white variety, is particularly sensitive to drought whereas Jaguar (Kelly et al., 2001), a black variety, is considered relatively drought tolerant (Mukeshimana, 2012). 95 Methods The methodology for the limited root study was adapted from research conducted by Muchero et al. (2008) who used similar screening methods to identify drought tolerance in cowpea (Vigna unguiculata L.). Cowpea is recognized as a highly drought tolerant grain legume, and diversity for drought tolerance has been observed between genotypes. Seedling drought tolerance in cowpea has been characterized to identify shoot phenotypic responses that could be used in high-throughput screening of numerous breeding lines for future use in breeding (Muchero et al., 2008). This study is considered a root-limiting experiment as genotypes were planted in small pots where roots become pot bound after a few weeks of growth. This experiment was initiated in the greenhouse, but due to lack of control of irrigation and temperature systems, the experiment was moved to a growth chamber. The experiment was conducted in nine random groups due to the limited growth chamber space and the large number of genotypes. Each group consisted of 13 genotypes and three check varieties (Jaguar, Fuji and B98311) except for the final experiment which had only six genotypes including the checks. Five replicates of each genotype were evaluated within each of the nine groups. Five blocks were created and genotypes were randomly placed in each of the five blocks. The experiment was conducted during the winter from August 2011 through April 2012. Growing conditions in the growth chamber were set at 26°C, 16 hr day with 8 hr night. Each genotype was planted in 9x9cm pots in 150 g of potting soil at a seeding rate of three seeds per pot. Pots were watered to field capacity and allowed to drain at planting. After germination, plants were thinned to one seedling per pot. Drought stress was induced by discontinuing watering until data was collected around two weeks after planting. No additional water was provided between planting and data collection. Ratings for wilting, greenness, and unifoliate senescence were taken when the most sensitive genotypes lost unifoliate leaves and the number of completely senesced unifoliate 96 leaves was recorded. Ratings for wilting were based on a scale from 0 to 5 where 0 has no sign of wilting and 5 was completely wilted (Figure 3.1). Stem greenness ratings were on a scale from 0 to 5 with 0 being completely yellow and 5 being completely green. Following the collection of these data at approximately 21 days, watering was resumed every other day for two weeks. At the end of the two week recovery period, genotypes were rated on their survival. Those plants recovering from the apical meristem received a 1 rating. Those recovering from the basal meristem received a 0.5 rating and those that showed no recovery or death were rated 0. 97 5 0 1 3 4 2 5 Figure 3. 1 Wilting scores in dry bean after 14 days without watering. 0 is no sign of wilting. 5 is completely wilted Statistical Analysis Data was analyzed using SAS 9.3 (SAS Institute Inc., Cary, NC, 2010). Analysis of variance was conducted for each measurement including wilting, stem greenness, unifoliate senescence and survival using PROC MIXED with repeated measures. Mean comparisons were performed using the least significant difference (LSD) test with a significance level at α=0.05. 98 Pearson Correlation analysis was conducted using PROC CORR to determine associations between wilting, stem greenness, unifoliate senescence and survival with one another as well as with yields from the previously discussed field season in 2011. Growth chamber grouping was a significant factor detected during statistical analysis. Results and Discussion Wilting All degrees of wilting were observed among the 98 genotypes (Figure 3.1). Wilting was the first sign of stress and in some genotypes was apparent approximately a week after planting. Average wilting ratings for all 95 genotypes plus checks were determined (Table 3.1-3). The highest value for wilting was 4.8 and the lowest was 1.6 with an overall mean of 3.5 and there was significant differences between genotypes (p<0.0001). The highest average wilting score was observed in Marquis (GN), ROG 312 (pink) and Sierra (pinto) with a score of 4.8. The lowest average wilting score was seen in Eclipse (black) at a level of 1.6. Checks fell near the middle of the range with Jaguar at 3.3, B98311 at 3.4 and Fuji at 3.7. 99 Table 3. 1 Average scores for wilting, stem greenness, unifoliate senescence after 21 days without water, and survival after 14 day recovery period of 28 Mesoamerican genotypes grown in the growth chamber. Seed Type black black black black black black black black black black black black black black black black black navy navy navy navy navy navy navy navy navy tan tan carioca Stem Greenness 2.8 2.6 3.0 2.3 2.4 3.4 2.6 1.8 3.0 2.8 1.8 3.0 2.0 3.0 2.4 2.6 2.8 3.6 3.6 3.8 4.0 4.6 3.6 3.6 3.4 4.6 3.1 4.0 3.2 3.1 Unifoliate Senescence 0.4 0.8 0.0 1.0 0.8 0.0 0.0 1.6 0.4 0.6 1.6 1.8 0.4 0.0 2.0 0.4 0.0 0.9 0.4 0.0 0.8 0.0 1.2 2.0 2.0 0.4 1.6 2.0 1.6 0.9 Genotypes Wilting Survival 115M 3.4 0.8 A-55 3.8 0.8 Aifi Wuriti 3.0 1.0 B98311 3.4 0.5 Domino 4.4 0.6 Eclipse 1.6 1.0 F04-2801-4-1-2 4.6 1.0 I9365-31 4.4 0.6 Jaguar 3.4 0.8 Jaguar-Check 3.3 0.7 Midnight 4.2 0.2 PR 0443-151 3.8 0.2 Raven 4.6 0.9 Shania 1.8 1.0 Shiny Crow 3.0 0.0 T-39 3.0 0.8 Zorro 2.6 1.0 Fuji 3.7 0.7 Verano 2.6 1.0 Avalanche 2.2 1.0 C-20 4.0 0.8 Mayflower 3.4 0.8 Medalist 4.2 0.6 Navigator 3.8 0.3 Schooner 4.2 0.1 Seafarer 3.6 0.8 BAT 93 3.5 0.3 SEA 10 4.2 0.0 A285 3.2 0.2 3.5 0.6 Mean LSD 0.05 1.2 0.9 0.8 0.4 26 24 72 55 CV% Wilting: 0=no wilting, 5=completely wilted, Stem greenness: 0=yellow, 5=completely green, Unifoliate senescence=count of dead leaves (0-2), Survival: 0=death, 1=recovery from apical meristem, 0.5=recovery from basal meristem 100 Table 3. 2 Average scores for wilting, stem greenness, unifoliate senescence after 21 days without water, and survival after 14 day recovery period 50 Durango genotypes grown in the growth chamber. Genotypes BelNeb-RR-1 Beryl R CDC Crocus Coyne Gemini GN#1Sel27 GN9-1 GN9-4 Marquis Matterhorn Orion Sawtooth UI-425 UI-59 US-1140 Weihing ABCP-8 Bill Z Buckskin Buster Chase Common Pinto Croissant Fisher Kimberly Kodiak La Paz Lariat Maverick Seed Type GN GN GN GN GN GN GN GN GN GN GN GN GN GN GN GN pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto Wilting 3.2 4.4 3.4 3.6 1.8 2.6 4.6 4.0 4.8 4.4 2.6 4.0 2.8 4.2 3.8 4.4 4.4 3.6 3.4 2.6 4.0 3.8 3.8 4.0 3.2 4.0 2.4 3.6 3.6 Stem Greenness 3.4 4.0 3.4 3.8 4.2 4.0 3.4 3.8 3.8 4.0 4.0 4.6 3.6 4.0 3.8 3.8 3.2 3.0 2.8 4.0 3.0 2.4 2.4 2.4 3.4 4.2 3.4 3.0 3.2 101 Unifoliate Senescence 2.0 2.0 0.0 1.2 0.0 0.4 2.0 2.0 2.0 0.0 0.0 1.6 0.0 0.8 0.6 1.2 1.5 2.0 0.4 0.0 0.8 0.4 2.0 0.0 2.0 0.0 0.0 1.6 1.2 Survival 0.0 0.0 1.0 0.4 1.0 1.0 0.0 0.2 0.0 1.0 1.0 0.4 1.0 0.6 0.6 0.4 0.4 0.0 0.8 1.0 0.8 1.0 0.0 1.0 0.0 1.0 1.0 0.4 0.4 Table 3.2 (cont’d). Seed Type pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto pinto Stem Greenness 3.0 3.0 3.4 4.4 3.2 3.4 2.0 2.2 3.4 2.2 2.8 3.6 3.8 3.6 4.0 3.0 3.2 3.4 3.0 2.0 2.4 3.3 Unifoliate Senescence 0.4 2.0 0.0 0.8 0.0 0.0 1.2 1.4 0.0 1.2 2.0 1.6 0.0 0.4 0.4 2.0 2.0 0.4 0.4 1.6 2.0 1.0 Genotypes Wilting Survival Medicine Hat 3.8 0.8 Montrose 3.8 0.0 ND-307 2.4 1.0 NE2-09-3 3.4 1.0 Nodak 3.4 1.0 NW-590 3.4 1.0 NW-63 2.8 0.8 Othello 3.6 0.7 Poncho 2.6 1.0 PT7-2 3.0 0.6 PT9-17 3.8 0.0 Quincy 3.8 0.5 San Juan 2.6 1.0 Santa Fe 2.4 0.8 Sierra 4.8 0.8 Stampede 4.4 0.1 TARS-VCI-4B 4.2 0.2 Topaz 2.4 1.0 UI-114 3.6 0.9 USPT-CBB-1 3.2 1.0 USPT-CBB-5 4.0 0.8 3.5 0.6 Mean LSD 0.05 1.2 0.9 0.8 0.4 26 24 72 55 CV% GN=Great Northern, Wilting: 0=no wilting, 5=completely wilted, Stem greenness: 0=yellow, 5=completely green, Unifoliate senescence=count of dead leaves (0-2), Survival: 0=death, 1=recovery from apical meristem, 0.5=recovery from basal meristem 102 Table 3. 3 Average scores for wilting, stem greenness, unifoliate senescence after 21 days without water, and survival after 14 day recovery period for 18 Jalisco genotypes grown in the growth chamber. Stem Greenness 3.4 3.0 3.0 2.0 2.8 3.4 2.4 2.4 2.0 2.8 2.4 2.2 2.0 4.0 2.2 3.2 2.6 3.4 2.7 Unifoliate Senescence 2.0 1.6 2.0 1.6 0.6 0.0 2.0 0.2 1.4 0.4 2.0 2.0 1.2 0.6 2.0 0.0 2.0 0.2 1.2 Genotypes Seed Type Wilting Survival Gloria pink 3.6 0.0 Harold pink 3.8 0.2 Pink Floyd pink 3.6 0.3 ROG 312 pink 4.8 0.3 Roza pink 3.6 1.0 Sedona pink 2.8 1.0 UI-537 pink 4.6 0.0 Victor pink 2.6 0.8 Viva pink 3.2 0.5 Yolano pink 3.4 1.0 CENTA Pupil small red 3.4 0.4 Common Red Mexican small red 4.4 0.0 DOR 364 small red 3.4 0.4 F07-449-9-3 small red 4.4 1.0 IBC 301-204 small red 3.0 0.4 Merlot small red 3.0 1.0 UI-239 small red 4.4 0.0 USRM-20 small red 3.8 0.8 3.7 0.5 Mean LSD 0.05 1.2 0.9 0.8 0.4 26 24 72 55 CV% Wilting: 0=no wilting, 5=completely wilted, Stem greenness: 0=yellow, 5=completely green, Unifoliate senescence=count of dead leaves (0-2), Survival: 0=death, 1=recovery from apical meristem, 0.5=recovery from basal meristem 103 Marquis and ROG 312 are private varieties developed under irrigated conditions in the western U.S. Sierra was the first upright pinto bean variety released from Michigan State University where architectural traits from race Mesoamerican genotypes were introgressed into the Durango race pinto (Kelly et al., 1990; Kelly and Adams, 1987). Sierra was developed under rainfed conditions and is known to possess an extensive root system. With such diverse seedtypes and genetic backgrounds exhibited by these three bean varieties, it is interesting that they exhibited the same average wilting score. Eclipse is a black bean variety from North Dakota State University developed under rainfed conditions. Of the seed types used in the study, pinto, pink and GN seeds are larger in seed size than the small-seeded black beans which had the lowest overall wilting score. Pinto and GN seed types of the Durango race are known from field studies to perform superior to other races under drought (Muñoz-Perea et al., 2006; Teran and Singh, 2002). Pink beans from the Jalisco race have also been shown to be more drought resistant when compared to other bean races grown under stress in the field (Singh, 2007). From this study, it appears that those pink, pinto and GN varieties utilize their extensive roots to help survive under water-limited conditions as they performed poorly in this study which limited the size and depth that roots could reach even at early growth stages. Those genotypes with lower wilting scores may have a mechanism to slow their transpiration rate and not deplete their soil moisture reserves as quickly as genotypes that have a high wilting score as was observed in soybean (Ries et al., 2012). Unifoliate Senescence The number of unifoliate leaves that senesced ranged from 0 to 2 and the average number of senesced leaves for each genotype is shown in Tables 3.1-3.3. Twenty-three different genotypes did not exhibit any senescence at the time of data collection in all five replications. 104 Twenty-four genotypes had senesced both unifoliates in all five replications at data collection whereas the remaining genotypes had various degrees of senesced unifoliates per pot. Stem Greenness Stem greenness ranged from 1.8 to 4.6 at the lowest with a mean rating of 3.2 and was significantly different between genotypes (p<0.0001). Stem greenness was intended to measure stem health and viability after the drought stress. The expectation was that the greener the stem, the more likely that a genotype would recover once watering resumed. Genotypes with the lowest stem greenness scores (1.8) were I9365-31 and Midnight, both black seeded genotypes that possess a purple stem (Table 3.1). I9365-31 is a black bean variety from the USDA-ARS program in Prosser, WA. Midnight is a black bean variety from Cornell University released in 1980. The highest stem greenness scores were observed in white seeded varieties Mayflower, Sawtooth and Seafarer. Mayflower was developed by Michigan State University as a high yielding upright navy and released in 1987 (Kelly et al., 1989). Sawtooth is a GN developed at the University of Idaho. Seafarer is a popular old variety with determinate plant growth habit released in 1968 from Michigan State University. Data suggests that there may be bias with the stem greenness ratings as all the black seeded genotypes had the lowest ratings and the white seeded genotypes had the highest rating. Black seeded genotypes have purple pigmentation in the stem which masks the green color which may have caused the evaluator to inadvertently rate those genotypes lower. The greenness rating was meant as an indicator of cell death and hence reduced survival. But upon investigation, this is not always the case with the stem greenness ratings, in fact some genotypes with low ratings had high rates of survival whereas, some of those with high stem greenness ratings had lower rates of survival. This rating system for stem greenness will need to be refined so that stem pigmentation is not confounded with necrosis 105 associated with stem death in order for it to be used as an effective tool for seedling drought tolerance. Survival All three recovery reactions were observed among the 95 bean genotypes which include recovery from the apical and basal meristems as well as no recovery. Thirty-one genotypes showed complete recovery after the drought treatment and fifteen genotypes showed no recovery (Figure 5). The LSD0.05 for the average recovery for each genotype was 0.92. The only significant differences were observed between those genotypes that recovered completely and those genotypes that showed little (0.1) or no recovery and plant death. One interesting observation was the nature of genotypic survival after both unifoliates had senesced. In most cases, no leaves remained or a very small trifoliate leaf survived. Some of these plants still recovered from the apical meristem, while others recovered from a basal meristem. Genotypes that exhibited this behavior in at least two of the five pots were Fuji, B98311, IBC 301-204, Pink Floyd, Navigator, Quincy, I9365-31, Othello, PT7-2, USPT-CBB-1, and USPT-CBB-5. These genotypes represent many seed types hence genotypic variation exists with no known genetic similarities among the genotypes. Interestingly many of these genotypes in this group were bred under irrigation. These same genotypes showed incredible survival, but none of the genotypes showed this reaction for all pots in the study. Those genotypes with complete recovery after extreme wilting represent important specimens for survival and more testing would be needed to elucidate underlying mechanisms. If these genotypes do possess stable recovery reactions to drought stress, they could be valuable genotypes in breeding for drought tolerance. 106 Genotypic Responses One of the checks chosen for the study was B98311, known for its deep rooting capabilities and drought tolerance (Frahm et al., 2004). In this study, B98311 behaved much differently than expected. Throughout the study, B98311 consistently wilted first and had an average survival score of 0.5, despite being known as a drought tolerant line. One possible explanation for its poor performance may be due to the restricted rooting area in 9×9 cm pots used in this study. Rooting space was controlled in order to elucidate whether drought tolerance was a shoot response. The superior drought tolerance of B98311 appears to be based on its extensive rooting capabilities not on any shoot trait. SEA 10, developed for drought tolerance, (Singh, 1995) exhibited the same behavior as B98311. SEA 10 had an average wilting score of 4.2 and showed no recovery. The drought tolerance of SEA 10 appears to be attributed to its root rather than shoot traits. Sawtooth, which was identified as possessing superior root traits in the field, had a high wilting score 4.0 but low recovery of 0.4 as would be expected in a limited rooting situation. The experiment clearly negated the importance of root traits among genotypes and could help identify genotypes that possess specific shoot traits functional under severe water stress. Matterhorn (Kelly et al., 1999) performed well in the field (chapter 1) and in growth pouches (chapter 2) and is reported in the literature to be drought tolerant (Singh, 2007; Urrea et al., 2009). In this study, Matterhorn had a high average wilting score of 4.4, but completely recovered every time. Matterhorn appears to induce wilting as a mechanism to survive drought. A number of genotypes that behaved similarly to Matterhorn were Chase (pinto) with a wilting score of 4 and survival of 0.8; landrace Common Pinto with a wilting score of 3.8 and complete recovery; Kodiak (pinto) with wilting score of 4 and complete recovery; Medicine Hat (pinto) 107 with a wilting score of 3.8 and recovery score of 0.8; Roza (pink) with wilting score of 3.6 and complete recovery; F07-449-9-3 (small red) with wilting score of 4.4 and complete recovery; USRM-20 (small red) with wilting score of 3.8 and recovery score of 0.8; A-55 (black) with wilting score of 3.8 and recovery of 0.8; Raven (black) with wilting score of 4.6 and recovery of 0.9; C-20 (navy) with wilting score of 4.0 and recovery of 0.8. Matterhorn combines both root traits and apparently shoot traits for superior performance in drought stress. Genotypes with this recovery mechanism could be combined with genotypes possessing desirable root traits to produce new bean lines with improved drought tolerance. Seafarer was the only genotype in this study with determinate growth habit other than the Fuji check variety. Determinate plants grow to a genetically determined size, flower and develop seed from those nodes. If those nodes were damaged or killed by drought stress, the plant does not produce any more vegetative growth so productivity is reduced. Genotypes with indeterminate growth habit continue to develop new leaves and nodes that produce flowers and can compensate and recover more easily when drought stress occurs. The expectation was that Seafarer would perform poorly under a terminal stress situation, but in this study it scored high (0.8) for survival and may have mechanisms for survival under water stress as compared to the determinate check variety Fuji. Correlation Analysis Two sets of correlations were performed. The first included all the parameters plus the random term group. The random term group signifies the grouping that was used for testing subsets due to limited growth chamber space and large population size. A significant correlation (r = -0.9**) was observed between unifoliate senescence and survival. This is not unexpected as number of senesced unifoliate leaves increases, survival decreases. Another important 108 interaction was observed between unifoliate leaf senescence and wilting (r = 0.5**). As wilting increases so does unifoliate leaf senescence. Correlation between survival and wilting (r = 0.5**) was negatively correlated suggesting that if wilting is high, survival is low. Correlations of group to other factors as well as stem greenness to other factors were low and non-significant. In the second correlation analysis, the random term group was not included and data was compared to field yield data for all genotypes in 2011. The expectation was that if the factors could be correlated to field conditions they would be of interest to bean breeders. No significant correlations were observed between wilting, unifoliate senescence, stem greenness, survival and yield data collected under drought stress in the field. Breeding Strategies to Enhance Drought Resistance in Dry Bean Important genotypic differences in shoot and root traits of dry bean were identified in the three studies described in this thesis. The findings from the study in chapter 1 identified basal root number and taproot diameter as root traits associated with drought tolerance in dry bean. Yield performance and harvest index were also reported for genotypes and as discussed are the cumulative measures of a genotype’s performance under stress conditions. The findings from the study in chapter 2 indicated that root length in the pouch and basal root whorl numbers were important traits that could be associated with drought tolerance observed in the field. In chapter 3, wilting and survival were identified as shoot traits associated with drought tolerance in a restricted pot experiment. The ultimate goal is to use the knowledge gained from these three studies to identify those genotypes that could be used in breeding specific root or shoot traits to enhance drought tolerance in dry bean. Rosetta (Kelly et al., 2012) was identified in chapter 1 as an important genotype when breeding for taproot diameter and root score. Matterhorn also performed well for basal root 109 number, average taproot diameter and overall root score. Medicine Hat also performed well for taproot diameter, basal root number, overall score, and harvest index. Several genotypes PT7-2, Merlot, and Matterhorn were shown to be important for yield and those for high harvest index were US-1140, Buster, Merlot and Rosetta. In the pouch study (chapter 2), Sedona (Kelly et al., 2006) and BAT 477 had the highest number of basal root whorls. US-1140, Gemini, Matterhorn and Santa Fe, PT9-17 had the greatest total root length when grown in pouches. Genotypes Matterhorn, Medicine Hat, GN9-4 and Merlot appeared to have superior root traits based on data from chapters 1 and 2. These genotypes along with the above mentioned genotypes for specific root traits from chapter 1 are important sources for improving root characteristics when breeding for drought tolerance in dry bean. Genotypes possessing shoot characteristics that may contribute to improved drought tolerance were identified in chapter 3. Those genotypes include Eclipse for wilting and Matterhorn, Chase, Common Pinto, Kodiak, Medicine Hat, Roza, F07-449-9-3, USRM-20, A55, Raven and C-20 for survival. It is of interest to observe genotypes such as Matterhorn and Medicine Hat that exhibit useful root and shoot traits. The identification of both root and shoot suggests that they possess superior drought tolerance. When breeding for increased drought tolerance in dry beans, these genotypes can be combined with other elite lines to improve root or shoot characteristics under drought stress. Many genotypes have been identified as possessing superior traits that are important to overall drought tolerance. The challenge for breeders will be to combine these traits in the best possible combinations to optimize genotypic performance under drought stress. Singh (1995) 110 cites the use of four-way crosses combined with single seed decent and bulk breeding methods as an effective strategy to increase drought tolerant in dry beans. Through this method diverse genotypes can be combined using interracial crosses to produce lines with greater drought tolerance. Another strategy for developing desirable root traits suitable for drought tolerance would be to use ideotype breeding. Extensive work in ideotype breeding at Michigan State University has produced high-yielding type II cultivars in Michigan under both rainfed and irrigated conditions (Kelly et al., 1987). Using a cyclic breeding method such as recurrent selection may be an ideal way to improve root ideotypes for drought tolerance (Kelly and Adams, 1987). Furthermore, the most successful strategy to employ would be that of elite by elite genotypic crosses that possess contrasting root and shoot traits. This strategy uses genotypes of the same seed size that possess genetic variation for specific traits like root characteristics. These crosses are advanced further using single seed decent when homozygous lines can be tested in replicated trials under stress and non-stress conditions (Kelly et al, 1998). Breeders need to employ geometric mean when measuring yield of new lines (Ramirez-Vallejo and Kelly, 1998; Schneider et al., 1997). Geometric mean is a better measure of individual genotypic performance when comparing performance under water stress versus non-stress. Schneider et al (1997) suggests a breeding strategy in which selections are initially made using geometric mean and then subsequent selections are made using highest yield under water stress. Those progeny with the highest geometric means would then be selected to develop future drought tolerant dry bean genotypes. Conclusions Ninety-five dry bean genotypes plus three checks were investigated for reaction to imposed drought conditions grown in small pots in the growth chamber. Ratings for wilting, stem greenness, unifoliate senescence and survival were assessed. Among these factors wilting 111 and survival were the most consistent across replications. We effectively demonstrated that genotypes like B98311 and SEA 10 known to be deep rooting and drought tolerant in the field performed poorly under the limited-rooting conditions in this study. We also identified several genotypes that had previously demonstrated superior performance in the field and also showed remarkable recovery in this study. Those genotypes possess both root and shoot traits that allow them to have superior performance under drought conditions. This rating system works well as a quick analysis of genotypic shoot drought tolerance, and combined with the pouch root study, could be used as an effective early screening method to identify genotypes for use in breeding of future drought tolerant bean lines. The study should be repeated as grouping of genotypes for testing was a confounding factor. The test has potential to serve as an early screening to identify bean genotypes with enhanced levels of drought tolerance in the shoot as roots were restricted in this study. 112 REFERENCES 113 REFERENCES Acosta-Gallegos J.A., Shibata J.K. (1989) Effect of water-stress on growth and yield of indeterminate dry-bean (Phaseolus-vulgaris) cultivars. Field Crops Research 20:81-93. BeanCAP. (2010) Dry Bean Genotypes BeanCAP. , www.beancap.org Broughton W.J., Hernandez G., Blair M., Beebe S., Gepts P., Vanderleyden J. (2003) Beans (Phaseolus spp.) - model food legumes. Plant and Soil 252:55-128. Bunce J.A. (2003) Effects of water vapor pressure differences on leaf gas exchange in potato and sorghum at ambient and elecated carbon dioxide under field conditions. Field Crops Research 82:37-47. Davies W.J., Zhang J., Yang J., Dodd I.C. (2011) Novel crop science to improve yield and resource use efficiency in water-limited agriculture. The Journal of Agricultural Science 149:123-131. Frahm M.A., Rosas J.C., Mayek-Perez N., Lopez-Salinas E., Acosta-Gallegos J.A., Kelly J.D. (2004) Breeding beans for resistance to terminal drought in the lowland tropics. Euphytica 136:223-232. Gepts P. and Bliss F.A. (1985) F1 hybrid weakness in the common bean: Differential geographic origin suggets two gene pools in cultivated bean germplasm. Journal of Heredity 76:447450. Kelly J.D. (1998) Bean Roots - A Plant Breeder’s Perspective. Annu. Rep. Bean Improv. Coop 41:214-215. Kelly J.D. and Adams M.W. (1987) Phenotypic recurrent selection in ideotype breeding of pinto beans. Euphytica 36:69-80. Kelly J.D., Adams M.W., Varner G.V. (1987) Yield stability of determinate and intdeterminate dry bean cultivars. Theoretical Applied Genetics 74:516-521. Kelly J.D., Adams M.W., Saettler A.W., Hosfield G.L., Varner G.V., Beaver J.S., Uebersax M.A., Taylor J. (1989) Registration of Mayflower navy bean. Crop Science 29:15711572. Kelly J.D., Adams M.W., Saettler A.W., Hosfield G.L., Varner G.V., Uebersax M.A., Taylor J. (1990) Registration of Sierra pinto bean. Crop Science 30:745-746. Kelly J.D., Judith M.K., Schneider K. (1998) Breeding for yield in dry bean (Phaseolus vulgaris L.). Euphytica 102:343-356. 114 Kelly J.D., Hosfield G.L., Varner G.V., Uebersax M.A., Taylor J. (1999) Registration of 'Matterhorn' great northern bean. Crop Science 39:589-590. Kelly J.D., Hosfield G.L., Varner G.V., Uebersax M.A., Taylor J. (2001) Registration of 'Jaguar' black bean. Crop Science 41:1647-1648. Kelly J.D., Varner G.V., Hosfield G.L., Uebersax M.A., Taylor J. (2006) Registration of ‘Sedona’ Pink Bean. Crop Science 46:2707-2708. Kelly J.D., Varner G.V., Roman B., Long B. (2009) Registration of 'Fuji' Otebo Bean. Journal of Plant Registrations 3:223-225. Kelly J.D., Varner G.V., Cichy K.A., Wright E.M. (2012) Registration of 'Rosetta' Pink Bean. Journal of Plant Registrations 6:229-232. Muchero W., Ehlers J.D., Roberts P.A. (2008) Seedling Stage Drought-Induced Phenotypes and Drought-Responsive Genes in Diverse Cowpea Genotypes. Crop Science 48:541-552. Mukeshimana G. (2012) Dissecting the genetic complexity of drought tolerance mechanisms in common bean (Phaseolus vulgaris L.). Doctoral Dissertation, Michigan State University, East Lansing MI. 203pp. Muñoz-Perea C.G., Terán H., Allen R.G., Wright J.L., Westermann D.T., Singh S.P. (2006) Selection for Drought Resistance in Dry Bean Landraces and Cultivars. Crop Science 46:2111-2120. Porch T.G., Ramirez V.H., Santana D., Harmsen E.W. (2009) Evaluation of Common Bean for Drought Tolerance in Juana Diaz, Puerto Rico. Journal of Agronomy and Crop Science 195:328-334. Ramirez-Vallejo P., Kelly J.D. (1998) Traits related to drought resistance in common bean, Euphytica. pp. 127-136. Ray J.D., Gesch R.W., Sinclair T.R., and Allen L.H. (2002) The effect of vapor pressure deficit on maize transpiration response to drying soil. Plant Soil 239:113-121. Ries L.L., Purcell L.C., Carter T.E., Edwards J.T., King C.A. (2012) Physiological Traits Contributing to Differential Canopy Wilting in Soybean under Drought. Crop Science 52:272-281. Rosielle A.A., Hamblin J. (1981) Theoretical aspects of selection for yield in stress and nonstress environments. Crop Science 21:943-946. Sadok W., Sinclair T.R. (2011) Crops yield increase under water-limited conditions: review of recent physiological advances for soybean genetic improvement, in: D. L. Sparks (Ed.), Advances in Agronomy, 113: 313-337. 115 Schneider K.A., Rosales-Serna R., Ibarra-Perez F., Cazares-Enriquez B., Acosta-Gallegos J.A., Ramirez-Vallejo P., Wassimi N., Kelly J.D. (1997) Improving common bean performance under drought stress. Crop Science 37:43-50. Singh S.P. (1982) A key for indentification of different growth habits of Phaseolus vulgaris L. Annu. Rep. Bean Improv. Coop 25:92-95. Singh S.P. (1995) Selection for water-stress tolerance in interracial populations of common bean. Crop Science 35:118-124. Singh S.P. (2007) Drought Resistance in the Race Durango Dry Bean Landraces and Cultivars. Agron. J. 99:1219-1225. Teran H., Singh S.P. (2002) Comparison of sources and lines selected for drought resistance in common bean. Crop Science 42:64-70. Urrea C.A., Yonts C.D., Lyon D.J., Koehler A.E. (2009) Selection for drought tolerance in dry bean derived from the Mesoamerican gene pool in western Nebraska. Crop Science 49:2005-2010. White J.W., Castillo J.A. (1992) Evaluation of diverse shoot genotypes on selected root genotypes of common bean under soil-water deficits. Crop Science 32:762-765. 116