M 1:43: .8". It!" ,. ‘V I 411:» ‘ N » “I ‘ 32.1"». . _ _- a . _ . . 4 --J_ 15/; . I "3 , 1 1 s v . r‘r. . . . ., . \ 33,3 1'0. - $941334) . ’ " ”'29:; 4 A 0 t 0 . I mg ‘L 4 \ (I E44,:f .132; Si? . "15%;” . 3?; {2'84 15):} “affirm“ . " 5v I:- 9‘" 1.44%.”. 'unypv't ‘ 3:31:31, ,1 "1( "5“?2‘3} :.';/h M. 4 1r.»- ' fish-11' $5; é: 43%;?! v {2-1 ‘45,"; ’53. ’{I‘é/ {‘14: 3;! 'gg‘fg 5'3“" 4:4? :4:/)$I_§; I" I ’4‘ “I: ([52), ‘IMN' .I. ‘- N >44..;’: 4" [:6 #}?}:’ glz'fi I nfhlk’sg.u/J 1-3,...Q‘Io If” '3 'a’v‘u/ “46/7/11. v . QT 75339314»: 44:...» ~: v\ ‘. 14v. 5",, 92.4% “i“w v 4: 4 = ‘ I I .4. N.» .‘ ‘ I- v 1 o 'y/H :29! f.“ :3? ’- 51:41! W :’ 0-31-31"! 7 3’) :4 4:51;, j}: 7,”"’ {'FI‘}.; 1’1. I {0" -* '- i If."~,...’1.., y‘ . , ' A‘f . 544:22541 f' ‘;’{(7 {’41: fig',l'f/€lr41§..4. [/7 if”: ”,3? $14,444.“, ~"I '='~".I ”" .1] . ~ 1? '1’ 'Z{)>44245?5.~¢ .. -- Mr. ‘ 1/” 4:15 '. .{fi'f J“. If ‘3') I: n M. '4’, 4 g. ’N. 2";2" I Minty. " if, :0}? [Ilia ' (rial/:31”! f" 71., '4'? l’ 5", fiffijfiq§fio }::'Z I)", ’ Z;::';~:,/-{ rIr-ifr? $42"; . .' ’ 4 {In I 0} WW: ”ii-”.7; ‘4. .. ,9 ,9_ 4 "21,121 1/.” [13'2" ,i’i {Iv/,1 - 1’ "" I. 1319521 ' 1/ ,I £701.71“ " ' I I | . ' " ' - 5")!” % I f [9.4. 1.): 7?; I; 14);); ¢/.,'/.-J:r t: 4”: [4;]: 5/?9’ . ’ I ~ 11"" .f 3’. .4 3/" {ll/1"": ’ #11111 1’ ' ’I’ ~: 9’5?" :'-’~ . ._,. -:~I. 1/1 ’4 1;}, . x 4 4 [4444421444 :'.4,.":; [4' I . l ’ ' .5], {0111;1".,/ 1: ,7, "4" “4::32/9' 1:4" ‘lfrc ~ "I” ”/4.“ ~l - . Mfr; "" , ,44/ 4,4 #1,! )4):_rr‘ 942m? 4'" 5‘, u’ :4. 51/131421? {’vfi5'1&§1r;‘/rf’f r-é't’; J I; v‘"'~’fl I I!" 4": v 4‘12”; I). {‘1’ )4. $1,)". 1/647 I ‘ ' ’ ' I' a 2;! | _ . 4 '- I/’ '7 Zia" w J ,47114-45505-7 1- 2 - 4..., ,1 '44:” . 4.1; t,- if, 12,-1.1. III, WI!- ’33“; x,‘ ' :fi141;;;’:"2’z I. {4/ 1 1 0' J .. 4‘ “/1334 IIJ” [ u‘. r - ["411 ., '.' ./l . '.I 25-4,?y'1'4/ ’4' .' [yr or {I}, 2...! ';: 1 . I. I.- 4 p '5 . , . ‘ , I ' £36!" ""9”" I 9'3"" I . . ‘ ”r‘ 9' - I”; 1 . .1131342', 55% 5113;} §;.’: 31.25.11.192?! ' I" fly“? ’ 1v. '31} .4, .J."I-/ 1/ ”I"; .r > "I. ' Ill ‘0 I " j . '1“ "'u' r ' _ 1 4 A LU; I! :riv o'nIfl’J’l'h/ ”’3? I U and.“ '1 I ' V‘ ”1' /" '.’ - 4‘ . (:Jr, 1'39; 12:, .1145]; . U. i”? ... my", $.19" .34! 1 I”? .4341] #14qi‘f 41":."’):'- c ;, v/v'v.’ I #111141 "#1117 1191,15 3m» 2005$ 70 5 LIBRARY Michigan State University This is to certify that the dissertation entitled Selection of common bean (Phaseolus vulgaris L.) genotypes with enhanced drought tolerance and biological nitrogen fixation. presented by Jorge Alberto Acosta Gallegos has been accepted towards fulfillment of the requirements for Ph.D. degmeh1 Plant Breeding and Genetics - Crop and Soil Sciences % #ngé. Major professor Date @047: [49? MSU is an Affirmative Action/Equal Opportunity Institution 012771 * MSU LIBRARIES RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. i AUG 2241 2.902 t 5’. SELECTION OF COMMON BEAN (EfllfifiQLQ§_1HL§BRIS L.) GENOTYPES WITH ENHANCED DROUGHT TOLERANCE AND BIOLOGICAL NITROGEN FIXATION BY JORGE ALBERTO ACOSTA GALLEGOS A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Plant Breeding and Genetics Program - Crop and Soil Science 1988 fixatic invest: £5 a: biolc 1 111 gr: Was in ABSTRACT SELECTION OF COMMON BEAN (EHASEQLQS_IHLGARIS L.) GENOTYPES WITH ENHANCED DROUGHT TOLERANCE AND BIOLOGICAL NITROGEN FIXATION BY Jorge Alberto Acosta Gallegos The potential for improvement for drought adaptation and N2 fixation concurrently in the same population of dry beans was investigated in this research. Parental genotypes, Lef-Z-RB, II900-5-M- 45 and N81017, showed an acceptable level of drought tolerance and biological nitrogen fixation (BNF). Hexican parental genotypes (type III growth habit) proved to be photoperiod sensitive. Their Sensitivity was initially expressed as a delay in flowering, and after flowering, as a low rate of partitioning to the developing fruits. Segregating populations were produced from crosses involving Michigan and Wisconsin parents (type II growth habit) and unadapted (type III) Mexican cultivars. Selection on the basis of nodule mass was conducted for two generations in these populations in the greenhouse by growing inoculated bean plants in a N-free medium. The segregant genotypes were compared to the check cultivar (UW 21-58) with superior BNF. Two additional generations of evaluation and selection were conducted under moderate water stress in the field where the primary bases for selection were seed and biomass productivity per unit land area. During selection in the field, segregants with the characteristics of the type II's were more adapted to the environment in Michigan. Nineteen selected F7 families, six parental genotypes and a check cultivar were evaluated for BNF under controlled conditions, and for f arr. n}. e ‘u Cs .s‘ .H‘ ‘F~ Lu 2. Jorge Alberto Acosta Gallegos drought tolerance in the field at two locations (Durango, Mex. and Michigan). The results under controlled conditions indicated the presence of promising families with BNF superior to the parents and the check cultivar POE 152. The best nitrogen fixers were those families displaying a longer vegetative phase. It seems that the genetic system governing photoperiod sensitivity, which sets the developmental stage for assimdlate partitioning in beans, also affects BNF. A close association between biomass and nitrogen assimilated (N2 fixed) per plant suggested that preliminary selection for BNF can be done by using an estimation of biomass of segregant families grown in a N-free medium. The results obtained in the field in Michigan indicated that recombinant families were produced with adaptation to this environment and with enhanced drought tolerance. In general, under stress, the metrical values of the recorded and calculated variables decreased. It was observed that morphologically different genotypes responded differently) to water stress. The grouping of the genotypes in a final evaluation for drought adaptation upon the basis of their phenology and growth habit is likely to facilitate data interpretation. Under drought stress and non-stress, the type III genotypes proved to be the most productive at the location in Mexico. In conclusion, recombinant families which displayed adaptation to the Michigan environment had enhanced BNF and drought tolerance. Further enhancement of those two traits into a single genotype is believed to be feasible, but since BNF is readily decreased_ by water stress, these two traits must be expressed at different developmental stages, i.e. BNF during the early vegetative phase and drought tolerance during the reproductive phase. RE 4; u- ‘c‘ ~v~ q is al: Stu me a: C 3. ACKNOWLEDGMENTS I wish to express my gratitude to Dr. M. Wayne Adams for his guidance and encouragement during my stay in Michigan State University. His helpful suggestions through the different phases of this research along with his patience were instrumental for the completion of my studies. His critical reviews of the manuscript are deeply appreciated. I would also like to thank Dr. Jim Kelly for serving in my committee and for his positive criticisms during the conduction of this reseach. The long and illustrative discussions we had on been breeding matters were very helpful. I am truly grateful to the other members of my guidance committee Drs. Frank Dazzo, Amy Iezzoni and Irwin Widders, for their helpful suggestions and good will, much thanks The assistance of Mr. Jerry Taylor, and Mrs. Betsy Bricker is greatly appreciated. The work in Durango, Mexico, was made possible through the cooperation of the technical and field personnel of the bean program of the Center for Forestry and Agriculture research in Durango. Special thanks to Mireille Khairallah and Susan Sprecher, for their unselfish help and moral support. I also appreciate the help of Barb Recchio during my first year at Michigan State. I gratefully acknowledge Mrs. Darlene Johnson for all her valuable administrative help. I extend thanks to many precious friends, who made the hard times at Michigan State more bearable. ii from t I deeply appreciate the financial support from the National Institute for Forestry and Agriculture Research (INIFAP) in Mexico and from the Bean-Cowpea Collaborative Research Support Program (MSU). iii VHF ‘s‘. Y'f' A. N'v‘ (v2:- TABLE OF CONTENTS LIST OF TABLES vi LIST OF FIGURES x INTRODUCTION 1 LITERATURE REVIEW 5 Drought effects on plant growth 5 Drought effects on nitrogen fixation 16 Concluding remarks 18 References 20 CHAPTER 1. EVALUATION OF A GROUP OF BEAN (Ehaagglng ynlgazifi L.) FOR BIOLOGICAL NITROGEN FIXATION AND DROUGHT TOLERANCE Introduction 26 Materials and Methods 28 Results and discussion 33 Conclusions 50 References 52 CHAPTER 2. SELECTION IN SEGREGATING BEAN POPULATIONS FOR DROUGHT TOLERANCE IN THE FIELD Introduction 55 Materials and Methods 58 Results 63 Discussion 78 Conclusion 83 References 84 CHAPTER 3. EVALUATION OF BEAN (Phaseglns,gnlgazis L.) GENOTYPES GROWN IN A NITROGEN-FREE MEDIUM FOR BIOLOGICAL NITROGEN FIXATION Introduction 87 Materials and Methods 89 Results and Discussion 91 Conclusions 99 References 101 iv CHAPTER 4. PLANT TRAITS RELATED TO PRODUCTIVITY IN BEANS (Phaseglns ynlgaris L.) UNDER MOISTURE STREES AND NON-STRESS CONDITIONS. Introduction Material and Methods Results and Discussion Conclusions References GENERAL DISCUSSION APPENDICES Appendix A Appendix B Appendix C 103 105 110 149 150 153 159 166 184 LIST OF TABLES CHAPTER 1 Characteristics of bean genotypes grown in green- house (G) and/or field (F) experiments. East Lansing, MI. 1985 Genotype average value for biomass per plant and nitrogen fixation related variables. East Lansing, MI. 1985 Correlations between biomass per plant and nitrogen fixation variables. East Lansing, MI. 1985 Average value for Physiological related variables of eight bean cultivars grown under two nitrogen sources at 45 days after planting. East Lansing, MI. 1985 Nedule score and total leaf nitrogen content of eight bean genotypes grown under two nitrogen sources at 45 days after planting. East Lansing, MI. 1985. Number of days to flowering, LAI, and root/shoot ratio of eight bean cultivars growing under two moisture treatmens at 70 days after planting. East Lansing, MI. 1985 Central leaflet length in mm of eight bean genotypes grown under two moisture regimes. East Lansing, MI. 1985 Seed yield under rainfed and stressed conditions, reduction in percentage, arithmetic and geometric means and drought susceptibility index (DSI). East Lansing, MI. 1985 Yield components of six bean genotypes grown under two moisture treatments. East Lansing, MI. 1985 List of been genotypes used as parents and crosses made at CIAT in 1983. East Lansing, MI. 1984 Recorded data and procedure. East Lansing, MI. 1985 Matrix of correlations among seed yield and some plant characteristics of 256 bean genotypes grown under drought stress in a rainout shelter. East Lansing, MI. 1985 vi Page 29 34 35 39 40 44 45 47 49 59 62 66 CHAPTER 2 (CONTINUED) 4. Mean seed yield and some agronomic characteristics of promising FA families and two parental genotypes grown under drought stress in a rainout shelter. East Lansing, MI. 1985 Mean seed yield and some agronomic variables of promising F6 families and two parental genotypes grown under drought stress in a rainout shelter. East Lansing, MI. 1986 Matrix of correlations among average seed yield and some plant characteristics of 121 bean genotypes grown under drought stress in a rainout shelter. East Lansing, MI. 1986 Mean seed yield and some agronomic variables of promising bean F6 families and two parental genotypes grown under irrigation. East Lansing, MI. 1986 Matrix of correlations among average seed yield and some plant characteristics of 121 bean genotypes grown under irrigation. East Lansing, MI. 1986 Origin and pedigree of the evaluated bean genotypes. East Lansing, MI. 1987 Analyses of variance for N assimilation and related variables in 26 inoculated bean genotypes grown in a N-free medium in the greehouse. East Lansing, MI. 1987 Values of seed yield per plant, yield components, HI, seed protein content and nitrogen harvest index (NHI) of 25 bean genotypes grown in a N-free medium in the greehouse. East Lansing, MI. 1987 Values of plant phenology, plant dry matter, total N per plant, biomass/plant/day (b/p/d), nitrogen accumulation mg/plant/day (N/p/d) and root/shoot ratio of 26 bean genotypes grown in a N-free medium in the greenhouse. East Lansing, MI. 1987 Correlation coefficients among nitrogen assimilation, plant biomass (TDM), seed yield and related variables in a group of 26 bean genotypes grown in a N-free medium in the greenhouse. East Lansing, MI. 1987 CHAPTER 6 List of plant traits measured in 26 bean genotypes grown under two moisture regimes in two locations. 1987 Seed yield of 26 bean genotypes grown under two moisture regimes in two locations. 1987 vii 68 70 73 76 77 90 92 94 95 98 109 112 if .14 11 CHAPTER 4 (continued) 3. 10. 11. 12. 13. 14. 15. 16. Mean seed yield and stability parameters of 26 bean genotypes grown under stress and non-stress conditions in Durango (Mexico) and Michigan (USA). 1987 Effect of two moisture regimes on plant traits of 26 bean genotypes. Kellogg Biological station, Battle Creek, MI. 1987 Drought tolerance indices of 26 bean genotypes grown under two moisture regimes. Kellogg Biological Station, Battle Creek, MI. 1987 Yield components, TDM and HI of 26 bean genotypes grown under irrigation. Kellogg Biological Station, Battle Creek, MI. 1987 Yield components, TDM and HI of 26 bean genotypes grown under stress. Kellogg Biological Station, Battle Creek, MI. 1987 Correlation coefficients of seed yield and TDM at maturity with some characteristics of phenology, plant growth and primary yield components of 26 been genotypes grown under two moisture regimes. Kellogg Biological Station, Battle Creek, MI. 1987 Effect of two moisture regimes on plant traits of 26 bean genotypes. Durango, Mexico. 1987 Leaf water content of detached leaves of 26 bean genotypes grown under drought stress and rainfed conditions at 50 DAE. Durango, Mexico. 1987 Drought tolerance indices of 26 bean genotypes grown under two moisture conditions. Durango, Mexico. 1987 Yield components, TDM and HI of 26 bean genotypes grown under rainfed conditions. Durango, Mexico. 1987 Yield components, TDM and HI of 26 bean genotypes grown under water stress. Durango, Mexico. 1987 Correlation coefficients of seed yield and TDM at maturity with some characteristics of phenology, plant growth and primary yield components of 26 been genotypes grown under two moisture regimes. Durango, Mexico. 1987 Combination of plant traits resulting in the best multiple regression model for seed yield and biomass at maturity for a group of 26 been genotypes grown under drought stress and irrigation. Kellogg Biological Station, Battle Creek, MI. 1987 Combination of plant traits resulting in the best multiple regression model for seed yield and biomass at maturity for a group of 22 bean genotypes grown under drought stress and irrigation. Kellogg Biological Station, Battle Creek, MI. 1987 viii 114 116 120 122 123 124 126 129 132 134 135 138 141 144 CHAPTER 4 (continued) 17. Combination of plant traits resulting in the best multiple regression model for seed yield and biomass at maturity for a group of 26 bean genotypes grown under drought stress and rainfall. 147 Durango, Mexico. 1987 LIST OF FIGURES CHAPTER 1 Dry matter of different plant perts of eight bean cultivars grown under two nitrogen sources (N= 50 kg/ha nitrogen, I= inoculant) at 45 days after planting. East Lansing, MI. 1985 37 Dry matter of different plant parts of eight bean cultivars grown under two moisture regimes (R= rainfed, S= water stressed) at 70 days after planting. East Lansing, MI. 1985 42 CHAPTER 2 Frequency distribution for seed and biomass yield of 256 bean genotypes grown under water stress (* indicates position of lowest and highest yielding parental genotype. East Lansing. MI. 1985 65 Frequency distribution for seed and biomass yield of 121 bean genotypes grown under water stress (* indicates position of lowest and highest yielding parental genotype. East Lansing. MI. 1986 72 Frequency distribution for seed and biomass yield of 121 bean genotypes grown under irrigation (* indicates position of lowest and highest yielding parental genotype. East Lansing. MI. 1986 75 CHAPTER 4 APPENDIX Average leaflet expansion rate of 26 bean genotypes grown under irrigated and water stressed conditions. Kellogg Biological Station, Battle Creek, MI. 1987 118 Average leaflet expansion rate of 26 bean genotypes grown under rainfed and water stressed conditions. Durango, Mexico. 1987 128 Effect of water stress during the reproductive phase on the yield components of 26 bean genotypes grown in Michigan (USA) and Durango (Mexico). 1987 137 C Soil moisture content during the growing season at three different depths, A= O - 15 cm, B= 16 - 30 cm, and C= 31 - 45 cm (FC= field capacity, -0.03 bars, UP= wilting point, -15 bars. Durango, Mexico. 1987 188 INTRODUCTION Pulses, such as coumon beans (W mlgaris L.), are a major source of proteins and calories throughout the world. Comon beans are especially important in developing countries, where they are construed by people of all social strata. The crop is generally grown for its mature seeds but its inmature pods and leaves are also consumed as a vegetable in some African countries (Adams et a1. , 1985). Cannon beans originated in the Americas (Gentry, 1969) , however, they are grown and construed in all continents . They are particularly important in Latin America and Eastern Africa. In these bean producing areas, the bean crop is constrained by different sets of biotic and abiotic factors. Some of those factors are widespread, like high or low temperatures, diseases, insects and weeds, while other factors are more site specific, such as marginal soils with a low content of essential plant nutrients or the lack of moisture at different times during the growing cycle. In semi-arid areas of North-Central Mexico, where the cannon bean is an important crop, shallow soils with poor nutrient and organic matter content increase the chance for drought damage to occur. The bean crop is grown during the portion of the year when precipitation is expected to occur, but alternate wet and dry periods of varying lengths which affect production can be expected. An examination of climate data for 110 production areas in Latin America indicates that almost 60 per cent of the crop experiences Par brat Yie} inc: Stab envi “npr moderate to severe water deficits after flowering (Laing et al., 1984). In spite of this, 2.,ynlgaris is considered to be a crop with poor tolerance to severe water deficits (Laing et al., 1984). Although during the past decade, common beans have been studied extensively, little research has been directed towards breeding for drought resistance. An old concern in bean research, to which much effort and resources have been lately channeled, is to overcome the yield plateau which experimentally ranges from three to five tons/ha (Temple and Long, 1980; White, 1987). Yields of common beans have been static for many years, whereas notable yield increases have been realized in several cereal crops. Increased cereal yields were attributed largely to the modification of plant morphology, improvement in grain/straw partitioning and an increased use of fertilizer (Coyne, 1980). Traditionally, advances in crop yields have been obtained through breeding and crop management. However, in some instances, particularly in developing countries, bridging the gap between actual and potential yields in variable environments can be more valuable than efforts to increase the yield potential of the crop. In other words, yield stability achieved through breeding for adaptation to adverse environmental stress is a more realistic approach to increase yields in unpredictable environments. The bean crop in most producing areas in developing countries is often confined to marginal lands where available soil nitrogen is limited and nitrogen fertilizer is either difficult to obtain or too expensive to purchase (Grahm, 1981; Bliss, 1985). Beans are generally considered poor in nitrogen fixation (Piha and MMnns, 1987) and show surprisingly variable response to inoculation in field experiments. The val wax to pla est is fol bec nat cer to Moi, Hit} eff1 brei is variable responses obtained with field experiments are due in part to variable enviromental conditions (Sprent, 1982) , and genetic variability among both W strains and bean genotypes for ability to efficiently fix biological nitrogen (Grahan, 1981). Cannon beans grown in rainfed areas of the Mexican plateau are planted after soil moisture is considered adequate to assure the establislunent of the crop. Biological nitrogen fixation in cannon beans is favored during vegetative growth peaking at the flowering stage (Graham and Roses, 1977; Rennie and Kemp, 1981a,b). The peak is followed by a sharp decline thought to be due to the developing fruits becoming a strong competitive sink for photosynthates . Thus , it is likely that bean genotypes which could readily enter into symbiosis with native or introduced rhizobia W and which possess a certain degree of drought tolerance would be the most suitable genotypes to use in drought prone environments on marginal soils. It is axiomatic that the productivity of a crop grown under moisture stress will be much less than its productivity when it is grown with an anple supply of water. Therefore, biological inmunity to the effects of drought is not a possibility (Quisenberry, 1982). However, breeding for enhanced biological nitrogen fixation and drought tolerance is an attractive approach to stabilizing and/or increasing bean yields without increasing inputs . As a part of the National Institute for Forestry and Agriculture Research INIFAP (Mexico) - Michigan State University, Bean/Cowpea CRSP project, this research was conducted to determine whether it is possible to select for drought tolerance and the ability for high BNF concurrently in the same population, and determine whether genotypes V3 V3 va‘ to pla est (Gr fol like natj cert to r moi: Wit}- effe brEE is with Proj COnc variable responses obtained with field experiments are due in part to variable environmental conditions (Sprent, 1982) , and genetic variability among both W strains and bean genotypes for ability to efficiently fix biological nitrogen (Graham, 1981). Cannon beans grown in rainfed areas of the Mexican plateau are planted after soil moisture is considered adequate to assure the establishment of the crop. Biological nitrogen fixation in common beans is favored during vegetative growth peaking at the flowering stage (Graham and Roses, 1977; Rennie and Kemp, 1981a,b). The peak is followed by a sharp decline thought to be due to the developing fruits becoming a strong competitive sink for photosynthates. Thus, it is likely that bean genotypes which could readily enter into symbiosis with native or introduced rhizobia We and which possess a certain degree of drought tolerance would be the most suitable genotypes to use in drought prone environments on marginal soils. It is axiomatic that the productivity of a crop grown under moisture stress will be much less than its productivity when it is grown with an ample supply of water. Therefore, biological inmunity to the effects of drought is not a possibility (Quisenberry, 1982). However, breeding for enhanced biological nitrogen fixation and drought tolerance is an attractive approach to stabilizing and/or increasing bean yields without increasing inputs . As a part of the National Institute for Forestry and Agriculture Research INIFAP (Mexico) - Michigan State University, Bean/Cowpea CRSP project, this research was conducted to determine whether it is possible to select for drought tolerance and the ability for high BNF concurrently in the same population, and determine whether genotypes superior in both characteristics simultaneously could be produced. 5“:* 3Dec: of t} dCVe: PreVQ tolel diScL Sulli LITERATURE REVIEW W Water deficits which occur during the growth of common beans affect many physiological and morphological characteristics associated ultimately with seed yield. The species is considered vulnerable to moisture stress (Laing et al., 1984). In terms of seed yield, in 21 ynlgazig, as well as in most pulse crops, the duration and intensity of the moisture deficit as well as the phonological stage of the crop at the time the stress occurs will determine the amount of damage done to the crop. There are essentially three mechanisms used by different plant species to overcome or survive periods of low water availability. Two of those involve the avoidance of large water deficits, either through developmental timing or by the plant surviving moisture deficits by preventing tissue dehydration . The third mechanism involves the tolerance of low tissue water potentials. These mechanisms are discussed in detail elsewhere (Blum, 1979: Turner, 1979; Levitt, 1980; Sullivan, 1983; Boyer, 1985). The gradual development of water deficits by plants growing in most field situations allows slow and continual adjustment in physiological processes, eventually manifested as alterations in growth and development (Jordan, 1983). This review discusses, from a practical point of view, morphological, phenological, partitioning, and economic yield responses of cannon beans and related pulse crops to drought stress and the effects of soil moisture stress on biological nitrogen fixation in the legume-rhizobia symbiosis. Women: Leaf area index, the most often used measure of canopy development, is very sensitive to water deficit, which results in a decrease in leaf initiation and expansion and an increase in leaf senescence and shedding, or both (Hsiao, 1973; Karamanos, 1978; Elston and Bunting, 1980; Hebblethwaite, 1982; Sheriff and Muchow. 1984). Leaf expansion is more sensitive to water deficits than are stomatal closure or photosynthesis (Hsiao, 1973). Leaf senescence does not appear to be as sensitive as leaf expansion (Sheriff and Muchow, 1984). Hewever, in the field, the most obvious morphological response to a sudden or prolonged water stress is leaf loss by accelerated senescence. ' The reduced rate of leaf area accumulation usually associated with growth in dryland environments may be associated with a.smaller size of individual leaves or with production of fewer leaves (Jordan, 1983). Bonnano and Mack (1983) evaluated the effect of differential irrigation on plant growth and development of two snap bean cultivars. They observed that the difference between treatments in total leaf area per plant occurred earlier in the season than the difference between treatments in total plant weight. The observed decrease in leaf area was due to a decrease in area per leaf rather than by a reduction in leaf number. In field beans, Karamanos (1978) found that after a period of 46 days the total leaf area of well-watered plants was about double that of the non-watered ones. He showed that the difference between treatments in total leaf area.was mainly produced by the mechanisms determining leaf size rather than those associated with leaf initiation and maintainance, namely leaf production, unfolding and death. Thus, it seems that the reduction in leaf area in grain legumes is due mainly to the reduction in size of individual leaves. With respect to leaf senescence, many researchers consider this accelerated senescence as a drought avoidance mechanism. in plants (Kraner, 1983). In regions where an extended growing season is feasible, cultivars with an indeterminate growth habit.may compensate later for the loss of leaf area by producing new leaves. However, for a short lived crop, such as the bean crop in the semi-arid highlands of Mexico, which rarely displays optimum LAI values of 4.0 (Laing et a1, 1983) at bloom stage, a heavy loss of leaves may be counter-productive. Differences in leaf abscission rates and yield among soybean cultivars growing under differential degrees of water stress were reported by Caviness and Thomas (1980). Vidal and Arnoux (1981), in a screening program involving 15 soybean cultivars and utilizing 19 morphological, physiological, and biochemical responses to drought stress, found that reduction in leaf expansion and petiole growth were the characteristics most highly correlated to the reduction in seed yield and canopy. Acosta and Kbhashi (1988) found that the responses in yield of indeterminate bean cultivars to water stress, imposed at late vegetative and early bloom. stages, could be explained by the decrease in LAI measured at bloom stage. Elston and Bunting (1980) pointed out that in faba bean, dry episodes affect final yield not by decreasing the rate of assimilation per unit area so much as by decreasing the rate of expansion of leaves; as a consequence, the total leaf area duration of the crop is reduced. In an experiment involving nine species of grain legun inter PTO?C ontos watel revic of w; SOHO! lowe- COVe CVap beco Pass via Flam to g Spre fi+ 1‘83: and legumes, Laing et a1. (1983), showed that leaf area duration (LAD, the integral of time course of LAI) alone explained a. remarkably large proportion (R2-0.99) of the variation in seed yield. The rate of natural leaf senescence is likely to change during crop ontogeny. An increase in natural leaf senescence is a common effect of water deficit during grain filling. As pointed out earlier in this review, the reduction in yield is dependent on the timing and duration of water deficit and the growth stage and type of crop. Increased leaf senescence caused by water deficit near physiological maturity has only a small effect on yield since senescence tends to occur first in the lower part of the canopy (Sheriff and Muchow, 1984). Usually, early water deficits reduce yield only when full ground cover is not achieved before flowering. Thus, soil water loss by direct evaporation occurs when crops are building up leaf area, and the losses become small once the leaf area index exceeds about 2.5 (Ritchie, 1983). Passioura (1986) indicates that the best prospect to avoid water loss via direct soil evaporation is to have a vigorous establishment of plants with a postrate, rather than erect, growth habit. It is important to establish an LAI approaching 2 as fast as possible, with leaves well spread out. Type III bean and indeterminate prostrate cowpea cultivars fit this description and are the most widely used in drought prone regions in the Mexican highlands and in semiarid zones of Africa (Hall and Patel, 1985; Acosta and Kohashi, 1988). Various mechanisms, particularly increses in leaf angle, can reduce the solar irradiance absorbed by leaves, so that stomatal closure does not result in metabolic damage. In beans, adjustment in leaf angle occurs rapidly in response to stress conditions (Dubetz, 1969). A which brand: :eascfi 0mm. Pinto second. mechanism in reducing the solar irradiance absorbed by leaves is by a decrease in leaf absorptivity. Leaf hairs, waxes and leaf color can reduce leaf absorptivity leading to lower leaf temperatures and transpiration rates (Schultze et al., 1987). Variation for all these traits seems to be present in beans, but does not appear to have been studied in relation to drought stress. W The time between sowing and maturity may be shortened or lengthened depending on the intensity and timing of water deficits. For example, intermediate maturing cowpeas and common beans flower and mature earlier under moderate levels of water deficits, but severe water deficits delay reproductive activity (Turk and Hall, 1980; Samper, 1984). This provides those crops with two possible adaptive responses. Under moderate water deficits, mid-season cultivars produce grain which may mature before the soil water is depleted. Such early maturity may be advantageous in dry years (Hall and Patel, 1985) and yet permit a longer season and larger yields under wetter years. With early severe water deficit, the crop remains in a quiescent vegetative stage but has the ability to continue reproductive activity if the water deficit is removed as long as lethal deficits are not reached (Turk and Hall, 1980). Other forms of developmental plasticity observed in legume crops which may be advantageous under water deficits include indeterminacy and branching. Where periodic water deficits occur during the growing season, the indeterminacy of certain grain legumes permits fruiting to occur in flushes during favorable periods. Such is the case of Colorado pinto bean cultivar San Juan Select (Adams, 1984; personal 10 communication) , selected in a region with spOradic rainfall during the growing season. Cowpeas recover after drought with a flush of flowers which produce a significant yield provided environmental conditions are then conducive to growth (Turk and Hall, 1980). The climatic characterization of the environment of the target area is basic to determining the kind of cultivar to be produced in a plant breeding program. Severe moisture stress occurring later in the growth cycle favors earliness, while mild stresses relatively early will favor later maturing cultivars with capability for recuperative growth (Singh and White, 1988 ) . Wide adaptation in cannon beans, as in most crops, is one of the aims of been breeders . Here , it is considered that genotypes with wide adaptation are those which possess individual or populational physiological and/or phenotypic plasticity that help them to cope with variable environments. Studies with common beans grown in variable environments have demonstrated that indeterminate growth types exhibit more yield stability than determinate ones (Beaver et a1. , 1985; Kelly et al., 1987). The latter types are also less productive (Laing et al. , 1984) . Developmental plasticity facilitates the matching of crop growth and development to the constraints of the environment, especially in minimizing the occurrence of the critical reproductive phase during periods of severe water deficit. In agronomic terms , in deveIOped countries, it may raise difficulties of uneven maturity of the crop in capital-intensive agriculture. In labor- intensive agricultural systems of the tropics, such plasticity substantially reduces the risk of complete crop failure for subsistence farmers (Sheriff and Muchow, 11 1984). comparing grain legume species, Wien et al. (1979) found that cowpeas which mature 17-20 days after anthesis were more likely to avoid environmental water deficits than soybeans, which.mature 40-60 days after anthesis. One disadvantage of matching crop cycle length by using early cultivars is that they are conservative, so that higher yields are not produced in moist years. Where a species like common bean has colonized a range of habitats it seems reasonable to expect to find various adaptations to those environments. Since habitats are subjected to continual fluctuation and change, adapted species must have considerable developmental plasticity (Summerfield, 1981). Furthermore, it is unlikely that any single attribute can provide all the adaptations that may be required to meet the complex array of possible ecological conditions that occur from.time to time and from place to place (Bunting, 1985). It is clear that if drought tolerance exists in common beans, it is unlikely that it would be due to a single attribute. E lll' . I 1.]. I' In most species only a distinct part of the plant, often a storage organ, is the economic yield. Economic yield (Ye) is the function of total dry matter production, the biological yield (Yb) and the harvest index (HI), so that ‘16 8 Yb X HI. Therefore, problems of partitioning of assimilates and the use of photosynthates for growth and storage (source-sink relationship) must be considered in the final yield (Apel, 1984). Furthermore, the size as well as activity of the photosynthetic apparatus and the pattern of assimilate distribution are genetically determined and also depend on changing environmental conditions and 12 adaptation to a given environment . In indeterminate grain legume crops , reproductive and vegetative growth phases overlap. When the lower nodes begin to produce flowers and set pods, the plant is still expanding and producing vegetative growth which competes for assimilates with flowers and young pads in the lower parts. From a review of the subject in field bean (219,13 faba), El-Faudty (1982) concluded that: a. During the entire flowering period pods and young seeds are competing with vegetative parts . b. From the middle of the flowering period, pads are competing with each other within the sane inflorescence . c . During maturity, there is competition between pods according to their position on the plant. Acosta and Kohashi (1988) mentioned that in cannon beans inter and intra-avary competition takes place under both stress and nonstress conditions. Under stress, this competition may be responsible for the reduction in the nmnber of pads per plant and seeds per pad of certain cultivars. A widely used index of photosynthate partitioning , the harvest index (HI), was first defined by Donald and Hanblin (1976) as the ratio of seed yield to biological yield. Although it has been claimed that the rise in cereal yields in the last several decades is due largely to an improvement in HI , in cannon bean contradictory results have been reported (Wallace and Monger, 1966; Laing et al. 1984; White, 1987; Acosta and Kohashi, 1988). 13 For many crops, it appears that further substantial improvements in HI are unlikely (Sinclair et al., 1984). White (1987) recently pointed out that increase of the yield potential in beans through an improvement in the HI is not a feasible approach. He considers that cannon been already possesses a high HI and reported that dry bean cultivars growing in the tropics have shown a lack of variability for this trait. He mentioned that in most yield trials a lack of correlation between HI and yield has been found. In addition, he considers that this lack of correlation is actually worse due to the statistical artifact introduced by correlating two components of the same end product. Increases in HI when the bean crop has been exposed to moisture stress have also been reported (Tosso, 1979; Couto, 1978). Plants produce many storage compounds that can be changed back to forms that can be translocated to other parts of the plant. The movement of compounds from a site where they were deposited to a site where they can be utilized is referred to as "remobilization" (Gardner et al., 1985). Water deficit during grain filling reduces grain yield through stomatal control of transpiration which reduces photosynthesis. Thus the demand for grain filling requires the use of stored assimilates, which results in a much higher proportional contribution by remobilization. In many crops, the economic yield is only a part of the total biomass. There is evidence in the literature that, for a number of crops, reallocation of carbohydrates produced before a stress period can partially alleviate the effects of the water deficit in terms of seed yield (Johnson and Moss, 1976; Bidinger et al., 1977; Gallagher and Biscae, 1982; Aparicia and Boyer, 1983). 14 Samper et al. (1984) grew cannon beans in a greenhouse study with incorporated 1" CO2 and, by monitoring dry weight changes of different plant parts, demonstrated that assimilates from storage organs (mainly leaves) were remobilized to developing fruits. They showed evidence for differences in remobilization due to drought treatment versus nan-stress and to genotype , with remobilization being more pronounced under water deficit. Economically, only the usable portion of a crop plant is important; however, in a biological sense, all plant dry matter is made through photosynthesis. Therefore, the production of total dry matter determines the response of a genotype to drought stress (Quisenberry, 1982). Quisenberry (1982) considers that under drought conditions, partitioning should be a second objective for improvement. As Adams (1986) pointed out, in common beans, high yields under moisture stress result from partitioning of a greater biomass not merely from a high partitioning ratio per :3. Thus, it is the ability to accumulate biomass and to partition it to the seed, under stress, that distinguishes top yielders from.law yielders. Although photosynthesis and nitrogen fixation during growth are two different processes, they cannot be separated since, in most legumes, remobilization of stored starch and nitrogen compounds from the leaves to the protein rich seeds plays an important role in the final yield and composition of the seeds (Summerfield, 1981). X' Ii 3 . 1: I It is well established that in any crop the effect of water stress on growth and yield depends on the degree of the stress, the stage of growth at which the stress occurs and the duration of the stress period. 15 In the case of common beans, there is general agreement in the literature that the reproductive stage is the most sensitive to water stress, affecting seed yield by reducing the number of pods set and single ...a weight (Robins and Domingo, 1955; Dubetz and Mahalle, 1969; Stoker, 1974; Stansell and Smittle, 1980; Bonnano and Mack, 1983; Samper, 1984, Ibarra, 1985, Elizondo, 1987, Acosta and Kohashi, 1988). Reductions in seed yield of up to 80 x for drought sensitive bean cultivars as compared to 40 % reduction for tolerant ones were reported by Sponchiado (as cited by Singh and White, 1988). Stoker (1974) obtained a yield reduction of 20 %.when water stress occurred at early or late vegetative phases of growth, and a reduction of 50 1 when stress was applied at early pod filling. Similar figures for a group of three indeterminate bean cultivars were reported by Acosta and Kohashi (1988). The yield of common beans may be considered as the product of its components: number of pods per plant, number of seeds per pod and individual seed weight. According to Adams (1967), there is no genetic interdependence among yield components in beans since correlations between components was essentially zero under non-competitive conditions. However, yield reductions in common beans due to water stress can be attributed to its effects on one or’ more components according to the stage of growth of the crop, and intensity and duration of the stress. Multiple reports in the literature have shown that if water stress occurs during vegetative growth, number of pods per plant is reduced; if water stress occurred during flowering, number of pods and number of seeds per pod are significantly reduced; if water stress occurs late during the pod filling stage, seed weight is reduced (Robins and 16 Domingo, 1956; Dubetz and Mahalle, 1969; Acosta and Kohashi, 1988). This would indicate that different components are being laid down sequentially during development. BEE l E i ll I ll 1 . 1 'l ii I' In temperate climates, variability for nitrogen fixation is due mainly to the environment and evidence is accumulating that. nitrogen fixation is more sensitive to stress than is the uptake of mineral nitrogen (Sprent, 1982). Graham (1981) pointed out that moisture stress is one of the environmental factors affecting nitrogen fixation in common beans; however, little research has been conducted on the matter. Sprent (1981) and Finn and Brun (1980) have suggested that water stress reduces nitrogen fixation in soybeans by a direct effect on nodule physiology, but may be aggravated by the inability of stressed leaves to supply photosynthates to the nodules. Other evidence suggests that reduction in photosynthates during water deficits causes the observed reductions in nitrogen fixation (Huang et al., 1975a,b). Nitrogen fixation has also been shown to decrease as nodule number and leaf water potentials decrease (Pankhurst and Sprent, 1975; Finn and Brun, 1980). Bennett and Albrecht (1984), working under greenhouse conditions, found that after 10 days of withholding water, nitrogen fixation was reduced, and nitrogenase activity declined to zero as stress became progressively more severe. Their data indicate the sensitive nature of nitrogen fixation to reductions in the water content of the soil and further suggest that nodules surrounded by dry soil may dessicate to water potentials lower than those observed for leaf tissue. Nitrogen fixation appeared to be more sensitive to drought stress than was 17 photosynthesis (Weisz et al. 1985), suggesting that some drought yield reductions may reflect the effects of nitrogen deficiency. Saito et al . (1984) conducted a glasshouse experiment with the bean cultivar Carioca and found that soil water content affected nodulation, N fixation and the utilization of mineral nitrogen by the plants. 2 Plants grown in wet soil produced twice as much biomass as those grown in dry soils . Nodule weight and activity were five to ten times greater than those from dry soil. Decreases in soil water content were accompanied by decreases in growth, and in nitrogen accunmlation by the plant. Recently, Durand et al. (1987) studied the effects of water deprivation on the activity of nodules of soybeans . During a seven-day period of water deprivation there was a close relationship between decreases in leaf and nodule water potential . Nitrogenase activity showed a 70 1 decrease during the first four days, whereas photosynthesis declined by only 5 1. They suggested that water stress exerts an influence on nitrogenase activity which is independent of the rate of photosynthesis; it acts directly on nodule activity through increases in the resistance to oxygen diffusion to the bacteroids. The data suggest that the linear relationship between oxygen diffusion resistance and water potential is more important than any reductions in photosynthate supply . Abdel-Ghaffar et al., (1982) conducted an experiment in which cannon beans were planted, then irrigated every 7, 12, 17, or 22 days during the growing season. They found that water stress inhibited nodulation, depressed nitrogenase activity and decreased the yield of been plants. Maximum yields were obtained when plants were irrigated 18 every 7-12 days. Briefly, it is believed that the combination of drought tolerance and high ability to fix nitrogen in a single bean cultivar can occur, but optimum expression of each will depend upon differential timing in development . MW: In the grain legumes in general , and with conlnon beans in particular, there is no single factor or unique characteristic which is sufficient alone to account for "adaptation to drought-prone enviromnents" . Traditional bean cultivars in Latin America seem to be adapted to cropping systems which imply sacrificing yield potential in exchange for reduced risk, production costs and other problems (White, 1987 ) . Improvement in plant production need not rest solely on increases in genetic potential but should also emphasize ways of bringing productivity closer to the existing genetic potential through management techniques . Plant types that are productive with lower imputs should be readily accepted by the farmers (Boyer, 1982) . Therefore, as pointed out in the introduction, selection for increased stability in variable environment should be assigned high priority in research centers . Where unpredictable water deficits occur, maximum productivity should be the goal , whereas conservation of water using a shorter growing season crop should be the goal where terminal water deficits occur. Maximn productivity will be achieved where leaf expansion and senescence are relatively insensitive to water deficit, where leaf expansion recovers quickly upon relief of water deficits , and where minimum dry matter is partitioned into inlnobile root reserves. 19 Crop water use efficiency is in nature a conservative approach (Sinclair et al., 1984), and selection for wide adaptation and/or yield stability has been considered to be defensive breeding. The improvement of been cultivars for ability to fix atmospheric nitrogen and display tolerance to soil moisture deficits will undoubtely benefit peasant farmers in those regions where the bean crop relies an uncertain rainfall patterns for its growth. REFERENCES Acosta Gallegos, J .A. and J. Kohashi Shibata. 1988. Effect of water stress on growth and yield of indeterminate dry bean (W W L.) cultivars. Submitted : Field Crops Research. The Netherlands Abdel-Ghaffar, A.S., El-Altar, H.A., El-Halfawe, M.H. and A.A. Abdel Salan. 1982. Effects of inoculation, nitrogen fertilizer, salinity and water stress on smbiotic N fixation by 11.9.13 faba and mm. In: P..H GrShanandS.C. Harris (Eds. ).BNF Technology for Tropical Agriculture. CIAT, Cali, Colombia pp. 153- 159 Adans, M. W. 1967. Basis of yield component compensation in crop plants with special reference to the field bean Phenom: mlgans L. Crop Sci. 7: 505- 507 Adams, M.W., Coyne, D.P., Davis, J.H.C., Grahan, P.H. and C.A. Francis. 1985. Cannon bean (Phamlus mm L.). In: R.J. Simmerfield and E.H. Roberts (Eds. ), Grain Legmne Crops. Willian Collins and Sons 6 Co. Ltd pp. 433-476 Adans, M.W. 1986. Improving resistance to enviromental stress in beans through genetic selection for carbohydrate partitioning and efficiency of biological nitrogen fixation. Bean-Cowpea-CRSP Report on the INIFAP (Mexico)-MSU Project p.24 Aparicia, T.P.M. abd J .8. Boyer. 1983. Significance of accelerated leaf senescence at low water potentials for water loss and grain yield in maize. Crop Sci. 23:1198-1202 Apel, P. 1984. Photosynthesis and assimilate partitioning in relation to plant breeding. In: P.V. Bose and 8.6. Blixt (Eds.), Crop Breeding. A Contemporary Basis. Perganon Press pp.163-182 Beaver, J.S. Paniagua, C.V., Coyne, D.P. and G.F. Freytag. 1985. Yield stability of dry bean in the Dominican Republic. Crop Sci. 25:923- 926 Bennett, J .M. and S.L. Albrecht. 1984. Drought and flooding effects on N fixation, water relations and diffusive resistance of soybean. on. J. 76:735-740 Bidinger, F., Musgrave, R.B., and R.A. Fischer. 1977. Contributions of stored preanthesis assimilate to grain yield in wheat and barley. Nature 270:431-433 20 21 Bliss, F.A. 1985. Breeding for enhanced dinitrogen fixation potential of common bean (Phasealns_ynlgaris L.). In: P.W. Ludden and J'Eh Burris (Eds.). Nitrogen Fixation and CO Metabolism. Proc. 14 Steenbock Symposium, 17-22 June 1984, University of Wisconsin, Madison. Elsevier Science Publishing Co., pp. 303-310 Blum, A. 1979. Genetic improvement of drought resistance in crop plants: A case for sorghum. In: H. Mussell and R.C. Staples (Eds.), Stress Physiology in Crops Plants. John Wiley and Sons Inc. pp. 429-445 Bonanno, A.R., and H.J. Mack. 1983. Yield components and pad quality of snap beans growth under differential irrigation. J. Amer. Soc. Hort. Sci. 105:869-873 Boyer, J.S. 1982. Plant productivity and environment. Science 218:443- 448 Boyer, J.S. 1985. Water transport. Ann. Rev. Plant Physiol. 36:473-516 Bunting, A.H. 1985. What is this thing called drought 7 Workshop on drought, Bean-Cowpea CRSP, Durango, Mexico, 26-28 August p.9 Caviness, C.E. and J .D. Thomas. 1980. Yield reduction from defoliation of irrigated and non-irrigated soybean. Agron J. 72:977-980 Coyne, D.P. 1980. Modification of plant architecture and crop yield by breeding. Hortscience 15:244-247 Couto, L. 1978. Effects of water stress on growth, reproductive development, dry matter partitioning and yield components of beans (Phasealns_ynlgaris L.) in the field. Ph.D. Thesis. University of California, Davis Donald, C.H. and J. Hamblin. 1976. The biological yield and harvest index of cereals as agronomic and plant breeding criteria. Adv. Agron. 28:361-405 Dubetz, S. 1969. An unusual photonastism induced by drought in Phaseglns ynlgaris. Can. J. Bot. 47:1640-1641 Dubetz, S. and P.S. Mahalle. 1969. Effect of soil water stress on bush beans W L. at three stages of growth. J. Amer. Hort. Sce. 94:479-481 Durand, J.L., Sheehy, J.E. and F.R. Minchin. 1987. Nitrogenase activity, photosynthesis and nodule water potential in soyabean plants experiencing water depravation. J. Exp. Bot. 38:311-321 El-Fouty, 1M.M. 1982. Flower and pad drop. In: G. Hawtin and C. Webb (Eds.), Faba Bean Improvement. ICARDA. Martinus Nijhoff, The Netherlands pp.177-184 22 Elizondo-Barron, J. 1987. Characteristics related to yield of dry bean (Rhasgglns_ynlgaris L.) under water stress conditions. M.S. Thesis Michigan State University, East Lansing p. 73 Elston, J. and A.H. Bunting. 1980. Water relations of legume crops. In: R.J. Summerfield and A.H. Bunting (Eds.), Advances in Legume Science. Royal Botanic Garden, Kew pp.37-42 Finn. C.A. and W.A. Brun. 1980. Water stress effects on CO assimilation, photosynthate partitioning, stomatal resistance and nodule activity in soybean. Crop Sci. 20:431-434 Gallager, J.N. and P.V. Biscoe. 1982. A physiological analysis of cereal yield. II. Partitioning of dry matter. Agricultural Progress 53:51-70 Gardner,, F.P., Pearce, R.B. and.RlL. Mitchell. 1985. Physiology of Crop Plants. Iowa State University press, Ames pp. 58-75 Gentry, H.S. 1969. Origin of common bean, 2haseglns,ynlgaris. Econ. Bot. 23:55-69 Graham, P.H., and J.C. Rosas. 1977. Growth and development of indeterminate bush and climbing cultivars of Phasealns,yn1garis L. inoculated with Rhigghium. J. Agric. Sci. Camb. 88:503-508 Graham, P.H. 1981. Some problems of nodulation and symbiotic nitrogen fixation in Phaseolns,ynlgaris L. : a review. Field Crops Res. 4:93-112 Hall, A.E. and P.N. Patel. 1985. Breeding for resistance to drought and heat. In: S.R. Singh and K.O. Rachie (Eds.), Cowpea:Research, Production and Utilization. John Wiley & Sons pp 137-151 Hebblethwaite, P. 1982. The effects of water stress on the growth, development and yield of Vicia faba L. In: G. Hawtin and C. Webb (Eds), Faba Bean Improvement. ICARDA. Martinus Nijhoff, The Netherlands. pp. 165-175 Hsiao, T.C. 1973. Plant responses to water stress. Ann. Rev. Plant Physiol. 24:519-570 Huang, C.Y., Boyer, J.S. and L.N. Vanderhoef. 1975a. Acetylene reduction (nitrogen fixation) and metabolic activities of soybean having various leaf and nodule water potentials. Plant Physiol. 56:222-227 Huang, C.Y., Boyer, J.S. and L.N. Vanderhoef. 1975b. Limitation of acetylene reduction (nitrogen fixation) by photosynthesis in soybean having low water potentials. Plant Physiol. 56:228-232 Ibarra, P.F. 1985. Effects of water stress on the relationships between root and shoot characteristics in dry beans. M.S. Thesis, Michigan State University, East Lansing p. 114 23 Johnson, R.R., and D.N; Moss. 1976. Effect of water stress on 1“C0 fixation and translocation in wheat during grain filling. Crag Sci. 16:697-701 Jordan, van. 1983. Whole plant responses to water deficits: an overview. In: HAM. Taylor, W.R. Jordan and T.R. Sinclair (Eds.), Limitations to Efficient Water Use in Crop Production. ASA, CSSA, SSSA. pp.289-317 Karamanos, A.J. 1978. Water stress and leaf growth of field beans (Iigia__faha L.) in the field: leaf number and total leaf area. Ann. Bot. 42:1393-1402 Kelly, J.D., Adams, MAW. and G.V. Varner. 1987. Yield stability of determinate and indeterminate dry bean cultivars. Theor. Appl. Genet. 74:516-521 Kraner, P.J. 1983. Water relations of plants. Academic Press, Inc. pp.342-389 Laing, D.R., Kretchmer, P.J., Zuluaga, S. and P.G. Jones. 1983. Field bean. In: wgag Smith and S.J. Banta (Eds), Symposium on Potential Productivity of Field Crops Under Different Environments. Los Banos, Philippines, 1980 (Proceedings) IRRI pp.227-248 Laing, D.R., Jones, P.G. and J.H.C. Davis. 1984. Common bean (Phaseolns ynlgaris L.)In: P.R. Goldsworty and N.M. Fisher (Eds), The Physiology of Tropical Crops. John.Wiley & Sons Ltd. pp.305-351 Levit, J. 1980. Responses of Plants to Environmental Stress. 2nd. ed. Vol. 2 Academic Press, New York Pankhurst, C.E. and J.I. Sprent. 1975. Effects of water stress on the respiratory and nitrogen-fixing activity of soybean root nodules. J. Exp. Bot. 26:287-304 Passioura, J.E. 1986. Resistance to drought and salinity: avenues for improvement. Aust. J. Plant Physiol. 13:191-201 Piha, M.I., and D.N. Munns. 1987. Nitrogen fixation potential of beans (Phaseglns ynlgaris, L.) compared with other grain legumes under controlled conditions. Plant and Soil 98:169-182 Quisenberry, J.E. 1982. Breeding for drought resistance and plant water use efficiency. In: J.W. Christiansen and C.F. Lewis, (Eds.), Breeding Plants for Less Favorable Environments. John Wiley and Sons, Inc. pp. 193-212 Rennie, R.J., and C.A. Kemp. 1981a. Selection for dinitrogen-fixing ability in Phaseglns ynlgaris L. at two law-temperature regimes. Euphytica 30:87-95 Rennie, R.J., and C.A. Kemp. 1981b. Dinitrogen fixation in pea beans (Phaseglus ynlgazis) as affected by growth stage and temperature 24 regime. Can. J. Bot. 59:1181'1188 Ritchie, J .T. 1983. Efficient water use in crop production: discussion on the generality of relations between biomass production and evotranspiration. In: H.M. Taylor, W.R. Jordan and T.R. Sinclair (Eds. ), Limitations to Efficient Water Use in Crop Production. ASA, CSSA, SSSA. pp.29-44 Robins, J .R., and C.E. Domingo. 1956. Moisture deficits in relation to the growth and development of dry beans. Agron. J. 48:67-70 Saito, S.M.T., Nazareth, M., Montanheiro, 3., Victoria, R.L. and K. Reichardt. 1984. The effect of N fertilizer and soil moisture on the nodulation and growth of We maria. J. Agric. Sci. Camb. 103:87-93 Samper, C. 1984. Effects of water stress imposed at mid-pod filling upon yield and dry matter partitioning in dry beans (Phenom mm L.). M.S. Thesis, Michigan State University, East Lansing p. 131 Sanper, C., Adams, M.W., Hanson, A.D., and E. Watt. 1984. Contribution of stored assimilates to the seed following a drought stress in dry beans. Agronomy Abstracts 76:87 Sheriff, D.W. and R.C. Muchow. 1984. The water relations of crops. In: P.R. Goldsworthy and N.M. Fisher, (Eds. ), The Physiology of Tropical Field Crops. John Wiley and Sons Ltd. pp.39-83 Shultze, E.D., Robichaux, R.H., Grace, J., Rundel, P.W., and J.R. Ehleringer. 1987. Plant Water Balance. BioScience 37:30-37 Sinclair, T.R., Tanner, 0.8. and J.M. Bennett. 1984. Water-use efficiency in crop production. BioScience 34:36-40 Singh, S.P. and J.W. White. 1988. Breeding cannon beans for adaptation to drought conditions. In: G. Hoogenboom, F. Ibarra, S.P. Singh, J .W. White and S. Zuluaga (Eds. ), Research on Drought Tolerance in Cannon Beans. CIAT, Cali, Colombia. In press Sprent, J .I. 1981. Nitrogen fixation. In: L.G. Paleg and D. Aspinall (Eds. ), The physiology and Biochemistry of Drought Resistance in Plants. Academic Press pp.131-142 Sprent, J .I. 1982. Nitrogen fixation in grain legumes in the U.K. Phil. Trans. R. Soc. Lond. 296:387-395 Stansell, J .R., and D.A. Smittle. 1980. Effects of irrigation regimes on yield and water use of snap bean (Wm mm L.). J. Amer. Soc. Hort. Sci. 105:869-873 Stoker, R. 1974. Effect on dwarf beans of water stress at different phases of growth. N. 2. J. Exp. Agric. 2:13-15 To; Vid Val 25 Sullivan, C.Y. 1983. Genetic variability in physiological mechanisms of drought resistance. Iowa State J. Res. 57:423-439 Summerfield, R.J. 1981. Adaptation to environments. In: C. Webb and G. Hawtin (Eds.). Lentils. ICARDA, Page Brass (Narwich) Ltd. pp 91- 110 . Temple, S.R. and L. Long. 1980. Crop improvement and genetic resources in Phasgglns ynlgaris for the tropics. In: R.J. Summerfield and A.H. bunting (Eds.), Advances in Legume Science. London, HMSO pp.365-373 Tosso, J.E. 1979. Effects of different levels of applied water on the vegetative growth, yield and water production function in dry beans (Phaseolng ynlgaris L.). Ph.D. Thesis, University of California, Davis p. 125 TUrner, N.C. 1979. Drought resistance and adaptation to water deficits in crop plants. In: H. Mussell and R.C. Staples (Eds.), Stress Physiology in Crop Plants. John Wiley and Sons, New York pp.343- 372 Turk, K.J. and A.E. Hall. 1980. Drought adaptation of cowpea. III. Influence of drought on plant growth and relations with seed yield. Agron. J. 72:428-433 Vidal, A. and M. Arnoux. 1981. Drought tolerance processes in soybean. Biol. Plant. 23:434-441 Wallace, D.H. and H.M;‘Munger. 1966. Studies of the physiological basis of yield differences. II. Variation in dry matter distribution among aerial organs for several dry bean varieties. Crop Sci. 6:503-507 Weisz, P.R., Denison, R.F. and T.R. Sinclair. 1985. Response to drought stress of nitrogen fixation (acetylene reduction) rates by field grown soybeans. Plant Physiol. 78:525-530 White, J.W. 1987. Strategies at CIAT for increasing yield potential of conmon beans -- finishing the domestication process. In: Symposium on Breeding for Higher Yields in Common Beans. XXXIII meeting of the PCCMCA, Guatemala, C.A. CIAT-B/C CRSP-ICTA p. 34 Wien, H.C., Littleton, E.J. and A. Ayanaba. 1979. Drought stress of cowpea and soybean under tropical conditions. In: H. ZMussell and R.C. Staples (Eds.), Stress Physiology in Crop Plants. John Wiley and Sons, New York pp.283-301 “J 3U in; K01 CHAPTER 1 BIOLOGICAL NITROGEN FIXATION AND DROUGHT TOLERANCE OF A GROUP OF BEAN GENOTYPES. Phenom malaria L. INTRODUCTION Leguminous plants use two principal sources of nitrogen in their nutrition : soil nitrate and atmospheric nitrogen. Legume symbiosis is governed by factors and processes independently inherited in the host and bacterium: these interact to produce a joint phenotype (Nutman, 1981) . The involvement of host genetic factors in root nodule symbiosis was first suggested by Wilson in 1939 (Nutman, 1981). He showed that the host ranges of different strains of W was not only related to properties of the bacteria but also to the hosts' characteristics (Nutman, 1981) . Most cultivated legumes are able to fix nitrogen, but differences in the efficiency of nitrogen fixation between species of legunes has been observed (Schubert and Evans, 1976: Piha and Munns, 1987). Differences have also been observed within species. Measurements of nitrogen fixed and various paraneters of nitrogen fixation suggest sufficient variability among cultivated dry bean gernplasm to allow improvement through selection (Graham and Rosas, 1977: Westerman and Kolar, 1978: Rennie and Kemp, 1981a,b: Mc Ferson, 1983: Rosas, 1983: Felix et al., 1984: Pacowsky et al., 1984: Bliss, 1985: St. Clair, 1986) . 26 of red biol pro< Zabj 1986 meta nit: to (Alb exis‘ 1984; 27 In. most legume crops it has been demonstrated that the occurrence of a water deficit drastically reduces nitrOgen fixation. This reduction can be either by a direct effect on the physiology and/or biochemistry in the nodule tissue or indirectly by affecting the production of photosynthates in the host plant (Sprent, 1981: Zablotowicz et al., 1981: Beadle and Long, 1985: IMyers Jr. et al., 1986), since the energy for nitrogen fixation is derived from. root metabolism of photosynthates. Recent findings seem to indicate that nitrogenase activity is much.more sensitive to nodule dehydration than to reduction in photosynthesis due to a fall in leaf water potential (Albrecht et al. .1984). On the other hand, it has been demonstrated that genetic variation exists in cannon beans for traits related to drought tolerance (Sanper, 1984: Elizondo, 1987). Elizondo (1987) studied the response of a group of 11 conmon bean cultivars to a mild water stress periodlimposed after anthesis. A principal factor analysis and stepwise multiple regression analysis performed on 27 traits showed that biomass at physiological maturity made the largest contribution to seed yield variance. He concluded that genotypes best adapted to drought were those possessing the greatest biomass at physiological maturity. The objective of this research was to rank a selected group of common bean genotypes for BNF related variables and drought tolerance. Since the literature shows that plant growth and development as well as BNF are affected by water deficits, the main objective was to determine if drought tolerance and ability for BNF could be combined in a single genotype. In order to achieve those objectives BNF related variables were evaluated early during the life cycle of the crop and water stress 28 related variables were evaluated after flowering, when the stress was imposed. MATERIALS AND METHODS To accomplish the stated objective, two experiments were conducted: a greenhouse experiment where variability in fixing atmospheric nitrogen was assessed within a selected group of bean genotypes, and a field experiment where a subgroup of selected cultivars was evaluated for drought tolerance and BNF Capacity. Greenhouse___gxpgziment. 'A group of 11 cultivars (Table 1) previously selected as drought tolerant, capable of good nitrogen fixation, or high yielding, were planted in January, 1985 in a greenhouse at Michigan State University (MSU), East Lansing, MI. Uniformly sized seeds were over-inoculated.with one of two strains of RAM. our 899 (isolated by p. Grahan at our) and J025 (isolated by J. Maya, provided by F. Dazzo from the Brazil-UW-MSU, CRSP project). In addition to treatments inoculated with two Rhingium strains, two other treatments without inoculant were included, one with nitrogen fertilizer (positive check) and the other without nitrogen fertilizer (negative check). For each cultivar, two seeds per pot were planted in 8x12 cm plastic pots containing a sterilized pure silica sand medium. After emergence, seedlings were thinned to one per pot. Depending on treatments, pots were irrigated each week day with a Hoagland solution with or without nitrogen. The four treatments obtained by testing the two Rhinginm strains and both checks on the 11 cultivars, were distributed in a Completely Randomized Design (CRD) with eight replicates. Each experimental unit was a single pot containing one plant. At 25 and 50 days after planting 29 Table 1. Characteristics of been genotypes grown in greenhouse (C) and/or field (F) experiments. East Lansing, MI. 1985 . . . 3 Seed Plant . Genotype Pedigree Origin color type Experiment UW.21-58 P. sint/Pue 152 UW black II CF UW.23-61 Ex-Rico 23/Pue 152 UW white II F A411 BAT 461/ (G879x62337) CIAT brown II C BAT 336 51052/Cacahuate CIAT cream II C N81017 Kent/NepZ/lPijao/Bunsi MSU white II GF N81064 Bunsi/NepZ MSU white II GF B76001 Nep2/BTS MSU black II G LEF-Z-RB (Ver 10/Chis 143)/Pue 144 INIFAP black III C striped Mex . 1213 - 2 Unknown INIFAP pinto I I I GF Dgo.222 Durango 222 INIFAP white III G Bayo Madero Bayo RGZ/C-102 INIFAP crean III GF II900-5-M-45 Pinto Amer./C-14 INIFAP brown III CF striped Pinto Nal.1 local Durango INIFAP pinto III F * UW II University of Wisconsin MSU = Michigan State University CIAT = Internatinal Center for Tropical Agriculture, Colombia INIFAP = National Institute for Forestry and Agriculture Research, Mexico 1- Type II = indeterminate-bush, erect stem and branches. Type III 8 indeterminate-bush, postrate main stem and branches . 30 (DAP) , plants were harvested by cutting them at the sand level. Shoots °c were placed in paper bags and dried in a forced dry air oven at 60 for 72 hours . Dry matter per plant was determined. Roots were carefully washed and the nitrogenase activity of plants in each treatment, at 25 DAP, except those under positive check, was determined by the acetylene reduction assay. Each root system with attached nodules was placed in a 50 cc flask which was sealed. A 10 ml volune of air was withdrawn with a syringe and replaced by 10 ml of acetylene. After 30 min incubation, 2 ml of gas were taken from the flask and injected into a calibrated gas chromatograph using nitrogen as a carrier gas at a flow rate of 25 ml/min at 80 °C. After nitrogenase activity was determined, nodules were detached from the roots and their fresh weight was recorded. Analyses of variance were performed on the recorded variables following a CRD as a two factor factorial 4 x 11. Simple correlations were calculated among recorded variables by using genotype averages over treatments . W. This experiment was planted in June 1985 on a Mariette fine-loamy soil (mixed, mesic, Glassoboric Hapludolfs) (USDA, Soil Conservation Service) at the Crops Research Farm of MSU in East Lansing, MI. A group of eight cultivars previously selected as drought tolerant or with outstanding BNF ability were tested in the field by using a Randomized Complete Block Design arranged as a split-split plot with three replications. Each experimental plot consisted of four rows of 4m length, spaced 50 cm apart. Plant density was 20 plants per m2. Treatments were as follows: Factor A: 1- drought stress imposed at flowering time 1'. II I row that were (DA; Of: Blis matt PhYS fill 31 2- rainfall and complementary irrigation Factor B: 1- commercial inoculant (no N fertilizer) 2- N fertilizer, 40 kg/ha Factor C Genotype: 1 UW 21-58 2 UW 21-54 3 N81017 P I N81064 U" I Mex.1213-2 0‘ I Bayo Madero 7 II900-5-M-45 8- Pinto Nal.1 Planting was done following rain, when soil moisture was considered adequate for germination (Table 1, Appendix A). In order to establish the treatments, planting was done using a hoe to open rows and depositing the seeds and abundant commercial inoculant in the bottom of the furrow then immediately covering them with moist soil. Nitrogen fertilizer was applied in a similar fashion 10 cm apart from.the seeded row. To create the stressed treatments, advantage was taken of the fact that the experimental site had a relatively strong slope (3 1). Rows were designed to be parallel to the slope and at 45 days after planting (DAP), black plastic stripes (0.4 x 6 m) were placed between the rows of stressed treatments. At 45 DAP, a visual nodulation score as described by Rosas and Bliss (1986) (Table 2, Appendix A) was recorded and total plant dry matter taken at two sampling times 45 and 70 DAP, was collected. Physiologically related variables were measured at flowering and mid-pod filling stages which correspond with the initiation and 25 days after l: we at we Fi: SUE 9XP and out. Coll hon: weig anal' fiel 382‘: Seas. 32 drought treatment was imposed. The recorded variables were: leaf water potential (LWP), using the pressure chamber technique: leaf conductance (LC), using a diffusion parameter: leaf area index (LAI), with the use of a leaf area meter. The relative crop growth rate and leaf specific weight were also calculated. Leaf elongation in each treatment was determined after drought treatment.was imposed by tagging three young leaves and recording the length of the central leaflet every day for two weeks. Leaf nitrogen content.was determined by micrOKjeldahl procedure at 45 DAP. At physiological maturity, seed yield and yield components were recorded and a harvest index (HI) was calculated. The method of Fischer and.Maurer (1978) (Appendix A) was used to calculate the drought intensity index (DII) for the site and the individual drought susceptibility index (DSI) per genotype. Soil moisture was monitored regularly in six different plots in the experiment at four depths: 0-15, 16-30, 31-45, and 46-60 cm. The soil moisture content of each plot was recorded at two different depths: 0-15 and 16-30 cm on the same day that pressure bomb readings were carried out, at two different depths: 0-15 and 16-30 cm” Soil samples were collected, weighed and immediately placed into an oven at 110 0C for 24 hours. The soil moisture content was expressed as a percentage on dry weight basis (Table 3, Appendix A). All measured variables were analyzed, following the design used to distribute the treatments in the field, with the aid of the MSTAT microcomputer statistical package for agricultural sciences. Data on rain and temperature during the growing season are presented in Table 1 of Appendix A. 33 RESULTS AND DISCUSSION In both experiments, genotype UW’ 21-58, which possessed an increased ability to fix atmospheric nitrogen (Bliss, 1985), was utilized as a check cultivar for comparison when looking at nitrogen fixation related variables. Greenhouse_experimentg In this experiment, results were recorded for samples taken at 25 DAP and emphasis was put on the variation of the 11 bean genotypes. Samples taken at 50 DAP were discarded duel to Rhizahinn contamination. The genotypes A 411 and N81064 showed the highest and lowest values, respectively, for nodule fresh weight per plant. For plant dry matter, cultivar Dgo. 222 showed the highest value for this trait, whereas N81064 showed the smallest value (Table 2). The activity of the enzyme dinitrogenase quantified by the acetylene reduction (AR) assay at 25 DAP is presented in Table 2. As we were mainly interested in the BNF ability of the bean genotypes, which accounted for’ most of the observed variation for AR, 'we will not present results for the treatments. Thus, on a per plant basis Lef-Z-RB was the genotype which fixed more N: with respect to nodule specific activity. Lef-Z-Rb, N81017 and the check cultivar UW 21-58 showed the highest values for this trait. Values for phenotypic correlations between pairs of variables are presented in Table 3. Plant biomass showed a significant positive correlation with all variables studied. At this stage of development (25 DAP), nodule fresh weight was highly correlated with nodule specific activity. Nodule fresh weight values showed the highest r value when correlated to plant dry matter. 34 Table 2. Genotype average value for biomass per plant and nitrogen fixation related variables. East Lansing, MI. 1985 Nadule AR per Nadule Plant fresh plant/ specific Genotype biomass(1) weight(2) hour(3) activity(4) aw 21-58 1.14 be 410 be@ 3593 abe 8.8 A 411 1.11 b 560 a 3336 abc 5.9 BAT 336 0.79 c 310 de 1837 e 5.9 N81017 1.00 b 370 had 3324 abc 9.0 N81064 0.75 c 270 e 1824 e 6.8 876001 1.04 b 430 bc 2877 had 6.7 LEF-Z-RB 1.04 b 440 be 4140 a 9.4 Mex.1213-2 1.09 b 350 cde 2479 cde 7.1 Dgo-222 1.35 a 460 b 2524 cde 5.5 Bayo Madero 1.05 b 380 had 2303 de 6.0 II900-5-M-45 1.08 b 450 b 3202 had 7.1 Average 1.04 403 2859 7.1 1- g, average of 32 pots. 2- mg/plant, average 24 pots. 3- nm/plant/hour, average 24 pots. 4- nmvmg nodule tissue/plant/hour, average 24 pots. 0 Duncan's Multiple Range Test (0.05). 35 Table 3. Correlations between biomass per plant and nitrogen fixation variables. East Lansing, MI. 1985 Nadule Nitrogenase Nadule fresh activity specific weight per plant activity (1) Plant biomass g/plant 0.57"")I 0.36** 0.45** Nadule fresh weight mg/plant ---- 0.47** 0.82** Nitrogenase activity ---- ---- 0.45** per plant nm/p/h ** highly significant (P<0.01). (1) nm/mg nodule tissue/plant/h 36 Field__experiment. In this experiment, the moisture stress treatment was not started until 45 DAP, thus analyses of variance for samples taken at 45 DAP included only the effects of nitrogen rate, genotype and their interaction (Tables 5, Appendix A). The addition of 50 kg/ha of nitrogen fertilizer increased significantly the weight of all above-ground plant parts (P<0.05), thus, the growth of the plant was N limited. The effect of the genotype was significant only for root and stem dry weights (P<0.05). No significant effect was found for the nitrogen x genotype interaction (Table 5, Appendix A). In Figure 1, total dry matter and dry matter of different plant parts under two nitrogen sources at 45 DAP are presented. The genotypes UW 23-61, Bayo Madero and Pinto Nal.1 showed a significant respOnse to the added nitrogen fertilizer. Genotype II900-5-M-45 gave the highest total dry matter (TDM) yield without added fertilizer. In the treatment without nitrogen fertilizer a commercial inoculant along with phosphorus fertilizer was uniformiy applied to the whole experiment: thus, assuming a low N content in the soil, dry matter yield and total leaf nitrogen content under this treatment could be used as an indicator of BNF. In general, at this stage of development, the type III unadapted cultivars were superior in total dry matter yield to adapted type II cultivars under both nitrogen sources. Cultivar II900-56M-45 was outstanding under inoculation. Graham. and Rosas (1977) previously reported the superiority in BNF of indeterminate type III growth habit versus type II cultivars. It was noticed when digging roots in the sampled strata, that type III genotypes showed a.much.more branched root system.than type II's. 37 Dry matter 8"“ D Leaves 200 .l W Stems IIIII Roots 150 -— FIT 1J 11 T] 33 T j] 100 -A SO N I N I. N I N I N I 21-58 23-61 017 064 11900 BM 1213-2 Cultivars Z H N I '62 H Figure 1. Dry matter of different plant parts of eight bean cultivars under two nitrogen sources (N=SOkg/ha nitrogen, I=inoculated) at 45 days after planting. East Lansing, MI. 1985. 38 Hewever, the latter genotypes showed a strong main root with few secondary roots going deep into the soil profile. Thus, it is likely that the root system of erect type II genotypes is more suited to support a taller plant under Michigan conditions: whereas the root system displayed by type III genotypes seems more suited to explore more soil volume, mainly in the upper soil strata. Probably this latter type of root system.plays an important role in the adaptation of type III cultivars to sporadic rainfall patterns of the semiarid highlands of Mexico. At this stage of development (45 DAP), the leaves accounted for approximately two-thirds of shoot biomass (Figure 1). In the analysis of variance for leaf dry weight, no significant difference was found between genotypes: however, a highly significant difference among genotypes was found for LAI. Therefore, it is clear that the genotypes possess different leaf specific weights (Table 4). In general, type III cultivars were superior to type 11's for LAI. With respect to leaf water status at this stage of development (45 DAP), genotype Bayo Madero showed the least negative leaf water potential (LWP), which was significantly different (LSD 0.05) from the rest of the genotypes (Table 4). The same genotype gave the smaller value for stomatal conductance. Since the moisture stress treatment had not been applied at this time in development, genotypic variations for these physiological variables were probably obtained in response to the relatively high temperatures (Table 1, Appendix A) observed at mid-day. Analyses of variance for nodule score and some physiological variables did not show significant differences for the effect of nitrogen rates nor its interaction with genotypes (Table 5, Appendix A). 39 Table 4. Average values for physiological related variables of eight bean cultivars under two nitrogen sources at 45 days after planting. East Lansing, MI. 1985 Stamatal Conduct. LWP* LSW* Root/ Genotype LAI* cml s bars mg/cm LAR* Shoot UW 21-58 1.41 0.278 -6.3 4.77 122.3 0.202 UW 23-61 1.28 0.239 -7.2 5.01 111.9 0.211 N81017 1.15 0.204 -6.7 5.76 97.6 0.226 N81064 1.31 0.209 -6.2 5.15 110.2 0.228 II900-5-M-45 2.32 0.196 -6.1 3.69 157.1 0.138 Bayo Madero 1.82 0.168 -3.8 4.97 124.6 0.119 Mex.1213-2 1.84 0.209 -6.4 4.46 134.1 0.158 Pinto Nal.1 2.14 0.272 -6.1 3.82 147.4 0.151 LSD 0.05 0.51 ____ 2.3 ____ 32.3 0.041 * LAI = leaf area index. LWP = leaf water potential. LSW 8 leaf specific weight. LAR - leaf area ratio. Table 5. Nadule score and total leaf nitrogen content of eight bean genotypes grown under two nitrogen sources at 45 days after 40 planting. East Lansing, MI. 1985 Nadule Score (1) Leaf Nitrogen Cont. (2) Genotype Inoculant N fertilizer Inoculant N fertilizer UW 21-58 3.0 2.2 2.8 2.9 UW 23-61 1.5 1.5 2.5 3.4* N81017 2.2 2.5 2.6 3.0* N81064 1.5 1.8 2.8 3.3* II900-5-M-45 2.3 3.0 3.5 3.5 Bayo Madero 2.0 1.5 2.9 4.4* Mex.1213-2 2.3 2.5 3.5* 3.3 Pinto Nal.1 2.7 2.3 2.8 3.9* Average 2.2 2.1 2.9 3.5* * LSD 0.05 between nitrogen sources. (1)- Following Rosas and Bliss (1986). (2)- Percentage on a dry weight basis. 41 Thus, average genotypic values across treatments are presented in Table 5. Scores for nodule mass were determined by comparing all cultivars to UW 21-58 check. Cultivar II900-5-MP45 displayed maximum average value for nodule score. Genotype N81017 and all type III cultivars were statistically similar to UW 21-58. A second set of samples was taken at 70 DAP, stage R5 for adapted cultivars, which was 25 days after the moisture stress treatment was applied. The dry matter yield of different plant parts at 70 DAP is presented in Figure 2. The data were analized for plant grown under two different water regimes. A significant effect of moisture treatment was observed (P<0.05) for the dry weight of stems, leaves and developing pods: a significant effect of the genotypes for the same variables and dry weight of roots was observed (Table 6, Appendix A). Nana of the interactions among evaluated factors was significant . With the exception of genotype UW 21-58, all genotypes showed a significant reduction in dry matter yield under stress (Figure 2). At this stage of development, type II cultivars were reallocating dry matter into developing pods at a much higher rate than type III cultivars (II900-5-M-45 and Bayo Madero) which started flowering about the same time as adapted cultivars. Cultivars Mex.1213-2 and Pinto Nal.1 did not start flowering at this time. Unadapted genotypes showed a striking variation in daylongth sensitivity with II900-5-M-45 being the least sensitive (Table 6). Daylength sensitivity of unadapted genotypes was expressed as an increase in the number of days needed to reach R3 stage (50 %. bloom) and a decreasing rate of dry matter partitioning into developing pads in those genotypes which reached R3 stage of development at the same time as adapted genotypes. 42 — . ///////////////, r ////////////////. m w. 7///////////////,. s R /////////////////// V///////////. V//////fl////A V/////////////////. MW nu nH nu he R S R S R S R S 064 II9OO BM 1213-2 PNI Cultivars East Lansing, MI. 1985. R S R S R S 21-58 23-61 017 R S under two moisture regimes (Rarainfed, S=water stressed) at 70 days after planting. Figure 2. Dry matter of different plant parts of eight bean cultivars 43 Higher yields of total dry matter of type III cultivars seemed associated ‘with large plant growth, as measured at 45 DAP. This was associated with a slower reproductive development (Figure 2) where the allocation of assimilates to the developing pods procceded at a reduced rate as compared to type II cultivars. Leaf area index of all genotypes, except UW 21-58, was significantly reduced under stress. No response of root/shoot ratio was observed for water and nitrogen treatments: however, the effect due to genotype was highly significant (P<0.01). Observed values for root/shoot ratio of different genotypes at this stage of development (70 DAP), were smaller than those observed at 45 DAP (Table 6). Daily measurements taken of central leaflet length per cultivar are presented in Table 7. Central leaflet length measurements on all cultivars were initiated at 50 DAP, five days after black plastic stripes were placed between the rows to establish the moisture stress treatment. Under stress, genotypes N81017 and II900-5-M-45 showed the highest accumulated values for leaflet elongation. A significant difference (P<0.05) between stress and nonstress treatments was found for this trait during most of the days that measurements were recorded. The above results are in agreement with findings reported for other legume crops, particularly soybeans and field beans (Karamanos, 1978: Myers et al., 1986). A relationship between LWP, leaf area expansion, and relative growth rate was demonstrated by Karamanos (1978). The relationship of leaf enlargement to turgor prompted Boyer and MCPherson (1975) and Hsiao and Acevedo (1974) to propose that leaf elongation rate could be used as a negative index of drought sensitivity in plants. 44 Table 6. Number of days to flowering, LAI and root/shoot ratio of eight bean cultivars growing under two moisture treatments at 70 days after planting. East Lansing, MI. 1985 LAI Root/ Days to ---------------------- shoot Genotype Flowering Rainfed Stressed ratio UW 21-58 52 3.24 3.02 0.107 UW 23-61 46 3.21 1.68* 0.071 N81017 48 2.81 1.75* 0.098 N81064 46 2.43 2.04* 0.103 11900-5-M-45 51 4.21 3.36* 0.112 Bayo Madero 46 2.82 2.338 0.125 Mex 1213-2 90 4.48 3.87* 0.134 Pinto Nal.1 90 3.99 3.44* 0.133 LSD 0.05 1 0.851 0.851 0.093 * LSD 0.05 between moisture treatments 8 0.275 45 Table 7. Central leaflet length in.mm of eight bean genotypes under two moisture regimes. East Lansing, MI. 1985 Days after planting Genotype 51 52 53 54 55 56 57 58 59 Rainfall UW 21-58 7.4 14.4 22.5 33.2 38.4 41.9 46.6 51.7 54.7 UW 23-61 6.0 14.2 22.8 29.9 32.1 32.7 34.2 36.3 38.5 N81017 7.1 16.5 26.7 36.1 39.7 41.2 44.1 47.5 49.1 N81064 6.0 13.7 20.7 30.9 34.2 34.6 38.7 42.0 44.3 II900-5-M-45 5.7 11.9 23.2 32.2 37.6 38.2 44.1 48.7 53.3 Bayo Madero 4.3 9.9 15.7 22.5 24.5 24.5 24.9 28.4 30.4 Mex.1213-2 5.3 12.2 18.5 25.8 28.2 28.2 30.1 33.5 35.8 Pinto Nal.1 4.2 12.2 20.5 26.7 30.5 30.5 32.7 36.0 38.9 Average 5.7 13.1 21.3 29.7 33.1 34.0 36.9 40.5 43.1 Stress UW 21-58 4.7 9.15 15.8 23.2 25.6 26.0 29.1 33.0 35.3 UW 23-61 4.3 9.50 13.3 18.4 19.5 19.5 20.0 20.8 21.6 N81017 6.4 11.5 16.3 22.9 29.1 31.6 34.8 36.9 38.4 N81064 5.4 10.6 17.0 25.2 26.9 27.8 29.8 33.0 34.1 II900-5-M-45 4.7 9.90 16.3 24.3 28.1 28.9 32.7 36.8 40.4 Bayo Madero 4.4 6.75 10.1 16.2 17.8 17.8 18.8 21.3 24.1 Mex.1213-2 3.2 6.60 11.1 13.9 17.5 18.1 19.2 22.7 24.8 Pinto Nal.1 4.1 7.30 12.1 16.5 18.5 18.5 20.0 23.0 25.3 Average 4.6 8.9 14.0 20.1 22.9 23.5 25.5 28.4 30.5 46 Yield and yield components. Results do not include two of the unadapted cultivars, Mex.1213-2 and Pinto Nal.1, because they did not produce seed. These two genotypes started flowering at 90 DAP and were killed by freezing temperatures in late September. They proved to be highly sensitive to the extended daylength of the Narthern latitudes. The analysis of variance for seed yield showed significant differences for' moisture (P<0.05) and genotype (P<0.01) effects. No significant effects were detected for nitrogen source or any of the interactions (Table 6, Appendix A).' Therefore, in tables summarizing those results, data related to nitrogen rates is not presented. An average yield reduction of 27%.was observed when the genotypes were subjected to mild drought stress, equivalent to a drought intensity index of 0.27 (Fischer and Maurer, 1978). In table 8, individual seed yields per genotype under both moisture conditions are presented, as well as the reduction per cultivar in percentage, arithmetic and geometric means and the drought susceptibility index of Fischer and Maurer (1978). Genotypes UW 23-61 and II900-5-M-45 showed the highest yield reductions under stress. When looking at the arithmetic mean values, II900-5-M-45 together with UW 21-58, UW 23-61 and N81017 were statistically similar: however, the former showed the highest value for the geometric mean, which is considered.more appropriate for comparison of genotypes under stress and non-stress conditions (Samper and Adams, 1985). The percent seed yield reduction of a genotype when evaluated in favorable and unfavorable environments can mislead the interpretation of results (Table 8). Since different genotypes possess different yield 47 Table 8. Seed yield under rainfed and stressed conditions, reduction in percentage, arithmetic and geometric means and drought susceptibility index (DSI). East Lansing, MI. 1985 Yield g/mz Reduction Arith. Geom. Genotype Rainfed Stressed 1 mean mean DSI(1) UW 21-58 213* 161 24 187 185 0.81 UW 23-61 229* 117 49 173 164 1.63 N81017 181 163 10 172 172 0.33 N81064 136 125 8 131 130 0.27 II900-5-M-45 275* 150 46 213 203 1.51 Bayo Madero 166* 121 27 143 142 0.90 Average 200 139* 27 170 166 0.91 * LSD 0.05 between water regimes. (1) Drought suscep. index of Fischer and Maurer, 1978: Appendix A. 48 potentials. Those genotypes which do not show a high reduction of yield under stress, like N81064, are not necessarily drought resistant. Since such genotypes possess a limited yield potential even under non-stress conditions as demonstrated by the arithmetic and geometric mean yield values, caution must be exercised when utilizing low percent seed yield reduction as a criterion for drought tolerance. Within the adapted cultivars, N81017 showed a better than average seed yield under stress and an intermediate value for both the arithmetic and geometric means, and a low value for the drought susceptibility index. Therefore, those results would indicate that N81017 is the more drought tolerant of the evaluated adapted cultivars, which is in agreement with the finding of Samper (1984) and Sanper et al. (1984). Briefly, in using any criteria to select or decide which genotype(s) shows the maximum degree of stress tolerance, genotypes with values above the mean under both stress and nonstress conditions are the ones to select. Meisture stress significantly reduced all the yield components (P<0.05) and the effect of genotype also had a significant effect on the yield components (Table 6, Appendix A). Interestingly not all the genotypes responded in the sane way to the stress. For instance, cultivars N81017 and N81064 showed the same number of pods perm2 and the same seed weight, but they differed significantly in the number of seeds per pod (Table 9). As a consequence of their difference in this yield component, they were significantly different for final seed yield iiper m2. In the case of II900-5-M-45 and UW 23-61, which showed the highest percentage yield reduction under stress: the number of pods per 49 Table 9. Yield components of six bean genotypes under two moisture treatments. East Lansing, MI. 1985 Pods/1m2 Seeds/pod 100 seed weight. Genotype Rainfed Stressed Rainfed Stressed Rainfed Stressed UW 21-58 224 192 4.3 3.7 22.0 22.4 UW 23-61 290 200* 4.0 3.4 19.5 17.4* N81017 206 190 4.1 4.0 21.2 21.2 N81064 220 192 3.1 3.2 20.1 20.4 II900-5-M-45 235 156* 4.7 3.8 25.1 25.3 Bayo Madero 162 134 2.5 2.4 40.2 37.0* Average 223 177* 3.8 3.4 24.7 23.9* LSD 0.05 63 63 --- --- 1.5 1.5 * LSD (0.05) between moisture treatments. 50 m2 was the yield.component.most affected by the stress. UH 23-61 also showed a significant reduction in seed weight. Finally, Bayo Madero had a significant reduction in seed yield due to a decrease in seed weight (Table 9). The observed variability of the yield components between genotypes in response to the imposed mild water stress in this experiment could be due to the differences in. timing and duration of different phenological phases of this small sample of genotypes in conjunction with the timing and intensity of the moisture stress. In general, the number of pods per mg was the component most affected by the stress. These findings are in agreement with previous reports (Robins and Domingo, 1956; Stoker, 1973; Stansell and Smittle, 1980; Bonnano and Mack, 1983). A modified harvest index (HI) was calculated by dividing the economic yield ‘by the T0! at physiological maturity which did not include the weight of the leaves. Cultivars N81064 and Bayo Hadero showed the smallest values for this variable, and the latter was the only genotype showing a significant decrease for this trait under stress. CONCLUSIONS: 1. Evaluated genotypes showed variation for BNF related traits and some of them proved to be as good fixers as the check UW.21-58. i.e. Lef-z- RB, II900-5-M-45, and N81017. 2. Significant differences among genotypes grown under stress and non- stress treatments were observed for seed yield and yield components as H 51 well as for the different morphological and growth related variables recorded. 3. Genotypes from two different genetic backgrounds. namely adapted and unadapted to Michigan conditions, exhibited different morphological characteristics and they probably possess different mechanisms to partially avoid drought effects. 4. Mexican type III cultivars were photoperiod sensitive in varying degrees, as indicated by the number of days to initiate flowering and an extended reproductive phase. A relationship between larger total dry matter accumulation of unadapted type III genotypes seems to be associated with a slow development or partitioning into reproductive structures. 5. The results in this study indicated that genotypes N81017 and II900- 5-H-45, are drought tolerant and possess a superior ability for BNF. Ka REFERENCES Albretch, S.L., Bennet, J.N. and R.J. Boote. 1984. Relationship of nitrogenase activity to plant water stress in field-grown soybeans. Field Crops Res. 8:61-71 Beadle, C.L. and S.P. Long. 1985. Photosynthesis - is it limiting to biomass production 7 Biomass 8:119-168 Bliss, F.A. 1985. Breeding for enhanced dinitrogen fixation potential of coulnon bean(£has_e_olnsmlgari§L..) In: P..H Ludden andJ.§h Burris (Eds. ). Nitrogen Fixation and CO2 metabolism. Proc. Steenbock Symposium 17- 22 June 1984, University of Wisconsin. Madison. Elsevier Science Publishing Co. , pp. 303-310 Bonanno, A.R., and H.J. Hack. 1983. Yield components and pod quality of snap beans growth under differential irrigation. J. Amer. Soc. Hort. Sci. 105:869-873 Boyer J .S. and R.C. Hc.Pherson. 1975. Physiology of water deficits in cereal crops. Adv. Agron. 27:1-23 Elizondo-Barron. J. 1987. Characteristics related to yield of dry bean (W mm L.) under water stress conditions. M.S. Thesis Michigan State University, East Lansing p. 73 Felix, J.F., Obaton, M., Messiaen, C.H. and L. Salsac. 1981. Nitrate reductase and nitrogenase activities of comon beans (2mm mm: L.) from different geographic locations. Plant and Soil 63:427-438 Fischer, R.A. . and R. Maurer. 1978. Drought resistance in spring wheat cultivars. 1. Grain yield responses. Aust. J. Agric. Res. 29:277- 317 Graham. P. H. , and J. C. Rosas. 1977. Growth and development of indeterminate bush and climbing cultivars of W W L. inoculated with W. J. Agric. Sci. Camb. 88: 503-508 Hsiao, T.C. and E. Acevedo. 1974. Plant responses to water deficits, water use efficiency, and drought resistance. Agr. Meteorol. 14:59-84 Karamanos, A.J. 1978. Water stress and leaf growth of field beans (119,13 faba L.) in the field: leaf mmber and total leaf area. Ann. Bot. 42:1393-1402 52 RC 3&1 53 MeFerson. J. 1983. Genetic and breeding studies of dinitrogen fixation in coumon bean. (Phenom mm: L.). Ph. D. Thesis, University of Uisconsin, Madison p. 147 Myers, 0. Jr. Yopp. J.R.. and.M;R,S. Krishnamani. 1986. Breeding soybeans for drought resistance. In: J. Janick (Ed.), Plant Breeding Reviews Volume 4. Avi Publishing Co. Inc. pp. 203-243 Nutman, P.S. 1981. Hereditary host factors affecting nodulation and nitrogen fixation. In: A.H. Gibson and N.E. Neaton (Eds.), Current Perspectives in Nitrogen Fixation, Elsevier / Nbrth Holland Biomedical Press, Amsterdam. pp. 194-204 Pacowsky,R.S., Bayne, H.G., and G.J. Bethlenfalvay. 1984. Symbiotic interactions between strains of W phamli and cultivars of Ehaseglns ynlgaris L. Crop Sci. 24:101-105 Piha, M.I., and D.N. Munns. 1987. Nitrogen fixation potential of beans (Phaseglns ynlgaris L.) compared with other grain legumes under controlled conditions. Plant and Soil 98:169-182 Rennie. R.J., and C.A. Kemp. 1981a. Selection for dinitrogen-fixing ability in 2haseglns_yn1garis L. at two low-temperature regimes. Euphytica 30:87-95 Rennie, R.J., and C.A. Kemp. 1981b. Dinitrogen fixation in pea beans (Phaseglns ynlgaris) as affected by growth stage and temperature regime. Can. J. Bot. 59:1181-1188 Robins, J .R., and C.E. Domingo. 1956. Moisture deficits in relation to the growth and development of dry beans. Agron. J. 48:67-70 Rosas, J. 1983. Partitioning of dry matter, nitrogen fixation, and seed yield of common bean (Phaseolns ynlgaris L.) influenced by plant genotype and nitrogen fertilization. Ph. D. Thesis. University of Wisconsin. Madison p. 127 Rosas, J.C. and F.A. Bliss. 1986. Host plant traits associated with estimates of nodulation and nitrogen fixation in. common bean. Hortscience 21:287-289 Samper. C. 1984. Effects of water stress imposed at mid-pod filling upon yield and dry matter partitioning in dry beans (Phaseglns ynlgaris L.). M.S. Thesis, Michigan State University, East Lansing p. 131 Samper, 0., Adams, M;H., Hanson. A.D., and E. watt. 1984. Contribution of stored assimilates to the seed following a drought stress in dry beans. Agronomy Abstracts 76:87 Samper, C. and M.H. Adams. 1985. Geometric mean of stress and control yield as a selection criterion for drought tolerance. Ann. Rept. Bean Improv. Coop. 28:53-54 54 Schubert, K.R., and H.J. Evans. 1976. Hydrogen evolution: A.maior factor affecting the efficiency of nitrogen fixation in nodulated symbionts. Proc. Natl. Acad. Sci. (USA) 73:1207-1211 Sprent, J.I. 1981. Nitrogen fixation. In: L.G. Paleg and D. Aspinall (Eds.), Physiology and Biochemistry of Drought Resistance in Plants. Acadomic Press Australia pp.131-143 Stansell, J.R., and D.A. Smittle. 1980. Effects of irrigationregimes on yield and water use of snap bean (2haseolns,ynlga:is L.). J. Amer. Soc. Hort. Sci. 105:869-873 St. Clair.19.A. 1986. Segregation, selection, and population improvement for N-determined dinitrogen fixation ability in comon beans (Phaseglns ynlgarisL.). Ph. D. Thesis, University of Wisconsin, Madison p. 90 Stoker. R. 1974. Effect on dwarf beans of water stress at different phases of growth. N. z. J. Exp. Agric. 2:13-15 Westerman, D.T., and J.J. Kolar. 1978. Symbiotic N2 (CZHZ) fixation by bean. Crop Sci. 18:986-990 Zablotowick, RoMg. Focht, D.D. and G.H. Cannell. 1981. NOdulation and nitrogen fixation of field grown cowpeas as influenced ”by well irrigated and droughted conditions. Agron. J. 73:9-12 CHAPTERZ SELECTION IN SEGREGATING BEAN POPULATIONS FOR DROUGHT TOLERANCE IN THE FIELD. INTRODUCTION There has been comparatively little breeding for drought tolerance in pulse crops. This is probably due to the complexity of the problem and the difficulty of measuring traits identifying drought tolerant genotypes. With the exception of soybeans, most pulse crops are not important in developed nations where much of the work on drought tolerance is conducted. During the last decade, research has been conducted with the aim of identifying the physiological and/or morphological plant attributes related to drought tolerance. attributes which could be readily used as selection criteria for breeding for drought tolerance in crops. So far, the more reliable approach to select for drought tolerance is the assessment of total biomass or economic yield produced under stress in the field (Hurd, 1976: Parsons. 1979: Quisenberry, 1982: Singh and White, 1988). Simmonds (1979) indicated that drought tolerance of well adapted varieties seems certain but he pointed out that it is difficult to give clear-cut examples. largely because such adaptation usually results from selection for average performance over seasons rather than for specific selection for drought tolerance per so. A clear example of the above statement is the reported performance of recently released soybean cultivars in Illinois (Boyer et al., 1980). which display a degree of $5 56 drought tolerance without having been specifically bred for tolerance, but rather they were selected on the basis of their performance in multilocational trials. A variety of approaches have been developed to identify drought resistance in crops (Clarke and Townley-Smith, 1984). Some of the most popular include the use of the pressure bomb and diffusive resistance porometers. An excellent review on.methods to determine plant water status and its applicability to physiological studies was presented by Turner (1981). Hewever, most of these approaches involve tedious procedures and/or sophisticated instrumentation that may have limited applicability for field comparisons involving numerous genotypes. In addition to leaf water potential and.osmotic potential, leaf water status can be evaluated by indices calculated from. the difference between turgid and fresh weight of leaves (Turner, 1981; Clarke and Townley-Smith, 1984). Water loss from excised leaves is another method which has been used to assess genotypic differences in water retention capacity of plant crops. In wheat, Dedio (1975) suggested that water retention capacity is simply inherited and that the ability to retain water is controlled by dominant gene action. In cowpea, walker and Miller (1986) found that the leaf water retention technique, used either in the greenhouse or in the field, was able to differentiate drought tolerant from.drought sensitive genotypes even in the absence of water stress. werking with common bean, Mkandawire (1987) combined data from.one maize-bean intercropping and one water Use Efficiency (RUE) experiment but did not find a significant relationship between total biological yield and leaf water retention 57 capacity. Hewever, cultivar differences in capacity to retain water were noticed. Yield under drought stress (and under non-stress) is a complex terminal outcome of growth to which there are probably diverse and interrelated paths. Two genotypes of similar yield may achieve their final yield by different routes. Plant breeding affects yield through the adjustment of biomass, or partitioning, .or both: however, it is surprising how poor is our information on the relative importance of these two major components of yield (Simonds, 1979). White (1987) has demonstrated that under tropical conditions at CIAT, the bean crop always seems to show strong limitations on assimilate availability while sink demand is usually adequate. In the current research. it is hypothesized that high yielding bean cultivars under water stress are capable of remobilizing reserve assimilates from storage organs. It was previously demonstrated, with the use of an iodine-potassium iodide starch indicator (IKI). that genotypic differences for levels of stored starch in different plant parts exists in common bean (Adams et al.. 1978). Later, Sebasigari (1981) demonstrated that some legume crops with high seed growth rates were correlated with a decrease in the level of non-structural carbohydrates in storage organs. More recently, Samper et al., (1984) have shown evidence of differential remobilization in two bean genotypes under water stress. The higher yielding genotype under stress displayed greater ability to remobilize stored assimilates. Therefore, a simple, non-destructive technique to evaluate bean genotypes under stress could make use of an IKI solution applied to reserve organs like the root and/or the stem. at physiological maturity. 58 This technique is considered here under the assumption that genotypes showing the smallest or no starch remaining in those organs are good remobilizers under water stress. A possible weak point of this technique is that genotypes which produce more assimilates than those demanded by the sink, and still produce acceptable yields, could be misclassified as inferior remobilizers. The study reported here was intended to evaluate during 1985 a group of 248 F4 families for drought tolerance and to select promising genotypes. The number of families was then further reduced in 1986 and a group of 113 F6 families was tested following the same objective. The F4 families were originally obtained from selection conducted in the greenhouse on the basis of a visual nodule score, a trait related to BNF ability (Rosas, 1983). MATERIALS AND METHODS Plant_matezial. In order to obtain the F4 families, a group of 13 genotypes was selected based on previous performance as either drought tolerant or demonstrating improved nitrogen fixation. These were crossed in specific combinations in the summer of 1983 at the International Center for Tropical Agriculture (CIAT), Cali, Colombia (Table 1). Eleven F1 hybrids were space planted in an unreplicated plot nursery in Sinaloa, Mexico during the winter of 1983-84. A population of 200 F2 seeds per cross and the genotype UW 21-58, considered as a check, were grown in the greenhouse facility of Michigan State University, East Lansing, HI., during the fall of 1984. Forty F2 and 10 UW 21-58 seeds were each planted in flat aluminum containers filled with pure silica sand as growth medium. Seeds were inoculated by using a commercial 59 Table 1. List 0 bean genotypes used as parents and crosses made at CIAT in 1983. East Lansing, MI. 1984 Seed Growth No. Genotype Source* type habit** 1 N81017 MSU navy II 2 N81064 MSU’ navy II 3 876001 HSU black II 4 Dgo.222 INIFAP white III 5 Bayo Madero INIFAP bayo III 6 II900-5-H-45 INIFAP striped III 7 Mex.1213-2 INIFAP pinto III 8 Lef-Z-RB INIFAP striped III*** 9 UW 23-61 UW navy II 10 UW 21-54 UW black II 11 UW 21-58 UW black II 12 A 411 CIAT brown II 13 BAT336 CIAT cream, II Cross Parents code -------------------------- 8 N81017 x Lef-Z-RB 9 N81017 x II900-5-M-45 25 N81064 x Dgo-222 31 UW 21-54 x Dgo-222 34 876001 x UW 21-54 38 UW 21-58 x Mex.1213-2 39 UW 21-58 x II900-5-M-45 41 UW 21-58 x A 411 44 BAT 336 x UW 21-58 48 UW 23-61 x Bayo Madero 51 UW 23-61 x Mex.1213-2 * CIAT = International Center for Tropical Agriculture, Columbia MSU = Michigan State University INIFAP = National Institute for Forest and Agriculture Research, Mexico UW = University of Wisconsin ** Type II = indetenminate-bush, erect stem and branches Type III = indeterminate-bush, prostrate main stem and branches *** Type II in Mexico 60 inoculant and irrigated with a nitrogen free nutrient solution and/or tap water as needed. A visual mild selection was practiced at 20 DAP by taking one seedling at a time and comparing it to the check genotype for nodule mass (nodule amber and nodule size). Those plants that had at least a similar amount of nodule tissue as the check were saved and transplanted, each in a 20 x 30 cm clay pot. For comparison, a small nmnber of plants which showed poor nodulation were also saved. Approximately 50 plants/cross were kept and their seeds were planted as an F3 fanily during the winter of 1984-85. In this generation, selection was practiced between families and individual plants within selected fanilies were scored following the procedure previously outlined. Again, approximately 50 plants per cross were saved and transplanted as before. At the-R3 stage, after the plants were considered to be well established, water was withheld for a 5 day period . Plants showing severe wilting symptoms were dis carded. The retained plants were allowed to recover and set seed. Different numbers of plants per cross were kept for further evaluation and from this stage forward, they were handled as families. In total, 300 families were retained. Here, it is convenient to clarify that some of the crosses were poorly represented, not because they were poor in nodulation but because they were highly photoperiod sensitive . W. 246 F4 families and 10 parents were planted on a Capac fine loamy soil (mixed mesic, Glassoboric Hapludolfs) (USAD, Soil Conservation Service) in June 1985. Planting was done under the rain- out shelter of the bean program at the Crop Science Research facility of Michigan State University (MSU), East Lansing, MI. Planting followed 61 rain when the soil moisture was considered adequate to allow crop establishment. No fertilizer was used and a commercial inoculant was applied to the seeds. From 25 DAP onward, the rain-out shelter was used to avoid rains during the remainder of the growing season. Entries were distributed in the field by using a Simple Lattice Design (16 x 16) in small row plots of 50 x 50 cm. Soil moisture was regularly monitored at two different soil depths, 0-15 and 16-30 cm. Soil samples were collected, weighed, and immediately placed in an oven at 110 0C for two days. Soil moisture content was expressed as a percent on a dry weight basis (Table 1, Appendix 8). Climate data during the growing season are presented in Table 4 of Appendix A. Recorded data and procedures are outlined in table 2. Experiment_12§§. The F5 selections from the preceding experiment were grown at Iguala, Mexico, during the winter of 1985-86. No selections were made in this generation. One hundred twenty one entries, consisting of 113 F6 families and eight parents, were planted at the Crop Research facility of MSU in small row plots (62 x 50 cm) in 1986. A simple Lattice Design (11 x 11) was used to randomize entries under the rain-out shelter. In this year, a similar planting was made adjacent to but outside the rain-out shelter which was sprinkler irrigated as needed. Planting of these two plots was done on June 16 and 17, 1986, respectively. With the exception of leaf nitrogen concentration, recorded data were the same as in the 1985 experiment, and additional data on the number of seeds per pod and 100 seed weight was collected. Soil moisture was again monitored and data are presented in Table 2 (Appendix B). Climatological data during the growing season are reported in Table 4 of Appendix B. 62 Table 2. Recorded data and procedure. East Lansing, MI. 1985 FLOW No. of days from planting to flowering, recorded when 50 x of the plants in a plot had at least one flower open. PHYMA No. of days from planting to physiological maturity, recorded when pods lose their pigmentation and begin to dry. TDM Total dry matter/m2 , calculated after harvested plants were oven dried. PODS No. of pods/m2, calculated by counting the total No.of pods with at least one normal seed per pod. SEEDS No. of seeds per pod, calculated from 50 pods per plot at physiological maturity . SM 100 seed weight in grams. YIELD Seed yield g/mz. HI Harvest index, calculated by dividing seed yield by total dry matter. LENIT Leaf nitrogen concentration, determined by micro-Kieldhal analysis on two leaves per plot at R4 stage. LWC Leaf water content, determined at R4 stage by excising two full expanded leaves and recording the difference between their fresh and oven dry weights. LWRC Leaf water retention capacity , determined at R4 stage by excising two leaves per plot and iulnediately weighing and placing them at normal room temperature , and weighing them at 24 hours after excision and after drying for 72 hours. IKI IKI visual score, determined at physiological maturity by cutting the base of stem and using two or three drops of an IKI solution prepared as described by Adams et al. (1978). Scale from 1-5, 1. no starch, 5= abundant starch. 63 Analyses of variance for all recorded data were performed with the aid of MSTAT, a microcomputer statistical package for agricultural sciences. RESULTS The rainout shelter used to conduct this research was built on a Capac soil. This class of soil characteristically has the soil water table close to the surface during part of the year. In this case, the soil water table level rose at the beginning of the fall. This rise of the water table was associated with the frequency and amount of rainfall in the surrounding area at that time of the year. During both years, 1985 and 1986, the bean genotypes tested were considered to be under at least a mild drought stress at the R3 and R4 stages of development (Tables 1 and 2, Appendix B). Experiment_128§. The genotypes were significantly different for all the recorded variables, but relatively high coefficients of variation were obtained for most of the variables (Table 3, Appendix 8). With the exception of LWC, average values for all variables showed a wide range in variation (Table 7, Appendix B). The range for seed yield extended from genotypes showing values as low as 20 g/m? to genotypes with yields higher than 250 g/m2. It is important to indicate that a few of the low-yielding genotypes performed poorly because of photoperiod sensitivity not because of drought treatment. These genotypes started flowering late in the season and were not killed by freezing temperatures because they were protected by the rainout shelter. At this late time in development, photoperiod sensitive cultivars had been already released from the drought stress by the rising of the soil water 64 table in the adjacent soil profiles. A frequency distribution for seed yield and biomass under stress is presented in Figure 1. This figure shows again the wide range in response to the effects of drought stress and other nondefined factors of the environment. Similar response was observed for seed yield and biomass. In Figure 1, the relative position of the parents N81064 and UW 21-54 is indicated (by position of *): among the parents, these two were the lowest and highest yielder, respectively. It was also noticed that some of the segregating families produced higher yields than the top yielding parent. Briefly, 22 F4 families were superior to the general mean value by more than one standard deviation, and 16 of them displayed values two standard deviations above the mean. Thus, clearly, experimental genotypes showed variation for seed yield under drought stress and some of them showed an ability to produce relatively well under the drought stress imposed in this experiment. The matrix of simple correlation values between pairs of variables is presented in Table 3. Seed yield, showed a significant association with TDM, and also with number of pods/m2 and HI: this was expected since all of these traits partially contribute to the same end product. TDM was significantly and positively correlated to the IKI score and number of days to flowering and physiological maturity, whereas HI showed a significant and negative association with the same traits. With the exception of a relatively important negative association between LWC and HI, in general, LWC, LWRC and leaf nitrogen concentration (LNC) did not show a relationship among themselves nor with the rest of the variables. 65 Frequency 70.. CDNI till H "0‘ H U1 & 60-J * ' ° 5... T— T" 40.. 30- T 10‘ .— mm 36 66 96 126 156 186 216 246 276 306 Seed yield g/m2 Frequency 70 _ §=380 6O - F ! 3.1). = 104 so- * 4O ‘ 30.. * a T— 20‘— 1° LELL-I 1.1L _L_D..=:_=_ 112 177 242 307 372 437 502 567 632 697 Biomass g/m2 Figure 1. Frequency distribution for seed and biomass yield of 256 bean genotypes grown under water stress ( * indicates position of lowest and highest yielding parental genotypes). East Lansing, MI. 1985. - 66 Table 3. Matrix of correlations among seed yield and some plant characteristics of 256 bean genotypes grown under drought stress in a rainout shelter. East Lansing, MI. 1985 YIELD PODS TDH HI LWC LWRC LNC IKI FLOW PODS .77** THU .82** .64** H1 .46** .34** -.09 LWC -.12 -.17** -.02 -.18** EURO .02 .02 .02 .09 .15* LNC -.03 -.04 .02 -.07 .03 -.01 IKI .00 -.09 .20** -.26** .11 .17 .01 FLOW -.ll -.24** .19** -.50** -.07 -.17* .05 .08 PHYHA .02 -.08 .36** .54** .09* -.15* .04 .31** .65** *,** Significant at P<0.05 and 0.01, respectively. 67 The average values for seed yield and other variables are presented for some of the promising F4 families and parents in Table 4. It can be seen that for each trait there are at least a few families which show higher values than the top parent. The grand mean for the experiment was slightly larger than the mean of the parents. With the exception of N81064, all genotypes showed a number of days to physiological maturity longer than desirable in agronomic terms. It seems that the release of the drought stress period at seed filling stage extended the length of growth cycle in.most of the genotypes. Under normal field conditions terminal drought stress usually tends to accelerate maturity in beans. Experiment_128§. The 1986 season in Michigan was characterized by fall rainfall considerably above the average (Table 4, Appendix B). An extremely heavy rainfall at the beginning of fall terminated the drought stress sooner than had been expected. In fact, the rainout shelter was always kept in place covering the experimental plots from the rains, but the rising of the soil water table in adjacent soil profiles was unavoidable. In addition, a September storm of more than five inches of rain in a short time period caused run-off from adjacent fields into the rainout shelter. This high soil moisture together with warm temperatures ended the stress period and caused a lengthening of the growth cycle. In the irrigation part of the experiment, the wet conditions resulted in the germination of seeds in the pods of standing plants of early maturing genotypes and a lengthening of the growth cycle of mid and full season genotypes. Due to the above mentioned factors, high CV's were associated with most of the recorded variables. CV's from the irrigation part of the experiment were essentially of the same magnitude as those from the 68 Table 4. Mean seed yield and some agronomic characteristics of promising F4 families and two parental genotypes grown under drought stress in a rainout shelter. East Lansing, MI. 1985 Days to Yield Pods TDM2 ------------ Pedigree g/m /m g/m HI IKI FLOW PHYMA 38-18-2 312 208 732 .42 1.8 60 110 48-32 302 272 708 .43 1.8 66 125 48-22 300 224 666 .44 2.3 61 125 48-94 278 268 604 .45 3.3 53 110 51-32-1 276 216 602 .46 3.0 44 110 48-10 264 176 542 .48 2.8 47 110 9-10 258 220 552 .45 3.5 52 110 38-16-2 250 172 556 .44 3.0 56 110 Mean * . 156 UW 21-54 ,224 224 454 .49 2.5 51 110 N81064 100 116 232 .42 3.5 42 88 Grand mean 161 167 380 .43 2.8 50 103 Std. dev. 54 45 104 .10 0.8 7 13 * Parents' mean, it does not includes Dgo-222 and Mex.1213-2 69 stressed part of the experiment. In spite of all constraints, highly significant (P<0.01) differences were found among genotypes for all variables except for LWRC under irrigation (Tables 5 and 6, Appendix B). The grand means for seed yield of both drought stressed and irrigated conditions were used to calculate the drought intensity index (DII) of Fischer and Maurer (1978) which gave an average value of 0.28. This value indicates that a mild drought stress period was imposed upon those genotypes grown under the rainout shelter. 7 D:gnght_ggnditign. A wide range was observed for most of the quantified. variables: average values for those variables are presented in Appendix B (Table 8). In Table 5, average values for seed yield and certain agronomic traits of a select group of promising families are presented, in addition to the top and bottom yielding parents. Differences in morpho-physiological characteristics can readily be observed in this table. For example, the top yielding genotype (41-39) shows a large biomass and a relatively acceptable HI, the following three genotypes (41-18-1, 41-48-1 and 8-30) show a better than average biomass and relatively high HI values, and finally, genotype 8-4 shows a large biomass and a smaller than average HI. Thus, it can be noticed that top yielders display different physio-morphological characteristics. As previously pointed out, it was shown that some genotypes started flowering early enough to reach physiological maturity at around 90 days after planting (DAP), however, their reproductive period was lengthened by the prevalent environmental conditions at the seed filling stage. A further observation is that most of the promising genotypes presented in this table belong to the crosses coded 8 and 41, crosses which include a Michigan and a Wisconsin adapted parent crossed to a Mexican unadapted 70 Table 5. Mean seed yield and some agronomic variables of promising F6 families and two parental genotypes grown under drought stress in a rainout shelter. East Lansing, MI. 1986 Yield Pods Seeds Seed TDMZ ............. Pedigree g/m /m /pod wt. g/m HI IKI FLOW PHYMA 41-39 398 319 6.3 24.3 893 .45 3.8 51 105 41-18-1 304 282 6.4 23.1 552 .55 2.5 43 108 41-48-1 288 281 5.7 21.8 570 .51 2.3 42 100 8-30 266 263 6.1 23.7 534 .50 3.0 51 102 8-4 260 281 7.0 18.5 761 .34 2.3 61 120 8-25-2 259 240 6.3 20.7 587 .44 2.8 41 98 8-21-2 257 203 6.7 22.9 560 .46 2.8 63 120 48-109 250 184 6.4 24.6 478 .52 3.5 35 100 mean * 144 Lef-Z‘RB 278 210 5.8 21.4 645 .42 3.5 42 98 II-900- 54 90 2.8 15.2 233 .24 2.8 63 125 5-M-45 (l) Grand mean 149 166 5.9 20.4 397 .37 3.0 51 110 Std. dev. 66 56 0.7 3.5 129 .10 0.6 11 12 1" Parents' mean, it does not include Dgo-222 and Mex.1213-2 ( 1) Photoperiod sensitive . 71 parent. Figure 2 is a frequency distribution for seed yield and biomass. This figure shows a wide range in variation with genotypes clustering around the mean value and exhibiting low and high yielders at the left and right ends of the distribution. The relative position of the top and bottom yielding parents is also shown (by position of *). In summary, in this distribution 18 genotypes achieved seed yield values one standard deviation above the mean and three of them showed values larger than two standard deviations above the mean. As expected for quantitative traits, environmental year to year variation did interact with genotypes influencing the response in seed yield and related agronomic traits. This year- to-year environmental- genotype interaction effect was responsible for an observed low correlation when comparing the yields of the genotypes in 1986 with the yields obtained in 1985 (r = 0.172). Simple correlation between pairs of variables are shown in Table 6. Seed yield was highly and positively associated with the number of pods/m2, TDM and HI. Seed yield and HI showed a relatively important negative correlation with LWC, IKI score, number of days to flowering and days to physiological maturity. The results obtained with simple correlation in this experiment are essentially the same as those obtained from the previous year's experiment. 1::1ga;gd__ggndi;ign. Beans under irrigation significantly outyielded those under mild drought stress. There were significant differences (P<0.01) among the genotypes for every recorded variable except LWRC (Table 6, Appendix B). As in previous experiments, wide variation was observed for most of the variables (Table 9, Appendix B). In the 72 Frequency 30..., 20 -‘ lO _. 1.: .. 149 _ 5.1). - 66 Frequency 30 '1 20 - 10 _. flflnfl 20 60 100 140 180 220 260 300 Seed yield g/mz a? . 397 q _ S.D. " 129 * Hm H_ 104 194 284 374 464 SS4 644 734 824 914 Biomass g/m2 Figure 2. Frequency distribution for seed and biomass yield of 121 bean genotypes grown under water stress (* indicates position of lowest and highest yielding parental genotypes). East Lansing, MI. 19862 73 Table 6. Matrix of correlations among average seed yield and some plant characteristics of 121 bean genotypes grown under drought stress conditions in a rainout shelter. East Lansing, MI. 1986 YIELD PODS SEEDS SEWT TDM HI LWC LWRC IKI FLOW PODS .89** SEEDS .27** .25** SEWT .30** .15 -.16 TDM .79** .77** .24** .06 HI .54** .41** .04 .43** .01 LWC -.30** -.32** .05 -.11 -.07 -.42** LWRC .21* .27** .20* -.11 .23** -.02 -.05 IKI -.21* -.20* -.25** .13 -.05 -.21* .04 -.09 FLOW -.36** -.31** .09 -.56** .00 -.56** .30** .01 .05 PHYMA -.30** -.l8* .24** -.38** .01 -.45**. .30** .02 .01 .66** *,** Significant at P<0.05 and 0.01, respectively. 74 frequency distribution for seed yield of Figure 3, a drift toward the right side of the distribution is observed. Under this condition, 20 of the genotypes showed seed yields higher than one standard deviation above the mean and eight of them were superior to the mean value by more than two standard deviations. In this case, two of the parental genotypes (UW 21-54 and N81017) were among the top yielders. This result was expected since we were selecting for drought tolerance under mild stress, therefore we had not expected many of the selected families to be superior to the adapted parents under irrigation. However, it is worth noting that the parent N81017, which had already been classified as drought tolerant, also seems to be a good yielding genotype under favorable conditions (Table 7). The same observation is valid for the family 8-4 which is included among the promising families under both stress and non-stress conditions. This family in particular, and the cross code 8 in general, seems to have produced superior recombinants. Results presented in Table 8 indicate that high yields are associated with a high biomass accumulation and/or a high partitioning ratio (HI). In this particular highly humid environment it was obvious that the type II families derived from.adapted parents possessing erect architectural traits were clearly superior to those families possessing a prostrate type III growth habit. 75 Frequency i=208 30- * S.D =1lO 20.. 10'- I I .—. * l:::.. 40 90 140 190 240 290 340 390 440 490 Seed yield g/m2 Frequency - * x = 560 S.D = 209 _ 30-# 20W — F 10'- L J * I I” 258 373 488 603 718 833 948 1062 Biomass g/m2 Figure 3. Frequency distribution for seed and biomass yield of 121 bean genotypes grown under irrigation (* indicates position of lowest and highest yielding parental genotypes). East Lansing, MI. 1986. 76 Table 7. Mean seed yield and some agronomic variables of promising F6 families and two parental genotypes grown under irrigation. East Lansing, MI. 1986 Days to . Yield Pods Seeds Seed TDM2 ------------- Pedigree g/m /m /pod wt. g/m HI IKI FLOW PHYMA 41-49 502 508 7.0 19.1 1315 .38 4.0 50 115 39-11-1 501 416 7.5 22.8 1114 .45 3.8 56 118 8-42 498 436 6.8 24.3 975 .51 3.0 43 102 8-47 465 486 7.0 18.5 767 .67 3.3 47 105 8-17 455 497 7.0 19.0 995 .46 3.3 54 125 N81017 433 452 6.5 16.9 873 .49 2.5 47 95 8-4 412 423 6.8 16.3 879 .47 2.0 52 112 8-26 412 387 6.9 21.3 804 .51 2.0 43 103 8-6-1 398 448 6.7 18.5 916 .44 3.8 52 112 Mean * 286 UW 21-54 505 527 7.5 16.9 1036 .49 2.8 56 112 Bayo Madero 150 155 5.0 29.5 441 .34 4.0 52 120 Grand mean 208 256 6.5 17.6 560 .37 3.2 52 115 Std. dev. 110 98 0.5 3.7 209 .14 0.6 9 10 * Parents' means, it does not include Dgo-222, Mex1213-2 and II900-5-M-45. ' 77 Table Matrix of correlations among average seed yield and some plant characteristics of 121 bean genotypes grown under irrigation. East Lansing, MI. 1986 YIELD PODS SEEDS SEWT TDM HI LWC LWRC IKI FLOW PODS .92** SEEDS .46** .49** SEWT .35** .11 -.25** TDM .81** .85** .49** .08 HI .61** .49** .13 .47** .14 LWC .14 -.10 .08 -.14 .02 -.07 LWRC .09 .21* -.09 -.06 .13 .05 -.27** IKI .34** -.28** -.14 -.15 -.12 -.40** .12 -.03 FLOW .17 -.04 .12 -.43** .24** -.45** .23* .09 .22* PHYMA. .15 .01 .33** -.45** .19* -.42** .25** -.02 .21 .61** *,** Significant at P<0.05 and 0.01, respectively._ DISCUSSION This research illustrates in part why breeding for drought tolerance is sometimes elusive. This is particularly true when we have only limited control over the environment and when the evaluation of genotypes is based on the final plant product, biomass or seed yield. After all, the final product is the accumulated output of genetic, environmental and interaction factors all controlling different plant responses in the field during the whole crop cycle. In spite of some constraints in getting reliable drought stress under ‘Michigan conditions, the results obtained with the aid of the rainout shelter indicated that the tested genotypes were under mild drought stress during the most critical portion of the reproductive phase (Tables 1 and 2, Appendix B). From past experience, we know that a crop can be severely damaged in terms of seed yield if the drought stress coincides with a critical phase of development. Thus, for the comparison of many genotypes a single stress period at the beginning of the reproductive period or in seed filling allowed us to detect differences in response to the mild stress encountered by the plants. However, for future evaluations, it seems sound to separate the different genotypes in groups according to their growth habit and phenology. Large differences in those traits increase the degree of difficulty in interpreting results. Although selection in more favorable conditions is largely inefficient in identifying the best genotypes for unfavorable conditions (Lewis and Christiansen, 1981: Boyer, 1982: Cecarrelli, 1987), the evaluations of early generation materials at more sites, or at the same site on different planting dates or under differential moisture supply 78 79 is advised. It can be valuable for two purposes: one, to expose the genotypes to a variety of stresses, in terms of timing and/or intensity, and two, it allows for selection for either arithmetic or geometric mean yield and the use of the drought susceptibility index of Fischer and Maurer (1978). Those data should give an opportunity to identify those drought tolerant genotypes which are likely to occur at a very low frequency, able to perform well under stress, and with sufficient plasticity to respond to improved conditions (Ceccarelli, 1987). If a population has a sufficiently broad genetic background, as we assume is our case, developmental plasticity will be displayed when an environmental stress is applied to that population. Results from.part one of this research .indicated that seed yield under stress is primarily dependent on or associated with pro-existing biomass and secondly with ability to partition (and/or remobilize) photoassimilates to the reproductive organs of the plant. Results obtained here confirmed the high association of biomass with seed yield: in this case, biomass at physiological maturity was important under both stress and nonstress conditions. Similar findings were reported in common. bean by Elizondo (1987) and with lime bean by Ziska et al. (1985). However, other traits must be taken into consideration when selecting for drought tolerance, in addition to biomass, since larger biomass in beans is often associated with a delay in.maturity (Rodriguez et al., 1987; White, 1987). Partitioning as indicated by the HI was also highly associated with seed yield, as indicated by the correlation values between these two traits. However, since H1 is a calculated ratio rather than a directly measured trait, and one cannot discern between partitioning of current 80 or stored assimilates, we believe it would not be a useful trait to select for under stress conditions. In this study some genotypes showed similar mean yields and different values for the HI. Thus genotypes possessing a greater biomass did produce relatively well under the conditions of these experiments without displaying a high HI value. Genotypes which produced high yields without much biomass, possess a high HI, and can be considered efficient in terms of partitioning, that is in terms of yield/unit area/day. However, in those genotypes, there is little room for genetic improvement as compared with genotypes producing a large biomass. High partitioning without high initial biomass will not lead to high yields unless photo-assimilation is extraordinarily high (Adams and Kelly, 1987). In any environment, the capacity for high biological yield sets the stage for manipulating plant photosynthate allocation (and reallocation) in such a way that the economically useful part could be increased only after obtaining a vigorous and healthy plant. Stunted bean plants are a common response under drought stress, therefore genotypes that grow more normally than others under stress are the primary target in a selection program. A primary concern in a breeding program for drought tolerance should be the selection of a field that has a stress factor at a level that will discriminate between less tolerant and more tolerant genotypes. As Lewis and Christiansen (1981) pointed out, there is no value to a test if it is located in an environment that is stress-free or if the stress is so severe that nothing survives. It can be argued that the primary objective in any agricultural situation is the economic yield, however, when crops depend on an unreliable rainy season, stability of yield can be as important as 81 potential yield. In some particular cases biomass is almost as important as the economic product: such is the case of the bean crop in Nerth-Central Mexico, where the crop is grown for seed and hay (Acosta and.Kohashi, 1988). A similar situation was described for cowpeas in the semiarid Sahel of Africa (Hall and Patel, 1985) and for beans in Eastern Africa (Adams et al., 1985), where fresh leaves of both crops are eaten as a vegetable. Thus, as indicated by Quisenberry (1982), the ability of the plant to accumulate biomass under stress should be the first criterion in selecting for drought tolerance. A second criterion should be yield per 5;, which ultimately embraces the ability to accumulate biomass and its subsequent allocation into the seeds. The ranking of the families selected on the basis of seed yield and biomass was inconsistent from year to year. However, progress can be expected if segregating populations are handled as families and evaluated in.more than one site. In those cases where seed is short, multiple sites would be prefered rather than replicates in a single site. Mean seed and biomass yields of the 1986 (F6 families) drought trial were compared with those obtained in 1985 (F4 families). An inconsistent response of the genotypes from one year to the other was observed due to the genotype x environment interaction. However, some good responders were also observed and one should expect that some progress could be made if those genotypes could be consistently identified in a selection program. It seems that in breeding for drought tolerance in beans a more consistent response could be obtained if selections are made based on biomass rather than on seed yield per :9. This is partially supported by the relatively greater correlation 82 of the genotypes when biomass was used to illustrate the response of the genotypes in both years (r = 0.326 *). The use of an nKI solution as an aid to select for ability in remobilization under stress is unclear in this research due to the poor relationship observed between IKI scores and seed yields. However, we have noted a consistent negative relationship between the IKI score and the HI. This may indicate the inability of some genotypes to remobilize stored starch or the presence of a larger source as compared to the sink. After the plants have been released from the stress, a small sink during pod filling could be due to a high pod drop during early flowering. It may be this circumstance which sometimes enhances size of the seed under stress as compared to non-stress. Nonetheless, beans in general possess an acceptable sink size and a.small source as indicated by flower and pod dropping and intra-pod competition under stress and non-stress conditions (Acosta and Kohashi, 1988). Drought tolerance from.an agronomic and plant breeding point of view means the ability for higher or at least more stable yield of the economic part of the plant grown under stress. We know that yield is the total sum of many biochemical and physiological processes and environmental factors and has low heritability. Therefore, yield should not be the sole criterion when selecting for drought tolerance: other traits can be considered, i.e. earliness, growth habit, grain filling length, etc. In beans, the chain of biochemical and physiological events between genes and final phenotypic expression under drought stress is not known but, if a key step in the chain is known to be highly correlated with drought tolerance, selection could be practiced for it rather than for 83 the final phenotype. Unfortunately, such key steps still remain to be discovered and much has to be learned with respect to the response of Ehaseglns ynlgazis to drought stress. Based on the observed response of the cultivars to the mild drought stress imposed in both years in. Michigan, and on preliminary observations from plots planted in Mexico during the summer of 1986, a final group of 19 families was selected for further testing in 1987. CONCLUSIONS 1. No relationship was found between the IKI scores and seed yield, however, a significant negative correlation between.IKI scores and HI may indicate the inability of some genotypes to remobilize. In spite of differences observed among genotypes, the use of the IKI solution has to be further tested because the photoperiod sensitivity of many of the genotypes in these experiments seems to affect the pattern of assimilate partitioning and remobilization. 2. Seed yield and biomass seemed to be affected in the same way by water stress. Since biomass showed a higher correlation value than other traits when comparing results from 1985 and 1986, it is suggested that more progress could be achieved by using biomass as a criterion for selection. Selection for increased biomass without modifying the HI may lead to increased seed yield under both stress and non-stress conditions. 3. For reasons inherent to their morphology, genotypes belonging to the type II growth habit proved to be superior to type III's under the moist environment caused by the heavy rains during fall of 1986. REFERENCES Acosta Gallegos, J.A. and J. Kohashi Shibata. 1988. Effect of water stress on growth and yield of indeterminate dry bean (Phaseglns_ ynlgaria L.) cultivars. Submitted : Field Crops Research. Netherlands Adams, M.W., Wiersma, J.W;, and J. Salazar. 1978. Differences in starch accumulation among dry bean cultivars. Crop Sci. 18:155-157 Adams, M;W. and J.D. Kelly. 1987. The role of architecture, crop physiology and recurrent selection in ideotype breeding for yield in dry beans. In: Symposium on Breeding for Higher Yields in Common Beans. XXXIII meeting of the PCCMCA, Guatemala, C.A. CIAT-BIC CRSP-IOTA p. 16 Boyer, J.S., Johnson, R.R., and S.G. Saupe. 1980. Afternoon water deficits and grain yields in old and new soybean cultivars. Agron. J. 72:981-986 Boyer, J.S. 1982. Plant productivity and environment. Science 218:443- 448 Ceccarelli, S. 1987. Yield potential and drought tolerance of segregating populations of barley in contrasting environments. Euphytica 36:265-273 Clarke, J.M. and T.F. Townley-Smith. 1984. Screening and selection techniques for improving drought resistance. In: P.B. Vose and S.G. Blixt (Eds. ), Crop Breeding, a Contemporary Basis. Pergamon Press pp.137-162 Dedio, W. 1975. Water relations in bean leaves as screening tests for drought resistance. Can. J. Plant Sci. 55:369-378 Elizondo-Barron, J. 1987. Characteristics related to yield of dry beans (Phaseolns ynlgaris L.) under water stress conditions. M.S. Thesis Michigan State University, East Lansing p. 73 Fischer, R.A., and R. Maurer. 1978. Drought resistance in spring wheat cultivars. I.Grain yield responses. Aust. J. Agric. Res. 29:277-317 Hall, A.E. and P.N. Patel. 1985. Breeding for resistance to drought and heat. In: S.R. Singh and K.O. Rachie (Eds.), Cowpea: Research, Production and Utilization. John Wiley & Sons pp 137-151 Hurd, E.A. 1976. Plant breeding for drought resistance. In: T.T Kozlowski (Ed.), Soil Water Measurements, Plant Responses and Breeding for Drought Resistance. Academic Press pp.317-353 84 85 Lewis, C.F. and M.N. Christiansen. 1981. Breeding plants for stress environments. In: R.J. Frey (Ed.), Plant Breeding II. Iowa State University, Ames, Iowa pp 151-164 Mkandawire, A.D.C. 1987. Productivity of Malawian landrace dry beans under intercropping and drought conditions. Ph.D. Thesis, Michigan State university, East Lansing p.127 Parsons, L.R. 1979. Breeding for drought resistance: what plant characteristics impart resistance? Hort. Sci. 14:590-593 Quisenberry, J.E. 1982. Breeding for drought resistance and plant water use efficiency. In: J.W. Christiansen and C.F. Lewis, (Eds.), Breeding Plants for Less Favorable Environments. John Wiley and Sons, Inc. pp. 193-212 Rodriguez, R.R, Miranda, C.S. and S.E. Beebe. 1987. Morpho-physiological characterization of early genotypes in cannon beans (mm ynlgaris L.). In: Symposium on Breeding for Higher Yields in Cannon Beans. XXXIII meeting of the PCCMCA, Guatemala, C.A. CIAT- B/C CRSP-IOTA p.30 Rosas, J.C. 1983. Partitioning of dry matter, nitrogen fixation, and seed yield of common bean (Phaseglns,ynlgaris L.) influenced by plant genotype and nitrogen fertilization. Ph.D. Thesis, University of Wisconsin, Madison p.127 Samper, C., Adams, MLW., Hanson, A:D., and E. Watt. 1984. Contribution of stored assimilates to the seed following a drought stress in dry beans. Agron. Abst. 76:87 Sebasigari, K. 1981. Patterns of partitioning and remobilization of non- structural carbohydrates in comon bean and other selected grain legumes. M.S. Thesis, Michigan State University, East Lansing p.152 Simmonds, N.W. 1979. Principles of Crop Improvement. Longman Inc. New York. pp. 40-65 Singh, S.P. and J.W. White. 1988. Breeding common beans for adaptation to drought conditions. In: G. Hoogenboom, F. Ibarra, S.P. Singh, J.W. White and S. Zuluaga (Eds.), Research on Drought Tolerance in Cannon Beans. CIAT, Cali, Colombia. In press Turner, J.C. 1981. Techniques and experimental approaches for the measurement of plant water status. Plant and Soil 58:339-366 Walker, D.W. and J,C, Miller. 1986. Rate of water loss from.detached leaves of drought resistant and susceptible genotypes of cowpea. Hort. Sci. 21:131-132 White, J.W. 1987. Strategies at CIAT for increasing yield potential of common beans -- finishing the domestication process. In: Symposium on Breeding for Higher Yields in Common Beans. XXXIII meeting of the PCCMCA, Guatemala, C.A. CIAT-B/C CRSP-IOTA p. 34 86 Ziska, L.H., Hall, A.E., and R.M. Hoover. 1985. Irrigation management methods of reducing water use of cowpea (ligna 31118111931113 [L.l Walp.) and lime bean (2mm 1mm: L.) while maintaining seed yield at maximum levels. Irrig. Sci. 6:223-239 CHAPTERB EVALUATION OF BEAN (2113mm: mm: L.) GENOTYPES GROWN IN A NITROGEN-FREE MEDIUM FOR BIOLOGICAL NITROGEN FIXATION. INTRODUCTION Variability in field responses to nodulation in beans has been ascribed to unsatisfactory host-microsymbiont interactions (Graham, 1981) , competition between introduced and native rhizobia and, in temperate climates, to the enviromnent (Sprent, 1981) including soil water deficits. In spite of inconsistent results with inoculation trials in the field, evidence is accumulating that the quantitative variation in nodulation and nitrogen fixation in beans is sufficient to attempt to increase nitrogen fixation through breeding .and selection (Rennie and Kemp, 1981a,b: McFerson et al., 1982: McFerson, 1983: Rosas, 1983: Bliss, 1985: Roses and Bliss, 1986: St.Clair, 1986). McFerson et al. (1982) and Bliss (1985) have recently demonstrated transgressive segregation for nitrogen fixation. Among backcross derivatives some families showed higher rates of nitrogen fixation than the recurrent parent. In selecting better genotypes for nitrogen fixation from large segregating populations in the field, indirect selection utilizing easy- to-measure correlated traits may allow the screening of numerous genotypes. Rennie and Kemp (1981a) found that the amount of nitrogen fixed in a phytotron trial with eleven bean cultivars was correlated with leaf area and leaf and shoot weight. They suggested that those 87 characteristics may help bean breeders in selecting bean plants with superior dinitrogen fixing abilities. The use of shoot nitrogen content has also been suggested as a criterion of ability for BNF (St.Clair, 1986). As we previously postulated, high ability to fix nitrogen. might contribute positively to the final response of the bean plant under drought stress at the seed filling stage. Chapman and.Muchow (1985) found marked differences in total nitrogen accumulated among six pulse crop species as well as a significant water regime x species interaction. Nitrogen accumulation and nitrogen accumulation rate decreased in response to prolonged water deficits and also the proportion of nitrogen partitioned to the seeds tended to decrease. In an irrigation experiment, Ziska et a1. (1985), working with cowpea and lime bean, found a low biomass production for treatments having longest irrigation intervals and lowest level of soil nitrogen. They indicated that these results may have been due to drought-induced reduction in nitrogen fixation during flowering and pod filling. Higher levels of soil nitrogen caused small increases in biomass, with the greatest increase occurring with the longest irrigation interval. In the introductory part of this report, we stated the hypothesis that been genotypes can partially avoid drought effects during pod filling by remobilizing previously stored nitrogen and carbohydrate compounds. It is also clear that in beans the vegetative nitrogen fixation period of early maturing beans is shorter than for late maturing genotypes (Rennie and Kemp, 1981b). In addition, the peak of fixation is around flowering time (Graham and Rosas, 1977: Rennie and Kemp, 1981a) followed by a decrease in fixation due to the strong sink 89 for assimilates of the developing pods. Thus, to assess the ability for nitrogen fixation of the bean families selected for drought tolerance in previously described field experiments, this part of the research reports a comparative study on the ability of a group of 26 been genotypes for biological nitrogen fixation (BNF) measured as response in biomass, seed yield and total nitrogen content under greenhouse conditions. MATERIALS AND METHODS The experiment was conducted at the greenhouse facility of Michigan State University (MSU), East Lansing, MI., during the spring of 1987. Nineteen F7 families, six parental genotypes and a check cultivar (Table 1) were used in this experiment. The F7 families had been previously selected on the basis of their adaptation to moisture stress. The genotypes were grown in 20 cm.diam. x 15 cm deep pots containing sterile silica sand:perlite (4:1). Three seeds were planted per pot and fifteen days after planting, seedlings were thinned to one per pot. After emergence, seedlings were inoculated several times with a. commercial inoculant specific for dry beans (Nitragin Co. Milwaukee, Wis.). Plants were fertilized three days per week with a nitrogen-free solution as described by Pacowsky et a1. (1984), and tap water was used the rest of the days. Ten pots were used per genotype and a Complete Randomized Design was used to distribute the pots on the greenhouse benches. MEASUREMENTS AND ANALYSES. When most of the families reached physiological maturity (80 % of the pods had lost their chlorophyll), the plants were cut at the sand level in the pot and separated into seeds, roots and shoots. Fallen 9O Table 1. Origin and pedigree of the evaluated bean genotypes. East Lansing, MI. 1987 Entry no Genotype Pedigree Origin 6 8-3 N81017 X LEF-Z-RB INIFAP-MSU 13 8-4 N81017 X LEF-Z-RB INIFAP-MSU 19 8-17 N81017 X LEF-2-RB INIFAP-MSU 7 8-15-1 N81017 X LEF-Z-RB INIFAP-MSU 3 8-25-2 N81017 X LEF-Z-RB INIFAP-MSU 9 8-42-1 N81017 X LEF-2-RB INIFAP-MSU 10 8-42-2 N81017 X LEF-Z-RB INIFAP-MSU 15 8-47 N81017 X LEF-Z-RB INIFAP-MSU S 9-39-1 N81017 X LEF-Z-RB INIFAP-MSU 8 39-17-1 UW 21-58 X II900- INIFAP-MSU 5-M-45 11 41-18-1 UW 21-58 X A411 INIFAP-MSU 12 41-39 UW 21-58 X A411 INIFAP-MSU 14 41-48-1 UW 21-58 X A411 INIFAP-MSU 1 48-10 UW 23-61 X BAYO INIFAP-MSU MADERO 2 48-66 UW 23-61 X BAYO INIFAP-MSU MADERO 4 48-94B-1 UW 23-61 X BAYO INIFAP-MSU MADERO 16 48-109 UW 23-61 X BAYO INIFAP-MSU MADERO 18 51-5-3 UW 23-61 X MEX1213-2 INIFAP-MSU 17 51-29-3 UW 23-61 X MEX1213-2 INIFAP-MSU * 20 N81017 KENT/NZ/IPIJAO/BUNSI MSU * 21 LEF-2-RB VER10/CHISI43//PUE144 INIFAP * 22 UW 21-58 PORRILLO SIN/PUE152 UW * 23 UW 23-61 EX-RICO/PUE152 UW * 24 BAYO MADERO BAYO RGZ/C-102 INIFAP * 25 II900-5-M-45 PINTO AM/C-14 INIFAP (1)26 PUB-152 LOCAL PUEBLA, MEXICO UW * Entries 20-25 are the parental genotypes (1) cultivar check INIFAP = National Institute for Forestry and Agriculture Research, Mexico. MSU = Michigan State University, USA. UW = University of Wisconsin, USA. 91 leaves were included. Roots were washed and together with shoots were oven dried at 60 0C for 72 hours. Daylength sensitive genotypes were harvested at 90 days after planting regardless of their phenological stage. The dry weight of different plant parts were recorded and micro-kjeldahl analyses were performed for seeds and a representative sample of the total dry matter per plant. Seed protein content (SPC), total nitrogen and nitrogen harvest index (NHI) were calculated for each genotype except Bayo Madero which never set pods. Other recorded and calculated data included : number of days to flowering and physiological maturity, seed yield per plant, yield components, HI, total dry matter and nitrogen accumulated per plant per day (TDM/p/d and N/p/d). Analyses of variance were performed on all variables following a Complete Randomized Design and simple correlations were performed among the variables with the aid of MSTAT, a microcomputer statistical package for agricultural research. RESULTS AND DISCUSSION The tested bean genotypes received no nitrogen fertilizer, therefore, we assume that all plant nitrogen was derived from. the planted seed and the N2 gas from the atmosphere through BNF. Even though the actual amount of nitrogen fixed per genotype was unknown, the accumulated nitrogen in the dry matter (including abscised parts) was used here to compare the relative ability of the different genotypes for BNF. If the average seed N content is substracted from. the total nitrogen content, the result gives a conservative estimate of nitrogen fixation (Pacowsky et al., 1984). Evidence of significant variation (P<0.01) among the genotypes was found for all recorded and calculated variables (Table 2). Some of the 92 Table 2. Analyses of variance for N assimilation and related variables in 26 inoculated bean genotypes grown in a N-free medium in the greenhouse. East Lansing, MI. 1987 Variable Mean square C.V. % Days to flowering 207.5 *** 9.34 Days to maturity 12.1 *** 4.31 Seed yield g/plant 9.76 *** 36.2 No. of pods/plant 9.69 *** 31.4 No. of seeds/pod 0.57 *** 16.0 Seed weight mg 1367.1 *** 19.7 Total dry matter g/plant 1078.3 *** 27.3 Harvest Index 0.034 *** 25.9 N assimilated mg/plant 532028 *** 28.4 Seed protein 7: ' 41.64 t" 17.6 Nitrogen Harvest Index 0.079 *** 22.7 TDM g/plant/day 0.113 *** 26.7 N assimilated.mg/plant/day 55.78 *** 27.8 Root/shoot ratio 0.082 *** 47.2 *** = significant at a P < 0.001 93 genotypes were daylength sensitive and did not set pods or the partitioning of assimilates to their reproductive parts was low: however, a comparison of all genotypes was possible using the traits total dry matter and nitrogen assimilated per plant. The seed yield of some families was superior to the yield obtained by the check, variety PUE 152. PUE 152 was chosen as a check for its superior expression in nitrogen fixation (Graham, 1981). Few families were equivalent or superior to the parental line UW 21-58, which was the higher yielding genotype among the parents. This was not unexpected, since UW 21-58 was bred for BNF ability (Bliss, 1985). The family 48-66 was the top yielder in the test and the families 8-3, 48-94B-1, and 51- 29-3, which were included as poor responders, were indeed among the low yielders (particularly 51-29-3)(Table 3). Family 9-39-1, which was daylength sensitive, showed an outstanding 30% protein content in the seed whereas the parental genotype Lef-Z-RB showed the lowest value for this variable (Table 3). The genotypes UW 21-58, Bayo Madero, and II900-5-M-45 showed the highest dry matter yields among the parents. The Mexican parents which were previously selected as drought tolerant probably had been indirectly selected for BNF ability since they are from. the Mexican Plateau where the agricultural soils are low in N content. The families, 39-17-1, 48-66, and 8-4 achieved maximum values for total dry matter (TDM). With respect to the total amount of nitrogen accumulated per plant, the parental genotype Bayo Madero and the previously mentioned families, 48-66, 39-17-1 and 8-4, showed maximum values (Table 4). It seems that the families with better ability for nitrogen assimilation (BNF) were also able to transform.it into biomass (primary 94 Table 3. Values of seed yield per plant, yield components, HI, seed protein content, and nitrogen harvest index (NHI) of 25 been genotypes grown in a N-free medium.in the greenhouse. East Lansing, MI. 1987 Seed Ent. yield Pods/ Seeds Seed 0 Seed No. Pedigree /plant plant /pod weight HI protein NH1 "'8" ..m.. ..x.. 1 48-10 7.9 8 4.34 235 0.26 24.0 0.39 2 48-66 16.5 14 4.96 237 0.37 23.0 0.59 3 8-25-2 9.6 12 5.19 165 0.34 21.8 0.53 4 48-94B-1 8.4 14 3.99 288 0.31 23.7 0.51 5 9-39-1 8.8 11 4.94 142 0.21 30.0 0.41 6 8-3 7.6 10 4.75 169 0.31 25.3 0.54 7 8-15-1 6.7 8 4.91 162 0.34 22.5 0.50 8 39-17-1 8.8 11 3.34 238 0.18 24.2 0.32 9 8-42-1 8.8 9 4.45 223 0.32 22.9 0.50 10 8-42-2 10.0 8 5.07 251 0.35 22.9 0.55 11 41-18-1 9.2 10 4.95 176 0.33 24.6 0.52 12 41-39 7.8 9 3.92 232 0.21 24.3 0.35 13 8-4 12.7 15 6.06 144 0.28 23.4 0.46 14 41-48-1 6.4 8 4.97 157 0.34 26.8 0.52 15 8-47 9.6 10 5.22 185 0.31 23.8 0.51 16 48-109 8.4 9 4.80 186 0.33 22.0 0.50 17 51-29-3 6.2 9 4.12 156 0.28 23.9 0.45 18 51-5-3 7.4 11 4.51 155 0.32 22.9 0.51 19 8-17 9.5 11 4.47 200 0.33 21.8 0.53 20 N81017 7.5 10 4.84 165 0.30 23.5 0.52 21 Lef-2-RB 8.4 8 4.84 209 0.36 18.4 0.52 22 UW 21-58 10.1 11 5.84 156 0.30 23.4 0.54 23 UW 23-61 4.8 7 4.72 139 0.28 23.6 0.44 25 II900-5- 6.4 6 4.70 204 0.14 24.8 0.23 M-45 26 PUE 152(1) 8.1 8 3.88 267 0.22 22.4 0.30 LSD (0.05) 2.7 2.7 0.67 33 0.07 3.7 0.09 * Bayo Madero was not included. 0 Roots and senesced leaves were included in the denominator. (1) Check genotype. 95 Table 4. Values of plant phenology, plant dry matter, total N per plant, biomass/plant/day (b/p/d), nitrogen accumulation g/plant/day (N/p/d) and root/shoot ratio of 26 bean genotypes grown in a N-free medium in the greenhouse. East Lansing, MI. 1987 Ent. _Days_tg_ Assimilated Root/ No. Pedigree FLOW MAT TDM N/plant b/p/d N/p/d shoot " 8 " " m8 " " 8 " " m8 " 1 48-10 43 80 33.17 724.8 0.41 9.09 0.31 2 48-66 49 85 44.60 1007.9 0.52 11.91 0.20 3 8-25-2 44 80 28.18 636.8 0.35 8.11 0.19 4 48-94B-1 4O 79 29.00 636.3 0.37 8.09 0.21 5 9-39-1 54 87 39.60 880.4 0.46 10.41 0.43 6 8-3 43 77 25.59 571.3 0.33 7.34 0.28 7 8-15-1 48 80 19.64 467.9 0.25 5.91 0.24 8 39-17-1 46 88 50.98 1071.3 0.58 12.24 0.34 9' 8-42-1 45 82 24.85 546.8 0.31 6.56 0.20 10 8-42-2 47 79 30.22 682.7 0.38 8.66 0.17 11 41-18-1 38 78 27.98 645.0 0.36 8.34 0.21 12 41-39 47 83 38.57 817.3 0.46 10.03 0.29 13 8-4 51 87 45.08 953.3 0.52 11.01 0.25 14 41-48-1 38 76 18.86 438.0 0.25 5.80 0.35 15 8-47 48 80 30.98 680.5 0.39 8.59 0.17 16 48-109 42 77 25.68 550.8 0.33 7.12 0.25 17 51-29-3 38 .77 22.45 565.3 0.29 7.42 0.23 18 51-5-3 37 75 23.87 . 506.5 0.32 6.83 0.27 19 8-17 45 78 29.57 616.9 0.38 7.99 0.23 20 N81017 44 78 25.49 560.2 0.33 7.24 0.23 21 Lef-Z-RB 42 78 23.14 415.9 0.30 5.46 0.18 22 UW 21-58 47 83 34.04 709.3 0.41 8.50 0.24 23 UW 23-61 38 75 17.34 385.0 0.23 5.17 0.35 24 BAYO MAD. -- 90 53.59 1343.8* 0.60 14.92 0.50 25 II900-5- 51 85 49.83 984.2 0.58 11.70 0.46 M-45 26 PUE 152(1) 42 85 41.14 882.1 0.48 10.22 0.32 LSD (0.05) 3.6 3.1 7.76 178.0 0.09 2.13 0.11 * Not corrected by N seed content, thus total N per plant. (1) Check genotype. 96 productivity): however, one cannot say that BNF is the direct cause of the increase in biomass but rather that there is a marked relationship between BNF and productivity. This high relationship between BNF and productivity has been repeatedly shown in the common bean and other legume crops. Rennie and Kemp (1981a) compared the N’ fixation ability of eleven dry bean cultivars at 56 days after planting and found that weight and area of leaves, plant dry matter and N yields were significantly correlated with N2 fixed. In a second study, to assess the amount of fixation of four cultivars under two temperature regimes, they found that the amount of N fixed was highly correlated with plant dry matter and N yield but not with acetylene reduction over the entire season (Rennie and Kemp, 1981b). Pacowsky et al. (1984) evaluated several bean-Rhinginm strain combinations for N2 fixation, N assimilation and biomass production. They indicated that assimilated nitrogen is a better, more direct measure of efficiency in the bean-rhizobia symbiosis than the estimate of nitrogen fixation by acetylene reduction and hydrogen production. Cultivars which came into flowering later were more productive in terms of seed yield, biomass and N accumulated than the early flowering genotypes. Thus, the best responders in this test were those genotypes with a relatively longer vegetative period than the low responders. It seems that the genetic system governing daylength sensitivity, which sets the developmental stage for assimilate partitioning .in beans (Wallace et al. 1987), is affecting BNF. This suggests that it may be advantageous to select beans that have a longer vegetative period to prolong the time of active N2 fixation within the limits of acceptable 97 season length to ensure normal maturity (Rennie and Kemp, 1981b). Since there were genotypes in this experiment with different plant types (i.e. types II and III) and phenology, two efficiency indexes were calculated to further evaluate the differences among genotypes, namely, biomass/plant/day and N assimilation /plant/day (Table 4). In general, the use of these indexes indicated the same results as using biomass and N assimilation per plant. In previous experiments in the greenhouse and in the field (Part 1, this research), the Mexican genotype Bayo Madero had shown an average ability to fix atmospheric nitrogen and a high response to the addition of nitrogen fertilizer: in this particular test, Bayo Madero showed the highest value for total nitrogen per plant. This result was due probably to its extended vegetative phase which is equivalent to a lengthened fixation period directly producing a larger biomass. In Table 5 some important correlation coefficients between N assimilation with biomass and other agronomic variables are presented. A marked relationship is observed between N- assimilation with the number of days to maturity, the weight of the root system and biomass (TDM). HI and NHI were highly associated, as was expected, since probably both variables depend on the same mechanism(s). With the exception of a non- significant positive correlation of HI and NHI with seed yield, these two indexes (HI and NHI) were negatively and significantly associated with all the other recorded traits. This result confirms the multiple-site ,competition for photoassimilates between the root nodules and the developing fruits in the pulse-rhizobia association where the seeds are the strong N and C sinks. Therefore, as indicated by Graham.(1981) and Pate and Minchin (1979), N2 fixation is 98 Table 5. Correlation coefficients among nitrogen assimilation, plant biomass (TDM), seed yield, and related variables in a group group of 26 inoculated bean genotypes grown in a N-free medium in the in the greenhouse. East Lansing, MI. 1987 Nitrogen Seed assimilation TDM yield NHI Days to flowering 0.65*** 0.66*** 0.44* -0.24 Days to maturity 0.91*** 0.91*** 0.47* -0.52** Root weight 0.91*** 0.90*** 0.11 -0.81*** Root/Shoot ratio 0.53** 0.51** -0.42* -0.74*** Harvest Index (HI) -O.61*** -0.64*** 0.32 0.94*** TDM 0.97*** 1.00 0.49* -0.58** Seed yield 0.54** 0.49* 1.00 0.38 Nitrogen Harvest -0.53** -0.58** 0.38 1.00 Index (NHI) *.**.*** = P<0.05, 0.01, 0.001, respectively 99 influenced by N source-sink relationships among plant organs. A positive non-significant correlation between seed yield and seed protein content was observed. The seeds of grain legumes are strong sinks for nitrogen due to their high protein content (Sinclair and de Wit, 1975) and part of their requirement is partially met by remobilization of N from other plant parts. Later in development, senescing leaves can supply large quantities of the nitrogen needed by protein rich seeds (Pate and Minchin, 1979: Izquierdo, 1981). Thus, as pointed out by Summerfield (1981), the death of leaves during reproductive growth cannot be divorced from their photosynthetic activity and interrelationships with symbiotic dinitrogen fixation during the pre-flowering period. CONCLUSIONS The evaluated families showed variation for traits related to BNF and some promising genotypes were identified. In addition, a high association between total dry matter and total nitrogen assimilated per plant may suggest that selection for BNF can be done by selecting vigorous plants from segregating populations. Since BNF is a quantitative trait, a portion of the segregating population, say the best 5 or 10 percent, could be taken to be parents of the next generation in which selection can continue. A reduction in the number of crosses by careful selection of parents will allow the handling of relatively large segregating populations. Rennie and Kemp (1981b) suggested selection on the basis of leaf area, assuming that a plant with a larger leaf area may have greater photosynthetic capability and should support greater amounts of nitrogen fixation. Growing bean 100 plants in a N-free medium. forces total dependence on seed and atmospheric N2 thus providing a convenient screening procedure for BNF ability before field experimentation (Graham, 1981) . The inherent ability of the beans to support nitrogen fixation should not be altered under field conditions although the efficiency of using fertilizer or soil nitrogen, or both, may alter the absolute amount of nitrogen fixed (Rennie, 1979). Recently, St. Clair (1986) pointed out that shoot N can be used as an effective selection criterion for measuring nitrogen fixation in plants with similar plant type and.maturity grown on a low-N soil. Thus, as stated by Graham (1981) and Bliss (1985), the selection and breeding of high N2 fixing bean genotypes is an important step toward increasing seed yields in both optimal and suboptimal conditions. A bean genotype displaying a quick vigorous seedling establishment with ability for high BNF may have more chance of success in a stressed environment. The experiments conducted so far in this research partially support this hypothesis, however, a further critical study in the greenhouse and in the field involving a few good and poor N fixer genotypes is needed to test the hypothesis. REFERENCES Bliss, F.A. 1985. Breeding for enhanced dinitrogen fixation potential of common bean (Phaseglns_ynlgazis L.). In: P.W. Ludden and J.Eh Burris (Eds.). Nitrogen Fixation and.CO metabolism. Proc. 14 Steenbock Symposium, 17-22 June 1984, 2University of Wisconsin, Madison. Elsevier Science Publishing Co., pp. 303-310 Chapman, A.L. and R.C. Muchow. 1985. Nitrogen accumulated and partitioned at. maturity by grain legumes grown under different water regimes in a semi-arid tropical environment. Field Crops Research 11:69-79 Graham, P.H., and J.C. Rosas. 1977. Growth and development of indeterminate bush and climbing cultivars of Phaseglns,yn1garis L. inoculated with Rhizghium. J. Agric. Sci. Camb. 88:503-508 Graham, P.H. 1981. Some problems of nodulation and.symbiotic nitrogen fixation in 2haseglns_ynlga:is L. : a review. Field Crops Res. 4:93-112 Izquierdo, F.J.A. 1981. The effect of accumulation and remobilization of carbon assimilate and nitrogen on abscission, seed development, and yield of coumon bean (2113199111: Maria L.) with differing architectural forms. Ph.D. Thesis, Michigan State university, East Lansing p. 188 McFerson, J., Bliss, F.A., and J.C. Rosas. 1982. Selection for enhanced nitrogen fixation in common beans (Phaseglns ynlgaris). In: P.H. Graham. and S. Harris (Eds.), Biological Nitrogen Fixation Technology for Tropical Agriculture, CIAT Series No. 03E-5 (82) pp. 39-44 McFerson, J. 1983. Genetic and breeding studies of dinitrogen fixation in common bean, (Phaseglns ynlgazis L.). Ph. D. Thesis, University of Wisconsin, Madison p. 147 Pacowsky,R.S., Bayne, H.G., and G.J. Bethlenfalvay. 1984. Symbiotic interactions between strains of Rhizobinm.phaseoli and cultivars of Phaseolns ynlgaris L. Crop Sci. 24:101-105 Pate, J.S. and F.R. Minchin. 1979. Comparative studies of carbon and nitrogen nutrition of selected grain legumes. In: R.J. Summerfield and A.H. Bunting (Eds.), Advances in Legume Science, London HMSO pp. 105-114 Rennie, R.J. 1979. Comparison of 15N-aided methods for’ determining symbiotic dinitrogen fixation. Rev. Ecol. Biol. Sol. 16:455-463 101 102 Rennie, R.J., and C.A. Kemp. 1981a. Selection for dinitrogen-fixing ability in Bhaseglns ynlgaris L. at two low-temperature regimes. Euphytica 30:87-95 Rennie, R.J., and C.A. Kemp. 1981b. Dinitrogen fixation in pea beans (Phaseolns ynlgaris) as affected by growth stage and temperature regime. Can. J. Bot. 59:1181-1188 Rosas, J. 1983. Partitioning of dry matter, nitrogen fixation, and seed yield of common bean (Phaseglus,ynlgaris L.) influenced by plant genotype and nitrogen fertilization. Ph. D. Thesis, University of Wisconsin, Madison p. 127 Rosas, J.C. and F.A. Bliss. 1986. Host plant traots associated with estimates of nodulation and nitrogen fixation in common bean. Hertscience 21:287-289 Sinclair, T.R. and C.T. de Wit. 1975. Photosynthate and nitrogen requirements for seed production by various crops. Science 189:565-567 Sprent, J.I. 1981. Nitrogen fixation. In: L.G.Paleg and D. Aspinall (Eds.), Physiology and Biochemistry of Drought Resistance in Plants. Academic Press, Australia pp.131-143 St. Clair,lg.A. 1986. Segregation, selection, and population improvement for N-determined dinitrogen fixation ability in common beans (Phaseolns ynlgarisL.). Ph. D. Thesis, University of Wisconsin, Madison p. 90 Summerfield, R.J. 1981. Adaptation to environments. In: C. Webb and G. Hawtin (Eds.). Lentils. ICARDA, Page Bross (Nerwich) Ltd. pp 91- 110 Wallace, D.H., K.S. Yourstone and P.N. Massaya. 1987. Daylength modulation of partitioning to yield and its control by a single photoperiod gene of been (Phaseolns ynlgaris L.). Agron. Abst. 79:102 Ziska, L.H., Hall, A.E., and R4M. Hoover. 1985. Irrigation management methods of reducing water use of cowpea (Vigna nngnicnlata [L.l Walp.) and lima bean (Phaseolus lunatns L.) while maintaining seed yield at maximum levels. Irrig. Sci. 6:223-239 CHAPTER 4 PLANT TRAITS RELATED TO PRODUCTIVITY IN BEANS (Phaseolus IDlBflIifi L.) UNDER STRESS AND NON-STRESS CONDITIONS. INTRODUCTION The Mexican Plateau is the largest bean producing area in Mexico. In this area, the bean crop depends on unreliable natural rainfall to meet the moisture requirements of the crop. Drought is a severe problem during the early flowering and pod filling stages of the bean crop. Usually, farmers plant beans following the onset of the rainy season during June or early July. In.most years, the crop meets its moisture requirements during the vegetative phase but the frequency of occurrence of drought increases later during the reproductive stage. The breeding of crop plants for drought tolerance has been the goal of breeders for a long time, although progress in this area of research has been slow. In fact, breeding for drought resistance has been considered elusive and frustrating. For example, Arnon (1980) pointed out that breeding for drought resistance was probably the greatest source of wasted breeding effort in the whole field of plant breeding. Drought occurs in many different ways and one of the difficulties in improving the yield of been plants exposed to drought under uncertain rainfall patterns, as is the case in the Mexican Plateau. The difference in timing and intensity of drought stress can influence the final crop yield in various ways. In other words, drought is multifaceted, varying greatly over different production regions and 103 104 often interacting with other detrimental factors such as high temperatures, pathogenic soil fungi and the use of marginal soils to grow beans (Singh and White, 1988). Hanson (1980), has listed the various types of drought adaptive mechanisms found in sorghum: mechanisms which probably are also important in most crop plants: phenological, morphological, physiological and biochemical . However, these adaptations should not be considered singly since they are largely interrelated: for example, many biochemical and physiological responses to drought stress influence crop growth and the final yield through changes in phenology and morphology. Although progress has been achieved in the understanding of basic physiological and biochemical responses of plants subjected to drought stress, much more remains to be learned. At the present time, in the improvement of any species for drought tolerance, breeders have to rely on two of the adaptive mechanisms indicated by Hanson (1980), namely, phenological and morphological attributes. During the evolution of higher plants, successful colonization of terrestrial environments was largely due to a wide array of phenological and morphological adaptations, whereas the basic biochemical apparatus remained essentially unchanged (Begg, 1980). Modifications of plant form and structure and changes in the number of plant parts may affect reproductive fitness. Thus, it can be argued that progress might be made in breeding for drought tolerance with the aid of phenological and morphological traits that are related to drought response in crop plants. In cowpeas, some of the advances made toward increased productivity in moisture stress environments are through a decrease in the length of 105 time for a crop to reach maturity (Hall and Patel, 1985). However, earliness as a trait is not favored among breeders because in many crops it has consistently been shown that a negative correlation exists between yield and early maturity. Beans grown in rainfed areas in Latin America are expected to be under moisture deficits during their reproductive phase (Laing et al., 1983). Therefore, it seems that indeterminate, mid-season cultivars with developmental plasticity are the most appropriate to develop for the sporadic rainfall patterns of the semi-arid Mexican Plateau. It is assumed that mid-season cultivars may be able to produce high yields in those years when above normal moisture is available and may also be able to accelerate their cycle and produce relatively well in the face of terminal drought. In this study, research is concentrated, among other traits, on those adaptative features which are intermediary in character, i.e. leaf area, (quantified at a particular time, not the integral of leaf area duration over time), or which have an overall integrative effect such as shoot biomass and seed yield per se. The objectives of the research reported herein. were : 1) To determine growth and yield responses to drought stress in 26 bean genotypes grown in two different locations, and 2) To evaluate the responses in order to determine whether any traits other than yield per 3e are amenable for use by plant breeders during selection for drought tolerance. MATERIALS AND METHODS Two experiments were conducted during sunmer of 1987 , one at the Centro de Investigaciones Forestales y Agropecuarias (CIFAP) field station in F.I. Madero, Durango, Mexico, and the other at the Kellogg 106 Biological Station.(KBS) of Michigan State University near Battle Creek, MI. The Durango site is located at 24° 20' North and 1040 20' West, at an altitude of 1932 masl, and the soil is a Luvic chernozem (Typic Argiudoll, FAO classification) from. volcanic ash with overlaying fragmented rock and low organic matter content. The KBS site is located at 42° 25' Nerth and 85° 30' West, at an altitude of 250 masl, and the soil is a Kalamazoo sandy loam (Fine-loamy, mixed mesic, typic Hapludolf, FAO classification). A separate randomized complete block design experiment with three replications was used with 26 genotypes (Table 1, previous Chapter, except PUE 152) at two moisture levels in each location, with differential drought stress induced by use of a rain-shelter and natural rainfall plus irrigation. Experimental plots consisted of a single row 4-6 m in length per genotype, depending on space under rain-out shelters th and 10th, in both locations. Planting dates in.Michigan were June 6 for stressed and irrigated portions of the experiment, respectively. In Durango, both parts of the experiment were planted on July 4th, however, at this site water for irrigation was not available and thus it was a stressed and a rainfed experiment. Plant stand in Michigan was approximately 20 plants/m2 and was 10 plants/m2 in Durango. At both locations, plant stands used were similar to those used by the farmers in commercial fields. In the stressed part of the experiment at both sites, plots were protected against rains from the beginning of flowering (most of the genotypes) to physiological maturity. This protection was achieved with the aid of an automated rain-out shelter at KBS and with a set of minishelters consisting of polyethylene sheets over gabled metal frames (1.5 m height at the mid point) at Durango. 107 ELBNI_5AHELINGLBND_HEBSEREHENIS Plant samples were taken at three times during the growth cycle as follows : 40 days after emergence (40 DAE), 57 or 60 DAE and at physiological maturity (PM). For the first two sampling dates, the plants of a 0.5 m section of row were cut at the soil level and dissected to determine leaf area index (LAI) with the use of a leaf area meter (LI-3000, LICOR instruments, Lincoln NB.) and the leaves, stems and branches oven dried and weighed. Abscissed and yellow leaves in the sample were not included, however, during the second sampling date, their weights were recorded to quantify the amount of abscission in response to moisture treatments. At the second sampling date, the weight of developing pods was also included. Five days after drought treatment was established, leaf expansion rate (LER) was estimated daily for two weeks from the product of leaflet length and width measurements using a linear regression equation. The equation was determined with the use of the actual area of leaves measured directly with the leaf area meter and regressed against the length times the width of the same leaflets. A regression equation was calculated for each genotype. In the Durango experiment, during the second sampling time, three fully expanded leaves, uniform in size, were cut from different plants in a plot and used to determine the leaf water content (LWC) and leaf water retention capacity (LWRC) at 24 and 48 hours after excision. At physiological maturity, a 2 m section of row per genotype was harvested to quantify the amount of total dry matter (TDM) excluding the leaves, which were removed in late maturing genotypes for uniformity of comparisons. Weights of remaining aerial parts are highly correlated with the actual total biomass (Wallace et al., 1987). In the same 108 sample, the seed yield, yield components, harvest index, percent of seed protein and total amount of protein produced in seeds/m2 were determined. In the Durango experiment, three plants exhibiting a uniform degree of maturity were used to estimate visually the amount of carbohydrates in the stem.with the use of an iodine-potassium iodide (IKI) indicator as described by Adams et al. (1978) and Izquierdo (1981). ELBNI_IRBIIS From plant sampling and measurements, a total of 33 trait-variables were calculated. Different numbers of traits per location and water regime are presented in detail because they (i) were recorded after stress treatment was imposed, and/or exhibited (ii) significant variation among genotypes and/or (iii) significant correlation with seed yield or TDM at maturity. The traits were grouped in sets according to the method of measurement and/or agronomic basis of the trait (Table 1). 5! I' l' J 1 Data for each trait were analyzed within moisture treatment for each location, and a combined analysis of variance including both locations was performed following McIntosh (1983). A stability analysis was performed by considering each moisture treatment-location as a distinct environment. For each moisture level per location, phenotypic correlations using means of three replications were estimated and used to study the association of seed yield and total dry matter (excluding leaves) at maturity with the rest of the variables. Although simple correlations are not indicative of cause and effect, they show the degree and direction of associations between two traits. Table 1 . 109 two moisture regimes at two locations. 1987 List of plant traits measured in 26 bean genotypes grown under Variable Abbreviation Phenomenal No. of days from.sowing to 50 % flowering. Flow No. of days from sowing to physiological maturity. PhyMa Length of reproductive phase (days) RePha * Dry weight of leaf lamina at 40 DAE2 * (g/m?) Leth Dry weight of stems at 40 DAE £g/m.) sthl Shoot dry weight at 40 DAE (g/m.) *** 2 TDM1 Dry weight of leaf lamina at 57 and 61 DAE (g/m ) LeWt2 Dry weight of stems at 57 and 61 DAE (gm ) SmWTZ Dry weight of developing pods (g/m ) 2 Poth Shoot dry weight at 57 and 61 DAE (g/m ) 2 TDMZ Shoot dry weight at physiological maturity (g/m ) TDM Dry weight of senesced leaves (g/m ) WtSeLe El . J . 1 Leaf area index at 40 DAE 2 LAIl Crop growth rate at 40 DAE (g/m /d) 2 CGRl Net assimilation rate at 40 DAE (g/m /d) NAR1 Leaf weight ratio at 40 DAE LWR1 Leaf area index at 57 and 61 DAE 2 LAIZ Crop growth rate at 57 and 61 DAE (g/m./d) CGR2 Leaf weight ratio at 57 and 61 DAE 2 LWR2 Net assimilation rate at 57 and 61 DAE (ghm /d) NAR2 Leaf expansion rate (cm /leaflet/d) LER Leaf water content (%.dry wt. basis)(1) LWC Leaf water retention capacity (1 dry wt. basis)(1) LWRC Crop Growth Rate average CGRAV fl . J 'l' Seed protein in per cent SePro N seed yield (g/m ) Nitro IKI visual score (1) IKI We Seed yield (g/mzl Yield Number of pods/m Pods Number of seeds per pod Seeds 100 seeds weight (g) SeWt Harvest Index HI * Above-ground plant parts. ** Days after emergence. ‘** At 57 and 61 DAE in Durango and Michigan, respectively. (1) Traits only recorded in Durango. 110 Principal Component Analysis (PCA) was used to identify the characteristics representing the components accounting for most of the variation associated ‘with each set of data. Varimax rotation was employed to accentuate the traits with larger loadings and facilitate the biological interpretation in each component. Two sets of data were formed with average values of 28 and 25 characteristics measured on the twenty six genotypes under both water regimes at Durango and Michigan, respectively. All traits, except those recorded at maturity, were subjected to Multiple Regression Techniques to determine which trait(s) could account for the largest proportion of the variation among the genotypes for seed yield and biomass at maturity. Variables resulting in the highest R2 as single traits or in combination with other traits were then selected to be used in the development of the best model for seed yield and biomass at.maturity in each location and water regime. Plant traits were used as independent variables and seed yield and biomass as dependent variables. RESULTS AND DISCUSSION Significant variation was found for moisture treatments and among bean genotypes per location and averaged over locations (Tables 1, 2, and 3, Appendix C). In the analyses per location, more variables showed a highly significant genotype X water regime interaction at KBS (Table 2, Appendix C) than at Durango (Table 3, Appendix C). It seems that the genotype X environment interactions at the KBS-site were due largely to highly contrasted environments encountered by the genotypes between the irrigated and stressed plots. At this site, the genotypes were able to express more completely their genetic potential for yield under the irrigated regime. The calculated drought intensity index (DII) at KBS 111 was equal to 0.58, while at Durango it was 0.38. In this case, however the lower DII does not mean that the stress imposed at this site was less than the one imposed at KBS. Instead, it means that in Durango the difference between the stressed and non-stressed plots was smaller than at KBS due to the lack of irrigation and the inability of the genotypes to express their maximum.genetic potential. In Table 2, the mean yield per water treatment and site indicates that the stressed environment in KBS can be considered moderate, and equivalent, in terms of moisture stress, to the rainfed environment of Durango. From.this table, it can also be seen that the range for seed yield under severe stress is limited in comparison to the range observed at KBS under irrigation. It is important to point out that the top yielders under stress include parental as well as recombinant genotypes, whereas under irrigation the top yielders are only recombinant progeny. This situation indicates that the level of drought tolerance of the parents was incorporated among some of the progenies, however, it seemed difficult to obtain superior transgressive segregants for drought tolerance. As indicated by the irrigation experiment at KBS, segregants with superior yield potential were also obtained, even though the evaluated genotypes were previously selected under mild drought stress conditions. The above results imply that selection for drought tolerance at higher yield potentials may be difficult. In addition, the fact that the genotype 8-47 was among the top yielders under stress and non-stress in Michigan and one of the low yielders in Durango indicates that the association between yield potential and drought tolerance becomes weaker as the stress level increases. A further complication is that some other 112 Table 2. Seed yield of 26 bean genotypes grown under two moisture regimes at two locations. 1987 Durango Michigan (KBS) Genotype stressed rainfed stressed irrigated ................... 8,,“2 ------------------ 48-10 78.9 110.1 119.9 219.5 48-66 78.7 106.1 134.0 251.8 8-25-2 75.1 103.5 80.2 186.9 48-94B-1 70.8 109.4 127.3 293.1 9-39-1 64.9 86.6 85.0 184.7 8-3 53.5 117.8 77.9 187.5 8-15-1 62.0 120.0 86.9 318.5 39-17-1 78.7 131.4 83.3 189.7 8-42-M-1 69.6 133.1 118.4 299.7 8-42-M-2 69.6 141.7 125.8 301.0 48-18-1 61.4 104.8 111.4 268.5 41-39 50.3 95.8 101.0 363.1 8-4 59.6 81.4 102.2 213.9 41-48-1 64.6 98.7 123.4 238.0 8-47 41.8 69.1 136.7 371.5 48-109 50.0 115.4 130.5 258.5 51-29-3 45.3 122.8 105.8 251.7 51-5-3 57.0 111.2 88.6 282.8 8-17 62.3 128.9 102.7 326.0 N81017 65.8 111.2 109.8 231.2 Lef-Z-RB 65.2 104.2 106.5 295.1 UW 21-58 52.3 93.4 95.2 270.9 UW 23-61 67.8 125.0 136.9 259.2 Bayo Madero 79.2 158.3 83.4 236.3 II-900-5-M-45 73.1 114.7 139.7 293.6 Katolika 85.1 113.6 123.9 291.0 Site mean 64.7 111.8 109.1 264.8 Range 45.3-85.1 69.1-158.3 80.2-139.7 184.7-371.5 LSD (0.05) 23.8 34.2 24.3 47.3 G.V. % 22.6 18.6 18.5 13.8 113 environmental factors and interactions become apparent when evaluating genotypes in different environments. Furthermore, it is only under severe water stress that the drought tolerance mechanisms are totally expressed and yield maintained in the most tolerant types. In general, the rank of the genotypes in the different test environments was markedly different: the exception was the family 9-39-1 which was among the poor responders, and the parental genotypes N81017 and II900-5-M-45 which were among the better performers (Table 2). The genotypic mean yield and some stability parameters were calculated to further assess the adaptability of the genotypes (Table 3). A range of responses in adaptation was obtained as indicated by the regression slope values of the genotypes. The majority of the genotypes displayed slopes stastistically equal to 1.0, even though many of them. are arithmetically different: this occurs because the associated error is relatively large. The five top yielders include four of the F7 families and one of the parental genotypes. Some genotypes such as 8-42-M-2, 8- 42-M-1, 48-18-1, II900-5-Mé45 and check cultivar Katolika, seem. to respond relatively well to the test environments. Families 8-47 and 41- 39 showed a high response to favorable moisture conditions. Results shown by genotypes like 8-47, which exhibited a large Sd2 value, should be viewed with caution. Since the climatic conditions in Michigan were more favorable than those in the Mexican Plateau (Tables 4 and 5, Appendix C), this evaluation for wide adaptation may imply the risk of selecting for mediocrity for the Michigan environment. This statement is reinforced by an observed highly significant positive relationship among the regression slopes of the individual cultivars and yield under irrigation in Michigan (R2 = 0.96 ***). 114 Table 3. Mean seed yield and stability parameters of 26 been genotypes grown under stress and non-stress conditions in Durango (Mexico) and Michigan (USA). 1987 Mean 2 2 Genotype yield b(1) Sd (2) R (3) g/m2 8-42-M-2 159.5 I8 1.13 95.4 0.994 II-900-5-M-45 155.2 | 1.09 201.9 0.985 8-42-M-1 155.2 I 1.14 38.9 0.997 8-17 155.0 I 1.34* 147.1 0.993 8-47 154.8 I 1.68 1394.0 0.959 Katolika 153.4 I 1.06 122.5 0.991 41-39 152.6 I 1.62* 262.1 0.991 48-94B-1 150.1 II 1.12 114.8 0.992 UW 23-61 147.2 II 0.91 215.3 0.978 8-15-1 146.8 II 1.32 346.3 0.983 Lef-Z-RB 142.7 Ill 1.12 75.2 0.995 48-66 142.6 III 0.86 229.5 0.974 Bayo Madero 139.3 lllI 0.78 1339.9 0.838 48-109 138.6 IIII 0.98 357.0 0.969 48-18-1 136.5 IIIII 1.04 22.4 0.998 51-5-3 134.9 IIlll 1.15 128.2 0.992 48-10 132.1 Illll 0.69* 40.7 0.993 51-29-3 131.4 IIIII 0.98 258.0 0.997 41-48-1 131.2 lllll 0.85 203.3 0.976 N81017 129.5 IIIII 0.81* 20.8 0.997 UW 21-58 128.0 IIIII 1.11 33.9 0.998 39-17-1 120.7 (Ill 0.55 545.5 0.864 8-4 114.3 III 0.78 136.7 0.981 8-25-2 111.4 (II 0.58* 152.6 0.962 8-3 109.2 II 0.64 444.4 0.914 9-39-1 105.3 I 0.61** 18.5 0.996 9 Duncan's Multiple Range Test (0.05). * b is different from 1 at a P<0.05 (1) Regression coefficient. (2) Sum of squares from deviations. (3) Coefficient of Determination. 115 Evidently, the bean crop in Michigan encounters different environmental conditions than those in the Mexican highlands, not just with regard to the drought stress, but other factors determined by the physical environment and/or by cultural practices. The varieties better adapted and ‘widely grown in these two distinct places display strong differences in morphology (Kelly et al., 1987: Acosta and Khohashi, 1988) and they probably possess different attributes with which to cope with drought. In addition, as Singh and White (1988) pointed out, much research has been conducted for unspecified drought conditions, implicitly assuming that drought is a single, well defined condition. This is most assuredly not the case. Therefore, before starting a breeding program. for drought tolerance in any region, the soil and climatic conditions should be characterized to determine the level and kind of tolerance that is needed. Given the difference in drought situation between Durango (Mexico) and Michigan, in the rest of this report results from each location will be presented and discussed separately. E . l . ".1. (KBS) With the exception of the weight of the leaves at 61 DAE, there was a significant detrimental effect of the water stress treatment for all the recorded traits and calculated variables (Table 4). The stress slightly reduced the length of the reproductive phase and therefore the whole cycle was reduced as indicated by the number of days needed to reach maturity. Those traits most affected by drought were those related to growth, the dry weight of the developing pods at 61 DAE and TDM at maturity. Since there was no difference between treatments in the time at which 116 Table 4. Effect of two moisture regimes on plant traits of 26 bean genotypes. Kellogg Biological Station, Battle Creek, MI. 1987 Reduction Variable Rainfed Stressed in.% Phenolosical PhyMa 91.2** 87.3 4.3 RePha 47.3** 43.5 8.0 B' @ LeWtZ 103.5 NS 102.7 0.8 smwtz 175.0 * 157.0 10.3 Poth 115.0 ** 72.5 37.0 TDM2 357.8 * 313.3 12.4 TDM (1) 520.9 ** 315.8 39.4 E] . 1 . J LAI2 3.38 * 2.98 11.8 CGR2 5.87 * 5.14 12.4 LWR2 0.30 * 0.33 -9.1 LER 2.87 ** 1.03 64.1 CGRAV(2) 6.02 * 3.45 42.7 SePro 21.9 ** 29.4 -35.5 Nitro 8.87 ** 4.87 45.1 Went: Yield 264.8 ** 109.1 58.8 Pods 265.2 ** 178.3 32.8 Seeds 4.61 ** 3.00 34.9 SeWt 21.7 * 21.0 3.2 HI 0.51 * 0.35 31.4 *,** Significant at P<0.05 and 0.01, respectively. @ Above-ground plant parts. (1) Leaves not included. (2) Average of 21 days after stress was established. 117 the genotypes commenced flowering, the weight of developing pods can be considered as a gross estimate of partitioning to the economic plant parts. Although translocation is considered one of the less sensitive physiological traits affected by water stress (Hsiao, 1973) and since partitioning depends on translocation, its reduction is important and very similar to the calculated index for partitioning at maturity i.e. HI. TDM at maturity, as measured in this study, was reduced in the same proportion as the developing fruits (Table 4). Water deficit affects many aspects of crop growth, develOpment and yield. Hsiao (1973) stated that almost any trait of crop growth is affected by water stress, provided that the stress is severe and lasts long enough. Of the physiologically related traits, the LER and the average CGR were the most affected by water stress. The large decrease in LER indicated the marked sensitivity of leaf expansive growth to water deficits (Figure 1). From Figure 1, it is apparent that the stress was building up slowly as indicated by the slight reduction in leaf expansion at 8 days after it was imposed. As a consequence of the reduced expansion in leaf growth, the OCR, measured as a daily average of 21 days during the stress build up, was also markedly reduced (Table 6). Under both stress and non-stress, LER was positively and significantly associated with all the phenological variables while negatively associated with dry weight of the developing pods. Since LER and the weight of the developing fruits were measured at the same growth stage, this negative association between them at the beginning of pod filling suggests competition for assimilates between these plant organs. 118 Leaflet ex ansion rate cm /3 days 13 - A—A Irrigated o O Stressed 7 .— 6 -I 5 _. " <5 3 - I 2 - % 1 _ “ \§ “4 T T T I T T I I I I T 8 9 10 11 12 13 14 15 16 17 18 Days after stress was imposed Figure 1. Average leaf expansion rate of 26 bean genotypes grown under irrigated and water stressed conditions. Kellogg Biological Station, Battle Creek, MI. 1987. 119 Since expansive growth can be viewed as an integrator of the metabolic and environmental events that influence over-all plant productivity (Braford and Hsiao, 1982). it may be justifiable to determine the amount of genetic variation for LER available to the breeders and its usefulness as a selection criterion. LER was previously suggested as a negative index of drought sensitivity in several crops (Boyer, 1970: Hsiao and Acevedo, 1974: Boyer and McPherson, 1975: Sammons et al., 1978). Vidal and Arnoux (1981) found in soybeans a high correlation between the reduction in leaf expansion and petiole growth on the one hand, and the reduction in biomass and seed yield under stress, on the other. The percentage of protein in seeds was highly reduced under irrigation, demonstrating the negative association between seed protein content and yield. However, the total amount of protein harvested in the seeds per unit area was severely reduced by the stress since yield was also significantly reduced. In the case of seed yield, an average reduction of 58.8 1 under stress showed the accumulated effect of drought stress upon the economic product. Using the individual yield per genotype under both stress and non-stress, different drought susceptibility indices were calculated (Table 5). Genotypes 8-47, 8-42-M62, 48-94B-1 and 41-39 showed a good yield response under both drought stress and non-stress conditions as indicated by their geometric mean values, while genotypes 8-3, 8-25-2, 9-39-1 and 39-17-1 obtained the lowest values with the use of this index. In general, the percent reduction of seed yield under stress did not agree with the ranking obtained with the use of the geometric mean: the best genotypes as identified by the geometric mean showed higher 120 G Duncans' Multiple Range Test (0.05). (1) Fischer and Maurer (1978), Appendix A. Table 5. Drought tolerance indices of 26 bean genotypes grown under two moisture regimes. Kellogg Biological Station, Battle Creek MI. 1987 Yield Drought Arith. Geom. reduct. suscep. Genotype mean mean % index(1) ............ 8,m2------------- 8-47 254.1 '8 225.3 63.2 1.1 41-39 232.0 ll 191.5 72.2 1.2 48-94B-1 220.2 III 193.2 56.6 1.0 II-900-5-M-4 216.6 IIII 202.5 52.4 0.9 8-17 214.3 IIII 183.0 68.5 1.2 8-42-M-2 213.4 IIII 194.6 58.2 1.0 8-42-M-1 209.0 IIII 188.3 60.5 1.0 Katolika 207.4 IIII 189.9 57.4 1.0 8-15-1 202.7 (Ill 166.3 72.7 1.2 Lef-Z-RB 200.8 IIII 177.3 63.9 1.1 UW 23-61 198.0 IIII 188.4 47.2 0.8 48-109 194.5 |I|| 183.7 49.5 0.8 48-66 192.9 IIII 183.7 46.8 0.8 48-18-1 189.9 lllI 172.9 58.5 1.0 51-5-3 185.7 llIl . 158.3 68.7 1.2 UW 21-58 183.0 Illl 160.6 64.9 1.1 41-48-1 180.7 IlIll 171.4 48.2 0.8 51-29-3 178.7 IIIIII 163.2 58.0 1.0 N81017 170.5 llllI 159.3 52.5 0.9 48-10 169.7 IIIII 162.2 45.4 0.8 Bayo Madero 159.8 lIII 140.4 64.7 1.1 8-4 158.0 llll 147.9 52.2 0.9 39-17-1 136.5 III 125.7 56.1 1.0 8-25-2 135.5 II 122.4 57.1 1.0 9-39-1 134.8 II 125.3 54.0 0.9 8-3 132.7 I 120.9 58.5 1.0 Average 187.0 169.2 58.8 1.0 121 than average yield reductions under stress. Nevertheless, the percent reduction agreed 'with the drought susceptibility index of Fischer and Maurer (1978) (r=0.98 ***). Therefore, these latter two indices should be used to separate genotypes which display yields above the average of the test, particularly if the degree of the stress in the target region is not severe. Some genotypes, e.g. 48-66, seem to display true drought tolerance as indicated by the values obtained with all the calculated indices. The relationship between seed yield under stress and irrigated conditions indicated the presence of genotypes which displayed both drought tolerance and good response under irrigation (r=0.43 **I. The number of pods/m? and the number of seeds/pod were markedly and equally affected by the stress, while the weight of 100 seeds remained almost unchanged (Tables 6 and 7). In spite of the relative mildness of the stress, it was enough to diminish the setting of pods at the beginning of the reproductive phase and continued with the same intensity during the rest of the cycle, creating a strong intra-ovary competition, leading to reduction in the number of seeds/pod. Stress markedly decreased HI values (Tables 6 and 7): although some genotypes did not show a significant decrease, e.g. 48-948-1 and N81017. Thus, it appears that differences in partitioning under stress exist among genotypes. The data confirm. work reported that N81017 is a good remobilizer of stored assimilates towards the economic product under drought stress (Samper et a1, 1984). Simple correlation values of seed yield and TDM at maturity with other traits are presented in Table 8. Under stress and non-stress, yield was essentially associated with TDM, HI and two of the primary 122 Table 6. Yield components, TDM and HI of 26 bean genotypes grown under irrigation. Kellogg Biological Station, Battle Creek, MI. 1987 100 Pads Seeds Seed Genotype /m /pod weight TDM HI ....... 8 ------- 48-10 209.7 4.4 25.2 410.0 0.54 48-66 251.0 4.7 22.1 517.7 0.49 8-25-2 277.3 4.1 17.4 368.7 0.50 48-94B-1 300.7 4.2 19.6 496.7 0.59 9-39-1 220.0 4.9 18.1 404.7 0.46 8-3 252.3 4.8 15.9 424.4 0.44 8-15-1 276.7 4.5 20.7 584.0 0.54 39-17-1 301.7 3.7 23.4 437.9 0.44 8-42-M-1 229.7 4.8 24.7 585.3 0.51 8-42-M-2 222.7 5.3 27.4 588.1 0.51 48-18-1 295.0 4.3 21.8 532.1 0.50 41-39 311.7 3.8 23.3 680.1 0.55 8-4 222.0 5.2 17.4 519.6 0.41 41-48-1 283.7 4.6 19.3 441.9 0.54 8-47 289.3 5.5 19.0 800.3 0.46 48-109 278.0 4.6 21.3 467.0 0.55 51-29-3 296.0 3.9 23.1 436.0 0.58 51-5-3 332.0 3.8 22.4 425.4 0.68 8-17 278.7 5.0 20.8 610.3 0.53 N81017 305.0 4.3 18.8 524.6 0.44 Lef-Z-RB 261.3 4.7 25.7 516.4 0.57 UW 21-58 239.3 6.4 18.7 579.0 0.47 UW 23-61 334.7 4.6 18.7 488.6 0.53 Bayo Madero 172.7 3.3 32.7 552.4 0.43 II900-5-M-45 231.7 4.6 23.4 531.5 0.56 Katolika 223.0 5.5 23.1 621.9 0.48 Average 265.2 4.6 21.7 520.9 0.51 123 Table 7. Yield components, TDM and HI of 26 bean genotypes grown under stress. Kellogg Biological Station, Battle Creek, MI. 1987 100 Poda Seeds Seed Genotype /m /pod weight TDM HI ........ 8 -------- 48-10 177.3 3.0 22.5 325.4 0.37 48-66 189.3 2.9 26.5 414.1 0.32 8-25-2 174.0 2.0 19.6 374.1 0.21 48-94B-1 240.3 3.1 18.0 253.9 0.51 9-39-1 82.0 4.1 19.8 237.8 0.36 8-3 154.7 3.1 16.9 239.9 0.32 8-15-1 145.3 2.8 20.7 295.5 0.30 39-17-1 140.0 2.6 25.2 359.8 0.23 8-42-M-1 192.7 2.8 24.1 355.7 0.33 8-42-M-2 137.0 3.5 27.7 341.7 0.37 48-18-1 205.3 2.8 20.2 337.6 0.34 41-39 148.3 2.6 22.7 304.4 0.33 8-4 164.3 3.1 19.7 363.2 0.29 41-48-1 212.3 3.5 15.8 272.6 0.45 8-47 , 197.3 3.3 20.3 348.2 0.39 48-109 180.7 3.5 19.9 318.1 0.41 51-29-3 196.0 2.8 18.3 240.5 0.43 51-5-3 219.7 2.3 16.6 267,5 0.33 8-17 199.7 2.7 20.5 300.7 0.34 N81017 201.0 2.9 18.5 266.8 0.41 Lef-Z-RB 189.0 2.5 22.2 274.4 0.39 UW 21-58 144.0 3.4 19.6 279.1 0.34 UW 23-61 284.0 3.1 15.1 301.4 0.45 Bayo Madero 100.3 2.8 31.0 342.0 0.24 II900-5-M-45 168.3 3.9 21.2 387.5 0.36 Katolika 192.3 3.0 24.3 410.3 0.31 Average 178.3 3.0 21.0 315.8 0.35 124 Table 8. Correlation coefficients of seed yield and TDM at maturity with some characteristics of phenology, plant growth and primary yield components of 26 bean genotypes grown under two moisture regimes. Kellogg Biological Station, Battle Creek MI. 1987 Seed yield TDM Variable Stressed Irrigatede Stressed Irrigatede TDM at 40 DAE -.13 .54** -.05 .48** LAI at 40 DAE .14 .52** .03 .45* LAI at 61 DAE .25 .36 .61** .39 CGR at 40 DAE .13 .54** .05 .48* Leaf exp. rate .18 .27 .41* .16 Days to flowering .00 .22 .49* .47* Days to maturity .02 .03 .53** .11 Length rep. phase .03 .24 .50** -.23 No. pods/m2 .56" .47 .05 -.04 No. seeds/pod .43* .21 -.08 .40* 100 seed weight .02 .23 .59*** .17 N in seeds (g/mz) .9M** .95*** .3o .eowt Seed yield (g/mz) --—— ---- .38 .33M Harvest Index .63*** .41* -.51** .16 *,**,*** significant at P<0.05, 0.01, and 0.001, respectively. 8 Rainfed and sprinkle irrigated. 125 yield components i.e. pods/m2 and seeds/pod. In addition, yield under irrigation was also associated with those growth traits recorded during the first sampling date, TDMl, LAI1 and CGRl, which indicates the importance of a good initial establishment of the crop to obtain high yields under favorable conditions. Significant correlation values of TDM with phenological traits were also observed (Table 8). Under stress, a significant and positive correlation between TDM and LAIZ and LER, both of which were measured after stress was established, point out the importance of the photosynthetic apparatus in the total accumulation of dry matter under mild stress conditions. The level of drought tolerance exhibited by some of the recombinant families appears to be enough to cope with the occasional water stress encountered in Michigan. The addition of further tolerance, especially that conferred by survival mechaniems, may not be useful because it may lead to reduced mean yields in non-stress environments. E . l . D I! . Rainfall data recorded during the growing season in this location (Table 5, Appendix C) revealed the occurrence of several drought periods during the season. Consequently, the potential yield of the genotypes grown under rainfall, which represented the non-stress environment, was not expressed. However, the severity of the stressed environment was large enough to significantly decrease yield along with almost all other recorded variables in this experiment (Table 9). Results obtained with phenological traits were similar to those observed in Michigan: the length of the reproductive phase was reduced and thus the number of days to reach maturity. All the traits related 126 Table 9. Effect of two moisture regimes on plant traits of 26 been genotypes. Durango, Mexico. 1987 Reduction Variable Rainfed Stressed in.% Rhenolosical PhyMa 90.5** 85.8 5.2 RePha 46.2** 41.8 9.5 . 8 Emma: LeWtz 98.9 ** 49.4 50.1 3th2 76.3 ** 46.7 38.8 Poth 60.5 ** 39.2 35.2 TDMZ 235.7 ** 135.2 42.6 TDM (1) 194.2 ** 119.2 38.6 E] . J . 1 LAI2 2.72 ** 1.36 50.0 CGR2 4.13 ** 2.37 42.6 LWR2 0.43 ** 0.37 14.0 CGR (2) 8.58 ** 2.70 68.5 LER 1.84 ** 0.55 70.1 LWC 79.3 NS 81.5 -2.7 LWRC 9.7 ** 49.4 -80.4 m . J 'l' SePro 28.6 NS 29.1 -1.7 Nitro 4.92 *4 2.88 41.5 IKI 2.86 * 2.62 8.4 Wis Yield 111.8 4* 64.7 42.1 Pods 139.5 ** 86.4 38.1 Seeds 4.30 ** 3.93 8.6 SeWt 19.2 NS 19.4 -1.0 HI 0.57 * 0.54 5.3 *,** Significant at P<0.05 and 0.01, respectively. 9 Above-ground plant parts. (1) Leaves not included. (2) Average of 17 days after stress was established. 127 to biomass which.were measured after the stress was established were severely affected by the stress as indicated by their percent reduction (Table 9). At this site, the LER and the LAI at 57 DAE were markedly reduced. Acosta and Kohashi (1988) previously reported a positive association between LAI at this stage in development and final seed yield: in their research (Acosta and Kohashi, 1988) the cultivars were released from stress at the beginning of pod filling. In the present study, the cultivars were not released from.the stress for the rest of the season and thus senescence was further hastened. The early onset of severe stress in this location was demonstrated by the decrease in LER measured one week after the stress was imposed (Figure 2). Under stress, a significant and positive correlation between LER and dry weight of senesced leaves was observed. An important consequence of reduction in leaf area and early senescence is a reduction in the rate of water use and a delay in the onset of more severe stress. Leaf water content was essentially the same under stress and non- stress (Table 9). However, in the case of leaf water retention capacity (LWRC) of detached leaves, leaves of genotypes under non-stress lost 80 x. more water in 24 hours than leaves from.stressed plants (Table 10). °c Detached leaves were kept at room temperature (24+2o C day, 17+2 night, and 50+5 % relative humidity) during LWRC determination. Under stress, significant differences were found among the genotypes at 24 and 48 hours after leaves were detached (P<0.1 and 0.05, respectively). As expected, LWC and LWRC were positively associated with LER. It appears that the genotypes under stress were pro-conditioned to retain their moisture, while, if grown under a more favorable soil water status they were not. Currently, the mechanism(s) the leaves used to 128 Leaflet ex ansion rate cm /3days 7 " 4r_____qg Rainfed .A O O Stressed /’/ \ 6 .4 /’ \\ .z I”! \ I \ x4 \ 5 - .r’ \ ,” \ / A’ \ 4 .p \ \ \ \\ 3 " A 2 -— .__ 6* °\ 1 - o\ (D L l l I l l I I I I I I 8 9 10 ll. 12 13 14 15 l6 17 18 Days after stress was imposed Figure 2. Average leaflet expansion rate of 26 been genotypes grown under rainfed and water stressed conditions. Durango, Mexico..l987. 129 Table 10. Leaf water content of detached leaves of 26 bean genotypes grown under drought stress and rainfed conditions at 50 DAE. Durango, Mexico. 1987 * Water percent on weight basis. Stressed Rainfed Genotype 0 hs 24 he 48 hs 0 hs 24 hs * ...................... % -------------------- 48-10 80.9 53.9 36.0 81.0 13.7 48-66 80.5 51.3 34.5 79.2 12.0 8-25-2 79.3 35.2 15.9 79.0 7.4 48-94B-1 81.4 50.9 33.0 79.9 11.2 9-39-1 82.1 46.0 26.8 81.4 20.6 8-3 77.4 43.3 23.0 76.9 10.6 8-15-1 80.2 50.3 26.1 79.5 8.0 39-17-1 79.3 46.1 26.3 79.6 13.0 8-42-M-1 80.1 47.6 28.2 80.5 9.9 8-42-M-2 80.4 49.3 26.5 78.1 7.2 48-18-1 83.2 57.1 37.5 78.4 8.1 41-39 84.3 54.6 37.7 78.0 6.5 8-4 83.5 45.0 23.1 79.5 8.5 41-48-1 80.6 51.0 29.7 79.7 7.3 8-47 79.9 41.0 20.6 81.0 7.4 48-109 80.9 49.4 32.3 79.0 9.1 51-29-3 83.0 53.4 37.8 78.2 5.4 51-5-3 82.6 54.0 33.3 79.0 7.4 8-17 83.7 54.6 33.1 79.0 11.1 N81017 79.8 44.5 22.9 80.0 7.4 Lef-z-RB 82.8 52.0 34.7 79.2 9.5 UW 21-58 78.9 45.8 26.5 81.1 14.1 UW 23-61 82.0 46.1 28.4 78.3 11.2 Bayo Madero 82.9 50.6 33.1 78.9 8.6 II900-5-M-45 86.2 58.2 41.0 80.7 12.7 Katolika 83.9 54.2 29.3 77.6 4.6 Average 81.5 49.4 29.9 79.3 9.7 LSD (0.05) 4.1 10.0 11.7 4.1 10.0 130 keep their moisture is not known. However, work conducted by Ibarra (personal conmunication, 1988) which included some of the same genotypes used in this study and the same water regimes, demonstrated that those genotypes under stress showed a reduction in stomatal conductance. Generally, stomatal closure has been reported to occur when a critical leaf water potential is reached, thus preventing a further decrease in potential. In addition, it has been suggested that increases in specific leaf weight in snap beans subjected to drought stress may be due, among other factors, to an increase in epicuticular wax (Bonnano and Mack, 1983), which could enhance cuticular resistance to water loss. Under stress and non-stress, LWRC showed a consistent, nonsignificant negative correlation with seed yield and TDM at maturity. Assessment of water loss from.excised leaves has shown some promise for differentiating between drought resistance of wheat cultivars (Dedio, 1975: Clarke and.McCaig, 1982; Clarke and Townley-Smith, 1984). In this study, the low values for LWRC of genotypes under stress may indirectly indicate the ability of the genotype to extract soil water .more efficiently. This is supported by negative association between LWRC with TDM and seed yield. High LWRC appears to be related to survival mechanisms. The use of the IKI solution to visually assess the amount of starch left in the stem.at.maturity indicated that starch amount was slightly reduced under stress (P<0.05): a significant difference was detected among the genotypes for this variable under both stress and non-stress conditions and for the interaction between water regime x genotype (P<0.01). This variable showed a negative non-significant correlation (r=-0.33) with seed yield under stress. Since LER, LWRC and the IKI 131 score showed differences among genotypes, it may be worth further work to assess their usefulness as a screening technique for drought tolerance in beans. The calculated drought tolerance indices (Table 11) showed that local genotypes, e.g. Bayo Madero, Mex 1213-2 and Dgo-222, were the most adapted to the stress conditions of this experiment in terms of seed yield. The data from Mex 1213-2 and Dgo-222 are not shown because they were excluded from. the test in IMichigan due to their daylength sensitivity. Within the tested families, 39-17-1 was among the top yielders under both stress and rainfed conditions exhibiting the largest geometric mean value. The results behaved in a similar fashion to those obtained in Michigan, the ranking of the genotypes by the geometric mean was markedly different than that shown using either the yield reduction (1) or the drought susceptibility index (DSI). The later two scales both essentially gave the same result as indicated by a high correlation (r=0.99 ***) found between the ranking of the genotypes by these two indices. Genotypes with average susceptibility or tolerance to drought have DSI values of 1.0. Values of DSI less than 1.0 indicate less susceptibility and greater adaptation to drought, with a value of DSI = 0.0 indicating maximum possible adaptation to drought (Hall and Patel, 1985). It seems that the use of either the yield reduction or the drought susceptibility index identifies those genotypes with true drought tolerance. Therefore, these indeces do not identify those genotypes which possess the ability to produce high yields under favorable conditions. In both locations, a positive relationship was observed between seed yield under non;st:e§s_ggnditign and the DSI. In 132 Table 11. Drought tolerance indices of 26 been genotypes grown under two (1) Fischer and Maurer (1978), Appendix A. G Duncans' Multiple Range Test (0.05) moisture conditions. Durango, Mexico. 1987 Yield Drought Arith Geom reduct. suscep. Genotype mean mean (X) index(1) ...... “al-“--- Bayo Madero 118.7 I0 112.0 50.0 1.2 8-42-M-2 105.6 II 99.3 50.9 1.2 39-17-1 105.0 II 101.6 40.1 1.0 8-42-M-1 101.3 III 96.2 47.6 1.1 Katolika 99.3 III 98.3 25.1 0.6 UW 23-61 96.4 II 92.1 45.8 1.1 8-17 95.6 III 89.6 51.7 1.2 48-10 94.5 III 93.2 28.3 0.7 II-900-5-M-45 93.9 III 91.6 36.3 0.9 48-66 92.4 lIlI 91.4 25.8 0.6 8-15-1 91.0 llII 86.2 48.3 1.1 48-94B-1 90.1 IIIII 88.0 35.3 0.8 8-25-2 89.3 IIIII 88.2 27.4 0.7 N81017 88.5 IIIII 85.5 40.8 1.0 8-3 85.6 IIIII 79.4 54.6 1.3 Lef-2-RB 84.7 IIII 82.4 37.4 0.9 51-5-3 84.1 IIII 79.6 48.7 1.2 51-29-3 84.0 IIII 74.6 63.1 1.5 48-18-1 83.1 IIII 80.2 41.4 1.0 48-109 82.7 (III 75.9 56.7 1.3 41-48-1 81.6 (III 79.9 34.6 0.8 9-39-1 75.7 IIII 75.0 25.1 0.6 41-39 73.0 III 69.4 47.5 1.1 UW 21-58 72.8 III 69.9 44.0 1.0 8-4 70.5 II 69.7 26.8 0.6 8-47 55.4 I 53.7 39.5 0.9 Average 88.2 84.7 42.1 1.0 133 general, those genotypes with small DSI were the less productive. Likewise Fischer and Wbod (1979) found that in cereals drought susceptibility increased with increased non-drought yield. However, Sojka et al. (1981) indicated that in wheat the given traits of any one cultivar which result in a high yield potential are likely to be expressed over a large range of environments and thus produce a high baseline yield under drought. This baseline must not be confused with drought resistance: drought resistance is better defined as the ability to minimize yield loss in the absence of optimal soil water availability. The challenge of applying these findings lies in transfering the ability to maintain percent yield under stress to cultivars with higher absolute yield levels. Nevertheless, with our present knowledge of the adaptation of crop plants to drought stress, it seems impractical to separate drought tolerance from agronomic performance or yielding ability. The relationship between yield under stress and non-stress showed a positive association indicating the presence of some genotypes which respond in a similar fashion to the different applied water regimes (r =0.43 *). Stress caused a severe reduction in pod number and a mild reduction in the number of seeds per pod, while seed weight remained unaltered (Tables 12 and 13). The number of pods/m2 or pods/plant has consistently been the component most affected by water stress in all the experiments conducted in this research. Since the number of pods set is largely determined in the first few days of flowering, in order to evaluate a group of genotypes for adaptation to drought using the criterion of seed yield, the genotypes must be sub-divided on the basis 134 Table 12. Yield components, TDM and HI of 26 bean genotypes grown under water stress. Durango, Mexico. 1987 100 Pods Seeds Seed Genotype /m /pod weight TDM HI ....... 8 ---_--__ 48-10 76.6 4.3 23 7 133.9 0.59 48-66 98.8 3.8 20 7 139.8 0.56 8-25-2 116.4 3.7 17 9 131.6 0.57 48-94B-1 101.2 4.1 16 9 119.6 0.59 9-39-1 88.0 4.7 15 9 119.3 0.54 8-3 82.2 4.3 14 9 101.5 0.52 8-15-1 92.1 4.3 15 8 136.3 0.45 39-17-1 94.4 2.9 27.8 129.8 0.59 8-42-M-1 94.4 3.6 20 4 128.7 0.54 8-42-M-2 87.4 3.3 24 1 139.5 0.50 48-18-1 82.7 3.8 19 3 112.6 0.54 41-39 67.0 3.6 20 5 103.2 0.48 8-4 87.7 4.3 15 2 116.4 0.50 41-48-1 79.8 4.7 17 2 117.0 0.55 8-47 72.5 3.4 17 1 98.5 0.43 48-109 71.3 4.1 17 0 94.7 0.53 51-29-3 71.3 3.6 17 9 82.7 0.54 51-5-3 88.9 3.9 16 3 102.0 0.56 8-17 98.8 3.4 18 7 120.8 0.51 N81017 89.5 4.1 17 7 118.7 0.55 Lef-Z-RB 87.1 3.5 20 8 119.0 0.54 UW 21-58 80.1 3.6 18 4 102.3 0.51 UW 23-61 103.5 4.3 14 9 124.0 0.54 Bayo Madero 79.8 3.0 33 8 135.7 0.58 II900-5-M-45 80.7 4.1 22 2 131.0 0.56 Katolika 73.4 5.6 21 5 142.1 0.61 Average 86.4 3 9 19.4 119.2 0.54 135 Table 13. Yield components, TDM and HI of 26 been genotypes grown under rainfed conditions. Durango, Mexico. 1987 100 Pods Seeds Seed Genotype lm /pod weight TDM HI ........ 8 -------- 48-10 124.3 4.3 21.1 194.7 0.57 48-66 154.6 3.5 20.0 195.8 0.54 8-25-2 133.3 4.2 18.1 177.9 0.58 48-94B-1 143.9 4.6 16.3 180.5 0.61 9-39-1 121.3 4.4 16.4 167.5 0.52 8-3 184.2 4.5 14.6 209.2 0.56 8-15-1 154.4 4.3 17.9 220.2 0.54 39-17-1 141.0 3.9 24.6 209.9 0.63 8-42-M-1 134.0 4.7 20.7 216.9 0.61 8-42-M-2 162.5 4.2 20.8 248.2 0.57 48-18-1 134.6 4.0 19.5 177.4 0.59 41-39 139.5 3.2 20.9 188.2 0.50 8-4 113.2 3.7 19.3 152.6 0.53 41-48-1 135.7 4.4 16.7 175.4 0.56 8-47 118.0 3.7 15.8 155.9 0.44 48-109 144.1 4.4 18.7 198.0 0.58 51-29-3 110.5 6.3 19.4 191.9 0.64 51-5-3 161.2 4.9 14.2 190.1 0.58 8-17 151.3 4.1 20.8 230.9 0.56 N81017 142.8 4.5 17.4 203.7 0.54 Lef-Z-RB 139.0 3.7 19.9 182.7 0.56 UW 21-58 135.3 4.2 17.2 161.6 0.58 UW 23-61 182.5 4.5 15 3 205.9 0.61 Bayo Madero 118.0 4.1 34.6 223.9 0.71 II900-5-M-45 124.6 4.5 20.6 205.9 0.56 Katolika 123.0 4.9 18.9 183.8 0.62 Average 139.5 4 3 19.2 194.2 0.57 136 of phenological characteristics in order to stress them evenly. The plants were under severe stress in Durango from the beginning of flowering and the number of pods/m2 was determined early during the reproductive phase (Figure 1, Appendix C). Therefore, since the plants retained few pods per plant, the number of seeds /pod was not affected by the stress in the same magnitude as the number of pods (Tables 12 and 13). At the Michigan site, the stress was less severe and it built up more slowly than in Durango. In Michigan, a continuous dropping of pods was observed throughout the reproductive phase and, at maturity , almost half of the pods had just one or two seeds each. In addition, in this site, some pods did not have a single normal seed. Thus, it is clear that each of those yield components was affected separately by the stress due to their sequential development and the timing, intensity and duration of the applied stress (Figure 3). Harvest Index was only slightly reduced by the stress in the Durango experiment. Indeed, some genotypes, e.g. 48-10, 48-66 and N81017, displayed equivalent or slightly larger HI values under stress (Tables 12 and 13). Similar correlation coefficients were found between seed yield and HI under stress and non-stress (Table 14). With respect to the use of the IKI solution, a significant negative association was found between it and HI under stress (r=-0.53 **I, which may indicate the inability of the poor yielders to carry out starch remobilization. If yield under severe stress at the pod filling stage depends to some significant extent upon stored assimilates, then cultivars with an early, vigorous establishment, and a relatively large biomass accumulated at the beginning of seed filling, may be the better adapted genotypes in this stressed environment. An important aspect of I... 3 2 .1 b _ b _ lllllll’llllIIIIIIIIlPrt IIIIIII‘IIII (111E ,/////////// _././ _/../ 9 ., _ m S O P fed ZZZ? Stressed _ _ no no no no no .5 no .5 3 during the re roductive base on the ts of 26 been enot es xico). 1987. 138 Table 14. Correlation coefficients of seed yield and TDM at maturity with some characteristics of phenology, plant growth and primary yield components of 26 bean genotypes grown under two moisture regimes. Durango, Mexico. 1987 Seed yield TDM Variable ;;;;;;a"‘.;;;;;a ';;;;;;é"'i;;;£;5 Days to flowering -.33 -.54** .00 -.28 No. pods/m2 .431: .30 .45* .47* No. seeds/pod .15 .30 .11 .12 100 seed weight .55** .59** -.48* .41* Seed protein (X) -.59*** -.07 .61*** -.03 N in seeds (g/mz) .9a*** .98*** .881!“ .87*** TDM at maturity (g/mz) .9o*** .88*** ---- ---- Harvest Index .73*** .77*** .37 .39 *,**,*** significant at P<0.05, 0.01, and 0.001, respectively. 139 developmental plasticity is the ability of plants to transfer assimilate accumulated prior to seed filling to the seed during the seed filling stage. Type III varieties are the most adapted to the semiarid Mexican highlands (Acosta and Kohashi, 1988). The type III genotypes used in this study were selected after the evalution of hundreds of local landraces and bred cultivars. Therefore, is not a surprise that they were the most productive in the Durango environment. Mighigan_expeziment. The PCA was used here to reduce the dimensions of multivariate data and to identify important variables to be used for prediction in multiple regression models. The first five principal components accounted for different cumulative amounts of variation under stress and irrigated conditions and different traits accounted for that variation in each condition (Tables 7 and 8, Appendix C). Under stress, the first PC accounted for 30% of the variation and was characterized by phenological traits, the second PC accounted for 17% of the variation and was characterized by early growth and weight traits at the beginning of flowering. Yield and yield related traits showed maximum loading in the third PC and explained 12% of the variation: the fourth PC accounted for 13% of the variation and was characterized by growth traits measured at the reproductive stage, and in the fifth PC the amount of senesced leaves showed maximum loading and accounted for 7% of the variation. On the other hand, under irrigation variables related to growth measured at different phenological stages were prominent in the first three PC's, primary productivity traits (seed yield and biomass) displayed maximum 140 loading values in the fourth PC, and the primary yield components in the fifth. Under irrigation the first five PC's accounted for 22, 17, 10, 12, and 7% of the total variation associated with the data, respectively. With ‘a representative trait from.each component and traits which have shown a significant association with seed yield and/or biomass, multiple regression.models were chosen to estimate the combination of traits which could account for the largest amount of variation for seed yield and biomass among the genotypes in each location and water regime. Traits measured. at. maturity were not included in order to avoid unrealistic estimation of R2, since, for example, biomass, seed yield, HI and the primary yield components are all part of the same end product. In addition, the presented models are not unique: better models might be generated by using other criteria and other statistical procedures. Data from. both water regimes were used to construct separate models. The traits selected for the models were subjected to a stepwise multiple regression procedure after which traits were removed or added until the best models, based on the magnitude of R2, were identified. Predictor variables that were intercorrelated were removed from. the models based on their tolerance values; tolerance values near zero indicate high intercorrelation among predictor variables (Wilkinson, 1986). Different combinations of plant traits resulted in the highest R2 values with seed yield or biomass in each water regime (Table 15). The best multiple regression models for seed yield under stress and non- stress included two traits each and accounted for 29 and 35 % of the 141 Table 15. Combination of plant traits resulting in the best multiple regression model for seed yield and biomass at maturity for a group of 26 been genotypes grown under drought stress and irrigation. Kellogg Biological Station, Battle Creek MI. 1987 Variable probability R2 --------------------- Regression Partial ---------------------- Dependent Independent coefficient R partial* cumulativeé ------------------------------ Irrigated ------------------------------- Seed yield constant 36.824 0.583 TDM1 0.784 0.289 0.004 2 LER 182.641 0.058 0.169 R 0.346 0.008 Biomass constant -115.347 0.426 Sth2 1.497 0.430 0.003 WtSeLe -3.361 0.090 0.004 Flow 9.193 0.089 0.043 2 Leth 1.294 0.042 0.077 R 0.665 0.000 ------------------------------ Stressed -------------------------------- Seed yield constant 184.607 0.000 NAR1 -43.632 0.192 0.037 2 CGRAV 3.243 0.146 0.080 R 0.295 0.042 Biomass constant 128.898 0.010 LWR2 299.443 0.366 0.028 2 Sth2 0.584 0.326 0.065 R 0.455 0.001 * T test (Two tail). @ F test for regression 142 genotypic variability associated with seed yield, respectively. The best models for biomass under irrigated and stressed condition included four and two traits and accounted for 66 and 45 % of the variability in biomass, respectively. It appears that better models can be generated for biomass than for seed yield as indicated by the R2 values. Forty three percent of the variability associated with biomass under irrigation was explained by a single trait, namely, the weight of the stems (StWt2) at 61 DAE. This indicates that genotypes with strong thick stems seem to be better adapted to favorable conditions in Michigan. The bean genotypes which display that morphological characteristic are the type II growth habit with the modified ideotype or architype as described by Adams (1982). It was previously indicated that type II architype cultivars are well adapted and exhibit high yield potential under the Michigan conditions (Kelly et al., 1987). Kelly and Adams (1987) and Acquaah (1987) found that the transfer of architectural traits, e.g. strength of main stem, is relatively easy within the same gene pool. Davis and Evans (1977) studied the importance of a set of phenological and morphological traits in the construction of selection indices for seed yield in navy beans. They indicated that the best . indices included the thickness of the stem (hypocotyl) which also showed a relatively high heritability. In this study, the evaluated families, which were derived from crosses involving type II and III parents from.diverse origin, were not selected based on architectural traits (see Chapter 2), but on the basis of seed and biomass productivity. Consequently, since it seems that the type II's are more adapted to the environment in Michigan, the Mexican genotypes, Bayo Madero, Lef-Z-RB and II900-5-M-45, together with 143 Katolika, which display the type III growth habit, were excluded, and principal component and multiple regression analyses were recomputed for the type II's only. Results of PC analysis excluding type III cultivars under stress showed that the first five components accounted for approximately the same amount of variation as in the previous analysis (80 %). The variables with the maximum loading coefficients were essentially the same in the first two components, but with opposite signs: while the variables ‘with maximum loadings in the remaining components were different (Table 9, Appendix C). Under irrigation, the variables showing maximum loadings were different than those in the previous analysis in all the PC's (Table 10, Appendix C). Even though the loading of the variables into the components were somewhat different than in previous analyses, most of the important variables remained the same, thus PCA was useful in reducing the number of variables to be used for prediction or description with or without all bean genotypes. With the elimination of the data from type III cultivars, the magnitude of the R2 estimates for the models to explain the variation in seed yield under irrigation and drought stress was increased (Table 16). 2 values were 46 and 48%, respectively. The R2 value for the Those R models with biomass at maturity as the dependent variable was slightly reduced under irrigation, but in the new model just two predictor variables were included versus four in the previous model. Under stress, the new calculated model kept the same variables, but its R2 decreased. It seems that under stress type III cultivars were more associated with the observed variation in biomass. 144 Table 16. Combination of plant traits resulting in the best multiple regression model for seed yield and biomass at maturity for a group of 22 bean genotypes grown under drought stress and irrigation. KBS, Battle Creek MI. 1987 Variable probability 1:2 ---------------------- Regression Partial --------------------- Dependent Independent coefficient R partial* cumulativefl ------------------------------- Irrigated ------------------------------ Seed yield constant -279.588 0.122 CGR1 15.743 0.274 0.221 CGR2 36.217 0.257 0.035 2 Flow 5.468 0.088 0.121 R 0.458 0.010 Biomass constant 264.450 0.002 Sthz 2.265 0.434 0.000 2 WtSeLe -3.233 0.128 0.008 R 0.614 0.000 -------------------------------- Stressed ------------------------------ Seed yield constant 279.374 0.000 NAR1 -46.492 0.138 0.048 CGRAV 4.339 0.171 0.015 2 RePha -2.287 0.116 0.019 R 0.479 0.007 Biomass constant 243.907 0.016 SthZ 0.837 0.226 0.170 2 LWR2 0.272 0.223 0.180 R 0.298 0.060 * T test (Two tail). @ F test for regression 145 In the case of seed yield, different variables related to early growth and phenology were included in all proposed models under both stress and non-stress, while for biomass, the weight of the stem at 61 DAE was important in all the calculated models (Tables 15 and 16). The importance of a closely related trait, namely, hypocotyl diameter, as an important component of plant architecture in beans was indicated by Acquaah (1987). Dnrangg_Expeziment. At this location the first five PC's accounted for similar amounts of the variance under rainfall and stress, 68.5 and 67.5%, respectively (Tables 12 and 13, Appendix C). In the first PC, traits related to early growth showed maximum loadings under both conditions. This response was expected, since at the time those variables were recorded, the stress treatment had not yet been imposed. The loadings on PC2 and subsequent PC's varied in sign and magnitude under rainfall as compared to those observed under stress. Under rainfall, PC2 was characterized by growth traits measured at 57 DAE (beginning of pod filling): in PC3 yield pangs: and related traits showed the highest loadings. Days to physiological maturity was the most important trait in PC4 and leaf weight ratio at 57 DAE (LWR2) in PCS (Table 11, Appendix C). Under stress, heavily loaded traits in PC's 2 to 4 were essentially the same traits as under rainfed conditions, but with opposite signs: and in the fifth PC, leaf water content (LWC) and leaf water retention capacity (LWRC) showed maximum loadings (Table 12, Appendix C). The observed differences for trait loadings under rainfall and stress indicates a different contribution to variability from the same traits. Thus, the bean genotypes expressed different traits to adapt 146 to the different environments. Data from both water regimes were used to construct separate models (Table 17). When seed yield under rainfall was the dependent variable, the number of days to start flowering (Flow) was the only important trait by itself (R2=0.29 M), and with the addition of two more variables to the model (WtSeLe and TDM1), it accounted for 48.6% of the variation associated with seed yield. The best model for biomass included four independent variables and accounted for only 37.1% of the variation. Under drought stress, the best models included different variables than under irrigation and the R2 estimates for seed yield and biomass were 35.0 and 27.2%, respectively: these were lower than those obtained under rainfed conditions. The weight of the stem.was again an important trait for both dependent variables studied. This result was probably due to the fact that 85% of the evaluated cultivars belong to the type II growth habit, in which most genotypes display heavier main stems than type III cultivars. Since in this site type III cultivars were the most productive and are the best adapted to this environment, it was considered impractical to remove all type II cultivars and recompute the models on the basis of only four cultivars. The IKI score was a trait measured at maturity and since it was considered as not being interrelated to seed yield or biomass, it was included as a predictor variable in the models. IKI score became an important predictor trait for seed yield under stress. The weight of the senesced leaves (WtSeLe) was used here as a measure of the rate of leaf senescence under stress. However, it was only included in the models for seed yield and/or biomass under 147 Table 17. Combination of plant traits resulting in the best multiple regression model for seed yield and biomass at maturity for a group of 26 been genotypes grown under drought stress and» rainfall. Durango, Mexico. 1987 Variable probability R2 --------------------- Regression Total -------------------- Dependent Independent coefficient R partial* cumulativeG --------------------------------- Rainfed ------------------------------ Seed yield constant 370.840 0.000 Flow -5.695 0.294 0.001 WtSeLe 1.189 0.081 0.022 2 TDM1 -0.364 0.002 0.019 R 0.486 0.002 Biomass constant 677.291 0.023 LWC -8.133 0.111 0.030 RePha 3.667 0.096 0.096 WtSeLe -0.403 0.040 0.051 2 TDM1 1.339 0.034 0.058 R 0.371 0.038 ------------------------------- Stressed ------------------------------- Seed yield constant 122.976 0.000 WtSt2 -0.486 0.142 0.043 IKI -8.163 0.113 0.024 2 NAR2 -13.305 0.059 0.142 R 0.350 0.022 Biomass constant 122.976 0.010 sthZ -1.527 0.134 0.028 LeWt2 1.165 0.029 0.115 2 LER 112.785 0.071 0.200 R 0.272 0.067 * T test (Two tail). ** F test for regression 148 favorable conditions, and its coefficient of regression was always negative. This suggest that under favorable conditions, leaf area duration (LAD) might be a key trait to manipulate in order to increase productivity. Laing et al. (1983) have pointed out the importance of LAD upon seed yield of several legume crops. The amount of senesced leaves as measured in this research may be of more practical use than determining the integral of LAI during the ontogeny of the crop. In designing an optimum model to predict yield or biomass under stress and/or non-stress, many aspects should be considered. From the present results, it seems obvious that the evaluation of cultivars belonging to the same growth habit will result in the identification of better models. The grouping of similar-type cultivars might diminish the complexity of the data by removing some of the multiple interactions that can be obtained with the use of phenologically and morphologically diverse genotypes. In general, when different models were compared, it was clear that the best independent variables varied between location and, to a lesser extent within locations. The plant traits act together in different ways in different genetic backgrounds and in different environments. It seems reasonable to assume that genotypes developed from different genetic backgrounds might use different physiological strategies in achieving their final productivity. The variation of traits included in the best models in the same location indicates that adaptation to drought stress and high productivity in this group of genotypes is conferred by different plant attributes. 149 CONCLUSIONS 1. A wide range in variation among the genotypes was observed for all evaluated characteristics, except for leaf water content. 2. A high and positive correlation between the drought susceptibility index and seed yield at the most favorable environment indicates that some adaptations are mutually exclusive. Therefore, the best genotypes under stress were different from those under non-stress conditions. 3. Genotypic drought response in two different locations, namely Durango (Mexico) and Michigan (USA), were not consistent. This suggests that other factors in addition to stress are important in adaptation. 4. In Durango (Mexico) reduction in yield was due mainly to a smaller number of pods per m2, while in Michigan (USA) both the number of pods per m2 and the number of seeds per pod were decreased by the stress. At both locations, individual seed weight was little affected by water stress. 5. It appears that no single variable investigated in this study can alone be reliably utilized to assess performance for seed or biomass productivity under stress and non-stress conditions. REFERENCES Acosta Gallegos, J.A. and J. Kohashi Shibata. 1988. Effect of water stress on growth and yield of indeterminate dry bean (Phaseglns ynlgaris L.) cultivars. Submitted : Field Crops Research. The Netherlands Acquaah, G. 1987. Genetic and multivariate statistical evaluation of a phenotypic recurrent selection program for recombining erect architecture and large seed size in Phaseglns,ynlgaris L. Ph.D. Thesis, Michigan State University, East Lansing p.317 Adams, MQW., Wiersma, J.W., and J. Salazar. 1978. Differences in starch accumulation among dry bean cultivars. Crop Sci. 18:155-157 Adams, M.W. 1982. Plant architecture and yield breeding. Iowa State J. Res. 56:225-254 Adams, M.W. and J.D. Kelly. 1987. The role of architecture, crop physiology_ and recurrent selection in ideotype breeding for yield in dry beans. In: Symposium on Breeding for Higher Yields in Comon Beans. XXXIII meeting of the PCCMCA, Guatemala, C.A. CIAT-BIC CRSP-ICTA p. 16 Arnon,stl. 1980. Breeding for higher yield. Co-ordinator's report on the 1 session. In: International Potash Institute (Ed.), Physiological Aspects of Crop Productivity. Proceedings of the 15 Colloquium of the International Potash Institute. Wagenningen. The Netherlands pp. 77-81 Begg, J.E. 1980. Morphological adaptation of leaves to water stress. In: N.C. Turner and P.J. Kramer (Eds.), Adaptation of Plants to Water and High Temperature Stress. Wiley, New York pp. 33-42 Bonanno, A.R., and H.J. Mack. 1983. Yield components and pod quality of snap beans growth under differential irrigation. J. Amer. Soc. Hort. Sci. 105:869-873 Boyer, J.S. 1970. Leaf enlargement and metabolic rates in corn, soybean and sunflower at various leaf water potentials. Plant Physiol. 46:233-235 Boyer J.S. and H.G. Mc.Pherson. 1975. Physiology of water deficits in cereal crops. Adv. Agron. 27:1-23 Braford, K.J. and T.C. Hsiao. 1982. Physiological responses to moderate water stress. In: P.S. Nebel, C.B. Osmond, and H. Ziegler (Eds.), Encyclopedia of Plant Physiology, new series, vol. 128 Springer- 150 151 Verlag, Berlin, FRG. pp. 264-324 Clarke, J.M. and T.N. MoCaig. 1982. Evaluation of techniques for screening for drought resistance in wheat. Crop Sci. 22:503-505 Clarke, J.M. and T.F. Townley-Smith. 1984. Screening and selection techniques for improving drought resistance. In: P.B. Vose and S.G. Blixt (Eds.), Crop Breeding, a Comtemporary Basis. Pergamon Press pp.137-162 Davis, J.H.C. and A.H. Evans. 1977. Selection indices using plant type characteristics in navy beans (Phaseglns ynlgaris L.). J. Agric. Sci. Camb. 89:341-348 Dedio, W. 1975. Water relations in wheat leaves as screening test for drought resistance. Can. J. Plant Sci. 55:369-378 Fischer, R.A., and R. Maurer. 1978. Drought resistance in spring wheat cultivars. 1. Grain yield responses. Aust. J. Agric. Res. 29:277- 317 Fischer, R.A. and J.T. Wood. 1979. Drought resistance in spring wheat cultivars . III. Yield association with morphophysiological traits. Aust. J. Agric. Res. 30:1001-1020 Hall, A.E. and P.N. Patel. 1985. Breeding for resistance to drought and . heat. In: S.R. Singh and.K.O. Rachie (Eds.), Cowpea: Research, Production and Utilization. John Wiley & Sons pp 137-151 Hanson, A.D. 1980. Interpreting the metabolic responses of plants to stress. Hort Science 15:9-15 Hsiao, T.C. 1973. Plant responses to water stress. Ann. Rev. Plant Physiol. 24:519-570 Hsiao, T.C. and E. Acevedo. 1974. Plant responses to water deficits, water use efficiency, and drought resistance. Agr. Meteorol. 14:59-84 Izquierdo, F.J.A. 1981. The effect of accumulation and remobilization of carbon assimilate and nitrogen on abscission, seed development, and yield of common bean (Phaseolnstnlgaris L.) with differing architectural forms. Ph.D. Thesis, Michigan State university, East Lansing p. 188 Kelly, J.D. and MLW. Adams. 1987. Phenotypic recurrent selection in ideotype breeding of pinto beans. Euphytica 36:69-80 Kelly, J.D., Adams, 2M.W. and G.V. Varner. 1987. Yield stability of determinate and indeterminate dry bean cultivars. Theor. Appl. Genet. 74:516-521 Laing, D.R., Kretchmer, P.J., Zuluaga, S. and P.G. Jones. 1983. Field bean. In: W.H. Smith and S.J. Banta (Eds), Symposium on Potential 152 Productivity of Field Crops Under Different Environments. Los Banos, Philippines, 1980 (Proceedings) IRRI pp 227-248 McIntosh, M.S. 1983. Analysis of combined experiments. Agron. J. 75:153- 155 Sammons, D.J., Peters, D.B. and T. Hymowitz. 1978. Screening soybeans for drought resistance. I. Growth chamber procedure. Crop Sci. 18:1050-1054 Samper, C., Adams, M5W., Hanson, A.D., and E. Watt. 1984. Contribution of stored assimilates to the seed following a drought stress in dry beans. Agron. Abst. 76:87 Singh, S.P. and J.Wt White. 1988. Breeding common beans for adaptation to drought conditions. In: G. Hoogenboom, F. Ibarra, S.P. Singh, J.W. White and S. Zuluaga (Eds.), Research on Drought Tolerance in Common Beans. CIAT, Cali, Colombia. In press Sojka, R.E., Stolzy, L.H. and R.A. Fischer. 1981. Seasonal drought response of selected drought cultivars. Agron. J. 73:838-845 Vidal, A. and M. Arnoux. 1981. Drought tolerance processes in soybeans. Biol. Plant. 23:434-441 Wallace, D.H., K.S. Yourstone and P.N. Massaya. 1987. Daylength modulation of partitioning to yield and its control by a single photoperiod gene of bean (Phaseolus__ynlgaris L.). Agron. Abst. 79:102 Wilkinson, L. SYSTAT. 1986. The system for statistics. Evanston IL. SYSTAT, Inc. GENERAL DISCUSSION Before discussing the results of this research, it is pertinent to recall that water is as important to drought adapted plants as to non- adapted ones. It is linked to productivity and is a substrate, a constituent and the medium in which all cellular processes take place. Water is also essential for evaporative cooling of the plant and is crucial for the flow of nutrients, hormones and other substances. Therefore, best yields can only be obtained with adequate water, even in the case of the most resistant plants. Maximum biomass production will be achieved only by supplying sufficient water to realize potential leaf area and meet evaporative demand during growth. In the present report, a modest breeding effort to improve drought tolerance and biological nitrogen fixation in beans was undertaken. The objective was to determine whether it is possible to select for drought tolerance and nitrogen fixation concurrently in the same population, and whether genotypes superior in both characteristics simultaneously could be produced. The results obtained during the evaluation of the parental genotypes indicated that nitrogen accumulation (under controlled conditions) in different plant parts was closely related to biomass production and the length of the vegetative phase of growth. In the field, increases in biomass in type III cultivars at the reproductive stage was closely associated with a low partitioning to the developing pods. In type III cultivars, the vegetative and reproductive phases 153 154 overlap, and a competition for assimilates among new expanding leaves and roots versus developing pods and seeds takes place. Mexican genotypes displayed a profusely branched root system which on a dry weight basis is lighter than the root system of type II Michigan cultivars. It seems that the root system of Mexican type III genotypes is designed to explore a larger superficial soil volume. The Michigan type II's possess a thick, less branched root system.which can go deeply into the soil profile, where a large amount of moisture accumulates during the winter or after prolonged rainy periods in the spring. All Mexican type III genotypes proved to be highly photoperiod sensitive. In spite of slow partitioning to the developing pods, some type III cultivars were as productive as type II's under stress. The combination of an acceptable ability for nitrogen fixation and drought tolerance seemed to be already present in some of the parental genotypes e.g. Lef-2-RB, II900-5-M-45, and N81017. A set of segregating populations was produced using Mexican (unadapted) and Michigan and Wisconsin (adapted) genotypes. During the advance of generations and selection (1984-1986), most of the segregant progenies with morphological and/or phenological traits derived from parental Mexican type III cultivars were indirectly eliminated. Two factors contributed most to that elimination: the sensitivity of those progenies to the long day photoperiod, even though some of them exhibited large biomass under stress, and the excesive amount of rainfall during the fall of 1986. Under the wet conditions of 1986, the morphological characteristics of type.II genotypes gave them advantage over prostrate type III genotypes. The final yield of type III 155 genotypes was reduced due to a close contact of their branches and pods with the wet soil and an enhanced incidence of white mold (Sglergtinia sclerotigrum). A group of nineteen genotypes (F6 families) was selected for further evaluation based on their response to mild drought stress encountered in Michigan. Field and greenhouse experiments were conducted in Michigan, USA, and Durango, Mexico. Under controlled conditions, promising genotypes with enhanced BNF were identified. A high association between total dry matter and total nitrogen assimilated (nitrogen fixed) per plant suggested that preliminary selection for BNF can be done by using an estimation of the total dry matter of segregating populations grown in a N-free medium. The results obtained in 1987 in the field indicated that depending upon the history of the crop during past ontogeny, stress can or cannot affect the proportion of dry matter allocated to the seed. In Durango, the stress treatment had no significant effect on the HI, while in Michigan it was drastically decreased by the stress. The crop in Michigan was under near optimum water supply before the stress, thus, these large plants when exposed to the stress were more affected than the smaller, already hardened, bean plants at Durango. A positive association observed between leaf expansion rate (LER) and the amount of senesced leaves indicated that under stress new leaves develop more slowly, and old leaves senesce more rapidly. Thus, there is likely to be a reduction in the photosynthetic area of stressed plants and selection could be practiced by evaluating genotypes on the basis of either an increased LER or a reduced senescence under stress. On the other hand, increased senescence under favorable conditions was associated with a larger biomass. Different selection criteria may 156 prove useful in breeding beans for drought adaptation. Leaf expansion rate has been proposed several times as a selection criterion for drought tolerance in soybeans (Boyer, 1970: Hsiao and Acevedo, 1974: Boyer and McPherson, 1975: Sammons et al., 1978). In the case of beans, LER could be used provided that the set of genotypes to be evaluated includes genotypes of similar growth habit. Since type III cultivars may be developing and senescing a larger number of leaves than types II's. The weight of the stem was a preponderant morphological trait among the genotypes evaluated. Since this trait is related to one of the main features of the architectural cultivars in Michigan, namely hypocotyl strength, it may be suitable for selection in Michigan under stress and non-streess conditions. The diameter of the hypocotyl or main stem at a certain internode could be used as an indicator for this trait. Morphologically different genotypes may display different attributes by which to cope with drought: and the timing of the stress coupled to different phenological stages may induce differential responses in final yield. In screening for drought tolerance, it is rarely possible to hold all other factors constant while screening for a few specific traits. Therefore, the less interactions the plants go through, the more straight forward the interpretation of data. Grouping the genotypes in a final evaluation for drought tolerance upon the basis of their phenology and growth habit is suggested. Drought affects many internal plant processes, plant morphology and phenology, therefore, it is not likely that a single plant trait can account for the response in yield or biomass under drought stress. The use of traits other than the end product (seed or biomass yield) may 157 speed the advance during selection in the field, provided that an extra nursery is planted under favorables conditions for seed security and increase. Breeders have to analyze the type of drought condition in combination with the characteristics of preferred and/or adapted cultivars in the target area. Where unpredictable water deficits occur, maximum productivity should be the goal. Maximum productivity will be achieved where leaf expansion and senescence are relatively insensitive to water deficit, where rate of leaf expansion recovers quickly on relief of water deficits, and where minimum dry matter is partitioned into inmobile root reserves. The semi-arid highlands of Mexico are characterized by extreme annual variations in total rainfall and its seasonal distribution. Especially at the lower range of rainfall, wide annual variation exist in the timing and the magnitude of stress imposed upon the bean plants. The environment of the semi-arid highlands exerts its effects on the genotypes with different intensities and in different directions from generation to generation. It is suggested that two locations at different representative sites in the target area should be used when testing early lines in preliminary yield trials. Since the natural conditions in the semi-arid highlands of Mexico during most years do not allow for full expression of the genetic potential of bean cultivars, yield and drought adaptation could be handled as separate genetic entities. Stable genotypes should be the goal in this environment. A negative association between potential yield and a drought tolerance trait (such as early maturity) need not exclude that trait, since potential yield is neither realized under the 158 actual conditions nor under severe stress. However, in.Michigan, where bean cultivars traditionally encounter a more favorable environment, such negative associations must be considered. In addition, since the drought stress level in these two regions is clearly different, it would not be expected that the cultivars display the same mechanisms of drought adaptation in both regions. In conclusion, recombinant families were produced which displayed good adaptation to the Michigan environment, with improved ability to fix atmospheric nitrogen and with enhanced drought tolerance. Further enhancement of these two traits in a single genotype is feasible, but since nitrogen fixation is readily depressed by water stress, these two traits must be expressed at different developmental stages, i.e. enhanced nitrogen fixation during the vegetative phase and drought tolerance during the reproductive phase. Many mechanisms relate to drought adaptation in plants. The importance of biochemical and physiological traits in adaptation to semi-arid regions cannot be neglected, however, the present understanding of such complex traits is incomplete and their utilization in the breeding of common been for drought tolerance is far from.reality. It seems inevitable that certain complex traits, such as yield under stress, are better handled as such, rather than by dissecting them into components for analysis or investigation. Nonetheless, a better understanding of the most important traits influencing seed yield and biomass under water deficits could, in the long run, lead to improved production practices and more efficient breeding programs. APPENDICES APPENDIX A 159 Appendix A THE METHOD OF FISCHER AND MAURER (1978) With this approach, yields of individual genotypes must be determinated under drought stress (Y3) and irrigated conditions (Yi). Data on the average yield of all genotypes under stress (Y5) and non- stress conditions (Yi) are used to calculate the drought intensity index (DII). DII = 1 - Ys / Yi Then the drought susceptibility (DSI) of individual genotypes is calculated as follows: Y5 = Yi (1 - DSI*DII) DSI = [1 - (Ys / Yi)] / DII Varieties with average resistance to drought have a DSI value of 1.0. values less than 1.0 indicate less susceptibility and greater resistance to drought with a value of DSI = 0.0 indicating maximum possible drought resistance (Hall and Patel, 1985). 160 Table 1, Appendix A. Climatic conditions recorded at the Botany Farm of Michigan State University during the growing season. East Lansing, MI. 1985 Temperature 0C Rainfall Period maximum minimum in am May 1 - 10 22.0 6.9 22.1 11 - 20 23.5 10.5 16.2 21 - 31 22.8 7.5 32.0 June 1 - 10 24.5 9.8 00.0 11 - 20 19.6 9.9 28.7 21 - 30 28.0 12.7 4.3 July 1 - 10 28.1 15.0 12.9 11 - 20 27.8 15.1 39.1 21 - 31 26.8 14.4 14.0 August 1 - 10 27.2 13.9 26.7 11 - 20 25.5 13.0 42.9 21 - 31 24.0 14.1 23.6 September 1 - 10 28.2 18.4 49.0 11 - 20 22.1 8.5 0.5 21 - 30 19.2 7.3 34.3 161 Table 2, Appendix A. Method to determine visual rating of root system for Biological Nitrogen Fixation from Rosas and Bliss (1986). Visual Nodulation rating class subjective description 1 poor nodulation < 50 % nodulation of control 2 below average 51-80 % of control 3 average level 81-120 % of control 4 higher than average 121-150 % of control 5 highly superior > 150 %Iof control 162 Table 3, Appendix A. Soil moisture content in percent at four different depths during the growing cycle. East Lansing, MI. 1985 Soil depth in cm. Days After Planting 0-15 16-30 31-45 46-60 -5 2.5 8.0 9.7 10.0 2 20.7 21.0 17.0 17.6 10 13.5 12.5 10.6 11.3 17 17.3 16.2 12.9 13.6 24 9.6 14.2 10.9 11.9 31 11.5 13.3 10.7 9.3 38 10.4 10.2 10.2 10.2 * 48 8.7 8 1 ---- ---- * 58 Rainfed 9.2 8.2 ---- ---- Stressed 6.9 7.3 ---- ---- * 70 Rainfed 11.1 8.1 ---- ---- Stressed 9.1 7.2 ---- ---- * each value is an average of 48 determinations. 163 Table 4, Appendix A. Mean squares and significance of F value for nodule fresh weight, plant biomass and acetylene reduction of 11 bean genotypes evaluated under four treatments. East Lansing, MI. 1985 Mean Squares Acetylene Reduction1 Nodule Source of fresh Plant variation df weight biomass plant g nodule/plant Treatment 3 1.87** 1.54** 0.97 1.14 Genotype 10 0.06** 0.42** 6.48** 2.71** T X G 30 0.01 0.11** 0.64 0.21 Error 132 0.01 0.04 1.05 0.53 ** F is significant at a = 0.01 1- Treatment with nitrogen fertilizer was not included, resulting in different degrees of freedom for sources of varietion. 164 Table 5, Appendix A. Analyses of variance and significance of F value of different traits of eight bean genotypes under two nitrogen sources at 45 days after planting. East Lansing, MI. 1985 mean squares Source of Variation df roots stems leaves shoot TDM ................... g/mz -----------—--------- Nitrogen 1 0.067 20.8* 94.1* 203.4* 210.8* Genotype 7 1.805** 14.7** 34.2 87.4* 68.1 N X G 7 0.253 3.6 11.2 26.5 29.9 Error 30 0.573 3.8 16.2 33.7 39.3 CV % 16 22 21 21 19 . Stomatal Source of conduc. Root/ Variation df LAI cm/s LWP LSW LAR shoot Nitrogen 1 0.38 2.07 0.02 1992 327 0.006* Genotype 7 1.13** 3.78* 5.90 6789** 2366** 0.011** N X G 7 0.35 0.67 4.73 3153 884 0.001 Error 30 0.19 1.42 3.82 2212 751 0.001 CV % 26 46 32 22 22 19 *,** significance of F at a = 0.05 and 0.01, respectively. * LAI = leaf area index. LWP = leaf water potential. LSW = leaf specific weight. LAR = leaf area ratio. 165 Table 6, Appendix A. Analyses of variance and significance of F value of different plant traits of eight bean genotypes grown under two moisture regimes and two nitrogen sources at 70 days after planting. East Lansing. MI.1985 mean squares Source of variation df Root Stem Leaves Shoot Pods Meisture 1 12.2 362* 2533* 7209 242* Error "a" 2 0.69 194 113 736 7.45 Nitrogen 1 0.92 130 92 1384 262* M X N 1 0.03 35 80 60 50 Genotype 7 13.2** 155* 678** 581 2601** M X G 7 3.5 15.4 50 285 84 N X G 7 2.4 58.5 60 353 25 M X N X G 7 0.8 58 55 389 82 Error "8" 60 3.0 62.4 73 387 53 CV % 25 36 27 29 46 Mean squares Source of Seed Pods Seed Harvest Variation df TDM yield /plot wt. Index Meisture 1 436178 264992* 149696* 9.98 0.009 Error A 2 218994 13161 9245 0.68 0.004 Nitrogen 1 115392 4050 13421 1.33 0.001 M X N 1 1120 12747’ 2952 1.56 0.003 Genotype 5 44669 41979** 46753** 48.1 0.019** M X G 5 26970 27474 11509 7.1 0.003 N X G 5 24630 10467 9451 7.0 0.003 M X N X G 5 22192 4997 1457 6.5 0.004 Error B 44 32185 12021 5907 3.5 0.004 CV % 29.5 32.3 19.2 7.7 10.9 *,** F is significant at a = 0.05 and 0.01, respectively. APPENDIX B 166 Table 1, Appendix B. Percent soil moisture at two soil depths during the growing cycle. East Lansing, MI. 1985 Soil depth in cm. Days After Planting 0-15 16-30 31-45 46-60 -5 2.5 8.0 9.7 10.0 2 20.7 21.0 17.0 17.6 10 13.5 12.5 10.6 11.3 17 17.3 16.2 12.9 13.6 24 9.6 14.2 10.9 11.9 31 11.5 13.3 10.7 9.3 38 10.4 10.2 10.2 10.2 48 8.7 8.1 ---- ---- 58 6.9 . 7.3 ---- ---- 70 9.1 7.2 ---- ---- 167 Table 2, Appendix 8. Percent soil moisture at two soil depths during the growing cycle. East Lansing, MI. 1986 Soil depth in cm. 0 - 30 31 - 60 Days after Planting Stressed Irrigated Stressed Irrigated -5 15.2 19.6 18.9 18.6 .10 19.1 17.5 20.4 18.7 25 12.2 14.1 19.3 20.2 32 11.0 11.4 15.0 19.7 39 11.0 12.0 15.4 16.1 46 10.2 12.5 17.4 17.0 53 '8.7 14.5 11.5 16.4 60 9.5 19.6 12.6 21.2 67 8.3 15.6 13.1 12.4 74 8.1 13.3 11.2 13.2 81 7.2 19.7 12.0 20.0 92 10.0 19.3 16.8 16.2 104 15.0 19.0 17.0 18.5 168 Table 3, Appendix B. Analyses of variance and significance of F-value of seed yield and some plant characteristics in 256 bean genotypes grown under drought-stress conditions in a rainout shelter. East Lansing, MI. 1985 Mean Squares Seed IKI Pads Source df yield TDM HI Score m. LWC LWRC Rep. 1 760 3719 0.0 0.47 338 43.2 4327 Treat. 255 4759** 21878** 0.01** 1.42** 4138** 5.0** 76** Unadj. Treat. 255 (1) 21725** 0.01** (1) (1) 4.9** 77** Adj. Error 255 2340 10022 0.005 0.66 2030 1.5 62 R08 ' Effec. 225 --- 9946 0.005 --- --- 1.3 61 Error CV % 30.2 26.2 16.1 29.1 1.4 26.9 26.4 *,** F is significant at a = (1) No lattice corrections 0.05 and 0.01, respectively. 169 Table 4, Appendix B. Climatic conditions recorded at the Botany Farm of Michigan State University during the growing season. East Lansing, MI. 1986 Temperature 0C --------------------------------- Rainfall Period maximum minimum. ,in mm May 1 - 10 20.7 5.8 24.2 11 - 20 20.5 9.6 134.0 21 - 31 21.8 11.5 35.0 June 1 - 10 24.1 10.6 168.2 11 - 20 24.4 12.8 239.7 21 - 30 25.5 13 7 40.0 July 1 - 10 28.0 15.6 55.5 11 - 20 29.2 18.7 72.0 21 - 31 29.0 16.3 3.2 August 1 - 10 25.3 14.5 43.2 11 - 20 26.0 12.2 12.7 21 - 31 23.4 9.4 184.7 September 1 - 10 23.0 7.5 24.2 11 - 20 20.2 8.4 74.7 21 - 30 25.1 15.7 314.2 170 Table 5, Appendix B. Analyses of variance and significance of F-value of seed yield and some plant characteristics in 121 bean genotypes grown under drought stress conditions in a rainout shelter. East Lansing, MI. 1986 Source Mean squares Seed Pods Seeds Seed df yield /m /pod wt. TDM HI IKI LWC LWRC Rep. Treat. Unadj. Treat. Adj 0 RC8 Error Effec. Error 1 50664 24496 .25 109 246633 .003 .20 15 251 120 1048262** 6297** .90 24** 33569** .023** .64** 8** 126** 120 982595** 6062** (1) (1) 32765** .022** (1) 7** 130** 120 474184 2850 .46 6.2 17405 .008 .33 4 63 100 383193 2721 --- --- 16308 .007 --- 4 59 CV % 41.4 31.4 11.7 12.2 32.1 23.1 18.9 2.3 12.2 ** F is significant at a = 0.01 (1) No lattice corrections 171 Table 6, Appendix B. Analysis of variance and significance of F-value of seed yield and some plant characteristics in 121 bean genotypes grown under irrigation. East Lansing, MI. 1986 Mean squares Seed Pods Seeds Seed Source df yield /m /pod wt. TDM HI IKI LWC LWRC Rep. 1 34694 45622 .51 1.8 106004 .001 1.8 134 1298 Treat. 120 24180** 19290** .62** 28** 87586** .03** .67** 5** 36 Unadj. Treat. 120 22543** 18440** .62** (l) (l) (l) .66** 5** 34 Adj. RCB 120 7777 7715 .35 6.4 38094 .02 .23 3 29 error Effec. 100 6885 7075 .34 --- --- --- .22 3 27 CV % 39.8 32.9 9.0 14.4 34.8 35.7 14.6 2.1 6.5 ** F is significant at a = 0.01 (1) No lattice corrections 172 Table 7, Appendix B. Average values of some agronomic traits of 256 bean genotypes grown under a rainout shelter. East Lansing, MI. 1985 Days to Family Pods IKI Yield TDM code Flower. Mat /m* score g/mi g/m* HI 48-13 52 110 156 2.5 172 400 .43 48-20 42 88 208 2.0 142 310 .46 48-7 57 110 144 3.0 166 462 .36 48-22 61 125 224 2.3 300 666 .44 48-4 46 96 200 2.8 158 384 .41 48-21 58 110 188 2.8 242 562 .43 48-12 43 100 200 2.3 . 202 446 .47 48-10 47 110 176 2.8 264 542 .48 31-95 55 105 106 2.5 122 300 .41 48-16 48 112 174 2.8 180 458 .40 48-17 47 100 184 2.0 140 306 .44 48-19 46 120 220 2.8 198 490 .40 48-15 52 100 252 3.8 256 502 .51 48-14 48 100 126 1.5 94 230 .41 48-8 46 88 172 2.5 160 334 .48 48-11 48 100 148 1.8 146 312 .45 31-50-2 58 110 164‘ 2.0 144 330 .45 31-26-1 56 100 158 2.0 126 286 .44 31- 29- 2 57 105 164 3.3 144 394 .37 31-33-1 51 93 172 1.8 188 372 .50 31- 48- 1 57 99 192 2.0 '156 410 .39 31-30-1 53 95 108 1.0 94 252 .37 31- 52- 2 47 88 164 2.5 128 274 .47 31- 31- 1 48 98 144 3.3 132 308 .42 31- 29- 1 62 105 138 3.8 126 386 .33 31- 54- 1 51 98 230 3.3 218 476 .46 31- 40- 1 53 99 218 2.5 186 400 .48 31- 51- 3 48 100 180 1.5 158 354 .44 31- 48- 2 43 110 172 2.5 156 394 .39 31- 33- 2 56 98 154 2.5 186 378 .49 31-23-1 58 110 114 3.5 114 258 .44 31-24-2 51 120 140 3.0 108 338 .29 48-105 40 100 162 3.0 184 428 .43 48-102 44 88 146 3.5 76 214 .35 48-107 46 92 172 2.3 174 348 .49 48-109 41 90 202 3.0 200 474 .41 48-99 48 110 116 3.8 130 286 .45 48-101 46 95 118 4.3 154 264 .57 48-104 51 112 150 2.8 166 352 .50 48-101 40 105 180 3.0 148 318 .47 48--93 46 85 192 2.5 166 370 .43 48-106 40 85 160 4.3 156 342 .45 48-106 47 88 232 1.3 146 316 .46 48-107 48 95 172 3.3 112 274 .43 173 Table 7, Appendix B (Cont.) 48-102 58 125 154 3.8 138 398 .35 48-110 46 88 150 1.5 140 262 .53 48-109 53 110 122 1.3 100 222 .45 48-111 41 84 202 2.8 186 360 .52 25-13-1 40 100 132 3.3 134 292 .46 25-23-1 48 100 160 4.3 184 432 .45 25-31-2 47 96 72 2.0 42 120 .35 25-16-1 37 89 154 2.0 188 336 .56 25-14-2 41 84 94 1.3 126 250 .59 25-42-1 44 100 152 3.5 162 374 .48 25-42-2 46 100 146 1.8 120 266 .45 25-1-2 52 94 70 3.5 48 108 .44 25-3-1 43 84 176 2.0 146 322 .45 25-40-2 40 95 150 3.8 132 306 .44 25-33-1 40 102 142 4.0 98 300 .35 25-31-1 54 98 30 1.0 22 80 .24 25-50-1 53 125 138 4.3 160 490 .32 25-40-1 37 88 138 3.8 132 232 .58 25-53-2 36 100 158 4.0 170 318 .53 25-1-1 43 88 140 3.8 150 302 .50 48-72 52 128 296 3.5 218 586 ‘ .37 48-99 46 110 162 4.0 140 368 .38 48-76 48 120 254 2.3 176 408 .42 48-75 47 98 218 4.0 190 454 .40 48-86 41 96 290 2.8 242 488 .50 48-71 48 110 166 4.0 234 602 .38 48-91 48 105 158 1.0 144 362 .40 48-74 50 120 138 2.5 170 416 .41 48-79 48 100 182 3.3 172 340 .53 48-100 53 112 108 3.5 114 348 .29 48-87 56 105 186 2.8 172 392 .46 48-94 53 110 268 3.3 278 604 .45 48-78 60 125 210 3.5 208 532 .38 48-90 57 110 212 3.8 186 474 .40 48-84 44 100 184 3.0 176 422 .42 48-98 40 90 200 2.5 202 396 .51 51-14-1 37 80 140 1.5 92 160 .59 51-19-1 37 90 152 3.0 108 252 .43 51-8-1 42 88 232 1.5 192 384 .50 38-18-2 60 110 208 1.8 312 732 .42 51-16-1 80 150 72 3.8 42 220 .19 51-5-3 40 83 236 1.3 180 360 .50 51-8-3 38 84 140 3.3 148 278 .54 51-16-2 57 104 76 2.5 84 190 .46 51-6-1 38 84 196 2.8 220 7 370 .59 51-8-2 38 84 154 3.3 194 382 .51 38-16-2 56 110 170 3.0 250 566 .44 51-18-1 67 120 214 3.0 202 462 .44 51-1-1 61 100 62 2.5 50 136 .31 51-1-2 41 98 226 3.5 192 478 .41 174 Table 7, Appendix B (Cont.) 38-18-1 58 110 144 3.0 192 458 38-16-1 60 110 184 1.8 248 588 51-32-1 44 110 280 3.0 276 602 51-40-1 58 120 118 2.8 144 354 51-46-1 40 83 112 2.5 100 234 51-23-3 44 86 144 2.3 122 282 51-47-1 43 120 214 3.0 224 478 51-29-2 43 100 178 4.3 142 386 51-24-2 43 84 168 1.0 170 332 51-22-2 44 88 220 1.8 144 324 51-22-1 42 86 240 3.3 148 414 51-44-1 38 80 236 1.3 114 252 51-26-1 80 150 108 3.3 64 270 51-19-2 64 120 52 2.5 38 116 51-20-1 46 86 194 2.3 138 292 51-29-3 38 110 234 4.0 196 452 51-37-1 45 88 132 2.3 118 238 51-23-1 38 88 204 2.8 124 264 31-58-2 56 110 154 2.8 132 336 31-60-1 54 125 122 4.3 158 444 31-85-2 58 120 112 4.0 132 572 31-94 58 110 72 2.5 78 290 31-87-1 55 120 142 4.3 102 296 31-93 58 110 116 2.8 128 318 31-62-2 42 80 138 2.5 112 288 31-63-1 52 98 118 2.3 114 290 31-65-1 45 98 174 3.5 188 480 31-67-1 46 96 110 4.0 120 294 31-59-1 69 120 94 4.0 70 554 31-56-4 46 100 142 1.3 146 290 31-58-1 58 110 196 2.0 152 368 31-54-2 52 107 200 3.8 166 382 31-61-1 44 98 176 2.3 120 300 31-90 52 90 170 2.0 150 326 31-17-1 48 120 218 4.0 172 512 31-4-1 57 102 162 4.3 160 436 31-5-1 57 102 206 3.0 160 390 51-49-1 38 120 294 2.8 212 586 31-8-1 48 96 140 2.0 118 342 31-11-1 52 96 156 2.3 182 396 31-6-2 47 88 196 3.5 164 344 51-53-1 45 120 124 2.5 96 328 31-7-1 50 96 198 4.0 184 404 31-20-1 43 92 116 2.8 116 286 31-22-1 46 85 174 3.3 134 300 31-13-1 52 100 132 3.5 140 338 31-18-2 52 94 192 1.3 176 404 31-20-3 52 98 128 1.5 106 296 31-7-2 47 91 242 2.3 216 474 31-3-1 52 99 154 3.8 154 388 175 Table 7, Appendix B (Cont.) 48'48 48 105 164 3.5 168 448 .34 48'64 48 120 136 3.0 130 368 .35 48'55 41 125 136 4.3 142 350 .40 48'63 48 110 136 4.0 140 382 .37 48'66 52 110 202 2.3 222 530 .42 48'60 55 105 170 3.8 146 362 .41 48'65 59 125 186 2.0 162 416 .39 48'61 43 100 194 2.0 168 442 .38 48'56 55 125 138 3.3 142 332 .45 48'70 48 110 162 3.8 168 400 .42 48'62 51 110 142 4.0 136 328 .42 48'53 47 115 168 3.5 132 312 .42 48'58 45 90 214 1.8 210 510 .42 48'49 48 125 150 2.8 140 354 .39 48'57 51 110 180 3.8 162 442 .36 48'52 41 85 190 2.5 164 300 .53 DGO'222 80 150 46 4.3 30 530 .06 8'26 46 85 202 1.8 194 416 .47 UW'23'61 46 80 184 1.8 148 320 .46 8'14 48 92 180 2.8 190 432 .43 8'28 46 92 176 3.3 162 368 .43 1213'2 80 150 84 4.0 58 486 .12 8'25 57 98 260 1.5 242 494 .49 II900'5'M'45 54 110 114 3.3 150 392 .38 UW 21'58 48 94 136 2.0 148 344 .42 UW 21'54 51 110 224 2.5 224 454 .49 LEF'Z-RB 44 88 140 3.8 180 376 .48 8'23 48 96 158 3.0 164 344 .47 8'27 50 90 114 3.3 134 310 .44 BAYO MAD 48 100 144 4.3 176 414 .43 N81064 42 88 114 3.5 100 232 .42 N81017 47 88 126 1.8 122 240 .51 9'18 51 110 212 2.8 228 510 .45 9'6 50 100 112 3.8 128 298 .44 9'15 47 90 166 3.5 188 418 .45 9'20 52 100 226 2.8 232 492 .47 9'8 52 110 228 2.3 212 490 .43 9'3 50 112 196 2.5 172 430 .41 9'12 47 96 186 2.0 178 424 .41 9'4 48 100 174 2.5 178 438 .41 48'94 55 110 122 3.5 108 278 .38 9'2 52 120 208 2.5 228 568 .40 9'10 52 110 218 3.5 258 552 .45 9'16 46 92 206 2.5 204 460 .45 9'17 48 110 160 3.8 198 482 .41 9'7 43 98 232 2.0 220 512 .43 9'11 47 105 158 2.0 234 440 .53 9'21 51 101 172 2.5 160 410 .39 48'43 61 125 192 3.0 244 556 .44 48'32 66 125 272 1.8 302 708 .43 Table 7, Appendix B (Cont.) 176 48'34 48'35 48'45 48'24 48'46 48'27 48'36 48'44 48'30 48'23 48'37 48'42 48'40 48'26 9'39 9'24 9'43 9'26 9'22 I I I I I I I I I I GHQHNHHWHNOHU‘HHNNl-‘QNMHNQ I PGHONPWN I I I I I I I I I OOHU‘I \IN N (D l wwwwcocommmmmmmmmmmmmmwomoomoo I «cocoon I 05>me I HHHN 39 63 45 52 36 58 47 46 41 51 48 52 48 44 58 46 46 53 48 48 44 51 51 58 52 45 56 48 55 43 45 47 48 50 47 48 44 50 53 47 52 48 51 48 46 51 55 57 51 52 105 125 85 125 110 110 100 105 105 125 105 110 100 88 110 89 100 120 105 96 110 110 110 110 110 110 105 ' 110 118 105 88 88 88 96 100 92 90 96 100 90 100 100 96 95 96 110 115 105 98 98 230 200 142 140 204 228 168 206 138 218 184 190 170 220 194 208 144 162 178 164 120 178 200 174 132 200 196 204 212 180 192 128 146 182 190 152 130 186 84 224 158 162 116 140 186 154 104 96 156 248 commencedwoocowmoulocowcowmcomcowcocomommomwwommommowmmmwwoooo O MHNNIPOONHHNNHHHQNHwNNwNQNMNIPNWPNNQHNQUQNPQNNNNNUNNP 204 164 122 116 190 182 150 178 142 214 192 176 148 220 206 184 192 170 222 160 118 212 174 184 132 170 204 236 190 206 208 122 138 174 188 158 114 184 56 214 122 130 116 128 216 164 120 124 190 196 484 366 266 322 406 454 416 398 288 486 476 338 370 454 574 450 378 408 380 408 320 490 464 410 296 440 472 532 472 464 378 270 272 386 418 382 222 346 180 434 300 312 278 296 454 346 274 272 416 476 .42 .46 .43 .36 .47 .40 .35 .44 .46 .44 .40 .53 .40 .48 .36 .41 .48 .41 .58 .39 .37 .43 .37 .45 .45 .38 .43 .45 .41 .44 .55 .45 .51 .45 .45 .42 .51 .53 .32 .50 .41 .42 .42 .43 .48 .50 .45 .48 .46 .41 177 Table 7, Appendix B (Cont.) 39'6'1 53 96 100 2.5 120 302 .41 39'11'1 52 100 160 2.8 246 508 .48 39'3'1 54 98 116 3.5 146 404 .36 38'10'1 55 98 222 1.3 198 438 .46 38'4'2 58 105 128 2.0 170 284 .60 38'7'1 59 105 180 3.3 198 484 .42 39'4'3 50 96 180 2.8 162 378 .43 38'4'1 56 98 120 3.5 144 328 .47 38'2'1 57 105 200 2.3 240 488 .49 39'11'2 53 105 132 3.8 158 288 .55 39'17'1 35 100 104 3.5 104 374 .30 39'13'1 52 101 126 3.5 168 398 .42 * Square meter . 178 Table 8, Appendix 8. Average values of some agronomic traits of 121 bean genotypes grown under drought stress in a rainout shelter. East Lansing. MI. 1986 L Seeds L W Days to Family IKI pods --------- TDM Yield W R ------- code score /m# /pod wt g/m# g/m* HI C C FL PM 48-21-M 3.5 153 7.3 16.2 444 134 .29 83 61 47 125 48-21-M 3.5 137 6.3 20.1 321 127 .36 82 57 56 120 48-21-1 2.3 190 5.8 20.1 476 166 .35 83 64 64 125 PLOT-13 3.0 211 5.5 21.1 364 153 .42 81 67 37 114 PLOT-15 3.3 118 5.7 26.6 259 124 .49 81 52 39 105 48-22-1 4.0 106 6.3 23.4 321 46 .11 84 72 68 125 48-10-M 3.8 205 6.4 20.0 532 199 .37 82 63 40 122 48-15-2 3.5 142 6.6 18.2 316 103 .33 80 64 56 122 48-14-1 2.5 210 5.5 21.3 460 170 .37 78 71 38 120 48-14-2 2.3 160 6.1 21.6 411 145 .35 81 69 38 98 48-86-1 3.3 123 5.5 18.7 229 98 .42 80 70 41 93 48-86-2 3.8 181 6.0 21.1 403 167 .41 80 71 36 90 48-71-M 3.3 189 6.3 20.7 542 166 .30 82 73 58 125 48-72-1 2.5 45 5.9 19.6 60 35 .57 81 62 58 105 48-94-1 2.5 163 6.3 17.0 371 118 .31 87 55 63 122 48-78-M 3.8 166 5.8 16.6 510 82 .15 80 65 69 125 48-78-1 4.3 163 6.3 15.3 419 68 .16 81 74 67 125 38-18-2-M 2.5 260 6.3 19.4 690 216 .31 83 69 56 125 38-18-2-1 2.8 184 5.7 19.7 482 165 .31 81 65 54 125 51-6-1-1 3.3 127 4.7 24.6 323 119 .37 79 54 42 92 38-16-2-M 3.3 160 6.9 21.2 461 150 .32 83 56 54 125 51-32-1-3 4.3 187 5.8 19.4 423 130 .32 82 59 54 125 51-32-1-2 2.3 147 6.0 20.8 281 136 .46 82 58 37 120 51-32-1-1 4.3 174 5.3 21.3 370 121 .32 85 60 53 125 51-47-1-1 3.5 152 5.1 21.3 332 111 .34 78 53 39 102 31-94-M 3.3 203 6.8 20.2 536 237 .44 81 55 52 100 51-49-1-1 3.8 135 4.9 26.4 180 93 .79 78 57 37 105 51-49-1-M 4.3 119 5.5 19.9 375 49 .12 82 53 51 102 31-7-2-1 3.8 145 5.7 19.5 308 128 .43 81 58 42 95 31-7-2-2 2.8 152 3.6 20.0 360 174 .51 79 58 53 118 48-66-M 2.8 152 6.5 21.1 418 119 .26 82 61 58 125 48-66-1 2.3 118 6.8 17.3 303 83 .27 81 55 66 118 48-58-1 3.0 127 6.0 17.3 321 90 .26 80 62 60 100 9-18-1-M 3.5 137 5.8 21.2 386 126 .32 82 64 58 110 9-18-1-1 3.0 102 5.8 19.8 287 65 .15 79 67 43 102 8-25-1 2.5 232 6.6 16.3 372 169 .45 77 71 56 110 8-25-2 2.8 240 6.3 20.7 587 259 .44 80 74 41 98 9-20-1 3.0 140 5.9 18.1 307 130 .40 81 65 52 108 9-20-2 2.5 118 5.4 17.0 386 105 .23 80 61 48 98 48-94B-1 3.3 121 6.1 16.5 224 77 .35 81 54 49 120 48-948-2 2.3 160 6.3 20.0 356 147 .40 84 64 58 107 \ “pi I I [a fit..— Table 8, Appendix B (Cont.) 179 I III I CDC): III-I :Itd I IHNQI—‘I—‘N H I 050:.)le :3 III Item I I II I y..- I I NHZHZHSHH HHWKONQNNSDIPNH I I I IHHUNNNNNNNNQ H0101 \‘II IHN 1'M 39'11'1'M PLOT'36 PLOT'36'1 8'42'M 41'25'1'M 8'40'M 41'18'1'M 41'10'2'M 41'12'1'M 41'39'M 8'46'M 45'5'2'M 8'30'M 8'4'M 41'48'1'M 41'18'3'M 41'52'2 41'6'1'M 8'35'M 8'47'M 41'32'1 41'48'1 41'37'1 8'29'M 41'14'1'M 41'36'3'M 41'58'3'H 41'39'3'M 48'109'M 'M 'M ammo:ocoommwommoowwooomwwmmwmomwoowomowmoomwwowwwowwo O O O O O O O O O O O O O O O O O O O O QNNN(a)N(.0NNUNUNNNNQUMQQQNNlewawNwFNwI-PNQQNUNWNUNQUUQ) O O 168 85 166 63 161 110 152 163 182 108 208 161 147 102 200 77 203 124 198 29 92 161 113 148 174 123 203 282 195 132 319 271 166 263 281 281 213 208 171 253 258 218 173 221 184 181 200 174 208 184 O‘MO‘O‘U‘WO‘U‘O‘O‘O‘U‘O‘U‘U‘QO‘U‘U‘O‘NU‘O‘“010503010(”O‘O‘O‘CBO‘O‘O‘O‘UIO‘O‘GO‘O‘MO‘O‘ONQUI C I C I O O O C O O C C O C O O O O O C C IPOJU'IOU'IIPNOI-‘UIONIUIQQOHNSDQONIIPHQU'IOUIU'IU'ICNHQOU'IQNIUIUIUPOUINIOQCDNG 20.7 19.5 15.2 15.3 18.0 21.7 19.1 21.6 14.9 18.5 19.3 16.2 16.5 22.2 18.4 18.8 22.9 19.6 21.5 22.0 23.6 23.3 23.5 26.4 25.2 21.6 20.3 23.1 21.8 20.6 24.3 20.7 22.7 23.7 18.5 21.8 22.6 18.9 23.5 22.2 18.0 20.3 13.8 22.3 23.2 18.6 18.4 18.1 17.6 24.6 384 305 368 152 400 234 377 389 403 310 500 397 353 352 509 190 560 317 461 81 311 629 239 368 511 284 492 552 432 339 893 546 427 534 761 570 463 510 371 551 565 403 295 437 386 394 406 468 311 430 145 93 112 42 115 98 129 163 146 111 148 112 113 94 198 74 257 116 197 27 87 177 62 129 232 97 217 304 187 138 398 249 179 266 260 288 229 195 185 240 246 175 113 203 180 170 167 135 128 250 .37 .29 .30 .28 .27 .41 .34 .41 .37 .34 .30 .27 .32 .27 .39 .40 .46 .41 .42 .34 .30 .28 .28 .35 .47 .34 .45 .55 .43 .41 .45 .45 .40 .50 .34 .51 .49 .38 .51 .42 .44 .43 .39 .46 .46 .44 .41 .30 .41 .52 80 8O 79 83 82 81 82 82 81 84 8O 84 81 84 82 81 80 77 80 84 80 82 82 80 80 81 81 77 78 83 79 79 81 82 80 80 78 78 82 77 78 81 73 81 81 78 79 82 81 83 50 72 64 73 64 59 64 59 65 65 67 65 54 68 57 60 60 62 74 56 50 63 64 53 40 71 6O 65 69 69 63 70 67 69 71 70 63 67 54 59 70 51 75 71 71 72 59 73 71 72 54 58 63 65 65 51 41 38 64 62 61 64 60 35 44 67 63 43 42 60 62 64 38 37 45 60 54 43 34 51 51 45 40 51 61 42 40 63 37 44 61 42 54 50 42 54 49 62 53 35 110 118 125 125 122 100 110 90 125 125 125 125 125 99 102 125 120 95 102 125 125 120 105 120 98 118 112 108 93 115 105 105 90 102 120 100 100 120 105 93 125 110 120 102 95 92 125 125 105 100 180 Table 8, Appendix B (Cont.) 51'1'1'M 2.5 227 4.7 23.6 439 198 .45 79 61 34 90 51'29'3'M 2.8 89 5.3 21.3 209 86 .44 77 55 45 100 48'19'1 2.3 139 6.6 19.3 328 134 .40 83 74 55 102 48'98'M 2.5 206 5.8 20.2 481 184 .38 81 64 39 87 9'24-M 2.5 219 6.0 21.5 566 206 .36 80 54 55 108 8'6'1 3.5 229 6.8 21.5 554 220 .38 81 77 52 116 48'105'1 2.3 185 7.0 23.0 445 203 .45 80 71 45 125 31'90'M 2.5 168 6.1 24.4 375 181 .48 81 57 55 118 51'8'2'M 3.0 144 4.9 26.3 289 116 .40 80 45 35 95 51'5'3'M 2.8 258 5.3 19.4 557 238 .43 81 53 44 90 8'17'M 3.0 231 6.4 21.6 504 234 .46 79 76 51 102 8'14'2 3.3 160 6.0 15.3 364 160 .42 82 68 44 102 48'48'M 3.0 176 5.5 20.0 367 127 .34 81 7O 50 108 48'34'1 3.8 98 5.3 25.4 201 87 .43 80 43 37 105 51'1'2'M 3.3 226 5.2 23.0 399 161 .38 81 60 39 100 48'36'M 2.3 165 6.0 22.7 390 132 .33 83 69 36 125 8'10'2 3.0 276 5.8 20.5 469 229 .47 80 57 43 100 8'26'M 2.3 155 5.6 22.1 497 198 .40 80 59 44 99 LEF'Z'RB 3.5 210 5.8 21.4 645 278 .42 79 74 42 98 II900'5'M 2.8 90 6.0 15.2 234 54 .24 82 67 63 125 N81064 3.3 124 5.5 20.7 197 108 .55 80 46 52 100 UW.23'61 3.3 166 5.3 21.0 268 134 .50 78 69 55 108 UW.21'54 3.0 181 6.5 16.1 329 137 .41 80 47 54 120 UW.21'58 2.8 160 7.1 18.9 361 158 .44 81 49 53 105 UGO-222 4.5 8 3.0 10.0 503 1 .01 84 54 90 125 BAYO MAD. 4.5 134 4.5 32.8 453 128 .29 81 58 63 125 1213'2 3.5 44 5.5 11.8 435 15 .03 85 56 90 125 N81017 3.5 148 5.9 22.0 292 152 .49 79 66 47 95 PINTO N.1 3.3 123 4.4 35.4 323 126 .39 82 69 33 85 39'11'1 2.5 61 5.0 15.4 149 34 .22 80 55 69 125 LWC = Leaf water content in %. LWRC = % of moisture lost after 24 hours. FL Days to flowering. PM Physiological maturity. * Square meter.. 181 Table 9, Appendix B. Average values of some agronomic traits of 121 bean genotypes grown under irrigation. East Lansing, MI. 1986 L Seeds L W Days to Entry IKI Pods -------- TDM Yield W R -------- code score lmfl pod wt g/m* g/m* HI C C FL MP 48-21-M 3.5 215 6.8 14.5 523 127 .25 84 79 63 125 48-21-M 2.5 448 6.8 13.0 937 331 .35 82 81 56 125 48-21-1 3.3 298 6.8 16.2 597 183 .31 85 83 60 125 PLOT-13 4.0 216 6.6 15.0 404 154 .38 83 82 50 118 PLOT-15 3.3 208 5.5 27.9 450 194 .43 82 72 40 107 48-22-1 4.0 119 6.8 13.6 366 40 .11 87 77 64 125 48-10-M 3.0 218 6.6 19.5 650 184 .31 85 74 56 125 48-15-2 4.0 224 7.0 14.5 458 145 .33 84 79 69 125 48-14-1 4.5 166 5.8 15.1 311 77 .25 84 84 55 125 48-14-2 3.3 166 5.7 22.5 320 163 .50 83 89 43 108 48-86-1 3.8 171 6.3 17.1 287 120 .42 86 74 47 118 48-86-2 3.0 235 5.8 16.2 417 176 .43 84 80 49 112 48-71-M 3.5 242 6.0 15.5 634 144 .23 83 80 62 115 48-72-1 4.0 103 6.0 15.7 228 52 .23 86 75 50 115 48-94-1 2.8 235 6.3 16.2 487 159 .33 84 82 62 125 48-78-M 4.3 292 6.8 12.4 677 114 .17 86 80 61 125 48-78-1 4.0 79 5.5 8.8’ 324 16 .05 83 77 65 125 38-18-2-M 4.0 292 7.1 18.1 746 217 .29 84 75 60 125 38-18-2-1 3.0 219 6.8 15.0 519 133 .23 87 75 56 125 51-6-1-1 3.5 234 6.8 20.5 479 192 .39 85 84 40 95 38-16-2-M 3.8 371 6.5 18.3 848 336 .39 83 77 62 115 51-32-1-3 4.0 237 6.7 13.4 457 117 .25 83 81 43 125 51-32-1-2 3.3 165 6.0 16.4 460 119 .22 84 82 39 125 51-32-1-1 4.0 168 5.5 15.1 568 82 .15 85 79 64 115 51-47-1-1 3.5 147 6.1 16.7 258 86 .34 84 80 43 95 31-94-M 2.0 239 7.2 17.6 590 227 .38 84 79 54 120 51-49-1-1 2.8 187 5.6 23.6 500 162 .33 82 82 47 95 51-49-1-M 3.5 252 6.0 16.4 251 139 .92 83 78 52 125 31-7-2-1 3.0 258 6.6 16.4 425 187 .41 85 77 64 112 31-7-2-2 2.5 310 7.1 17.0 654 280 .43 83 78 47 105 48-66-M 3.5 271 6.3 16.0 597 181 .30 83 75 58 125 48-66-1 2.5 218 6.7 15.0 538 173 .31 84 84 64 125 48-58-1 3.0 252 6.6 14.9 608 186 .32 83 84 59 125 9-18-1-M 3.8 116 6.7 16.1 338 61 .20 87 71 50 125 9-18-1-1 3.3 232 7.1 16.4 588 215 .37 87 72 47 125 8-25-1 2.3 303 7.0 13.8 541 225 .40 82 79 53 125 8-25-2 3.3 340 6.6 16.0 515 247 .47 85 81 50 115 9-20-1 2.8 348 6.6 16.3 859 254 .31 85 73 47 120 9-20-2 2.3 105 6.0 11.8 369 85 .23 83 74 52 118 48-94B-1 3.5 126 6.4 16.7 227 88 .40 82 75 43 118 48-948-2 3.3 258 6.9 14.9 465 179 .38 86 74 53 118 9-2-1 3.0 290 6.0 15.9 736 196 .26 83 81 61 125 9-10-1 2.3 150 7.4 16.9 368 139 .36 85 80 55 118 Table 9, Appendix B (Cont.) 182 III HQQH :IgIo I HI H 3 coco II 090) II P: I NHZI-AzI-IZHH HHSDNDQNINNOIPNH I I 3 I HHQNNNNNNNNQ’ Wr I-‘N 39'17'1'M 39'11'1'M PLOT'36 PLOT'36'1 8'42'M 41'25'1'M 8'40'M 41'18'1'M 41'10'2'H 41'12'1'M 41'39'M 8'46'M 45'5'2'M 8'30'M 8'4'M 41'48'1'M 41'18'3'M 41'52'2 41'6'1'M 8'35'M 8'47'M 41'32'1 41'48'1 41'37'1 8'29'M 41'14'1'M 41'36'3'M 41'58'3'M 41'39'3'M 48'109'M 51'1'1'M 51'29'3'M O O I O O O O wwwmoommmwowmomomowmwommwmmowwmwmwmwmwooowuwwmmwmo e e e o e e o e e e e e e o e e o e e o e e e e o e 194 171 384 248 240 168 176 197 337 140 302 265 277 244 313 150 218 323 144 416 166 258 436 155 271 292 245 198 508 395 281 281 423 342 316 386 287 318 486 192 226 287 321 286 342 374 323 227 84 137 O\O$O\05\I\IO\O\O‘\)O\O\ O O C QOU'INOOQIU‘O‘OOUI 0 m. \J wMQmOONWCflmU‘owOI-‘NWQQJHmome—‘QOQMOUNOQOQOF' O O O O O O O O O C C O O O O O O C O O 0901030\I\IO‘O‘IOVUIONNIQO‘\IO‘UO‘O‘Q‘DQO‘O‘QO‘O‘O‘O‘O‘JO‘O‘O‘O‘QO‘ O O O O O C C C O O C I O MO‘GUOU‘INOWONIQ O O O O C O O O C O O O O O O O O C O C O O O O C *wmIPOI-IQOQOWUUNNNOUG¢QHNHQW 22.8 1 1 400 624 847 520 506 351 425 524 834 307 702 607 624 464 762 354 459 913 335 114 474 567 975 268 686 608 422 392 315 711 579 586 879 642 557 829 505 695 767 419 348 573 577 559 712 650 499 549 201 382 159 125 285 236 148 172 110 133 281 68 288 172 239 175 384 124 203 246 166 501 110 184 498 96 210 311 216 138 503 331 288 275 413 356 267 279 223 309 465 87 153 203 238 165 280 276 201 218 65 139 .35 .24 .34 .44 .29 .48 .26 .25 .33 .22 .42 .29 .39 .35 .54 .37 .44 .28 .45 .45 .22 .32 .51 .34 .31 .51 .51 .35 .38 .47 .50 .47 .47 .54 .48 .31 .44 .44 .67 .21 .44 .36 .41 .34 .38 .68 .41 .40 .28 .37 85 83 84 83 85 83 83 86 84 87 85 83 86 88 84 84 86 86 84 86 85 84 84 83 84 83 84 84 85 85 84 87 84 83 84 83 84 85 83 86 83 85 84 85 83 83 83 81 84 84 71 76 83 79 81 74 79 77 78 78 82 84 76 76 77 80 80 75 64 68 76 81 77 76 83 74 81 75 80 83 78 79 78 8O 77 85 82 83 85 78 81 86 86 79 81 86 85 73 81 70 55 54 54 51 53 43 51 61 66 59 54 56 60 51 49 48 51 51 51 56 48 49 43 54 48 48 43 43 53 50 43 51 56 41 49 43 40 56 47 43 42 58 48 47 49 60 51 47 37 43 125 125 118 115 125 95 125 125 125 125 125 118 125 118 115 115 115 115 115 118 125 120 102 112 115 118 95 112 115 115 100 115 112 95 115 125 94 118 105 115 90 125 100 118 118 115 110 106 90 Table 9, Appendix B (Cont.) 183 48'19'1 48'98'M 9'24-M 8'6'1 48'105'1 31'90'H 51'8'2'M 51'5'3'M 8'17'M 8'14'2 48'48'M 48'34'1 51'1'2'M 48'36'M 8'10'2 8'26'M LEF'Z'RB II900'5'M N81064 UW.23'61 UW.21'54 UW.21'58 DGO'222 BAYO MAD. 1213'2 N81017 PINTO N.1 39'11'1'M QwNPFFWNQNQNNwNwNFNwUPNWQWQN O O C O O O O O . O O C C C C O . C C I O O o01010001ommmmmowmowomwmomwowwm 195 173 258 448 374 329 85 147 497 332 300 224 177 258 356 387 221 111 277 227 527 302 119 155 126 452 137 269 UPOOUO‘\JQO‘GO‘O‘O‘OQU‘O‘O‘VQU‘U‘O‘QGO‘O‘O‘ O O O O O O O O O O I O O C O O . mpmooocouu-dwwootocomqmcoooe'mowqmmo GQQUGONVQQQO‘NwUfimkaOPO‘ 369 348 575 916 764 757 225 258 995 712 632 399 338 458 780 804 478 302 521 396 1037 792 800 441 724 873 331 563 113 118 258 398 373 359 47 127 454 340 228 206 148 165 329 412 214 75 235 181 505 283 30 149 55 433 172 169 .31 .33 .45 .44 .48 .47 .20 .37 .46 .46 .36 .52 .41 .36 .51 .45 .23 .45 .48 .49 .37 .06 .34 .07 .49 .53 .30 87 84 82 86 83 86 83 83 85 84 85 78 85 86 86 82 84 85 86 81 83 83 89 83 84 82 86 82 81 78 75 80 84 77 82 75 79 84 77 74 81 81 83 80 84 79 78 82 85 80 85 81 87 84 85 85 49 57 44 52 47 44 43 43 54 46 50 47 44 41 52 43 43 61 43 42 56 58 90 52 90 47 37 63 112 115 98 112 115 110 115 90 125 118 125 108 104 118 125 103 115 125 103 97 112 125 125 120 125 95 90 125 LWC = Leaf water content in %. LWRC = % of water lost after 24 hours. FL PM * Square meter Days to flowering. Days to physiological maturity. APPENDIX C Table 1, Appendix C. 184 Combined analyses of variance for growth and yield related variables of 26 bean genotypes grown under drought stress and non-stress conditions at Durango (Mexico) and Michigan (USA). 1987 Significance of F-values ......................................... CV Variable LC G LG CG LCG % df 1 25 25 25 25 TDM 40 DAE *** NS ** ** *** NS NS 21.9 TDM 61 DAE *** *** *** *** * NS NS 20.9 Senes leaves *** *** NS ** *** *** 38* 32.9 LAI 40 DAE it! *t* *** it! t** NS NS 24,1 LAI 51 DAE *** 43* **3 ext xxx NS NS 24.1 CGR 40 DAE *** NS NS ** *** NS NS 22.0 CGR 61 DAE *** *** *** *** * NS NS 20.9 Leaf expans *t* It: It: 1*: *** *t 4* 39,3 Days flowering NS NS NS *** *** *** *** 3.3 Days maturity t** 88* NS *tt xxx 4*: *t* 2,7 Rep, phase **¥ *8* NS **¥ *** *** *** 6.0 Seed yield tit tit tea it: xxx *4 NS 23.4 Pods/m t** #18 st: tit tee NS NS 23,5 Seeds/pod ttt it: til *** *** NS *3 15,5 100 seeds wt. *** NS * *** *** * * 9.8 Seed protein2 *** NS *** *** NS ** NS 5.6 N seeds (g/m ) *** *** *** *** *** ** NS 22.6 TDM maturity xx: *** ex: ex: X** t** It 20,4 Harvest Index *tt xxx *** it: 8** ex: *8* 9,0 *,**, *** F is significant at P<0.05, 0.01, and 0.001, respectively. @ L= location, G= genotype 185 Table 2, Appendix C. Analyses of variance for growth and yield related variables of 26 bean genotypes grown under drought stress and rainfall conditions. Kellog Biological Station, Battle Creek, MI. 1987 Mean squares Variable "11;;5; """" e.g.; """ {If df 1 25 25 CV % TDM at 40 DAE 3096 3628* 2909 18.3 TDM at 61 DAE 77247 9959** 5954* 17.8 Senesced leaves 4675** 42h** 398** 30.1 LAI at 40 DAE 8.77* 1.37** 0.75 21.7 LAI at 61 DAE 6.12 2.48** 0.60 22.8 CGR at 40 DAE 1.93 2.27* 1.81 18.3 CGR at 61 DAE 20.7 2.68** 1.60* 17.9 Leaf expansion 132** 3.04** . 1.36** 35.3 Days to flowering 0.16 1799** 202** 3.0 Days to maturity 592** 418** 42** 2.8 Length reprod. phase 573** 185** 57** 6.3 Seed yield (g/mz) 945131** 5832** 3216* 22.4 No. pods/m2 294756** 8338** 1869 23.4 No. seeds/pod 100** 1.45** 0.54* 14.2 100 seeds weight 16.9 73.2** 8.73* 10.5 N seed yield (g/mz) 623** 7.2** 4.1** 22.0 TDM at maturity 16h0528** 22362** 13524** 18.7 Harvest Index 1.0** 0.018** 0.007** 11.6 *,** F-value significant at P<0.05 and 0.01, respectively. 186 Table 3, Appendix C. Analyses of variance for growth and yield related variables of 26 bean genotypes grown under drought stress and rainfall conditions. Durango, Mexico. 1987 Mean Squares Variable Regime Genotype C x G df 1 25 25 CV’% TDM at 40 DAE 37 3169** 40 30.1 TDM at 57 DAE 393462** 2883 2540 24.5 Senesced leaves 2632* 109** 96** 32.8 LAI at 40 DAE 0.011 1.12** 0.011 27.2 LAI at 57 DAE 71** O.54** 0.26 23.9 CGR at 40 DAE 0.023 1.98** 0.025 30.1 CGR at 57 DAE 121** 0.88 0.78 24.5 Leaf expansion 65** 0.28 0.17 43.6 Days to flowering 2.6 17.8** 2.96 3.7 Days to maturity 845** 38.8** 8.77** 2.5 Length rep. phase 754** 11.3** 8.80 5.6 Seed yield (g/mz) 86625** 1008** 454 19.9 No. pods/m2 110005** 1010** 491 16.8 No. seeds/pod 5.39* 1.24** 0.76* 16.7 100 seeds weight 2.64 97.2** 4.60 8.4 N in seeds (g/mz) 163** 1.81** 0.89 20.3 TDM at maturity 218997** 1636** 735 16.3 Harvest Index 0.46 .011** .002* 6.3 IKI score 2.92** 1.02** 0.43** 15.1 *.** F-value significant at P<0.05 and 0.01, respectively. 187 Table 4, Appendix C. Percent soil moisture at three soil depths during the growing season. Kellogg Biological Station, Battle Creek. MI. 1987 Soil depth in cm, Days 0 - 15 16 - 30 31 - 45 after ---; ----- ; ----------------------------------- planting I S I S I S 40 15.5 6.1 13.6 8.7 13.0 15.2 46 12.1 4.8 11.3 8.0 12.2 10.7 55 12.0 6.8 12.4 7.8 13.3 8.0 61 13.1 4.9 8.7 7.0 8.6 7.1 * I = Irrigated, S = Stressed. 188 20 ‘l A .Rainfed . ------ Stressed A FC 15 " . ’0 . - \\\\ ’ S /\/\ / . ,’ 10 - \a ’1 1' WP 5 .- \‘.-.——°-——o‘~~: I L , "‘r I I I I I I r I M o 20“ B PC I 15 4 S 1° " \ m» T \V/ U 5 - R 'hd‘ ' I I l I F I l l E 20 _! . C C , ////\\\ O 15 —I .\/,// \‘l.\ \ # FC \.\\ ./ . N o \o---. I, 10 \ ' \\\ I — \g/ T \/ WP E 5 _I N. T HY I F I I I I I T 10 20 30 40 50 6O 70 80 Z Days after planting Figure 1. Soil moisture content during the growing season at three different depths, A = 0-15 cm, B = 16-30 cm, C = 31-45 cm ( FC= field capacity, -0.03 bars and WP= wilting point, -15 bars). Durango, Mexico. 1987. 189 Table 5, Appendix C. Climatic conditions recorded at Kellogg Biological Station, during the growing season. Battle Creek, MI. 1987 Temperature 0C --------------------------- Rainfall Period maximum minimmn in am. May 16 - 20 24.7 10.8 22.0 * 21 - 25 22.3 10.1 00.0 26 - 31 30.8 18.2 2.0 * June 1 - S 25.9 13.8 3.0 * 6 - 10 26.8 12.3 0.0 +5 11 - 15 32.1 16.5 3.0 * 16 - 20 34.4 16.2 40.0 * 21 - 25 28.5 17.4 13.0 * 26 - 30 24.5 15.0 3.0 * July 1 - 5 26.9 14.6‘ 2.0 * 6 - 10 30.7 19.3 28.0 *8 11 - 15 26.0 15.6 21.0 16 - 20 30.6 16 1 8.0 21 - 25 32.0 . 20.2 19.0 26 - 31 30.5 15.4 11.0 August 1 - 5 31.3 18.9 12.0 6 - 10 27.4 16.3 23.0 11 - 15 30.2 17.2 7.0 16 - 20 27.5 14.9 55.0 *40 21 - 25 23.2 11.3 14.0 26 - 31 19.8 11.7 58.0 September 1 - 5 24.6 8.6 4.0 6 - 10 26.7 14.2 20.0 11 - 15 24.6 11.4 38.0 16 - 20 20.6 14.0 22.0 21 - 25 19.0 8.2 12.0 26 - 30 23.3 10.2 19.0 October 1 - 5 15.3 2.7 6.0 6 - 10 10.8 2.8 4.0 11 - 15 15.4 0.5 00.0 16 - 20 15.4 4.9 28.0 * Rainfall (+) plus irrigation under rain shelter. 190 Table 6, Appendix C. Climatic conditions recorded at "Los Llanos" Experimental Station, during the growing season. Durango, Mexico.1987 Temperature 0C --------------------------- Rainfall Period maximum minimum in am. June 1 - 10 25.7 10.9 33.5 * 11 - 20 29.8 13.7 0.1 * 20 - 30 30.4 14.2 28.5 * July 1 - 10 30.7 14.7 11.8 * 10 - 20 25.8 14.8 50.2 * 20 - 31 21.5 12.4 5.3 * August 1 - 10 24.5 13.3 52.5 *16.5 11 - 20 26.6 14.3 26.1 21 - 30 25.8 14.1 46.9 September 1 - 10 24.0 12.9 24.0 11 - 20 28.3 11.5 4.0 21 - 30 22.5 13.2 74.5 October 1 - 10 20.5 7.9 0.0 11 - 20 22.7 6.3 0.0 21 - 31 25.6 9.4 10.5 * Rainfall for stressed plots. 191 Table 7, Appendix C. Loading coefficients of 25 traits on five principal components for 26 bean genotypes grown under irrigation. Kellogg Biological Station, Battle Creek, MI.1987 Principal Components Trait 1 2 3 4 5 CGRAV 0.945 -0.259 -0.122 0.027 0.010 CGR2 0.927 0.306 -0.063 0.155 0.012 Poth 0.757 0.104 0.064 -0.170 0.169 LWR2 -0.707 -0.153 0.203 0.115 0.099 Flow -0.657 0.042 -0.064 0.270 -0.042 PhyMa -0.656 -0.351 0.193 0.131 0.073 NAR2 0.645 0.007 -0.304 -0.029 -0.008 CGR1 0.047 0.952 0.090 0.223 0.004 Stwtl 0.246 0.897 -0.198 0.175 -0.033 Leth -0.055 0.792 0.532 0.219 0.028 RePha -0.428 -0.555 0.339 -0.029 0.143 LWR1 -0.179 0.040 0.926 0.046 0.017 NAR1 0.288 -0.010 -0.695 -0.044 0.480 H1 0.212 0.032 0.576 0.281 0.493 Nitro 0.063 0.197 0.234 0.925 0.034 Yield 0.079 0.326 0.151 0.904 0.023 TDM -0.078 0.303 -0.200 0.821 -0.253 Seeds 0.013 0.070 0.051 0.204 -0.888 Sewt -0.114 0.128 -0.016 0.197 0.638 LeWt2 -0.095 0.206 0.151 0.236 0.080 StWtZ -0.020 0.195 -0.235 0.348 -0.342 WtSeLe -0.494 -0.057 0.210 -0.275 0.031 Pods 0.329 0.091 0.125 0.172 0.178 LER -0.393 -0.060 0.430 0.335 0.012 SePro -0.123 -0.488 0.172 -0.055 0.027 Variance % 21.976 16.821 10.366 12.108 7.262 Cumulative variance 21.976 38.797 49.193 61.271 68.533 192 Table 8, Appendix C. Loading coefficients of 25 traits on five principal components for 26 bean genotypes grown under drought stress. Kellogg Biological Station, Battle Creek, MI.1987 Principal Components Trait 1 2 3 4 5 LWR2 0.946 0.027 -0.077 -0.054 -0.150 PhyMa 0.942 -0.045 -0.016 -0.028 0.044 Flow 0.900 -0.132 -0.025 -0.138 -0.037 Poth -0.899 0.081 0.264 0.183 0.005 RePha 0.892 -0.008 -0.012 0.018 0.073 NAR2 -0.870 -0.000 -0.074 0.095 0.196 LeWt2 0.813 0.188 -0.007 0.455 -0.164 LER 0.739 -0.088 0.210 0.088 0.107 SeWt 0.655 -0.272 -0.204 0.042 0.116 StWtZ 0.630 0.224 -0.121 0.560 0.256 TDM 0.627 -0.094 0.172 0.237 0.330 Pods -0.589 0.243 0.551 0.249 0.122 HI -0.506 0.006 0.711 -0.025 -0.287 CGR1 -0.030 0.980. -0.078 0.060 0.026 Leth 0.005 0.934 0.040 0.111 -0.292 Stth -0.254 0.897 -0.175 0.070 0.182 Yield 0.027 -0.099 0.935 0.212 0.011 Nitro -0.074 -0.177 0.905 0.153 0.064 Seeds 0.310 -0.221 0.570 -0.215 -0.275 CGRZ -0.019 -0.019 0.146 0.921 -0.028 CGRAV 0.004 -0.426 0.200 0.856 -0.047 SePro -0.248 -0.107 -0.030 0.539 -0.184 LWR1 0.130 0.216 0.254 0.163 -0.797 VtSeLe 0.105 0.450 0.260 0.044 0.677 NAR1 -0.020 -0.298 -0.470 -0.134 0.440 Variance % 30.518 17.290 12.039 13.407 7.103 Cumulative variance 30.518 47.808 60.847 72.254 81.357 Table 9, Appendix C. 193 Loading of 25 traits on five principal components for 22 bean genotypes grown under irrigation. Kellogg Biological Station, Battle Creek, MI.1987 Principal Components Trait 1 2 3 4 5 CGR1 0.957 0.100 0.190 0.033 0.134 Stth 0.915 -0.056 0.188 -0.053 -0.130 RePha -0.805 -0.212 -0.201 0.059 0.319 Leth 0.789 0.116 0.163 0.120 0.547 PhyMa -0.732 0.477 -0.001 -0.159 0.197 SePro -0.517 -0.106 -0.020 0.006 0.198 LWR2 -0.005 0.924 0.003 -0.003 0.167 LeWt2 0.210 0.879 0.169 0.008 0.171 NAR2 0.018 -0.870 0.057 0.037 -0.283 Flow 0.180 0.828 0.258 -0.261 -0.181 Poth 0.032 -0.784 -0.186 0.326 0.073 StWtZ 0.176 0.757 0.366 -0.423 -0.141 UtSeLe -0.127 0.660 -0.347 0.040 0.131 Nitro 0.203 0.044 0.917 0.095 0.197 Yield 0.318 0.138 0.865 0.076 0.113 TDM 0.363 0.250 0.795 -0.233 -0.144 Seeds -0.015. 0.147 0.257 -0.833 0.147 Pods 0.145 -0.275 0.284 0.798 0.017 HI -0.079 -0.145 0.270 0.602 0.431 LWR1 -0.001 0.101 -0.010 0.143 0.930 NAR1 0.013 -0.423 -0.055 0.263 -0.764 LER -0.085 -0.067 0.391 -0.451 0.542 SeWt 0.099 0.055 0.140 0.068 0.036 CGRZ 0.330 -0.216 0.269 0.051 0.010 CGRAV -0.480 -0.296 0.104 0.022 -0.103 Variance x 20.870 20.710 12.348 9.147 10.290 Cumulative variance 20.870 41.580 53.928 63.075 73.365 194 Table 10, Appendix C. Loading of 25 traits on five principal components for 22 bean genotypes grown under drought stress. Kellogg Biological Station, Battle Creek, MI.1987 Principal Components Trait 1 2 3 4 5 LWR2 -0.924 -0.124 -0.046 0.240 0.022 Poth 0.891 -0.059 0.228 -0.001 -0.262 PhyMa -0.874 0.103 0.120 0.010 0.271 NAR2 0.848 0.063 0.147 -0.249 0.186 Flow -0.808 0.127 -0.061 0.029 0.064 RePha -0.746 0.069 0.205 -0.003 0.343 LeWt2 -0.741 -0.274 0.508 0.218 -0.059 SeWt -0.737 0.270 0.079 -0.158 -0.189 Pods 0.703 -0.206 0.159 0.029 -0.529 StWtZ -0.630 -0.230 0.563 -0.330 0.117 MI 0.519 0.028 0.020 0.237 -0.562 TDM -0.507 0.097 0.152 -0.209 -0.372 CGR1 0.020 -0.980 0.040 -0.038 0.066 Stth 0.135 -0.945 0.055 -0.220 0.075 Leth 0.037 -0.924 0.107 0.306 0.025 CGR2 -0.042 -0.315 0.917 0.009 -0.190 CGRAV -0.055 0.412 0.852 0.036 -0.230 SePro 0.327 0.149 0.619 0.304 0.176 LWR1 0.036 -0.143 0.202 0.848 -0.053 WtSeLe 0.215 -0.365 0.054 -0.753 0.010 NAR1 0.193 0.271 0.021 -0.533 0.514 Yield 0.130 0.142 0.206 0.041 -0.929 Nitro 0.280 0.244 0.161 0.008 -0.848 Seeds -0.204 0.214 -0.113 0.147 -0.316 LER -0.455 0.089 0.342 -0.198 0.015 Variance % 27.449 17.509 14.140 8.418 11.873 Cumulative variance 27.449 44.958 59.098 67.516 79.389 195 Table 11, Appendix C. Loading coefficients of 28 traits on five principal components for 26 bean genotypes grown under rainfall conditions. Durango, Mexico. 1987 Principal Components Trait 1 2 3 4 5 StWti 0.983 0.039 0.103 -0.005 -0.061 CGR1 0.971 0.049 0.148 -0.012 -0.050 Leth 0.956 0.055 0.176 -0.017 -0.042 LURl -0.889 -0.077 -0.076 -0.040 0.049 WtSeLe 0.609 -0.013 -0.147 0.246 0.084 CGR2 0.120 0.960 0.021 0.051 -0.226 StWtZ 0.238 0.913 -0.054 0.067 0.030 LeWt2 0.104 0.875 0.117 0.052 0.221 CGRAV -0.457 0.859 -0.069 0.037 -0.168 Poth -0.048 0.582 0.005 0.007 -0.745 Yield -0.157 0.017 -0.976 -0.075 -0.071 Nitro -0.189 0.034 -0.963 -0.078 -0.035 TDM -0.040 -0.102 -0.915 0.189 0.109 H1 -0.209 0.173 -0.686 -0.460 -0.248 SeWt -0.240 -0.176 -0.528 0.056 -0.215 PhyMa 0.043 0.109 0.226 0.912 0.234 RePha -0.059 0.183 -0.205 0.761 -0.129 LWRZ -0.181 -0.272 0.080 0.063 0.838 LER -0.106 0.046 0.019 0.040 0.701 LWRC 0.181 0.117 0.025 0.115 0.531 SePro -0.136 -0.032 0.037 -0.094 0.116 NAR1 0.191 -0.278 0.110 -0.079 0.125 NAR2 0.053 0.169 -0.016 -0.147 -0.470 IKI -0.087 -0.036 0.001 0.150 0.017 LWC -0.117 0.217 0.370 0.223 0.477 Seeds -0.241 0.447 -0.196 -0.465 0.032 Pods 0.331 -0.144 -0.423 0.094 0.086 Flow 0.112 -0.030 0.492 0.479 0.429 Variance %. 19.135 17.057 14.366 7.777 10.330 Cumulative variance 19.135 36.192 50.558 58.235 68.565 196 Table 12, Appendix C. Loading coefficients of 28 traits on five principal components for 26 bean genotypes grown under drought stress conditions. Durango, Mexico. 1987 Principal Components Trait 1 2 3 4 5 CGR1 0.978 0.040 0.134 0.019 0.030 Stth 0.976 0.023 0.151 0.034 0.082 Leth 0.971 0.050 0.122 0.009 -0.003 LURl -0.840 0.066 -0.050 0.006 -0.241 CGRAV -0.676 0.046 0.672 -0.035 0.051 TDM 0.008 -0.945 -0.134 -0.100 0.018 Yield -0.007 -.0934 -0.175 0.247 -0.019 Nitro 0.056 -0.902 -0.185 0.259 0.030 SePro 0.238 0.666 -0.005 -0.035 0.211 HI -0.021 -0.531 -0.082 0.717 -0.094 Pods 0.087 -0.502 0.133 0.080 0.522 LeWt2 0.103 0.051 0.963 -0.031 0.085 CGR2 0.201 0.100 0.913 -0.012 0.085 Stwtz 0.140 0.253 0.897 0.093 -0.021 PhyMa 0.001, 0.030 -0.010 -0.980 -0.006 Flow 0.097 0.062 0.138 -0.870 0.080 RePha -0.091 -0.008 -0.150 -O.849 -0.088 IKI -0.201 0.317 -0.091 -0.579 0.114 LWRC -0.102 -0.025 0.036 0.114 -0.890 LWC -0.148 -0.005 -0.214 0.073 -0.775 LER 0.008 -0.352 -0.044 -0.357 -0.610 LWR2 -0.205 -0.068 0.001 0.035 -0.020 Podwt 0.228 -0.029 0.479 -0.075 0.124 WtSeLe -0.040 0.147 0.251 0.032 -0.398 NAR2 0.037 0.152 0.005 -0.017 0.051 SeWt -0.203 -0.480 -0.106 0.006 -0.182 Seeds 0.161 -0.109 -0.139 0.289 -0.198 Variance % 18.075 13.964 14.991 12.279 8.292 Cumulative variance 18.075 32.039 47.037 59.316 67.608