.- . fl»: W” . C . .. I 0-— - fifi mdifl $3?" ifigg‘fi ' f1}. - . I . _ ._ -';:~”". 5". Dr. ’ ,tr. . «12:... ‘- u. (M 4 .4. w I-A “-0 nvv‘ .‘J.’ ..o— ‘ "J. ' .- ' . . :1 Q 6 t D ' . . '1' .044pr . ..4- -;::::Y. "t on r...” ‘10.“. . .. ‘l‘ 0‘... I I ___..~.... .Ju “:13? °~ “ ,_ ”o ‘ ~ ' -..-I ~ _" I :“1‘ If: .... , ,.. ‘.L‘.‘f:‘ I ' I I “.' "‘... .. 3o .2 ": ‘. ---c‘-“ “' ' fl .‘ . . . _vi . It . '9 f J. ‘F - O .o ' ”.o» 00. >‘-‘ _, ..u- l - ‘I _ ’, .4 ‘ - Ax. - —- .0 a ""'_.."rd.‘l-n - o .F‘DIIQ IIJI‘I‘lfi'f‘“ W J r', ,. m ”‘13." ~ .‘v '-a I': WNW; " l. .0 II III} .6“ "n21 1”Moat" 13313 4} mil {3‘ LIBRARY Michigan State University This is to certify that the thesis entitled A Study on Seed Filling and Dry Matter Partitioning Characteristics and Their Combining Ability Effects and Relationship to Yield among Dry Beans with Differing Growth Habits, Architectures and Maturities presented by Oscar Mario Paredes Carcamo has been accepted towards fulfillment of the requirements for M.S. degree in Plant Breeding and Genetics Date August 1, 1986 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution 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. A STUDY OF SEED FILLING AND DRY MATTER.PARTITIONING CHARACTERISTICS AND THEIR COMBINING ABILITY EFFECTS AND RELATIONSHIP TO YIELD AMONG DRY BEANS WITH DIFFERING GROWTH HABITS, ARCHITECTURES AND MATURITIES BY OSCAR MARIO PAREDES CARCAMO A THESIS Submited to Michigan State University in partial fulfilment of the requirements for the degree of MASTER OF SCIENCE Department of Crop and Soil Sciences Plant Breeding and Genetics 1986 -c-O ‘ ' I‘ I rub.- n‘noa .- .- AMWQ: I--v-. . v‘ ‘ -..~ 3 (u (I) ~VA._‘ c u"’. . . _"~— I. vd.. . -ha.‘~ ‘v Vuu . : I u" “ "w “ Os ~ 4 y. 'v ‘u. c ‘, F. b ‘- ‘ a.» _ \- s.‘ kn~§ u‘: in'A 'w_.- ‘V ‘- s — . '._-.‘1 - v ‘- ~ -.~ A 'u \: \- ‘ ‘ hh‘ ‘\ h. a - V‘. a H ‘A .1"-x )‘ QV ‘ ABSTRACT A STUDY OF SEED FILLING AND DRY MATTER PARTITIONING CHARACTERISTICS AND TREIR COMBINING ABILITY EFFECTS AND RELATIONSHIP TO YIELD AMONG DRY BEANS WITH DIFFERING GROWTH HABITS, ARCHITECTURES AND MATURITIES BY Oscar Mario Paredes Carcamo The present study was undertaken to evaluate whether variation in morphological traits among 12 dry bean genotypes was related to yield and estimate the combining ability and genetic effects of these traits. The genotypes were characterized by a Type I, II and III growth habit. Type I and II represented by two different architectural forms, traditional and architypes. Dry matter accumulation and seed filling parameters were evaluated. Significant variability existed among and within growth habits for some of the traits study. The combining ability study indicated that GCA is the major component controlling the expression of these traits and rapid progress could be made using breeding systems to concentrate primarily additive genes present in this population. Reciprocal effects were detected for some of the traits studied which indicated that the selection of a parents should be considered in a breeding program for these traits. Determinate genotypes had fewer days to 50% flowering, days to physiological maturity and seed weight compared with . 9‘!!- QB .uv 0“ n Ah! “I vbuou . «n! I . n v-JU Oo\ nn..-a. .- Uv' vv 0!. .34., a A G n.‘ .‘Vs H “\- ua r . ~.,”‘ 5‘ +- ‘43 ‘: .‘ ‘ Q Oscar M. Paredes the indeterminatecmuxa Plants of determinate growth habit accumulate just much total dry matter than indeterminate types. Type II genotypes outyielded Type III. This difference could be explained due to the higher number of pods per m2 and number per seed per pods. The lack of relationship between HI and seed yield suggested that a high HI itself is not a good selection criterion for a high yielding cultivar. Late maturing genotypes outyielded earlier ones. This suggested that a medium to full season dry bean cultivars should be an important criterion to insure a high final production. Grain yield was positively correlated with number of pods per plant, number of seeds per pod, biological yield, days to 50% flowering, physiological maturity and linear filling rate. Type I tended to have a shorter total filling duration, linear filling duration and linear filling rate compared with Type I and III. Architypes showed a similar seed filling duration than the traditional form and late maturity genotypes tended to have a longer total filling duration and seed filling duration than the earlier ones. Linear filling duration was positively correlated with yield, number of seeds per pod, biological yield and physiological maturity and total filling duration was positively correlated with 50% flowering and physiological Inaturity. To my parents, and all other members of my family ii Ina ~ OI-v o ~v-I. .. .- A - M-‘a-uu§. ... ,._‘_“ .- “ “b 5‘ ‘ U ' o to» u"“~¢:.' "v".“ u g \"‘ U-.. I ‘vr.,.'_ s.‘ n ‘ - 0 t..‘- a: , .- u‘_"“ - a . .- .,.. ‘ . ..'§.‘ - C1. AKNOWLEDGMENT The culmination of this study comes as a result of invaluable assistance and support rendered by many persons. My sincere gratitude and appreciation.to Dr.CLLn Hosfield, my major professor, for providing the opportunity of carry out this thesis project. His kindness, guidance and encouragement were invaluable and greatly appreciated. Sincere appreciation is extended to the members of my Committee, Dr. A” Iezzoni, Dr. MMA. Uebersax for their helpful suggestions when reviewing this manuscript. Special thanks to Dr. J.D. Kelly for his valuable advice and criticism during the design of this study, as well as for his support as a Committee member. I am thankful to Dr. C. Cress for his statistical advice in the analysis and interpretation of this results. I am also deeply greatful to Dr. TIL Islieb for his help, patience and time devoted to the genetic analysis of this thesis. To Mrs. Homyra Habibi and Ms. Sallie Wellso, thanks for their kind field assistance and data collection. Special thanks ischueto my wonderful family, wife and children for their love, patience, and moral support during the course of this study. Finally, I would like to thank the Instituto de iii Investigaciones Agropecuarias (INIA) of Chile and the US. Department of Agriculture for the financial support enabling me to undertake this graduate study. iv con- ~Q". O‘u—on .I‘.:A n"\ a ‘ - a- - Bee... ('2‘ . "“-OP ‘ he, - ‘ \ ":1: TABLE OF CONTENTS LIST OF TABLES INTRODUCTION CHAPTERl. CHAPTER 2 . CHAPTERB. THE EFFECT OF GROWTH HABIT,ARCHITECTURE AND MATURITY ON DRY MATTER ACCUMULATION, PARTITIONING EFFICIENCY AND THEIR INTERRELATONSHIPS WITH YIELD IN DRY BEANS Abstract Introduction Literature Review Materials and Methods Results Discussion Bibliography VARIABILITYINSEEDFTLLINGCHARACTERIS- TICS AND THEIR INTERRELATIONSHIPS WITH YIELD IN DRY BEANS WITH DIFFERING GROWTH HABITS,ARCHITECTURES AND MATURITIES Abstract Introduction Literature Review Materials and Methods Results Discussion Bibliography COMBINING ABILITY EFFECTS FOR YIELD, COMPONENTS OF YIELD AND MORPHOLOGICAL TRAITS IN A DIALLEL CROSS OF DRY BEANS WITH DIFFERING GROWTH HABITS AND MORPHOLOGICAL CHARACTERISTICS Abstract Introduction Literature Review Materials and Methods V Page vii 28 34 55 69 74 76 78 84 87 97 102 105 107 109 116 " “nu-g- un" , v . U a. o ~A an M" fin...- . — hilra‘ . ~ -". ‘ ~ I! H‘u~‘ CHAPTER 3 (continued) CHAPTER4. Results Discussion Bibliography COMBINING ABILITY EFFECTS FOR SEED FILLING TRAITS IN A DIALLEL CROSS OF DRY BEANS WITH DIFFERING GROWTH HABITS AND MORPHOLOGICAL CHARACTERISTICS Abstract Introduction Literature Review Materials and Methods Results Discussion Bibliography SUMMARY AND CONCLUSIONS APPENDICES BIBLIOGRAPHY Appendix A vi Page 120 137 140 143 145 147 152 152 160 163 166 172 175 V‘s- '1¢ e by n— v... 5‘ UL .- Ayn, C C‘" .0. .C \m .3 a - 2w 2 .c .t a. a u: 2 .c I .1 .1 Z .- A: Ow v. .n. a» a a» C.» re 0 ». Av «C ‘5“ 2n .3 C» a: .3 c. .. 0 .Va Vu Au 3. I C 9» 4n r. s 2.. s a a . ~ 5 . a «I. IV. .Kd fhh § I. a...“ LIST OF TABLES CHAPTER 1 JnGrowth habit, architectural form, loo-seed weight, commercial class and.days to physiological maturity of 12 dry bean genotypes grown in 1984 24Mean yield and components of yield of 12 dry bean genotypes grown in 1984 3.Mean squares from single degree of freedom orthogonal comparisons of yield and components of yield for 12 dry bean genotypes grown in 1984 4.Mean biological yield, days to 50% flowering and days to physiological maturity of 12 dry bean genotypes grown in 1984 5.Mean squares from single degree of freedom orthogonal comparisons for biological yield, days to 50% flowering and physiological maturity’of 12 dry bean genotypes grown in 1984 6.Total dry weight of 12 dry bean genotypes at four sampling dates after 50% flowering and grown in 1984 74Relative growth rate of 12 dry bean genotypes at four sampling dates after 50% flowering and grown in 1984 8.Leaf area ratio of 12 dry bean genotypes at four sampling dates after 50% flowering and grown in 1984 9.Leaf area per cm2 of 12 dry bean genotypes at four sampling dates after 50% flowering and grown in 1984 10.Leaf area index of 12 dry bean genotypes at four sampling dates after 50% flowering and grown in 1984 LLHarvest index, seed yield efficiency and biological yield efficiency of 12 dry bean genotypes and grown in 1984 vii Page 29 35 36 4O 41 43 44 45 46 47 49 CHAPTER I (Continued) 12.Mean squares from single degree of freedom orthogonal comparisons of harvest index, seed yield efficiency and biological yield efficiency of 12 dry bean genotypes grown in 1984 13.Simp1e correlation coefficients between yield, yield components, biological yield, days to 50% flowering and physiological maturity of 12 dry bean genotypes grown in 1984 IAuSimple correlation coefficients between total dry weight,bdologica1 yield, leaf area, leaf area index,and relative growth rate of dry beans at four sampling dates after 50% flowering grown in 1984 1548iologica1 yield, stem dry weight and pod wall dry wall weight of 12 dry bean genotypes and grown in 1984 16.Leaf dry weight of 12 dry bean genotypes atfour sampling dates after 50% flowering and grown in 1984 17.Stem dry weight, pod dry weight (pod wall and seed dry dry weight) and stem pod dry weight ratio for 12 dry bean genotypes grown in 1984 18.Pod wall dry weight at four sampling dates after 50% flowering of 12 genotypes grown in 1984 CHAPTER 2. JuRegression and determination coefficients from the cubic polynonial regression equation for seed dry weight on days after'50% flowering of 12 dry beans genotypes and grown in 1984 2.Total filling duration,linear fillingduration, linear filling rate, and mean filling rate for different bean growth habits,architectural forms and maturity groups 3.Tota1 filling duration, linear filling rate, and mean filling rate for 12 dry bean genotypes grown in 1984 viii Page 50 52 54 63 65 66 68 89 9O 91 uh -~ f". “at 00 e . . n4 14..q. 3. A: a: v. A. c- .n.. A: n. v... m: .1. .e. Mm 1.. a: . . ¢. v . . 4‘ h e t as . Q q . o A n I. AB A U ab 5 2h \ a a: s s so Q on. .01 l A\w ». 2 Av .1. m...“ 1.» av C4 \um Y. a c. mum HI; he I. . 1am vs C a . a T. c» h». 4 2 Th my. us. «3 P» s C D. It. :4 n.» a . «I. 1.. I? 2.. C. ~ w CHAPTER 2 (continued) Page 4.Weight of loo-seeds and days from 50% flowering to physiological maturity for 12 dry bean genotypes grown in 1984 93 5.Simple correlation coefficients between seed filling parameters and yield, and components of yield for 12 dry bean genotypes grown in 1984 95 CHAPTER 3 1.Growth habits,architectural forms,seed size, commercial class and days to maturity of six genotypes grown at two locations in 1985 117 2.Ana1ysis of variance for GCA, SCA and reciprocal effects for components of yield and seed yield for a six parent diallel cross grown at two locations in 1985 121 3.Estimates of GCA effects for components of yield and seed yield measured on F1 progeny of a six parent diallel cross grown at two locations in 1985 122 4.Analysis of variance for GCA, SCA and reciprocal effects of morphological traits for a six parent diallel cross grown at two locations in 1985 124 5. Estimates of GCA effects for morphological traits measured on F1 progeny of a six parent diallel cross grown at two locations in 1985 126 6.Estimates of SCA effects for components of yield and seed yield measured on F progeny of a six parent diallel cross grown at two locations in 1985 127 7hEstimates of SCA effects for morphological traits and seed and seed yield measured on F1 progeny of a six parent diallel cross grown at two locations in 1985 130 thAnalysis of variance and estimates of location x combining ability interactions for components of yield and seed yield of a six parent diallel cross grown at two locations in 1985 131 ix \ no. 1 C” U. 4’? 5‘. w 10 r“ '3‘ be 5.. v. un~ an! RU 0.. CHAPTER 3 (Continued) 9.Analysis of variance and estimates of location x combining ability interactions for morphological traits of a six parent diallel cross grown at two locations in 1985 CHAPTER 4 lnAnalysis of variance for GCA,SCA and reciprocal effects for total andlinear filling duration measured on F1 progeny of a six parent diallel cross grown at two locations in 1985 2.Estimates of GCA effects for TFD, LFD and yield measured on F1 progeny of a six parent diallel cross grown at two locations in 1985 3.Analysis of variance for GCA,SCA and reciprocal effects for linear, mean filling rate and predicted maximum weight measured on F1 progeny of a six parent diallel cross grown at two locations in 1985 4.Estimates ofGCA effects for linear filling duration, mean filling rate and measured on F1 progeny of a six parent diallel cross grown at two locations.in 1985 Page 136 153 155 156 151 INTRODUCTION In developed countries of the world, production in excess of consumers' needs occurs; hence, oversupply forms the basis for export and a source of foreing exchange. This is not the case in the poorer countries of the world. In these countries numerous constrains hamper a food production sufficient to meet consumers' needs. The first task of agriculture in any country of the world is to produce enough food to meet the needs of the population. Cereals and food Legumes are the most important foods for many developing countries in the world. Dry seeds of commom beans (Phaseolus vulgaris L.) plays an important nutritional role in many Latin American and African countries. The social, economic and cultural well being of inhabitants of many countries throughout the world are integrated tighly with the production of dry beans. Yields of dry beans have been static for many years, whereas notable yield increases have been realized in several cereal crops, particularly wheat and rice ( 73 ). Increased cereal yields were largely atributable to the modification of plant morphology ( 73 ) and an increased use of fertilizer ( 21). Traditionally, advances in yield in crops had been brought about largely through pragmatic advances in breeding 1 7'" p ..I in 4 .4 ’2“ _ ova... A fl ’ on. P ‘ 0-.- . q 9" "‘. v: ‘ s.- v. _. 5“:- (I’ .0 I t V. n. ‘c (I! 2 and crop management. However, Donald ( 29 ) espoused the concept of ideotype breeding plant breeders have attempted to design the architecture of Food crops to optimize the production and partitioning of assimilate in time and space. Dry matter production depends upon interception of solar radiation, uptake of carbon dioxide and water. However, photosynthetic partitioning plays an important role in Hue final seed yield, therefore a better understanding of growth and development of a plant and their interrelationships with yield is necessary. Since at least 90% of the dry matter of the plant is the product of photosynthesis, plant growth analysis as defined in dry weight terms essentially measures the performance of the plant as a producer of photosynthate ( 44 ). Ideally, this information could be used by plant breeders to break "yield barriers" in new cultivars. Adams ( 3 ) proposed a model for dry beans wich could maximize enviromental inputs to enhance yield. Adams ( 3 ) called his bean model and architype. Other studies ( 112, 113 ) in dry beans have confirmed that genetic variation existed among cultivars for morphological traits and physiological processes determining yield. Bunting ( 13 ) pointed out that most of the improvements achieved in plant physiology have come from research directed to components that influence the rate . una+~ n‘v ‘ 5=uhcvl I u Ana-I A “auaavuunc dw'b . ‘ V --n A'V ‘1 :n-u-U! U o A . . '- .‘nqiq F .- .VOI'.vla~ ’ : "“V-I- ~04 - A °|Mvotnu .-~ .28 C "--6L-‘, ‘ :...:. I,‘ ~:Pfl.ltn; 'uu'.'~ ' a. ' N ... 'R‘an" bye.~.u.“‘ d 3 rather the duration of accumulation of yield and more attention needs to be paid to the physiology of seed development. In this regard, several studies have been employed in cereals ( 3O ) and soybeans ( 36, 43, 46 ) to determine whether genetic variation for yield was associated relationship with physiological traits. However, this information is limited for dry beans. The objective of the present researh was to: l)evaluate whether variation in morphological traits among 12 been genotypes was related to seed yield and 2) estimate the combining ability and genetic effects of the traits. CHAPTER 1 THE EFFECT OF GROWTH HABIT, ARCHITECTURE, AND MATURITY ON DRY MATTER ACCUMULATION, AND PARTITIONING EFFICIENCY AND THEIR INTERRELATIONSHIPS WITH YIELD IN DRY BEANS ABSTRACT Knowledge of the variation in the physiological characteristics attributes of growth and their association with yield is important to improve the yielding ability of a crop. The objective of this study was to ascertain whether there was variabilility for several morphological traits among 12 dry bean (Phaseolus vulgarileJ genotypes and to study the interrelationships between the morphological traits and yield. The results indicated that bean genotypes with a determinate growth habit flowered earlier and took a shorter time to reach physiological maturity and had smaller lOO-seed weight compared to genotypes with indeterminate growth habits. The Type II genotypes used in this study had a higher 2 seed yield, number of pods per m , seed per pod and biological yield than Type III. However, Type III genotypes 4 :51 a A: v. Ca 3. m. .o. k,” V . ~§~ . a .-u v a nnv o . any any . . m2 ah. .fiu .NH an. n—v .—. a.» A\. n\~ .n. a Au 6 u .— u A V ~ ‘4 .. . a: .Pv an. “I ”I .a‘ v s .. « a~u -u H ks b \ a .v .~; New .e. H.M 5 had a higher 100-seed weight, fewer days to 50% flowering and shorter period to physiological maturity than Type II. Type II showed a lower harvest index but a higher biological yield efficiency than Type III. Growth habit I with the architype I form had a higher 2 and number of seed yield, a greater number of pods per m seed per pods than traditional Type I genotypes. The traditional form within Type I showed a higher lOO-seed weight. The architype II had only a high lOO-seed weight when was compared to traditional II genotypes. Late Type III genotypes outyielded in seed and biological yield production to earlier Type III genotypes. The dry matter accumulation reached a maximum weight approximately 16 to 24 days after 50% flowering and then decreased as the rate of abscission of leaves and petioles from the plant increased and /or exceeded the production of new leaves and petioles. In general, indeterminate types produced more total dry matter than determinate ones. However, Type II genotypes produced more leaves than Type I and Type III which could indicate a greater potential for photosynthetic production. The maximum LAI, LAR and RGR was reached at 4 or 12 days after 50% flowering depending on the genotypes, thereafter, these parameters declined as dry matter of leaves decreased. GrainIyield was positively correlated with biological V 5'5- a- “gut . - ti o» «I. .. 3. Au Co a. i m“ A c a. .: S .3 L . 3 . . .24“ ~ 3 ... ”J .. “I .. ~- ~.. wv. 2. :— . . L. . . .3 .. a 2. 2. .. u ”I .u n s . a a: . a .2 ..~ . ‘vI'I-Iv‘n “§ Q 6 yield, and days to physiological maturity. There was a positive association between number of pod per plant and yield, biological yield, days to 50% flowering, and physiological maturity. The number of seed per pod was positive correlated with yield, days to 50% flowering and days to physiological maturity, but negatively correlated with lOO—seed weight. The weight of 100 seeds weight was negatively correlated with grain and biological yield, while biological yield was positively associated with days to physiological maturity but negatively correlated to HI. The lack of association between HI and seed yield efficiency suggested that although.a genotype might have a high HI it may be inefficient in its daily production of dry matter. It seems that a high yielding cultivar should have both a high HI and seed yield efficiency than just a high HI. Total dry weight, leaf dry weight, and leaf area were positively correlated with yield at 12 days after 50% flowering. Leaf area index was positive associated with yield at 4, 12 and 20 days after 50% flowering. Leaf area ratio and relative growth rate were not correlated with yield at any of the plant growth stages considered in this study. -‘yo on. v- ' .4 . . I‘d A- ,. 1v .3 ... .1 fl ~\~ . . ~\~ v n.. :- .. g s ..n ,w a .q a u. a n. at. o h Hag 7—. has- “v- s U. r a e 3. . a T. sxw a.» u a . .. . a . .. a . 2.. . ‘ A u L» u.. t» .4 . .. . a a: . . ..... «h .1 .. .. k . a Q .4 s l-. ‘ u I O o I INTRODUCTION Selection of superior plants for economic yield, disease resistance, and tolerance to adverse climatic conditions in breeding programs has led to the development of most new crop cultivars ( 43 ). However, an increased understanding of growth and development, and the manipulation of the physiology of plants may be a more efficient approach to yield improvement than the breeding for yield per se. For this approach to be successful, however, genetic variability for the traits under study needs to exist in the population and the traits must have high heritabilities and should be favorably correlated with yield. Breeding for physiological an morphological traits may result in an increased biological yield because of a redistribution of assimilates to the seed. Remobilization of assimilate is one mechanism that enables the plant to counter environmental stresses( 43 ).Since controversy; existtas to whether or not yield is limited by the size of the source or sink of the plant ( 52 ) further study of the partitioning of assimilates and the interactions between source and sink is needed. T“ F In. \ na '1 .- .M gun»- C gau‘-.. ~q tI'U-'° ‘- J van- ' MN flue-v. ' libs. ‘Q q. --un .A a " 'v .‘u h 1. ~‘I- -3 «cu-c": U -u -l a ...e -- -~na .‘I 4 .1 a’ ‘ 'Uu 8 In dry beans, four growth habits are recognized( 98 ). In general, Type I, II and III are the most important growth habits grown under monoculture. However, in some areas Type II and III are planted in association with corn ( 75 ). Genotypes within each growth habit usually differ in yield, days to maturity and in other morphological traits. Type III is generally later maturing than Type I and II ( 4O ). The objectives of this study were to l)escertain the magnitude of variability for morphological traits in a population comprised of 12 dry bean genotypes, and 2)study interrelationships between the morphological traits and yield. "C rhit‘ sac—\A 4 o A sun..u: I O... ‘ \ . . .w . 4‘ «v a» a3 u .... r a.» 7. v . a: .n . u u ”w. i. a . .~. :~ «3 a.‘ La nos a... on .-.. a n - s Literature Review. Plant ideotype and plant architecture. In the past, plant breeders have given major attention to the selection for yield and quality of a crop. However, more emphasis could be placed on selection for physiological and morphological yield components related to yield because these components are involved in crop productivity and adaptation ( 12 ). The most notable breeding achievements using morphological features to improve yield was in the development of the dwarf wheat and rice by the International Maize and Wheat Center (CIMMYTW and the International Rice Research Institute (IRRI) ( 12 ). In general, the high yields of the new cultivars was due to their nitrogen fertilizer responsiveness and resistance to lodging ( 12 ). In breeding for morphological or physiological traits, two concepts have been used and need to be defined. These are plant ideotype and plant architecture. Donald ( l4 ) defined a plant ideotype as a biological model which is expected to perform or behave in a predictable manner within a defined environment. Moreover, he proposed that this concept could be used as a breeding strategy for improving wheat. He pointed out that a wheat ideotype under Australian conditions should 9 10 have a short, strong stem, a few small and erect leaves, a large and erect ear, awns,and a simple culm. Adams ( 3 ) defined plant architecture in terms of the number, size, shape, structural arrangement and display of particular plant parts, but he stated that architecture only has merit in a yield sense if it is associated or leads to improve physiological functions..Adams (:3) proposed an architectural design in grain legumes to increase yields. The architectural objectives were to 1) build a leaf-stem canopy capable of intercepting all incident light by absorbing, scattering, and transmitting the photosynthetically active radiation throughout the plant profile, 2) construct the canopy so that the profile consisted of as many phytomeric units (source-sink units) as possible auui 3) make each phytomeric unit as functionally efficient as possible. Adams ( 3 ) suggested that a high yielding dry bean architype should be tall with 12-15 nodes on the main stem and have two to three 3 basal branches. Other features of the plant model included: l.indeterminate growth habit with upper internodes longer and more numerous than basal internodes, 2.thick stem diameter, 3.narrow plant profile, 4.high values of first order yield components, 5.1eaf area index near four at flowering time, ll 6nsmall leaves capable of light induced orientation, 7.full season leaf area duration, 8.higher specific leaf dry weight, ELstem and root starch storage and remobilization during pod and seed filling, JILhigh rate of seed filling, and llllonger duration of linear phase of seed filling. Despite the potential advantages associated with breeding for a yield model, the ideotype concept has not been widely used. This may be due to a lack of information regarding the contributions and relative merits of morphological and physiological aspects related to yield components. In addition, several ideal types may be required to meet the various cultural and/ or environmental conditions under which grain legumes are produced ( 3, 12 ). Growth habit. Since the plant ideotype has been described in terms of morphological characteristics which may be related to yield production, the growth habit plays an important role in the ideotype concept. In dry beans four major growth habits are recognized ( 47 ). Plants of the Type I are determinate and have reproductive terminal buds on the main stem and branches. No further node development and leaf production occurs after flowering. The Type II is characterized by an indeterminate growth habit, and both main stem and branches are strong and I V unvs ~“t. .1.‘ it“. v . - .‘Qr— "~~ H ou‘av ~v5 n. . A .. ‘va 't-I n .5 Ow-.. {I '-V‘~- ‘ 4 .4 ‘~ eaten. .- ‘:,‘, l‘“‘“- at I n -. '“Rna v. ““HV‘V . o "fla‘. u. “v- fit on... ‘ --:.‘~“ A A " noe'v..:: 12 upright. A vegetative terminal bud occurs on the main stem and branches. Node and leaf production occurs after flowering. The Type III plants are characterized by an indeterminate growth habit with semi prostrate branches that are relatively weak and open. The pod set of Type III plants is largely concentrated in the basal part of the plant.fmue maximum yield of Type III genotypes occurs under conditions of monoculture. Plants with the Type IV growth habit have an indeterminate and weak and excessively long stems and branches. Type IV plants possess strong climbing ability, and a support using some kind of trellis is essential for maximum production. Growth Habit and Yield §Qy§g§g. Wilcox ( 56 ) evaluated 40 determinate and 40 semideterminate soybean genotypes and concluded that semideterminate soybeans were higher in yield than determinate ones. Chang et al ( ll ) tested isolines with different degrees of determinancy and found that semideterminate genotypes yielded equal to their indeterminate counterpart. Egli and Leggett ( 20 ) made comparisons of the dry matter accumulation between.ea determinate soybean genotype 'D66-5566' and the indeterminate cultivar 'Kent'. At time of first flower the stem dry weight of the determinate genotype Afiv V. O R“ g '04. ' fl gnu- vc- “ ~ "a...v-o .3 l o ..-p-- at v.-.~-u- . - o - ' in".-. ' hit-Oh.-. '0‘ Q A.- u‘ U. I v A an. why «x» ‘h n' .: ... .n. .. ufiH r .m A» . ‘NC .fi‘ LC ~ g u C. II . Q ‘ .... Cs u: 13 was 67 % of the maximum weight of the plant compared with only 30% for the indeterminate. They concluded that indeterminate genotypes accumulate more dry matter during the reproductive stage and have a greater potential for competition for photosynthates between reproductive and vegetative growth than the indeterminate ones. Severe lodging in different crops species usually reduces yield. Hicks EH: al ( 32 ‘) evaluated short determinate, tall determinate and indeterminate isolines of 'Clark' and 'Harasoy' soybeans at different densities. Lodging of tall indeterminate plants increased as plant density increased, but the short determinate plants did not lodge with increasing plant densities. Yields of the short determinate plants were not significantly different from those of indeterminate plants.1fimatall determinate plants yielded about 5% more than the indeterminate plants. Wilcox. and Sedimaya ( 55 ) studying the interrelationship among height, lodging, and yield in indeterminate and determinate soybeans found that the highest yielding indeterminate lines were tall and susceptible to lodging while the tall determinate lines combined both high seed yield and excellent lodging resistance. Bramel et al ( 8 ) using factor analysis studied several soybean lines selected from the F6 generation, and which were segregating for the degree of stem termination. They (E3) .- or 9v to 'v ll) (I) 'v. ‘3 ll" 7/! Hi l4 concluded that in the determinate stem types, the length of the seed filling period and number of nodes could be used to predict yield. In the semideterminate type some vegetative traits and length of the terminal raceme might be useful to predict yield. However, in the indeterminate types, all the variables analyzed had a very low predictive value for yield. Common bean. Kuenemam et al ( 39 ) studied yield responses of several dry bean genotypes with different growth habits and concluded that yields of genotypes with indeterminate habits (Type II and III) were significantly greater than those with the determinate habit (Type I). Robitaille ( 44 ) showed that determinancy and indeterminancy did not directly influence yield in dry beans if other factors such as maturity date and space planting were equal. Nienhuis and Singh ( 41 ) concluded that yield and seed weight in beans, could be increased indirectly through selection for long internodes on the main stem. They also suggested that the main reason for a low yield of medium and large-seeded compared with small-seeded genotypes may be due 1x>a reduction in internode length on the main stem of the former genotypes. However, direct selection for an increased internode and main stem length could lead to a weak and prostrate plant. Kueneman and Wallace ( 52 ) comparing different non climbing dry bean genotypes at three spacings in the tropics .A V. F- n. '5 \‘. 15 concluded that determinate genotypes have a fewer number of days to anthesis, fewer leaves, a smaller leaf area ratio (LAR), and harvest index (HI), a slower crop growth rate (CGR) prior to anthesis, fewer racemes per plant, fewer pods per plant, fewer seeds per pods and larger seeds. They ( 52 ) also found that indeterminate small vine genotypes had a high biological yield coupled with a high harvest index. Robitaille( 44 )studied three indeterminate dry bean cultivars, erect, semi erect and vine type.Ikispite of the variations in growth habit, the dry matter accumulation patterns were similar for the three genotypes suggesting that plant type may not be an important factor in determining yield potential. Beaver et a1 ( 6 ) studied dry beans grown at variable locations and reported that indeterminate types produced in the Dominican Republic exhibited more yield stability than determinate ones. Kelly et a1 ( 37 ) also concluded that the indeterminate upright growth habit (Type II) offered the best choice for breeding beans with high and stabilized yields under temperate climate growing conditions. Type I beans grown at the International Center for Tropical Agriculture (CIAT) have been lower yielding and less stable over locations and seasons than Type II and III. It seems that type I has a very limited phenological potential for yield. It only produces a few nodes before flowering, l6 leaf area is small and the duration of the reproductive growth period is short (IUD). Type II strains grown in the International Bean Yield and Adaptation Trials of CIAT have been consistently the highest yielders compared to Type I and III even under conditions of monoculture in temperate latitudes. Type III strains have good yields and adaptation world-wide while the prostrate plant habit makes interrow cultivation and harvesting difficult. Also seed quality is reduced sometimes becausezpods are in contact with the soil and absorb moisture ( 4O ). Err Matter Accumslstiea Total dry matter production may vary among crops because of variability in the size of the photosynthetic system or its activity and the length of the growing period during which photosynthesis occurs ( 27 ). Sgybgag. Hanway and Weber ( 27 ) studied the dry matter accumulation in eight soybeans cultivars. They observed that all the genotypes developed at the same rate through developmental stage 2 (three trifoliate leaves completely unrolled). After this, gradual differences in development began to occur among entries and some cultivars were in full bloom sooner than others. The differences in rate of development increased with time and were maximum at maturity. During early soybean vegetative growth and through stage l7 5 (9 to 10 trifoliate leaves completely unrolled and plants in the full bloom stage) total dry weight of plants consisted of 55% leaves, 14% petioles and 31% stems. After stage 5, dry weights of each genotype increase at a constant daily rate. Common bean and peas. Wallace and Munger ( 53 ) indicated that the variation in dry matter accumulation of different cultivars of beans might be related to factors such as leaf area, net accumulation rate, leaf area ratio, and relative growth rate. Eastin and Gritton ( 18 ) observed that the efficiency of dry matter production of peas was greatest between flowering and harvest rather than during the vegetative period. Flowering and Physiological Maturity. Sgyggan. Variation among soybean cultivars with respect to the time of flowering and maturity is generally quantitatively inherited and stem length is primarily dependent upon the time of cessation of apical bud growth ( 30). Hartung et al ( 3O ) evaluated near isogenic lines of soybeans developed from 'Clark' and 'Harasoy' which possessed various genes for stem termination and maturity. They concluded that the genes that delayed flowering at maturity significantly increased the number of main stem nodes, stem length, lodging, branches per plant, but reduced seed weight. l8 Genes that hastened flowering and maturity had the opposite effect ( 30 ). The th gene for indeterminate stem significantly reduced the number of main-stem nodes and stem length, but the reduction in lodging was not significant. The gene for determinate stem growth (dtl) severely reduced the number of main-stem nodes and substantially reduced stem length and lodging. There was no advantage with respect to yield because delayed flowering resulted in a shorter reproductive period which tended to reduce seed weight. Common bean. Indeterminate bean genotypes have been found to flower later than determinate types ( 4O ). Coyne ( 12 ) showed that this association is often due tx>la genetic linkage between indeterminancy and late flowering. Laing et a1 ( 4O ) showed that an increase in the length of the bean growth cycle increased the number of nodes formed and leaf area duration and yield. This suggested that by extending the duration of the growth period over which pods are formed, the demand of the pods and the current supply of assimilates from the leaves are more nearly balanced, hence, the pod set increase ( 40 ). Leaf Area and Leaf Area Index. Leaf area index (LAI) has been regarded as an indicator of the degree of competition for light by individual plants inithin a row and thereby should have a substantial influence Q! '1 I'- a. ‘V '01. 1 CI. u- I:- 0' ill 19 on crop yield ( 27 ). The photosynthetic rate of a plant canopy per unit of land is related to the area and leaf angles. However, Duncan (115) concluded that for leaf area indices of less than aproximately 3.0 differences in leaf area angle have little practical influence. Soybean. Emecz ( 22 ) suggested that the critical leaf area varies with the angle of the sun above the horizmnn He also indicated that the photosynthetic productivity is between limits of O to 95% of solar interception. Shibles and Weber ( 46 ) found that the percent solar of radiation interception and rate of dry matter production increased with increasing leaf area development. Dry matter production reached a maximun and remained constant with further increases in LAI. Since the rate of dry matter production did not decline when the LAI exceeded the value needed for maximum interception, it was concluded that there was no optimum LAI in soybean. Likewise Buttery ( 9 ) observed that there was no sign of an optimun LAI when crop growth rate was at optimum. He also reported that maximum LAI was obtained between 70 and 80 days after planting regarless of plant density in soybean. Fehr et al ( 24 ) studied the response of indeterminate and determinate soybean genotypes to 100% defoliation at three stages of growthl The indeterminate genotypes had an average yield loss across stages of 59% compared with 80% for the determinate genotypes. In general the most critical stage ll 20 for defoliation was either begining seed filling or full seed develOpment. Seed Growth and Development. Seed consists of three structurally different tissues that may differ in their basic genome constitution (N). These are the seed coat (2N), embryo (2N), and endosperm (3N). Seed size is regulated through the course of its development by interactions of the maternal and paternal genotypes and the environment. The net effect of these interactions at different stages is reflected in the size changes of different seed parts ( 33 ). The development of a seed can be separated into two distinct processes which are the formation of the basic cellular structures, or compartments, and the filling of the compartment with the storage materials( 27 L.In.legumes, there is little increase in the weight of seed at first, although the final number of cells in the embryo is determined.early( 27 L The subsequent increases in weight of the seed are the result of cell expansion and the accumulation of starch or oil or both and, later of storage protein ( 27 ). Therefore, the weight of individual seeds depends on the duration of the linear filling period and the rate of seed growth ( 27 ). Soybean. Unlike cereals, the flowering grain legumes extends 21 over a longer period of time. Pods originating from early flowers, begin filling earlier than pods originating from later flowers. Gbikpi and Crookston ( 28 ) reported that late-maturing seeds of the several soybean cultivars studied had a faster rate of dry matter accumulation, and an even faster rate of protein accumulation than earlier maturing seeds of the same cultivars. The rapid dry matter and protein accumulation in late-set soybean seeds may be related to differences in abscisic acid (hormonal) concentration between the top and the lower portion of the plant. Also, late setting seed may have a greater supply of assimilate because assimilate competition by roots and nodules is reduced, therefore, the supply for late-filling seed is greater. The growth rate of seeds has been shown to be determined by both genetic and environmental effects. Hedley and Ambrose ( 31 ) suggested that the testa and embryo in peas are sinks which both compete for the endosperm which may act as a common source, and this relationship accounts for variation in the endosperm volume. Egli et al ( 19 ) reported that differences in seed growth rates are controlled by the cotyledons, not by the supply of assimilate from the plant or the process involved in the transfer of assimilate from the phloem 1x) the developing cotyledons. Variations in the effective filling period (EFP) among cultivars were not significantly correlated with seed size, seed growth rate, or 22 the number of cells in the cotyledons, suggesting that the factors controlling the duration of the seed filling period are not closely related to the factors controlling the seed growth rate. Common bean. Hsu (1K3) concluded that differences between two dry bean genotypes possessing different sizes of seed could be attributed to differences in growth rate and not to the duration of development. Both embryo and seed coat contributed to the final seed size differences, although the embryo was clearly the major component. 39g Grgwth and Development. Pea and soybean. Tissues of developing fruits tissues are particularly rich in gibberelins, auxins and abscisic acid. Eeuwens and Schwake (2K3) reported that growth of the pea pod wall appeared to depend largely on hormones supplied by the seeds. A high concentration of abscisic acid in the embryo rather than the restriction imposed by the testa prevented precocious germination (ME seeds. competition between the pod wall and seed for available metabolites may explain the relatively slow growth rate of the pod during the rapid growth of the seed. Fraser et al ( 26 ) found that pod length and width were at a maximum when the soybean seed averaged 4% of their maximun dry weight. Length and width were correlated 23 positively with final soybean seed size. Pod wall weight per unit seed weight did not seem to be related to genetic differences in seed size, although there was a tendency for late-developing pod walls to exhibit:a greater weight than the earlier pods. Abscission 93 Flowers and Pods Abscission of flowers and immature pods varies among species and cultivars. Izquierdo and Hosfield (INS) reported that 50% of the theoretical yield in commom bean is lost by abscission and 61% of the total abscission was accounted for by small pods less than 10mm in length. Ethylene evolved from flowers of two genotypes, Sanilac and MSU #61618, was associated with a high rate of abscission. Competitive total non-structural carbohydrate (TNC) and nitrogen in the stem stimulated abscission in small bush type beans ( 35 ). Binnie and Clifford (17) reported that abscission of flowers and immature pods varied between 45 to 80% in seven dry bean genotypes studied. The ability of grain legumes to compensate for flower removal appears to depend upon the severity of flower removal in relation to the developmental stage and vigor of plants as well as environmental conditions. The fact that removing opened flowers resulted in the formation of new flowers indicated that bud primordia which would have failed to 24 develop into flowers are able to do so when given the opportunityu It seems, there exists no genetic barriers to further flower development. The removal.of opened flowers from alternate nodes led to pods being set on the remaining nodes ( 7 ). Therefore, flower abscission of young pods would appear to be related to a genetic control mediated by a limited supply of assimilates or other nutrients, hormonal influences or a combination of these factors ( 27 ). Estimates of photosynthetic contributions from seed heads or fruiting structures to the final harvestable yield range from negligible in some species to almost 100% in rapeseed. Soybean. Thorne ( 50 ) suggested that, in extremely short season cultivars, assimilate storage and later redistribution by pod walls to the developing seeds may serve to effectively lengthen the period of pod filling and thus contribute significantly to final seed yield in soybeans. Egli and Leggett ( 21 ) reported that the pod wall is the primary sink for photosynthates during seed development and consequently accumulation of dry weight in pods (pod wall and seeds). Studies have indicated ( 25, 26 ) that the width of Inature pods can be used effectively for indirect selection of seed weight in soybeans. Pods reached their maximum length and width about 15 to 20 days after pollination, but pod thickness increased until the seed was full sized. Frank and Fehr ( 25 ) reported that selection for pod 25 width and length based on measurements of full size green pods was as edfective as selection based on mature pod dimensions. Selection for pod length or width were more effective than pod thickness. Pod width was considered the most efficient character to LEN? for indirect selection because it can be measured without regard to number of seeds per pod and is faster to measure than pod length. Common bean. Pods of 'Redkote' kidney bean were not found to be an important source of photosynthate for dry matter of the developing seeds. It appears that pods, similar to the seeds are actually a storage organ for assimilates formed in the leaves ( 13 L Harvest Index. Donald (115) originally defined harvest index (HI) as the ratio of seed yield to biological yield, and later described it as a measure of partitioning of photosynthate. Donald in cereals ( 15 ) and Wallace ( 54 ) in dry beans emphasized that possible HI and biological yield were probably important to yield and breeders should utilize these two traits for improving the efficiency of breeding for seed yield. Cerealsy rape seed and legumes. Improvements in the yield potential of cereals such as wheat, rice and sorghum are often claimed to be largely due to improvements in HI. 26 Therefore, it is often suggested that similar improvement could be obtained in legumes and rape seed through modification of the HI. Sinha et a1 (4H3) compared the HI based on dry matter and energy in cereals, legumes and rape seed and concluded that the HI based on dry matter is not adequate for comparing partitioning of photosynthates in different crops. It is expected that on the basis of equal partitioning of photosynthate between grain and straw in cereals and rape seed, rape seed would have a lower grain weight because of a higher energy demand for the production of oil ( 48 L The true HI is difficult to measure unless one collects abscised leaves of the plant to calculate biological yield. However, Schapaugh and Wilcox ( 45 ) found a high correlation between actual and apparent HI. The apparent HI is plant weight of the mature plant (minus leaves). This indicated that the measurements of apparent HI could permit a valid comparison of the relative efficiency among genotypes. Common bean. Wallace and Munger ( 52 ) concluded that HI in 'the dry bean genotypes studied was neither correlated with ‘time to maturity, nor with determinate or indeterminate gixawth habits. Laing et a1 ( 4O ) reported that HI in dry txearicultivars grown in the tropics was negatively correlated with yield. 27 Sgybgag. Buzzel and Buttery ( lO ) reported that HI was negatively correlated with plant height, maturity, straw weight and seed yield in soybeans. On the other hand, Schapaugh and Wilcox ( 45 ) reported a negative correlation between actual and apparent HI with biological yield, straw and leaf weight, maturity and height. These authors ( 45 ) also found that the relative performance of the entries evaluated based on HI means was more stable than the relative performance based CHI seed yield across years. Therefore, measurements of the apparent HI should permit a valid comparison of the relative efficiency of a group of genotypes. Wilcox ( 56 ) measured HI of individual soybean plants and found that HI had a low variance estimates compared with the variance of absolute total leaf mass. Johnson and Major ( 36 ) found that HI of soybeans was independent of planting date except in the case of where a premature killing frost caused a reduction in dry matter. Spaeth et al ( 49 ) studying the variation of HI under different conditions in both determinate and indeterminate soybean cultivars concluded that HI was a stable characteristic within each soybean cultivar. Materials and Methods Twelve dry bean genotypes were used in the present study (Table 1). They were characterized by a Type I, II and III growth habit according to the classification of the International Center for Agriculture (CIAT), Cali, Colombia ( 47 ). Beans with these growth habits are commonly grown in monoculture cropping system ( 41 ). Types I and II were represented by two different architectures, namely the traditional form and the tall erect and narrow profile form refered to as architype ( 2 ). The traditional Type I (TI) is characterized by a small bush plant with a determinate growth habit and having few nodes per plant. The TI plants have a non erect canopy and a short flowering period. The TI genotypes used in this study were 'Seafarer', a navy bean and 'Brasil—2', a bean of tropical origin and having small beige colored seed. The Type I architype (AI) plants have a determinate growth habit; Other characteristics include plants have fewer branches, better lodging resistance, and a higher yield than TI plants ( 2 ). The AI genotypes used were 'Laker' and 'C- 49', navy beans. The traditional Type II (TII) plant is characterized by 28 29 Table 1. Growth habit, architectural form, lOO-seed weight, commercial class and days to physiological maturity of 12 dry bean genotypes grown in 1984. Days to 100— Commer- physio- Growth Archetypal seed cial logical Genotype habit+ form weight class maturity Seafarer Type I Traditional 17.8 Navy 68 Brazil-2 Type I Traditional 20.6 Undefined 82 Laker Type I Architype 17.2 Navy 82 C-49 Type I Architype 14.6 Navy 79 T-39 Type II Traditional 19.2 BTS 8O Nap-2 Type II Traditional 16.1 Navy 83 C-20 Type II Traditional 21.2 Navy 82 Swan Type II Architype 18.7 Navy 84 Valley Carioca Type III Prostrate 26.8 Undefined 82 Vine Viva Type III Prostrate 25.4 Sutter pink 69 Vine Harris Type III Prostrate 26.4 Great 7O Vine northern Valley Type III Prostrate 26.8 Great 8O vine northern +Growth habit classification used by the International Center for Tropical Agriculture (CIAT), Cali, Colombia. 30 an indeterminate growth habit with a short vine. Plants are semi-erect containing several branches and they lodge at maturity. 'Nep-2' and 'T-39fl a navy euui tropical black genotype, respectively were representative TII genotypes used in this study. The Type II architype (All) is characterized by a narrow profile which is supported by two to four strong branches vertically oriented and separated from one to another by an acute angle. This form has superior lodging resistance and was represented in this study by"C-20' and Tynan Valley', navy beans. Type III dry bean genotypes have an indeterminate growth habit, with several branches, and are semi-postrate. Type III plants lodge at maturity. The genotypes that characterized this growth habit were 'Cariocafl 'Vivafl 'Harris' and 'Valley'. The genotypes were selected based on their differences in days to maturitry, reproductive grown pattern, seed size and yield. The 12 entries were grown in 1984 on the Michigan State University Botany Farm at East Lansing. Seeds were drilled into eight row plots on June 4th with a tractor- mounted air planter ( 51 L Rows were 10 m long and 47 cm apart. Plant spacing within rows was 7 to 8 cm. The experimental arrangement utilized was a randomized block design with three replications. Row numbers 2, 3, and 7 were 31 used for sampling purposes using rows 1 and 8 as borders. Rows 4 and 5 were selected for yield estimation at harvest. Depending on the genotypes, the rows used for sampling were sub-divided into 8, 9 or 10 segments, each 1 m long. Within each sampling unit ( 1m segment), six plants were randomly selected to measure dry weight of plant components, and leaf area. Samples were taken at four day intervals beginning five days after 50% flowering and continued until physiological maturity was reached. Plants in a sample were separated into the leaf lamina, stem, petioles, pod walls and seeds. The various plant components were dried in a forced air oven at 75 C for 48 h and weighed. During the growing season and beginning at blooming, measurements were recorded for days to 50% flowering in a plot, days to physiological maturity and leaf area. Days to 50% flowering was taken as the number of days from planting to when 50% of the plants in a plot had one open flower. Physiological maturity occurred when bean pods had turned from their normal green color to a perceptible yellow and gxmds had started to dry. Leaf area was measured using a leaf area meter (Lambda Instruments, L—300) after removing all leaves from plants. Leaf area ratio (LAR) and relative growth Inate (RGR) were calculated using the equation proposed by Redford ( 42 ). 32 Mature plants from 2 m sections of individual plots were harvested and threshed to measure final seed yield. The number of pods per plot was the average number of pods from a 1 m sample of row at harvest. The number of seeds per pod was calculated by averaging the number of seeds from a 50 pod sample from each replication. Weight of 100 seed was determined as the average of two, lOO-seed samples per replication. Partitioning ratios were calculated to evaluate the production efficiency of the 12 genotypes. The HI was calculated as the ratio of economic to biological yield. Biological and yield efficiency were estimated as the ratio of biological and seed yield per genotype to the number of days to physiological maturity. Statistical analysis. All data were subjected to an analysis of variance appropiate to a randomized complete block design. Correlations and orthogonal single degree of freedom comparisons were conducted to compare the rates of dry matter productions between and among genotypes, growth habits, architectural forms and maturity groups. A second and third degree polynomial was used to fit the data to lines which explained the variation in dry matter production observed over time. 33 The equation was: Yij= (3+ Bx + BX2+...BXn+€ ij Where Y was the dry matter production per plant “3L and x was the number of days after 50% flowering. Bl,(52, [33 were constants calculated using the least squares technique. Results Growth Habit Carioca, Viva, Harris, and Valley are Type III genotypes and.had.a profuse pattern of branching; Most of the pods on these genotypes were born on the numerous plant branches. Numerous branches and a heavy pod set predisposed plants of these genotypes to lodge at maturity. The genotypes T-39, Nep-2, C-20, and Swan Valley are indeterminate and Type II genotypes with moderate to only a few branches. These genotypes did not lodge at maturity. Seafarer, Brasil—Z, Laker and C-49 are Type I genotypes with a short stem and internodes. Branches and pods were set mostly on the main stem. Seafarer and Brasil-Z lodged prior to physiological maturity and might show high yield losses under period of excessive rain during the harvest season. Plants of Laker and C-49 did not lodge at maturity. Grain Yield and Yield Components. Differences gmong types.Cmthogonal comparisons were made zmnong growth habits, architectural forms and early and late Inaturity for genotypes within the various categories (Table 2,.3 L When comparisons among growth habits were made for 34 35 Table 2 Mean yield and components of seed yield of 12 dry bean genotypes grown in 1984. Seed Yield Components of seed yield Growth lOO-seed habits and Dry B is Fresh asis pod/m2 seeds/pod weight gethPeS+ (g/m ) (g/m ) (No.) (No.) (9) I Traditional Seafarer 97.4 196.3 286 3.6 17.8 Brazil-2 105 . 5 245 . 6 296 4 . 3 20. 6 Mean 101.5 221.0 291 4.0 19.2 Architype Laker 126.2 249.3 380 4.9 19.2 C-49 130.1 269.2 434 4.5 14.6 Mean 128.2 259.3 407 4.3 15.9 II Traditional T-39 119 . l 236 . 5 388 4 . 9 l9 . 2 Nep-2 124.2 257.3 460 3.9 16.1 Mean 121.7 246.9 424 4.4 17.7 Architype C-ZO 121.8 247.8 500 3.6 21.2 S Valley 125.8 262.3 418 4.8 18.7 Mean 123.8 255.1 459 4.2 20.0 EYE III Early _ Viva 124.7 243.3 350 4.3 25.4 Harris 104.5 203.4 296 3.6 26.4 Mean 114.6 223.4 323 4.0 25.9 Late Carioca 100.9 209.9 236 3.6 26.6 Valley 104 . 8 218 . 7 274 3 . 7 26 . 8 Mean 102.9 214.3 255 3.7 26.7 +Growth habit classification used by the International Center for Tropical Agriculture (CIAT), Cali, Colombia. 36 Table 3. Mean squares from single degree of freedom orthogonal comparisons for yield, and components of yield of 12 dry bean genotypes grown in 1984. Orthogonal Pads/ Seed/ 100 seed Seed Zield/ Comparison m pod weight m Type I vs. Type II & III 533.56 0.01 198.34** 1.87 Type II vs. Type III 34960.67** l.65** 338.25** 292.81** Traditional I vs. Architype I 10050.08** 0.30** 32.34** 534.00** Traditional II vs Architype II 918.75 0.16 15.4l** 3.44 Late III vs. Early III l6l0.08** 0.52 0.61 109.63** Seafarer vs. Brazil-2 32.67 0.67** 12.33* 24.32 Laker vs. C-49 1148.17 0.28 10.40* 5.74 T-39 vs. Nap-2 1944.00* l.50** 15.04** 9.86 C-20 vs. S Valley 2521.50* 2.16** 9.38* 6.18 Viva vs. Harris 1380.17 0.01 0.04 4.81 Carioca vs Valley 2242.67* 0.54** 2.94 l48.60** *,** Significant at 5% and 1% probability level, respectively. 37 seed yield and yield components, differences among the Type I and Type II and III habits were noted for only the lOO-seed weight traitu One hundred seed weights for Type I, II, and III, were 17.6, 18.9 and 22.4 g, respectively. Type II and III differed from each other for seed yield and the components of yield. Type II had a higher seed yield, number 2 of pods per m number of seed per pod, and lower seed weight compared with Type III (Table 2, 3 ). Differences within types: The AI and TI genotypes were significantly different in number of pods per m2, seed per pod, lOO-seed weight, and seed yield, The AI genotypes had higher number of pods per m2 , more seeds per pod and a higher seed yield per plant than TI. However, the TI had a higher lOO-seed weight. The seed weight difference was mainly due to Brasil-2 which had a larger seed size compared with the navy genotypes comprising the AI group. There was a highly significant difference between TII and AII for 100 seed weight. The All group had a heavier seeds than the traditional group.‘This difference was mostly due to the seed weight of C-20 (Table 2, 3 ). Because of the absence of architectural forms in Type III, the genotypes were grouped into an early (Viva and Harris) and late category (Carioca and Valley). Examining the data, this contrast showed thatiflmalate genotypes yielded more and had more seeds per pod than the earlier maturing 38 genotypes (Table 2, 3 ). Differences within architectural forms. Differences between the TI genotypes indicated that Brasil-Z had a greater number of seeds per pod than Seafarer. Differences between TII showed that Nap-2 had a higher number of pods per m2 but a lower number of seeds per pod and lOO-seed weight than T-39. Between the AI genotypes, Laker differed from C-49 only for the lOO-seed weight trait but between AII genotypes, C-20 was significantly superior to Swan Valley in number of pods per m2 and lOO-seed weight. However, Swan Valley had a higher number of seeds per pod (Table 2 L Differences between the Type III early maturing class showed no differences between Harris and Viva for yield and yield components. Whereas within the late maturing class, Carioca had a higher number of pods per m2, seed per pod, and grain yield than Valley (Table 2, 3 ). Days 39 50% Flowering and Physiological Maturity Differences between types. Type I genotypes on average were earlier taking 51 days before 50% flowering occured while Type II and III genotypes reached the 50% flowering stage in 53.5 days. Type I genotypes reached physiological maturity 78 days after planting, but Type II did not reach physiological maturity until 82.5 days after planting. When type II and type III genotypes were compared for physiological maturity, 39 Type 11 was mature later ( 82.5 days after planting) than Type III (75i3days after planting) VTable 4, 5 L D££ferences E13512 Eypgm: The TI and T11 flowered and reached physiological maturity earlier than tin) architype genotypes of these growth habits. The early maturing genotypes within Type III, also flowered and matured earlier than the late maturing ones within the same growth habit e4}, Viva and Harris also flowered and matured earlier than Carioca and Valley (Table 4, 5 L 91££9£eaee§ 912519 glasses: The traditional genotYPes, Seafarer and T—39 were earlier maturing genotypes than Hrasil-Z and Nep-Z. The architype C-49 and C-20 were earlier maturing than Laker and Swan Valley. There were no statistical differences between Harris and.Viva in days to 50% flowering and physiological maturity, however, Valley flowered and matured earlier than Carioca (Table 4, 5). Dry matter accumulation. In the present work, data were used to select linear functions which described the'trendcofch3{ Iuatter accumulation in different plant parts over time. Polynomials of best fit were determined using the least squares procedure. Coefficients of determination were calculated and compared (Appendix A ). In general, the Inaxinuun weight of total dry matter accumulation was reached betyueen 16 and 24 days after 50% flowering and then decreased Table 4. Mean biological yield, days to 50% flowering and days to physiological maturity of 12 dry beans genotypes grown in 1984. (jrcwatli habits and Biological, yield (g/m‘) genotypes+ _.___._—.__._...1-L ..—......_-....._- . ‘-- IYPOUJ Traditional Seafarer 261. Brazil—2 222. Mean 242. Alcfliitype Laker 314. C-49 313. Mean 314. Triuiull Traditional T—39 318. Nap-2 308. Mean 313. Architype C-20 354. S Valley 316. Mean 335. TYF¥3HIEL Early Viva 275. Harris 232. Mean 254. Lair} Carioca 214. Valley 240. Mean 227. 0000 OU‘IU1 NbO UINCD I-‘COID 8 9 9 Days to 505:. f lower ing 51 54 53 53 54 54 49 49 49 54 50 52 Days to Physiological maturity 68 82 75 82 79 81 80 83 82 82 84 83 69 7O 7O 82 80 81 +CuTivth habit classificationnfisedwby the International Center for Tropical Agriculture (CIAT), Cali, Colombia 41 Table 5. Mean squares from single degree of freedom orthogonal comparisons for biological yield, days to 50% flowering and physiological maturity of 12 dry bean genotypes and grown in 1984. Days to physiolog- Orthogonal Biological Days to 50% ical comparison yield flowering maturity Type I vs. Type II and III 10.22 12.50** 10.13** Type II vs. Type III 9207.69** 28.17** 315.38** Traditional I vs. Architype I 387l.82** 6.75** 102.08** Traditional II vs. Architype II 781.18* 2.08* 8.33** Late III vs Early III 884.26* 30.08** 420.08** Seafarer vs. Brazil-2 562.99* 28.17** 308.17** Laker vs. C-49 0.12 10.67** l6.67** T-39 vs. Nep-Z 42.03 13.50** 10.68** C-20 vs. S Valley 558.93 2.67** 6.00* Viva vs. Harris 447.55 0.17 0.67 Carioca vs. Valley 120.33 24.00** 4.17* *,** Significant at 5% and 1% probability level, respectively. as the 1 A: (9“ do “'Qfi H .‘ ,."‘. A {“9“ ‘v I’ a“ 42 as the rate of abscission of leaves and petioles from the plants increased and /or exceeded the production of new leaves and petioles. Between 4 to 28 days after 50% flowering, Type I, II and III had produced 9.7, 13.5 and 10.7 g per plant of their total dry matter, respectively (Table 6 ). During the period of flowering and pod set, Type II and III genotypes produced more dry matter than Type I. However, Type II genotypes produced more leaves than Type I and III genotypes which indicated that Type II could have a greater potential for photosynthetic production. At harvest, the dry matter production was highly and signicantly different among genotypes. In general, Type II (AI and TII), AI and the later genotypes within Types III were significantly different from traditional I, and the Type III forms, respectively. When comparisons between genotypes were made, the only significant differences detected were between Seafarer and Brasil-Z (Table 4 ). The data on dry matter accumulation were also used to calculate for selected days during the growing season, instantaneous values of relative growth.rate (RGR)eumlleaf area ratio (LAR). Tables 7 to 10 present the results of RGR, LAR, LA and LAI for each cultivar. Similar trends in RGR and LAR were observed among genotypes. The range for maximum LAR and RGR was between 4 and 12 days after flowering, 1'A"“1 '.U' .t o . . H9 “GU S ”A ‘5 2' 4 ‘ui pe5* '7 6 i=41U E ‘ ~Anab 1 Hear. 43 Table 6. Total dry weight of 12 dry bean genotypes at four sampling dates after 50% flowering and grown in 1984. Growth habits and Days after 50% flowering genotypes+ 4 12 20 28 ---------- g / plant ----------- Type I Traditional Seafarer 9.6 12.4 16.7 19.8 Brazil-2 8.9 9.9 13.9 15.6 Mean 9.3 11.2 15.3 17.7 Architype Laker 8.2 11.2 14.5 18.8 C-49 9.4 10.9 16.4 20.7 Mean 8.8 11.1 15.5 19.8 Type II Traditional T-39 10.1 16.9 20.4 26.4 Nep-2 8.7 11.6 14.6 20.5 Mean 9.4 14.3 17.5 23.5 Architype C-20 10.6 14.7 17.9 22.8 S Valley 11.1 12.9 19.5 24.8 Mean 10.9 13.8 18.7 23.8 Type III Early Viva 7.5 8.4 12.1 14.9 Harris 8.6 10.7 16.0 17.9 Mean 8.1 9.6 14.1 16.4 Late Carioca 8.0 14.0 17.6 22.8 Valley 9.5 10.9 18.3 21.0 Mean 8.8 12.5 18.0 21.9 +Growth habit classification used by the International Center for Tropical Agriculture (CIAT), Cali, Colombia. 7 h as e b ‘1 O H.“ .fib A v .5 v . A J .3 .l .. . Z .3 n; .C .3 id 4‘ 5.3 . .G n... a u a .3 «d v. .3 .. . s 2. e 1.; .c a e a. ...n. 4 .t +1. .15 C. e ‘1. «L “V C.» v a .1 r 2.. r . s E C. .1 e r u... .. . a . u... Tl 4 . . e n... .. . . nu. » . u! .. a 5 n.. a G n... kn r S "J I .«u S B .3 v.~ C T; .«u T .\\. -Wu n y 8 Ti . a “I “P. Q. ~ . ml 4 . a: ‘s be .5 a» «a an rs s» u!- s. .. . A9 a: r. v . a: vs v; A: .C .3 A u . .. .n. d an my 7; \fi n. .. if. as .L . ~ 5. .2 a: an a: .. u. n. «J «. .hu an a: . . m . \ . s Table 7. Relative growth rate of 12 dry bean genotypes at four sampling dates after 50% flowering and grown in 1984. Growth habits and Days after 50% flowering genotypes+ 4 12 20 28 ---------- 9 / p1 / day -------- Type I Traditional Seafarer 0.0519 0.0564 0.0441 --- Brazil-2 0.0083 0.0427 0.0331 -0.0025 Mean 0.0301 0.0496 0.0386 -0.0001 Architype Laker 0.0340 0.0236 0.0280 0.0330 C-49 0.0092 0.0534 0.0224 -0.0130 Mean 0.0216 .0.0385 0.0252 0.0100 Type II Traditional T-39 0.0265 0.0297 0.0296 -0.0198 Nep-Z 0.0372 0.0872 0.0500 0.0405 Mean 0.0319 0.0585 0.0398 0.0104 Architype C-20 0.0257 0.0112 0.0363 0.0296 S Valley 0.0298 0.0645 0.0365 0.0129 Mean 0.0278 0.0379 0.0364 0.0213 Type III Early Viva 0.0445 0.0398 0.0187 --- Harris 0.0530 0.0727 0.0291 --- Mean 0.0488 0.0563 0.0239 --- Late Carioca 0.1095 0.0310 0.0230 0.0102 Valley 0.0629 0.3551 0.0210 0.0165 Mean 0.0862 0.1931 0.0220 0.0134 +Growth habit classification used by the International Cali, Colombia. Center for Tropical Agriculture (CIAT), 3 :6 n1 8 5uuoe ° vNM""\ ck!“ V“ 'p‘n‘n; 9 CU- HS '1“. :..d numb I» V. v o I H U oi Fannbvvh a a.» S 45 Table 8. Leaf area ratio of 12 dry bean genotypes at four sampling dates after 50% flowering grown in 1984. Growth habits and Days after 50% flowering genotypes+ 4 12 20 28 -------- cmZ/ g ---—-—------ Type I Traditional Seafarer 159.1 119.2 72.4 17.1 Brazil-2 134.6 133.1 77.9 32.8 Mean 146.9 126.2 75.2 25.0 Architype Laker 143.5 135.7 95.2 39.7 C-49 162.1 129.9 70.8 28.0 Mean 152.8 132.8 83.0 33.9 Type II Traditional T-39 154.5 135.2 66.2 51.0 Nap-2 160.3 134.7 65.1 62.3 Mean 157.4 135.0 65.7 56.7 Architype C-20 131.1 133.4 74.6 44.7 S Valley 159.1 123.3 79.0 39.2 Mean 145.1 128.4 76.8 42.0 Type III Early Viva 171.2 124.7 106.9 32.8 Harris 145.8 111.9 72.6 22.4 Mean 158.5 118.3 89,8 27.6 Late Carioca 165.5 122.7 68.2 26.1 Valley 157.6 131.2 103.1 61.6 Mean 161.6 127.0 85.7 43.9 +Growth habit classification used by the International Center for Tropical Agriculture (CIAT), Cali, Colombia. s.a‘ 9 OCUOe . 'A VI-- JUMP. h-‘In‘l on “:9. ~D AI” .ufi 3. us. r. an “14 at A5 «a #L k ab. 6 ..i «G . Va \hu L PL a» .M an ab .Q.‘ $.- Yu‘ .s‘ v& .«u .3 :- v. M» a A 9.. T‘39 e 4 -2 H U ‘. A Mean .H. a a nu. . . an n: NU «Q «C s y «A ”V as 4. . nvu .nu h,» n\u Atv v ‘hlk «Q n v uv‘ . a n u a G An. “,u .~. p c .6 r4 .0 ..2 ~ . .0 .AH £1 v a V vi 2. v. ~ 4 as v a y .. a .3 n... :u «G H... L u r; Y i . . NJ Hun a. n . Hi ‘ . L. v. a . at: a . A: «a .1 A a . .. n e nns . 5 I a a- n I A .8 A v y s 46 Table 9. Leaf area per cm2 of 12 dry bean genotypes at four sampling dates after 50% flowering and growing 1984. Growth habits and Days After 50% Flowering genotypes+ 4 12 20 28 Type I Traditional Seafarer 9095.9 8889.5 7216.1 2037.5 Brazil-2 7132.2 7878.9 6570.1 3104.0 Mean 8114.1 8384.2 6893.1 2570.8 Architype Laker 7030.3 9101.4 8306.2 4623.7 C-49 9081.6 8444.0 6953.8 3484.5 Mean 8056.0 8772.7 7630.0 4054.1 Type II Traditional T-39 7691.7 10535.2 9678.7 8089.0 Nep-Z 8291.1 9371.8 8536.9 7677.0 Mean 7991.4 9953.5 9107.8 7883.0 Architype C-20 8391.7 11785.4 11118.1 6114.9 S Valley 10614.6 9467.1 9662.3 5828.9 Mean 9503.2 10626.3 10390.2 5971.9 Type III Early Viva 7646.7 6202.4 7839.6 2991.9 Harris 7528.4 7202.2 7071.1 2635.5 Mean 7587.6 6702.3 7455.4 2813.7 Late Carioca 7940.7 10300.1 7207.2 3643.5 Valley 9019.4 8590.0 11303.1 7763.7 Mean 8480.1 9445.1 9255.2 5703.6 +Growth habit classification used by the International Center for Tropical Agriculture (CIAT), Cali, Combia. O. Q 1‘ Table and 9 an 9 9 rue b} Fe Sea a. «3 .0 in. e . n... C «In nu 9 . a 3 “Us e . at V. I ”Kn nu. V ¢ n "(.0 «a «Nu L1. 5‘ H11 Q.» .1 . n: kn bk 8 5V f. ‘IA F4 a. .nu “V huh F‘ c Av AN» PM 4‘ . a :M V‘ ‘1‘ a» «a «an III” e hL h..- ‘t- «o 47 Table 10. Leaf area index of 12 dry bean genotypes at four sampling dates after 50% flowering and growing in 1984. Growth habits and Days after 50% flowering genotype 4 12 20 28 Type I Traditional Seafarer 3.0 3.2 2.4 0.8 Brazil-2 3.4 3.1 2.4 1.4 Mean 3.2 3.2 2.4 1.1 Architype Laker 3.3 3.6 3.5 1.6 C-49 3.9 3.6 2.8 1.3 Mean 3.6 3.6 3.2 1.5 Type II Traditional T—39 3.6 4.9 3.7 3.0 Nep-Z 3.8 3.5 3.5 2.3 Mean 3.7 4.2 3.6 2.7 Architype C-20 3.2 4.5 4.3 2.0 S Valley 4.1 3.2 4.0 2.2 Mean 3.7 3.9 4.2 2.1 Type III Early Viva 3.1 2.5 2.8 1.7 Harris 2.6 2.9 2.6 1.1 Mean 2.9 2.7 2.7 1.4 Late Carioca 3.4 3.5 3.1 1.5 Valley 3.1 3.8 3.8 3.3 Mean 3.3 3.7 3.5 2.4 +Growth habits classification used by the International Center for Tropical Agriculture (CIAT), Cali, Colombia. . n a: “a . ' C | Q 3 no It 9 3. WV. VF“ Ar‘ J . .c .k. :u m. a y o A "In . . A: a“ e A u 1‘ u h:— . .. v . a. a 3 e .1 .1 A R D u A ‘. My. pun“ ”I“ ”u a ,v 6 AH. a: a.“ .\u a. y ‘ ‘§ Q§"~ V nAo‘ .;: V.“ 48 thereafter, both parameters declined (Table 7 and 8 ). Seafarer, Nep-Z and Carioca had the highest RGR value among all genotypes in the study. All these values were observed at the early flowering stage for each genotype (Table 7 ). Genotypes response for LA and LAI traits followed the same pattern as LAR and RGR. The maximum values were observed at 4 or 12 days after 50% flowering. C-20 and T-39 reached the maximum figures within the genotypes evaluated in this study (Table 7 and 10 L Partitioning ratios: The HI is probably the most popular and most often used index of dry matter partitioning. However, other indices such as seed yield and biological yield efficiency have also been proposed as a selection criterion for plant efficiency. In this study HI, seed yield and biological yield efficiency were used to evaluate partitioning efficiency. On average, the Type III had a higher HIcompared with Type II. However, within Type I, Brasi1-2 had a higher HI than Seafarer despite the fact that the seed yield efficiency trait of the two genotypes was not significant (Table 11, 12 ). C-49 showed the highest yield efficiency value (0.82) and Brasi1-2 and Valley had the lowest ones U164 Table 10,:L1L Biological yield efficiency differences were found between Type II and Type III, TI and AI and within AII. In general, the Type II and AI forms were more efficient in ..: 6.5 49 Table 11. Harvest index, seed yield efficiency and biological yield efficiency of 12 dry bean genotypes grown in 1984. Growth habit Harvest Efficiency and genotypes Index Seed yield Biological yield ----------- g/day ----------—- Type I Traditional Seafarer 37.3 0.72 1.93 Brazil-2 47.5 0.64 1.36 Mean 42.4 0.68 1.65 Architype Laker 40.5 0.76 1.91 C-49 41.6 0.82 1.98 Mean 41.1 0.79 1.95 Type II Traditional T-39 40.6 0.76 1.85 Nap-2 40.3 0.75 1.86 Mean 40.5 0.76 1.86 Architype C-20 34.4 0.74 2.15 S Valley 40.0 0.75 1.87 Mean 48.4 0.99 2.01 Type III Early Viva 45.1 0.76 1.72 Harris 47.4 0.71 1.54 Mean 46.3 0.74 1.63 Late Carioca 45.3 0.72 1.68 Valley 43.7 0.64 1.50 Mean 44.5 0.68 1.59 +Growth habit classification used by the International Center for Tropical Agriculture (CIAT), Colombia 50 Table 12. Mean squares from single degree of freedom orthogonal comparisons of harvest index, seed yield efficiency and biological yield efficiency of 12 dry bean genotypes grown in 1984. Orthogonal Harvest Seed Yield Biological Yield Comparison Index efficiency efficiency Type I vs. Types II and III 1.21 0.02 0.01 Type II vs Type III 256.56** 0.16 0.64** Traditional I vs Architype I 5.47 0.04 0.28** Traditional II vs Architype II 31.59 0.16 0.08 Late III vs. Early III 9.72 0.01 0.01 Seafarer vs. Brazil-2 157.08** 0.01 0.49** Laker vs. C-49 1.71 0.01 0.01 T-39 vs. Nep-2 0.20 0.01 0.01 C-20 vs. S Valley 47.99 0.01 0.12* Viva vs Harris 8.17 0.01 0.05 Carioca vs. Valley 3.84 0.04 0.05 *,** Significant at 5% and 1% probability level, respectively. aaV'D‘ “; A v...voaV“"’ ~—_ ‘Av “" 4 .vd‘ctev 23 FM .i. :crphclog Q: Q n.“ u... r; ‘K an EU; F. a u. re .3 a: ‘4 .~\ 51 producing dry matter than the TI and Type III genotypes, respectively. In addition, C-20 had the highest biological yield efficiency compared with the other genotypes (Table 10, ll ). Correlation studies. Simple correlation coefficients were calculated to study the relationships between pairs of morphological traits, and yield, and yield components (Table 13 ). Grain yield was positively correlated with biological yield, days to 50% flowering, and days to physiological maturity. There was a positive association between number of pods per m2 and yield, biological yield, days to 50% flowering, and physiological maturity. The number of seeds per pod trait was positively correlated with yield, days to 50% flowering and days to physiological maturity, but negatively correlated with 100-seed weight. The weight of 100 seeds was negatively correlated with seed and biological yield. Biological yield was positively correlated with days to physiological maturity but negatively correlated with HI (Table 13 L A simple correlation analysis was performed to study linear relationships between Some (HE the growth characteristics and seed yield. Leaf dry weight, total dry ‘weight and LA were significantly and positively related to yield at 12 days after 50% flowering. Total dry weight was “I Q '..'~ 'A ‘ mr‘ ”a“: .3, bdnfse . . A an nu ‘v- ‘- q Q n 2 HVV‘VI ‘ W! I ‘ 3... A“ 'wn u. ‘ o «1. ~4— “ . om ‘ ¢ ' ~— ‘ u. A . . ' 1. .. C ‘ V. v Ill- . _‘ ”v 'm h. O F“ A A -u.., I _ i ' l V. V 52 Table 13. Simple correlation coefficients between yield, yield components, biological yield, days.to 50% flowering and physiological maturity of 12 dry bean genotypes grown in 1984. Pods Seeds 100 Seed Seed dry Biological Days to Days to / m2 / pod weight weight yield 50% physiol . (No.) (No.) (9) (g) (g) flower. maturity No. of seed/pods 0.355 100 seed weight -0.629* -0.393* Seed dry weight 0.841** 0.586* -0.539 Biological yield 0.944** 0.336 -0.637 -0.824** Days to 50% flowering 0.600* 0.374* -0.133 0.725** 0.522 Days to physiological maturity 0.646* 0.512** 0.315 0.749* 0.587 0.839** Harvest Index -0.687 0.017 0.557 0.351 -0.812** -0.103 -0.187 *,** Significant at 5% and 1% probability level, respectively. 53 also correlated with yield at 12 days after 50% flowering. The trait LAI, was positively associated with yield at 4, 12 and 20 days after 50% flowering. The LAR and RGR were not associated with yield at any of the plant growth stages considered in this study (Table 14 ). 54 Table 14. Simple correlation coefficients dry beans, biological yield, leaf area, leaf area index, and relative growth rate of dry beans at four sampling dates after 50% flowering grown in.1984. Correlation Coefficient Days after Dry Biological Leaf Leaf Leaf Relative 50% weight Yield Area Area Area growth flowering leaves Index ratio rate 4 0.183 0.243 0.154 0.359* -0.059 0.119 12 0.444** 0.347** 0.471** 0.344* 0.233 0.320 20 -0.105 0.177 0.063 0.430** 0.160 --- 28 0.288 0.434* 0.276 0.111 0.117 --- *,** Significant at 5% and 1% probability level, respectively. Discussion Seed yield in dry beans can be described in terms of a multiplicative index of components of yield which are the number of pods per plant ( X ), seeds per pod ( Y ) and weight of seed ( Z )( 1 ). Since yield is determined by physiological events occurring during seed filling, one might gain insight into this complex character by dissecting yield into its components. The failure in) find significant differences in biological yield, seed yield and the components of seed yield between Type I growth habit and those genotypes with Type II and III indicated that plants with a determinate growth habit accumulated equivalent total dry matter as the indeterminate ones. This finding agrees with previous reports ( 32 ) in soybean and dry beans ( 44 ) which pointed out that some determinate genotypes could produce as much dry weight as indeterminate ones. The differences in 100—seed weight among growth habits was probably due to the use of genotypes with ‘varing seed weights. For example, seeds of the Great Northern cmmnnercial class have average seed weights of 26 9 while navy beans generally have seed weights ranging from 1735 to 20.5 g (Table 1, 2 L 56 Type II genotypes had higher seed yields and yields of vegetative parts than Type III genotypes. The higher seed yield of Type II compared to Type III could be due to the 2 and number of seeds per pod. higher number of pods per m Despite the fact that Type IIIlgenotypes had heavier seeds, this characteristic, apparently could not compensate for the 2 and seeds per pods. reduced number of pods per m 0n the average there were no significant differences in seed yield efficiency between Type II and Type III. However, Type III genotypes showed a higher HI than Type II. Besides, HI was not correlated with seed yield. This agrees with Schapaugh and Wilcox ( 45 ) who found no relationship between HI and seed yield in soybean. The lack of association between HI and seed yield efficiency suggested that although a strain might have a high HI it may be inefficient in its daily production of dry matteru The HI is determined by dividing the seed (economic) yield by the total biological yield but does not represent the daily production of seed dry matter (economic yield) as does the yield efficiency trait. A high yielding genotype should have a high HI and seed yield efficiency rather than just a high HI. It was interesting to note that within each growth habit there were significant differences amongifiuagenotypes for nearly EHJ. the traits studied. This indicated that 57 significant variability existed within the growth habits and suggested that characteristics normally associated with a specific growth habit may not be obligatorily associated in a physiological sense. If commonly associated traits are linked in the genetic sense such as prostrate growth habit and lateness then such linkages have been possibly broken. For example, early maturing genotypes (Viva and Harris) as well as late maturing (Carioca and Valley) characterized the Type III growth habit used in this studyu Genotypes with a Type III growth habit are traditionally considered to be late in maturity ( 40 ). Late flowering and maturing genotypes ( AI and AII and LIII) outyielded those that flowered and matured earlier. In general, Type I and III taken as a group, flowered and matured earlier than Type II (Table 4, 5). This could be another reason for the lower yield of genotypes within Type I and III growth habits compared with genotypes with indeterminate habits (Type II). This finding agrees with Dunphy et a1 ( 17 ) who reported that the number of days from emergence to flowering and to physiological maturity stages were positively correlated with seed yield in soybean. The late maturing and determinate genotypes (AI) outyielded the earlier maturing ones, but days to harvest in this group seemed to be less critical than for AII and LIII genotypes, which were generally later than TI (Table 4 L 58 These advantages, in dry beans would be more pertinent if adverse weather is prevalent during the harvest season. The use of a short season cultivar might be advantageous because it allows harvest to occur before weather becomes inclement in mid-autunun This is similar to the possible benefits of earliness in soybeans. Beaver and Copper ( 5 ) and Reicosky et al ( 43 ) indicated that earlier soybean cultivars could have different benefits. For example, the earliness characteristic could permit seed filling to occur during a period of greater light intensity and duration. It might also allow the period of seed fill to occur during a time of lower probability of moisture deficit and it provides a benefitial effect of reducing some of the reproductive leaf area considered superflous for soybean grown in narrow rows. A short season dry bean cultivar might also have another advantage over later maturing ones in some areas where the planting date must be delayed until late spring or early summer because other crops are more profitable and are planted first, or dry beans could be planted to complete a second crOp in a year; Otherwise, a medium to full season dry bean cultivar should be an important selection criterion to insure a high seed and biological yield. However, an earlier but a more efficient dry matter producing cultivar should not be discarded because it could compete with some mid-season cultivars. It seems that genetic variability for differences 59 in degrees of efficiency does exist.in the dry beans asxnas demonstrated by this sample of genotypes (Table 11 ). Other physiological important traits, such as the LAI, which measures the extension and duration of leaf surface is generally correlated with the phothosynthetic potential of the plant ( 27 L This was demonstrated in this study because dry matter accumulation was directly dependent on the leaf area per unit of land. It has been shown that many crops have an optimum leaf area index ( 27 ). In dry beans, Wallace et al ( 54 ) reported that the maximum LAI estimated was 3.9 in a sample of dry bean genotypes studied. Leaf area index was positively correlated with seed yield at 4, 12 and 20 days after 50% flowering (Table 10 L Since, Type II had a higher LAI at 4, 12 and 20 days after flowering than Type I and III, this may be one of the factors for the higher yield of Type II compared to genotypes with other growth habits..Although the LAI of Type I was high at 4 days after 50% flowering it was lower than Type II and III at 12, 20 and 26 days. The advantage of having a higher LAI at early stages seemed to be one of the factors for a higher yield of Type II and III compared with Type I. Normally for any crop the LAI at flowering is closely related with seed yield because leaf area is closely associated with production of carbohydrates needed for seed filling. Thus, an early development of a leaf canopy and maintenance of maximum leaf 60 area should be important for high yields. For example, AII and the late-maturing Type III genotypes had a high LAI through 16 days after 50% flowering (Table 10 ). In general, plants with the LAI below an optimum have a reduced dry matter production because not all the sunlight is intercepted and photosynthesis is not at a maximum rate. Plants with the LAI above an optimum have a reduced dry matter production because the lower leaves are shaded and the net assimilation rate is reduced ( 40 ). The optimum LAI occurs when leaves intercept 95% of the light energy ( 22 ). The ability to intercept light is also related to the distribution and size of the leaves. In this regard, growth of determinate plants with large leaves, which.overlaplone another due to the relatively short internodes characteristic of this growth habit may limit the available light which becomes more limiting under high plant density ( 40 ). The leaf area ratio of each genotype decreased with time during the reproductive stage (Table 8 ). This may be due to the fact that the dry matter accumulation was more rapid in non-leaf tissue than leaf tissue. The LAR was not correlated with.yield at any stage of growth in this study. Similar :results were observed by Wallace et2$l( 53 ) in dry beans. In this study, the CGR followed the same trend as LAR (Table 7, 8 ) and it was not correlated with yield.'Phis suggested that a possible competition for photosynthate between the 61 vegetative and reproductive organs occurred during flowering and seed production. Total dry matter production is the summation of crop growth rate over the entire growth period. Correlations clearly indicated that the association between seed yield and dry matter was positive and significant over a period of time (Table 13 ). Thus, the highest yields within each growth habit were associated with genotypes with the largest dry weight accumulations. Grain yield depends upon photosynthetic activity occurring during reproductive growth and.1flua remobilization of photosynthate into developing seeds. If one considers the dry matter accumulation of leaves of different growth habits, it is evident that Type II genotypes accumulated their highest dry matter at the early stages of growth. The possibility of having a high dry matter accumulation and long duration of a high LAI at flowering time could lead to a greater carbohydrate production. This might be another physiological basis for thejyield advantage for the Type II over Type I and III genotypesuThis hypothesis agrees with the positive and significant association observed between grain yield and dry weight of leaves at 12 days after 50% flowering (Table 13 L Normally, carbohydrates such as simple sugars, starch, and other polysacharides reach a maximum concentration in the 62 plant about the time of flowering after which time they decrease ( 4 ). The stored carbohydrates could also be remobilized into the seed thus contributing to seed filling production or it could be consumed by respiration or used for additional vegetative growth. Therefore, the loss of dry weight from vegetative parts (stems) during the seed filling period could give one approximation of the contribution of the stored carbohydrates to the weight of the developing seed. At the time of harvest, Type II‘genotypes produced the highest total dry matter among the three growth habits but had similar stem dry weights as Type I and Type III genotypes (Table 15 L This suggested that the Type II genotypes may have remobilized.carbohydrateanulother compounds from the stem. For example, the All genotypes had 33~4% and 28.6% dry matter in stems and pod walls respectively while TII genotypes had 27.4% and 29.8% in these structures, respectively (Table 15 L These results suggested that the TII genotypes remobilized stored carbohydrate to a greater degree than the AII. Adams et al ( 4 ) suggested that carbohydrate quantity varied significantly with the stage of development among dry bean genotypes but usually declined as seed filling was completed. However, these authors ( 4 ) cxmald not associate the change in plant starch that occurred Iflith seed yield. In the present study, All had a higher 63 Table 15. Biological Yield, stem dry weight and pod wall dry weight of 12 dry bean genotypes grown in 1984. Growth Pod habits wall and Biological Stem dry dry genotypes+ yield weight weight ________________ g/m4__-____—--------__- Type I Traditional Seafarer 261.3 80.4 83.6 Brazil-2 222.6 66.2 51.0 Mean 242.0 73.3 67.3 Architype Laker 314.5 102.2 85.6 C-49 313.5 94.6 88.8 Mean 314.0 98.4 87.2 Type II Traditional T-39 318.0 76.2 100.6 Nap-2 308.4 95.2 86.0 Mean 313.2 85.7 93.3 Architype C-20 354.8 126.8 106.2 S Valley 316.2 93.4 96.8 Mean 335.5 110.1 101.5 Type III Early Viva 275.4 79.6 71.0 Harris 232.8 74.2 54.0 Mean 254.1 76.9 62.5 Late Carioca 214.8 58.2 55.8 Valley 240.9 82.8 53.2 Mean 227.9 70.5 54.5 +Growth habit classification used by the International Center for Tropical Agriculture (CIAT), Colombia. 64 percentage of leaf dry weight than TII genotypes (Table 16 ) and therefore more source to produce a higher level of carbohydrate for seed production. The remobilization of carbohydrates from the stem may have been greater for AII. Nevertheless, the amount of carbohydrate that apparently was translocated from stems seemed enough to produce a high level of yield in AII. It has been suggested ( 1 ) that bean plants with the architype form (AI and AII) might obtain their high yields compared to the traditional genotypes because of remobilization from storage sites (stems, roots) and especially during periods of stress ( 1 ). When the maximum stem production and the stem dry weight at physiological maturity were compared, it was observed that the reduction in stem dry weight (remobilization) during this period was also less for the architype genotypes. For example, All and TII apparently remobilized 3.8 and 7.7% dry matter, respectively. Most of this is expected to be carbohydrate. Within the Type III genotypes the early maturing ones appeared to remobilize more carbohydrate than late maturing genotypes during the same period of time (Table 17 ). The stem dry weights of the 12 genotypes also illustrated the variability in the morphology and growth habits that existed in this experiment. The AII and AI were taller and less branching compared with TII and TI, 65 Table 16. Leaf dry weight of 12 dry bean genotypes at four sampling dates after 50% flowering grown in 1984. Growth habit and Days after 50% flowering genotypes+ 4 12 20 28 ----------- g / pl ----------- Type I Traditional Seafarer 5.5 5.8 5.1 1.3 Brazil-2 5.2 5.7 4.6 2.0 Mean 5.4 5.8 4.9 1.7 Architype Laker 4.6 6.2 5.2 2.8 C-49 5.4 5.1 3.9 2.3 Mean 5.0 5.7 4.6 2.6 Type II Traditional T-39 5.6 5.7 5.9 4.9 Nep—2 4.7 6.2 5.3 4.0 Mean 5.2 6.0 5.6 4.5 Architype C-20 6.0 7.2 6.5 4.0 8 Valley 6.5 6.9 5.8 3.9 Mean 6.3 7.1 6.2 4.0 Type III Early Viva 4.2 4.0 4.5 1.7 Harris 5.2 4.9 4.8 2.0 Mean 4.7 4.5 4.7 1.9 Late Carioca 4.3 6 7 4.9 2 7 Valley 5.6 5 7 7.7 4 9 Mean 5.0 6 2 6.3 3 8 +Growth habit classification used by the International Center for Tropical Agriculture (CIAT), Cali, Colombia. I ,- 66 Table 17. Stem dry weight, pod dry weight (pod wall and seed dry weight) and stem pod dry weight ratio for 12 dry bean genotypes and grown in 1984. Growth habits and Stem++ Pod++ Stemzpod genotypes dry weight dry weight ratio ---------------- g/pl ---------------- Type I Traditional Seafarer -1.0 8.7 11.5 Brazil-2 -0.5 2.1 23.8 Mean -0.8 5.4 17.7 Architype Laker -0.6 2.7 22.2 C-49 -0.7 9.6 7.3 Mean -0.7 6.2 14.8 Type II Traditional T-39 -0.4 12.4 3 2 Nep-2 -0.7 5.8 12 1 Mean -0.6 9 1 7 7 Architype C-20 -0.3 8.0 3.8 S Valley -0.3 7.7 3.9 Mean -O.3 7.9 3.9 Type III Early Viva -0.4 7.9 5.1 Harris -1.0 8.2 12.2 Mean -0.7 8.1 8.7 Late Carioca -0.7 14.3 4.9 Valley -0.6 12.6 4.8 Mean -0.7 13.5 4.9 +Growth habit classification used by the International Center for Tropical Agriculture (CIAT), Cali, Colombia. ++Differences between maximum and harvest dry weight. 67 respectively; Late cultivars produced more stem dry weight than.the early ones Crable 17 L It was also observed in this study that the development of pod walls was completed first and then a rapid increase in the dry weight of seeds occured (Table 18 L This suggested that the development of pod wall is a prerequisite for a rapid filling of the seed. This agrees with results published ixisoybeans( 25, 26 L Thus, across all genotypes, growth habit, architectural forms and maturity groups, each physiological factor made a relative large or small contribution towards the seed yield of each genotype but the importance of one morphological factor as compared to another varied between and within the groups of genotypes. 68 Table 18. Pod wall dry weight at four sampling dates after 50% flowering of 12 genotypes grown in 1984. Growth habit and Days after 50% flowering genotypes+ 4 12 20 28 ---------- g / pl -------- Type I Traditional Seafarer 0.0 0.3 0.8 1.2 Brazil-2 0.0 0.0 1.2 1.2 Mean 0.0 0.2 1.0 1.2 Architect Laker 0.0 0.0 1.0 1.2 C-49 0.0 0.4 1.0 1.5 Mean 0.0 0.2 1.0 1.4 Type II Traditional T-39 0.0 0.1 0.7 0.9 Nap-2 0 0 0.1 0.6 1.0 Mean 0.0 0.1 0.7 1.0 Architype C-20 0.0 0.3 0.8 1.2 S valley 0.1 0.2 0.9 1.2 Mean 0.1 0.3 0.9 1.2 Type III Early Viva 0.0 0.2 0.6 0.5 Harris 0.0 0.3 0.7 0.9 Mean 0.0 0.3 0.7 0.7 Late Carioca 0.0 0.5 1.0 1.6 Valley 0.0 0.2 0.7 1.3 Mean 0.0 0.4 0.9 1.5 +Growth habit classification used by the International Center for Tropical Agriculture (CIAT), Cali, Colombia I! BIBLIOGRAPHY lnAdams, NLW. 1967. Basis of yield component compensation in crop plants with special reference to the field bean, Phaseolus vulgaris . Crop Sci. 7:505-509. ZuAdams, NLW. 1981. Update: new bean architypes. Michigan Dry Bean Dig. 5:12-13. 3.Adams, M.W. 1982. Plant archytecture and yield breeding. Iowa State J. Res. 56:225-254. 4.Adams, M.W., J.V. Wierma, and J. Salazar. 1978. Differences in starch accumulation among dry bean cultivars. Crop Sci. 18:155-157. 5.Beaver, J.S. and R.L. Cooper. 1982. Dry matter accumulation patterns and seed yield components of two indeterminate soybean cultivar. Agron. J. 74:380-383. 6.Beaver, J.Sn (LV; Paniagua, D.P. Coyne and G.F. Freytag. 1985. Yield stability of dry bean in the Dominican Republic. Crop Sci. 25:923-926. 7.Binnie, RJL 1981. Flower and.pod production in Phaseolus yplgaris. J. Agric. Sci., Camb. 97:397-402. 8.Bramel, P.J.,P.N. Hinz, D.E. Green and R.M. Shibles. 1984. Use of principal factor analysis in the study of three stem termination in types of soybean. Euphytica 33:387-400. 9.Buttery, BLR. 1970. Effects of variation in leaf area index on growth of maize and soybean. Crop Sci. 10:9-13. 10.Buzze1, R.I. and B.R. Buttery. 1977. Soybean HI in hill plots. Crop Sci. 17:968-970. 11.Chang, J.F” ILE. Green and R. Shibles. 1982. Yield and agronomic performance of semi-determinate and indeterminate soybean stem type. Crop Sci. 22:97-101. 12.Coyne, DJA.1980. Modification of plant archytecture and crop yield by breeding. Hort. Sci. 15:244-247. 69 7O 13.Crookston, R.K” J. O'Toole and J.L. Ozbun. 1974. Characterization of been pod as a photosynthetic organ. Crop Sci. 14:708-712. 14.Donald, C.M. 1967. The breeding for crop ideotypes. Euphytica 17:385-403. 15.Donald, C.M. and J. Hamblin. 1976. The biological yield and harvest index of cereals as agronomic and plant breeding criteria. Adv. Agron. 28:361-405. 16.Duncan, qu. 1971. Leaf angles, leaf area and canopy phothosynthesis. Crop Sci. 11:482-485. l7.Dunphy, ELJ., J.J. Hanway and D.E. Green. 1979. Soybean yields in relation to days between specific developmental stages. Agron. J. 71:917-920. 18.Eastin, J.A. and E.T. Gritton. 1969. Leaf area development, light interception and the growth of canning peas (Pisum sativum L.) in relation to plant population and spacing. Agron. J. 61:612-615. 19.Egli, D.EU, J. Fraser, J.E. Leggett and C;H. Ponleit. 1981. Control of seed growth in soybeans (Glycine max L. MerrilL Ann.Bot.48:l71-176. 20.Egli, D.E. and J.E. Leggett. 1973. Dry matter accumulation patterns in determinate and indeterminate soybeans. Crop Sci. 13:220-222. 21.Egli, D18. and J;E. Leggett. 1976. Rate of dry matter accumulation in soybean seed varying source sink ratios. Agron.;L 68:371-374. 22.Emecz, T.I. 1962. Suggested amendments in growth analysis and.potentiality assessment in relation to light.)huL Bot. 26:517-527. 23.Eeuwens, C.J. and W.W. Schwabe. 1975. Seed and pod development in Pisum sativum L. in relation to extracted and applied hormones. J. Exp. Bot. 26:1-14. 24.Fehr, qu., B.K. Lawrence and T.A. Thompson. 1981. Critical stages for soybean defoliation. Crop Sci. 21:259- 262. 25.Frank, S.J. and W.R. Fehr. 1981. Associations among pod dimentions and seed weight in soybean. Crop Sci. 21:547-550. 71 26.Fraser, J., D.E. Egli and J.E. Leggett. 1982. Pod and seed development in soybean cultivars with differences in seed size. Agron. J. 74:81-85. 27.Goldsworthy, P.R. 1984. CrOp growth and development: The reproductive phase. p.163-212. In: P.R. Goldsworthy and N.M. Fisher (Eds). The physiology of tropic field crops. John Willy & Sons. New York. 28.Gbikpsi, P.j. and K. Crookston. 1981. Effect of flowering date on accumulation of dry matter and protein in soybean seeds. CrOp Sci. 21:652-655. 294Hanway, JZJ. and CLR. Weber. 1971. Dry matter accumulation in eight soybean varieties. Agron. J. 63:227-230. 30.Hartung, R.C., J.E. Specht and J.H. Williams. 1981. Modification of soybean plant archytecture by genes for stem growth habit and maturity. Crop Sci. 21:51-56. 31.Hedley, C.L. and M.J. Ambrose. 1980. An analysis of seed development in Pisum sativum L. Ann. Bot. 46:89-105. 32Hicks, D.R. J.W. Pendleton, R.L.Bernard and T.J. Johnston. 1969. Response of soybean plant type to planting patterns. Agr. J. 61:290-293. 33.Hsu, F.C. 1979. A development analysis of seed size in common bean. Crop Sci. 19:226-230. 34.Izquierdo, J.A., and G.L. Hosfield. 1983. The relationship of seed filling to yield among dry beans with differing architectural forms. J. Amer. Soc. Hort. Sci. 108:106-111. 35.Izquierdo, J.A. 1980. The effect of accumulation and remobilization of carbon assimilate and nitrogen on abscission, seed development, and yield of common bean (Phaseolus vulgaris L.) with differing architectural forms. PhD Thesis, Michigan State Univu, East Lansing. 36.Johnson, D.R. and D.J. Major. 1979. Harvest index of soybeans as affected by planting date and maturity rating. Agron. J. 71:538-541. 37.Kelly; J.DH,CLV3 Varner, M.W. Adams. 1986. The effect of different dry bean growth habit on yield stability. Bean Improv. Coop. 29:58-59. 38.Kueneman, E.A. and D.H. Wallace. 1979. Simplified growth analysis of non-climbing dry beans at three spacing in the 72 tropics. Expl. Agric. 15:273-284. 39.1(uememan, E.A., G. Hernandez-Bravo and D.H. Wallace. 1978. Effects of growth habits and competition on yield of dry beans (Phaseolus vulgare ) in the tropics. Expl. Agric. 14:99-104. 40.Laing, D.R., P.G. Jones and J.H.C. Davis. 1984. Common bean (Phaseolus vulgaris I“), pn305-351. In: PgR. Goldsworthy and NlM. Fisher (Eds.). The physiology of tropical field crops. John Willey & S Sons. New York. 41.Nienhuis, J. and S. Singh. 1986. Combining ability analysis and relationships among yield, yield components, and architectural traits in dry beans. Crop Sci. 26:21-27. 42.Radford, P.J. 1967. Growth analysis formulae-their use and abuse. Crop Sci. 7:171-175. 43.Reicosky, D.A., J.H. Orf and C.H. Poneleit. 1982. Soybean germplasm evaluation for length of the seed filling period. Crop Sci. 22:319-322. 44.Robitaille, H.A. 1978. Dry matter accumulation patterns in indeterminate Phaseolus vulgaris L. cultivars. Crop Sci. 18:740-743. 45.8chapaugh, JruW.T. and J.R. Wilcox. 1980. Relationships between.harvest indices and other plant characteristics in soybeans. Crop Sci. 20:529-533. 46.Shibles, R.M. and CLR. Weber. 1965. Leaf area, solar radiation and dry matter production by soybean. Crop Sci. 5:575- 577. 47.Singh, SJL 1982. A key for identification of different growth habits of Phaseolus vulgaris L. Bean Improv. Coop. 25:92-95. 48.Sinha, S.K., S.C. Bhargava and A.Goel. 1982. Energy as the basis of harvest index. J. Agric. Sci., Camb. 99:237-238. 49.Spaeth, S.C., H.C. Randall, T.R. Sinclair and J.S. Vendeland. 1984. Stability of soybean harvest index. Agron. J.76:482-486. 50.Thorne, J.H. 1979. Assimilate redistribution from soybean pod walls during seed development. Agron. J. 71:812-816. 51.Taylor, J. 1975. A modified air planter ("plantair") for 73 use in experimental plots. Bean Improv. Coop. 52.Wallace, D.H. and H.M. Munger.1966. Studies of the physiological basis for yield differences. 1. Growth analysis of six dry bean varieties. Crop Sci. 5:343-348. 53.Wallace, D.H. and H.H. Munger. 1966. Studies of the physiological basis for yield differences. II. Variations in dry matter distribution among aerial organs for several dry bean varieties. Crop Sci. 6:503-507. 54.Wallace, D.Hn, J.L. Ozbun and ILNL Munger. 1972. Physiological genetics of crop yield. Adv. Agron. 24:97-146. 55.Wilcox, J.R. and T. Sedimaya. 1981. Interrelationships among height, lodging and yield in determinate and indeterminate soybeans. Euphytica 30:323-326. 56.Wilcox, J.R. 1980. Comparative performance of semideterminate and indeterminate soybean lines. Crop Sci. 20:277-280. CHAPTER 2 VARIABILITY IN SEED FILLING CHARACTERISTICS AND THEIR INTERRELATIONSHIPS WITH YIELD IN DRY BEANS WITH DIFFERING GROWTH HABITS, ARCHITECTURES AND MATURITIES Abstract The objective of this research was to study seed filling in a population of dry beans with different growth habits, architectural forms and maturity dates and ascertain interrelationships among variation in morphological characteristics, seed filling and yield. Plants samples from 0.5m2 were taken at four days intervals for approximately 36 days during the filling period. A cubic polynomial model was fitted to the dry weight data. Coefficient of determination ( R2 ) varied between 0.936 to 0.992 for all genotypes. Grain filling rate and duration were calculated fitting the data to a linear model until the deviation from the mean square of the regression analysis was minimum. Correlation between the seed filling Characteristics indicated that linear filling duration was positive and significantly correlated to yield, number of 9093 Per m2, number of seed per pod, biological yield and phYSiOlogical maturity. Total filling duration was only 74 75 positively correlated with days to 50% flowering and physiological maturity.lwean filling rate was positive and significantly correlated with number of pods per m2, 100-seed weight and negatively correlated to number of seed per pod, yield, and biological yield. In general, Type I tuul a shorter total and linear filling duration, and a faster linear filling rate than Type II and III. Architypes presented either a similar or a slightly longer total and linear filling duration to traditional forms. However, late genotypes tended to have a longer total and linear filling and a slower linear filling rate than earlier genotypes. Introduction Yield in crops used for their seeds is a complex character that is generally correlated with the first order yield components e.i. pods or spikes per plant, seeds per pod or spike, and seed weight. Each component is the final product of certain physiological events that is controlled by several genes. A change in any of the components may bring about a change in final yield. Therefore, it can be perceived that if each yield component can be altered in a positive direction, a maximun expression of yield should be obtained. Genotypic variation in the physiological process governing the manifestation of these components such as seed growth rate and duration and canopy development should enable one to develop a strategy for improving yield. In this context, two process which are related to yield in a physiological sense are the effective filling duration ( EFD ) and effective filling rate ( EFR ). A study of these parameters and their relationships with morphological characteristics may provide insight to a breeder attempting to redesign plant architecture to maximize yield. Recent studies in dry beans (Phaseolus vulgaris L.) (17 ), soybean (Glygipe mgx ( L ) Merr ) ( 12, 20, 30 ), 76 77 wheat (Triticum aestivum L) ( 15, 25 ), corn (Zea mays L.) ( 4, 5, 7, 16 ), and rice (Oryza sativa L.) ( l9 ), have demonstrated that species and cultivars within species differ in both rate and duration of grain filling. The presence of genetic variation for seed filling period and rate and their positive correlation with yield, indicates that these parameters could be a useful as selection criterion for breeders to improve seed yield. The choice of the appropiate selection strategy for increasing EFD and / or EFR depends on the kind of genetic effects controlling these traits. If EFD and / or EFR are controlled by genes with largely additive or additive types of epistatic effects then it should be possible to improve these characters through recurrent selection. The objective of this research was to study grain filling in a population of dry beans with different growth habits, architectural forms and maturity dates and to ascertain interrelationships between morphological characteristics, seed filling traits and yield. Literature review Variation ip seed filling. The grain filling duration has been shown to influence yield in several crop species. A number of studies in corn ( 4, 5, 6, 16, 26 ) demonstrated that differences in cultivar yield were associated with differences in the length of the seed filling period» There is evidence suggesting that the grain filling duration is important in determining yield in corn. Studies have shown true presence of genetic variability for the effective filling duration (EFD) of kernels in corn ( 4, 5, 6, 15 L Moreover, thiAS trait responded to selection ( 3, 5, 18, 23, 27 ). :huiicating that the heritability was probably high. Poneleit and Egli ( 27 ) reported reciprocal differences among F1 hybrids in corn for both EFD and kernel growth rate. However, an analysis of the F2 indicated that the reciprocal differences noted in the F1 were not of cytoplasmic origin but were probably caused by a dosage effect in the kernel CYtOplasm. In small grains, the situation regarding seed filling is “0t Clear as for corn. Gebechow et a1 ( 15 ) found a positive Islaticnuship between grain filling duration and yield in durum wiieat (Triticum turgidum 1“). However, Nass and Reisser 78 p 79 ( 25 ), and Wych et al ( 35 ) found no relationship between grain filling duration and yield in spring wheat and oats. Similarly, Jones et al ( 19 ) found no significant association between grain filling duration and the components of yield in rice. In barley, Rasmusson et al (29 ) reported large differences in grain filling duration. Based on the heritability of the trait, they concluded that selection based on means of genotypes in replicated traits would be effective. Askel and Jonhson ( 2 ) found that barley cultivars with a long vegetative period tended to produce more kernels per spike and a higher yield than those with a long grain filling period. Metzger et a1 ( 24 ) studying different advanced populations of barley concluded that there was no significant association between grain filling duration and yield. Hence, they concluded that selection for seed filling duration does not appear to be a worthwhile objective in barley breeding. Watson et a1 ( 34 ) reported that recently developed cultivars of both winter and spring wheats yielded more than older cultivars, partly because their ears emerged sooner and carried on photosynthesis for a longer duration. This prolonged the period of grain growth. In small grains and corn, the increase in grain weight is approximately linear with time over the major period of seed filling ( 23 ) and the final seed characteristics can 80 be described as the product of the rate of grain filling and the duration of the EFD ( 6 ). On the other hand, seed filling in grain legumes occurs sequentially on fruiting branches that vary in number among genotypes. This makes the number of flowering nodes on branches a major determinant of seed filling ( 23 ). Thus, in legumes the rate and the duration of seed filling which determines the total seed weight are both strongly affected by the flowering pattern. This causes seed filling in legumes, to have periods of rapid growth separated by distinctive lag periods characterized by a slower growth. These changes were attributed to numrphological changes occurring during seed maturation ( 7 ). Ix: this regard, yield differences may not be indicative of the: existent of variability in the filling of individual seeds. Gay et al ( 14 ) compared the performance of traditional and new soybean cultivars and found that, higher yields of the rmew cultivars of maturity group III were associated with a Longer seed filling period. In addition, Mc Blain and Hume ( 22 ,) reported similar results working with early maturity soybean cultivars (group 0). Reicosky et a1 ( 30 ) reported a wide range in the seed filling duration among a number of soybean genotypes (maturity group III through group V). The length of the seed fithx; period of the genotypes was under genetic control and 81 was highly correlated with years. Egli et a1 ( 12 ) demonstrated that the seed filling period was related to yield in soybean and may be a useful selection criterion for breeding; These authors also found significant genotype x year interaction effects on seed filling which was attributed to moisture stress of unadapted cultivars. Egli et a1 (1J.) also studied the effect of seed size and position on the rate and duration of the filling period in soybean and concluded that seed growth rates were relatively constant across the early and late pods. The chiration of EFP was shorter for late pods on indeterminate Ctultivars leading to slightly smaller seeds. Mahon and Hobbs (1K3) studying several pea cultivars corncluded that the effective duration of seed filling was not influenced by the genotype or year of production. The final pod ineight was strongly correlated with rate of filling and whet: all genotypes in the study were considered, the rate of pod filling was correlated with lOO-seed weight, and final seed yield. Izquierdo and Hosfield ( 17 ), investigated the reLationship between seed filling and yield among dry bean cultivars of differing architectural forms and concluded that heavy seed of architypes (1J7) compared to the traditional determinate type appeared to be due to a longer filling duratton, and the ability to prolong the duration of 82 photosynthesis. Linear filling duration was significantly and positively correlated with yield and seeds per pod, seed per m2 and seed weight. On the other hand, the linear filling rate was negatively correlated with seeds per pod, seeds per m2 and yield ( l7 ). Methods o_f measuring seed filling. A number of methods have been used to measure the length of the seed filling period. The EFP measurement was proposed by Daynard et al (55) for use in corn.(geg mgyg L.L This parameter was calculated by dividing the final seed weight by the rate of accumulation of dry matter in the seed during the linear phase of accumulation. The EFD has been used to estimate the length of the seed filling period based on the total seed development over the entire plant ( 22 ). Johnson and Tanner ( 18 ) developed a technique for calculating rate and duration of dry matter accumulation in corn. This technique was based on a regression analysis of yield over time. These authors ( 18 ) concluded that 90% of the Zlinear accumulation of grain dry matter occurred within two and one half weeks after silking in corn. Hence, the duration of the grain filling period in corn can be estimated flnmn two or three sampling dates between 3 and 6 weeks after Silking. 131s growth stage system described by Fehr and Caviness ( 13 ) has been used to estimate the beginning and the end of 83 seed filling in soybeans on a whole plant basis. Both growth stages 4 and 5 have been used as indicators of the beginning of seed filling ( 8, 14 ) while growth stage 7 has been used to estimate the end of seed filling. Boote et a1 ( 3 ) defined the beginning of seed filling as detectable seed swelling in any pod on the soybean plant and the end as growth stage 8. Reicosky et a1 ( 30 ) defined the beginning of seed filling in soybeans as detectable seed swelling in the first pod to develop on the plant and the end of seed filling as either the first pod turning yellow or the first pod reaching its mature color. Spaeth and Sinclair ( 32 ) reported that the linear increase in harvest index during most of the soybean seed filling period provided another alternative for characterization the length of seed filling. Salado-Navarro et al ( 31 ) compared several visual estimates of seed filling duration and concluded that the usefulness of EFP, reproductive period duration and R5 (beginning seed formation) to R7 (beginning maturity) in soybean as selection criterion appears to be limited by the existence of highly significant genotype by environmental interactions. In dry beans, Izquierdo and Hosfield ( l7 ) estimated the seed filling parameters (rate and duration) by fitting curves to the data using equations appropiate for a first, second and third degree polynomial. Materials and methods Twelve genotypes were used in this study. The genotypes represented different growth habits, architectural forms and maturity groups which are described in detail elsewherel. The procedure used to study the seed filling parameters was that used by Izquierdo and Hosfield ( l7 ). Essentially all pods formed on the plant were removed and the seed separated from the pod wall. Seeds were dried at 75 C for 48 h. The rate and duration of seed filling, were calculated by fitting the data to a third degree polynomial using the least squares regression technique with time (days after 50% flowering) as the independent variable (x), and mean weight per seed as the dependent variable (y). The linear filling rate (LFR) was calculated using the linear regression coefficient determined from the line of best fit to the data points during the linear phase of seed growth. The linear regression was calculated using the predicted values after fitting the third degree polynomial to the data. However, in order to reduce subjectivity in choosing the limits of the linear phase a criterion based on the minimum change in the residual mean square was chosen. 1Chapter l. Paredes, O. M., 1986, M.S. Thesis, Department of Crop and Soil Science, Michigan State Univ., East Lansing, MI 48824. 84 85 Several data points in the middle of the linear phase of seed growth were selected, a least square fit was determined, and a coefficient of determination R2 for a linear regression model was calculated ( r7 L The number of data points was progressivaly reduced by deleting points one by one by starting at one end of the period and then at the other and refitting the curve and recalculating R2. This was continued until the residual mean square was at a minimum. The linear filling duration (LFD) was estimated by extrapolation of the line of best fit to its intersection with the x-axis. The maximum seed weight was calculated by setting the first derivative of the cubic function f'(tL used to estimate the seed filling parameters, to zero, and obtaining the corresponding root that maximized the cubic equation ( 17 ). The total filling duration (TFD) was then calculated by extrapolation of the maximum seed weight to the x-axis. The mean seed filling was estimated by calculating the ratio of maximun seed weight to total filling duration. The maximum filling rate was estimated by setting the second derivative f”(t) to zero and obtaining the value of t that maximized th). At harvest, mature plants from two, 2 m sections of individual plots were harvested and threshed to measure final seed yield. Biological yields and the number of pods per m2 were estimated by averaging the number of pods from a 1 m row ‘ V CHI 5 .‘w'SeE-Cl f weight c f uflfi EJEtc T; an W U. :licati 87‘ 'w ,- e 86 sample at harvest. The number of seeds per pod was obtained from a 50 pod sample from each plot in a replication. The weight of 100-seeds was determined as the average of two, 100-seed weight samples per replication. In this study differences among genotypes could not be ascertain by a analysis of variance because of the lack an appropiate error term brought about by the confounding of replications in data analysis. It was necessary to combine data from the three replications to estimate the seed filling parameters rather than analize data using individual replications as is the typical procedure. Hence, only trends will be presented and discussed in the text. Results The growth curves drawn for the 12 genotypes studied in 1984 are shown in Fig 1. In general, the lag phase associated with beginning seed growth was relatively short, and the seeds rapidly entered the linear phase of growth. Although the linear phase of seed filling varied among genotypes in all cases it extended to nearly the point where seed size was xnaximum. The regression coefficients R2 of the equations used to described the third degree polynomial and establish the limit of linear growth are listed.in Table L.The slope of the linear regression line indicated the growth rate during seed filling. The 12 genotypes were grouped according to their growth habit (Types 1, II, III), architectural forms (traditional and architype), and maturity (early and late) and differed for TLD, LFR, and mean filling rate (MFR), (Table 2 and 3 ). These results agree with previous work which showed the existence of phenotypic variability for seed filling parameters in dry beans ( 17 L Differences between types. Type I genotypes tended to have a shorter TFD than Type II and Type III genotypes. However, the Type I group had a more rapid LFR than Type II and Type III 87 asauuauopt. sueesuuueux. SEED DRY WEIGHT (mg/seed) I“! my Seafarer nau- tam saw an «a can-4 an gnoiufoaufia‘a mun I461. C-49 . ' i . «.04 ....J / na—I' . u! «i ”‘4‘ “T". uu'auieiouiaia Viva DAYS AFTER Fig 1. Seed filling of me 3'31””) . . au- Laker : / ‘u'i |m . . 1 e I‘M g u+ i ' . u‘ ”321 —3 f 3 (a no a a. u a 3 7o Olrzllelvbnhgd T-39 : “’1 1 Nap-2 I an ml 1 «i a “-1 /, . . 'E—ZQquozoInM a.) ”“1 Swan Valley wk. Carioca mu . 1 mm "£04 ‘ 1.0.0- mm new new 120.04 can an and «at «0.94 an- ‘r‘. 'a "izraaafififaa 1 Harris , w] Valley , tau-i . i ads-w an- an- I“ new I“ 'u M nun «ee- ado-4 " ' ‘ ”"T' ' ' ' ' a a a “4.;3‘3‘353333 0001230301010 50 % FLOWERING 12 dry bean strains grown in 1984. In 89 Table 2. Regression and determination coefficients from the cubic polynomial regression equation for seed dry weight on days afte 50% flowering of 12 dry bean genotypes grown in 1984. Genotypes 50 B1 B2 '33 R2 Seafarer -18.2892 2.0894 0.5290 -0.0140 0.984 Brazil-2 -5.9415 -0.2736 0.3790 -0.0066 0.975 Laker 4.9717 -2.9741 0.4870 -0.0083 0.936 C-49 -6.7434 0.5691 0.3447 -0.0070 0.990 T-39 -16.7674 2.2087 0.3615 -0.0080 0.966 Nap-2 -9.2348 2.4575 0.1986 0.1986 0.973 C-20 -6.5949 0.5292 0.2777 0.0040 0.981 S Valley -7.6441 1.4753 0.2470 -0.0042 0.992 Carioca -0.9759 -0.5303 0.3888 -0.0059 0.973 Viva -39.4576 8.0691 0.1602 -0.0062 0.972 Harris -32.4576 5.5391 0.4271 -0.0108 0.986 Valley -43.7256 9.8178 0.0844 -0.0034 0.965 90 Table 2. Total filling duration, linear filling duration, linear filling rate, and mean filling rate for different bean growth habits, architectural forms, and maturity groups. Seed filling characteristics Filling duration Filling rate Total Linear Linear Mean ----- days ----- --mg/seed/day-- Growth habit Type I 33.6 11 6.9 4.5 Type II 39.8 13 6.5 4.4 Type III 36.8 12 9.0 6.3 Architectural am Traditional+ 33.7 12 7.0 4.7 Architype++ 39.7 13 6.4 4.2 Maturity Group Earlyg 32.6 11 8.2 5.6 Late 88 40.9 13 7.6 5.1 +Mean of traditional I and II genotypes ++Mean of architype I and II genotypes 8 Mean of traditional I, II and early Type III genotypes SfiMean of architype I, II and late Type III genotypes \ ale 3 r-wth habit and genotype h ‘ mm k'v 'Ue I Tradit Sea I he 91 Table 3. Total filling duration, linear filling duration, linear filling rate, and mean filling rate for 12 dry bean genotypes grown in 1984. Growth habit and Filling duration Filling rate genotypes+ Total Linear Mean Linear ----- days------- ---mg/seed/day--- Type I Traditional Seafarer 26.9 9 5.5 8.8 Brazil-2 37.8 11 4.5 6.8 Mean 32.4 10 5.0 7.7 Architype Laker 36.0 13 4.0 6.4 C-49 33.5 11 4.0 6.1 Mean 34.8 12 4.0 6.3 Type II Traditional T-39 33.0 13 5.0 7.4 Nap-2 37.0 13 3.8 5.4 Mean 35.0 13 4.4 6.4 Architype C-20 47.0 13 4.6 6.8 S Valley 41.9 13 4.2 6.2 Mean 44.5 13 4.4 6.5 Type III Early Viva 31.1 10 5.8 7.7 Harris 31.9 11 7.2 10.9 Mean 31.5 11 6.5 9.3 Late Carioca 43.6 14 5.3 7.9 Valley 40.3 11 6.6 9.5 Mean 42.0 13 6.0 8.7 +Growth habit classification used by the International Center for Tropical Ariculture (CIAT), Cali, Colombia. H 92 genotypes. The LFD was almost similar for the three growth habits, but Type II genotypes had a slight longer LFD than Type I genotypes. The Type III genotypes tended to have a mean filling rate higher than Type II and III (Table 2). This difference in seed weight could be attributed to different seed sizes between the growth habits. The mean filling rate is directly proportional to the maximum seed weight and indirectly proportional to the TFD. Differences yithin groypp pgpipg. Values for the TI and TII genotypes and AI and All were averaged to analize their performance. Inigeneral, the architype genotypes tended to have a longer TFD compared to genotypes with the traditional form, regardeless of their growth.habita However, LFD, LFR and MFR were similar for both architectural forms (Table 2, 3). The later-maturing genotypes regardless of their growth habit or architectural form tended to have a longer TFD and LFD but a shorter LFR (Table 3 ). In this study, Seafarer, Brasi1-2, T-39, Nep-2, Harris, and Viva were considered as earlier genotypes compared with Laker, C-49, C-20, Swan Valley, Carioca, and Valley. Differences within architectural forms. In general, genotypes with a high LFR and MFR had a short LFD (Table 3) indicating an inverse relationship between filling rateanuiduration (Table 4 ). The LFD ranged from 9 days for Seafarer to 14 93 Table 4. Weight of 100 seeds, and days from 50% flowering to physiological maturity for 12 dry bean genotypes grown in 1984 Growth habits and 100-seed Days from 50% flowering genotypes+ weight (g) to physiological maturity Type I Traditional Seafarer 17.8 20 Brazil-2 20.6 30 Mean 19.2 25 Architype Laker 17.2 29 C-49 14.6 29 Mean 15.9 29 Type II Traditional T-39 19.2 29 Nep-2 16.1 29 Mean 17.7 29 Architype C-20 21.2 29 S Valley 18.7 28 Mean 20.0 29 Type III Early Viva 25.4 20 Harris 26.4 21 Mean 25.9 21 Late Carioca 26.6 28 Valley 26.8 30 Mean 26.7 29 +Growth habit classification used by the International Center for Tropical Agriculture (CIAT), Cali, Colombia. 94 days for Carioca. The experiment mean was 11.8 days. The TFD also varied among genotypes. Viva had the shortest TFD (31 days) compared with Carioca which had the longest (44 days). The mean of the TFD for the 121genotypes was 37 days (Table 3). The LFR varied from 5.4 to 10.9 mg per seed per day for Nap-2 and Harris, respectively. The experimental mean was 7.5 mg per seed per day. The maximum dry weight based on predicted values from the cubic model varied as expected among genotypes because of the presence of differential seed size brought about by the use of genotypes from different bean market classes. Viva, Harris, and Valley had the heaviest seeds and as a group, the navy beans had the lighest (Table 4 ). The percentage of time that the total filling duration represented over the life cycle of the plant varied from 39.6 for Seafarer to 57.2% for C-20. Only the three genotypes (Carioca, C-20, and Swan Valley) had seed filling periods greater than 50% of their total growing period (Table 4 ). Correlations. The mean values of the 3 replications were used to compute linear correlations between pair of traits evaluated (Table 5 ). The TFD inns positively and significantly correlated with days to 50% flowering and days to physiological maturityu However, LFD was positively and significantly correlated with seeds per pod, seed yield, and biological yield. PM) significant correlation.idas found 95 Table 6. Simple correlation coefficients between seed filling parameters and yield, and components of yield for 12 dry bean genotypes grown in 1984. Filling duration Filling rate Total Linear Linear Mean No. of pods /m2 0.497 0.625* -o.ooz -o.791** No. of seeds/pod 0.149 0.530** -0.068 -0.500 100-seed weight 0.152 -0.140 -0.193 0.834** Seed yield 0.502 0.784** 0.221 -0.749** Biological yield 0.454 0.612* 0.181 -0.723** Days to 50% flowering 0.782** 0.929** 0.126 -0.577* Days to physiological maturity 0.770** 0.849** 0.142 -0.645* *,** Significant at 5% and 1% probability level, respectively. H 96 between LFR and any of the traits studied. The MFR was positively'and.significantly correlated with the number of pods per m2 and 100-seed weight, and negatively correlated with seed yield, biological yield, days to 50% flowering and days to physiological maturity (Table 5 ). Discussion Examination of the seed filling parameters of the 12 dry bean genotypes studied, indicated that phenotypic variability existed for seed filling rate and duration in this population (Table 3 and‘4 L It should be pointed out that values for the variables studied are averages for a six plant sample because data was taken on seeds developing throughout the plant. In contrast, Egli et al ( 9, 10 ), in soybean and Mahon and Hobbs ( 23 ) in peas have shown that seeds in the same pod developed at different rates, depending on the position on the pod and Egli et al ( ll ) reported that soybean pods at different positions within the plant also had different filling durations but the same filling rate. The variation among genotypes with different growth habits, architectural forms and maturities for TFD and filling rate could have been due to differences in the efficiency of production and translocation of photosynthate. The difference in LFD observed between the Type I and III groups in this study was approximately 2J1nmgper seed per day, and 2&5 mg per seed per day between Type II and III. This variabilitily could be due to differences in the carbon and nitrogen requirements for fruit development. The 97 98 accumulation of seed dry matter (increased seed weight) should.be related to the ability of the plant to fix carbon during the filling period and /or the translocation of storage carbohydrates from other plant parts ( 11 ). The absence of a correlation between LFR and yield could be due to environmental factors impossing differential stresses during the non linear phase of seed filling or it could be that seed filling rate in dry beans is not affected by a source-sink relationship but is mainly controlled by the the seed itself. The lack of correlation between linear filling rate and yield agree with previous reports in beans ( 17 ), soybeans ( 12, 14 ) and some cereals ( 3, 4, 5, 15 L This indicated that LFR was less important that LFD in determining yield in this population. The importance of a long filling duration to maximize yield was mentioned by Laing et a1 ( 21 ) who showed that by extending the duration of growth and the reproductive period, the demand of the pods and the current supply of assimilates from the leaves were more nearly balanced. This was because a larger leaf area supported the extra fruit formed in the plant. The absence of a significant correlation between LFR and yield does not mean that yield could not be improved if a high seed filling rate was a genetic characteristic of a strain possessing other characteristict important for high yield. For example, a genotype with a long filling duration, 99 high seed filling rate and a high leaf area index may produce a higher yield than another genotype with only one or two of these characteristics selected for the maximum expression. The differences in seed filling parameters and yield between the AII and TII genotypes agrees with Izquierdo and Hosfield ( 17 ) who reported that All had a long LFD and short LFR. However, the differences in the current study could also be attributed to differences in growth habit and length of maturity instead of an improvement in physiological efficiency by the All. This hypothesis is supported by the fact that the All had similar seed filling characteristics as TII (Table 3 ). The results of this study showed the same trend when the architectural forms within growth habits I and II (architype vs traditional) were compared. When the data for each architectural form was averaged over the growth habits, the architypes had only a slighly longer TFD compared with the traditional forms. However, this difference could also be attributed to the lateness characteristic of this architectural form (Table 4 ). One run] to resolve this question is to develop isolines differing in days to maturity and having constrasting architectural forms. Izquierdo and Hosfield ( 17 ) suggested that the high yields of the architype compared to genotypes within the traditional forms was associated with a high linear filling 100 duration and heavy seeds brought about by the architypes ability to produce photosynthate later in the season“ This suggested that breeders during the selection of plants with the architype characteristic had indirectly selected for a long seed filling duration.TWu£;hypothesis agrees with the other results of this study which indicated that seed filling duration is positively and significantly correlated with days to 50% flowering and physiological maturity (Table 5 ). The significant correlation between IJW) and physiological maturity agrees with that found by Reiscosky et al ( 30 ) and in soybean which suggests that selection for a long seed filling duration might result in development of late maturing cultivars. However, one way to counter this association would be to select strains from segregating populations which.have variable seed filling durations but the same length of maturity ( 30 ). The positive correlation between days to 50% flowering and days to physiological maturity indicated that data from date of 50% flowering and physiological maturity could be utilized in a preliminary screening and selection of plants for a longer linear seed filling duration. There were no clear cut differences in LFR between large and small seeded strains, although, the Great Northern genotypes had a higher LFR than the navy and black seeded ones. This suggests that seed filling rate is controlled by a 101 genetic mechanism of the seed rather than the sink source relationship between plant and seed. This result also agrees with the finding that early genotypes had a higher LFR and lower biological yield compared with the late maturing genotypes. The negative and significant correlation between MFR and seed yield, biological yield, days to 50% flowering and days to maturity suggested a genetic linkage may exist between seed weight and days to maturity. These results suggested that a medium or a full season cultivar with a long LFD and a high biological yield, could have some more seed yield potential than early maturing genotypes with a shorter linear filling duration. B IBL IOGRAPHY luAdams, NLW. 1981. Uptade: New bean architype. Michigan Dry Bean Dig. 5:12-13. 2.Askel, R. and L.P.V. Johnson. 1961. Genetic studies in sowing-to heading and heading-to ripenning periods in barley and their relatopnship to yield components. Can. J. Genet. Cytol. 3:242-259. 3.Boote, K.J. 1981. Response of soybeans in different maturity groups to march plantings in Southern USA. Agr. J. 73:854-859. 4.Cavaleiri, A. J. and 0.8. Smith. 1985. Grain filling and field drying of a set of maize hybrids released from 1930 to 1982. Crop Sci. 25:856-860. S.Cross, H.Z. 1975. Diallel analysis od duration and rate of grain filling of seven inbred lines of corn. Crop Sci. 15:532-535. 6.Daynard, 128., and L.W. Kannenberg. 1976. Relationships between length of the actual and effective grain filling periods and the grain yield of corn. Can. J. Plan. Sci. 56:237-242. 7.Daynard, T.B., J.W. Tanner and W.G. Duncan. 1971. Duration of the grain filling period and its relation to grain yield in corn, Zea mays L. Crop Sci. 11:45-48. 8.Dunphy, E.J., J.J. Hanway and D.E. Green. 1971. Soybean yield in relation to days between specific developmental stages. Agr. J. 71:917-920. 9.Egli, D.E., J. Fraser, J.E. Leggett and C.G. Poneleit. 1981. Control of seed growth in soybeans (Glygipg mgx (L). Merril). Ann. Bot. 48:171-176. 10.Egli, D.B. and J.E. Leggett. 1976. Rate of dry matter accumulation in soybean seeds varying source-sink ratios. Agron. J. 68:371-374. 11.Egli, D.B.,J.E. Leggett and J.M. Wood. 1978. Influence cs 102 103 soybean seed size and position on the rate and duration of filling. Agron. J. 70:127-130. 12.Egli, D.E., J.H. Orf and T.W. Pfeiffer. 1984. Genotypic variation for duration of seed fill in soybean. Crop Sci. 24:587-592. 13.Fehr, W.R. and C.R. Caviness. 1977. Stages of soybean development. Iowa State Univ. Coop. Ext. Serv. Spec. Rep. 80. 14.Gay, S., D.E. Egli and D.A. Reicosky. 1980. Physiological aspects of yield improvement in soybeans. Agron. J. 72:387- 391. 15.Gebeyechou, (L,D.R. Knott and R.J. Baker. 1982. Relationships among durations of vegetative and grain filling phases, yield components and grain yield in durum wheat cultivars. Crop Sci. 22:287-290. 16.Hanway, J.J. and W.A. Russel. 1969. Dry matter accumulation in corn (Zea pays L.) plants: comparisons among single cross hybrids. Agron. J. 61:947-951. 17.Izquiedo, J.A. and G.F. Hosfield. 1983. The relationship of seed filling to yield among dry beans with differing architectural forms. J. Amer. Soc. Hort. Sci. 108:106-111. 18.Johnson, D.R. and J.W. Tanner. 1972. Calculation of rate and duration of the grain filling in corn (Leg pgyp L.) Crop Sci. 12:485-486. l9.Jones, D.E., M.L. Peterson and S. Geng. 1979. Association between grain filling rate and duration and yield components in rice. Crop Sci. 19:641-644. 20.Kaplan, S.L. and B.R. Koller. 1974. Variation among soybean cultivars in seed growth rate during the linear phase of seed growth. Crop Sci. 14:613-614. 21.Laing, D.R., P.G. Jones and J.H.C. Davis. 1984. Common bean (gpgggplps yplgggig L.). p.305-351. In: P.R. Goldsworsthy and N.M. Fisher. (Eds). The physiology of tropical field crops. John Willey and Sons. N.J. 22.Mc Blain, E.A. and D.J. Hume. 1980. Physiological studies of higher yield in new, early-maturiy soybean cultivar. Can. J. Plant Sci. 60:1315-1326. 23.Mahon, J.D. and S.L.A. Hobbs. 1983. Variability in pod 'l 104 filling characteristics of peas (piggy pppiygp L.) under field conditions. Can. J. Plant Sci. 63:283-291. 24.Matzger, DAL, SJL Czaplewski and DJ; Rasmusson. 1984. Grain filling duration and yield in spring barley. Crop Sci. 24:1101-1105. 25.Nass, HgG. and B. Russer. 1975. Grain filling period and yield relationships in spring wheat. Can. J. Plant Sci. 55:673-678. 26.Peaslee, D.E”,;LL. Ragland and W.G. Duncan. 1971. Grain filling period of corn as influeced by phosphorus, potassium and the time of planting. Agron. J. 63:561-563. 27.Poneleit, CLG. and D.B. Egli. 1983. Differences between. reciprocal crosses of maize for kernel growth characteristics. Crop Sci. 23:871-875. 28.Poneleit, CLG., D.B. Egli, P.L. Cornelius and D.A. Reicosky. 1980. Variation and associations of kernel growth characteristics in maize populations. Crop Sci. 20:766-770. 29.Rasmusson, D.C., I. McLean and T.L. Tew. 1979. Vegetative and grain-filling periods of growth in barley. Crop Sci. 19:5-9. 30.Reicosky, D.A., J.H. Orf and C.H. Poneleit. 1982. Soybean germplasm evaluation for length of the seed filling period. Crop Sci. 22:319-322. 31.8a1ado-Navarro, L.R., T.R. Sainclair and K. Himson. 1985. Comparisons among effective filling period, reproductive period duration, and R5 to R7 in determinate soybeans. Crop Sci. 85:1050-1054. 32.Spaeth, S.C. and T.R. Sainclair. 1985. Linear increase in soybean harvest index during seed filling. Agron. J. 77:207- 211. IILTaylor, J; 1975. A modified air planter (plantair) for use in experimental plots. Bean Improv. Coop. 34;Watson, DuJ. 1952. The physiological basis of variation in yield. Adv. Agron. 4:101-145. 35.Wych, RJL, R.L. Mc Graw and D.D. Stuthuman. 1982. Genotype x year interaction for length and rate of grain filling in cats. Crop Sci. 22:1025-1028. CHAPTER 3 COMBINING ABILITY EFFECTS FOR YIELD, COMPONENTS OF YIELD AND MORPHOLOGICAL TRAITS IN A DIALLEL CROSS OF DRY BEANS WITH DIFFERENT GROWTH HABITS AND MORPHOLOGICAL CHARACTERISTICS ABSTRACT Efficient breeding for higher yielding cultivars requires extensive understanding of the system leading to phenotypic expression of yield. On the other hand, the choice of an appropiate selection strategy for breeding depends on the type of genetic variance present in the population and its heritability. The objective of this study was to estimate combining ability and genetic effects of different genotypes for morphological components of plant architecture related to yield. The six parents and 30 F1 hybrids from a 6 x 6 complete diallel were grown in a randomized block design at two locations.'The combining ability determined by Griffing's Model 1, Method 1 showed that the GCA mean squares were significantly different for yield, and yield components, stem dry weight per plant, biological yield and days to 50% flowering. In general, the GCA mean squares were larger than SCA, maternal and reciprocal variances for all the traits 105 106 studied, indicating that mainly additive gene action was controlling the expression of these traits. The SCA mean squares was significant for number of seeds per pod, 100-seed weight and days to 50% flowering. There was considerable variation in the GCA contribution of each parent to the various traits analyzed. However, no one parent had all positive and negative effects for yield, components of yield and morphological traits studied. However, there were some parents which could complement each other by combining positive GCA for some traits. The presence of reciprocal and maternal effects indicated that the choice of a parent to be used as either male or female is going to affect the F1 outcome and special care should be taken in selection of parents. Although some traits showed significant location x GCA, location x SCA, location x maternal effects, these components of variance were small compared with the additive part indicating that F1 hybrids manifested a broad adaptation. In general, the genetic variability of each trait suggested that it could be possible to make rapid improvements using breeding systems to concentrate primarily additive gene action present in this population. Introduction Seed yield in corn (Egg _m__a_y§ L.), the cereals (Triticum sp. (Hpgggpm yplgggp L)., rice (Oryza satiyg L)., and the food legumes is a complex character that is generally correlated with specific first order components of pods or spikes per plant, seeds per pod or spike, and seed weight. Each component is the final product of physiological events that may be controlled by several genes. A change in any one of the components may bring about a change in final yield. Therefore, it can be perceived that if each yield component can be altered in a positive direction, an increased expression of yield should be obtained. In order to maximize the expression of one, or all yield components, redesigning the physiology and morphology of a crop may prove fruitful. This concept is consistent with the theory for yield breeding in wheat (Triticum aestivug L.) and dry beans (Phaseolus vulgaris L. ) via the ideotype approach expoused by Donald (:2) and.Adams (1.), respectively. Breeding for physiological and morphological traits may result in an increased biological yield by a redistribution of assimilates to the seed as a mechanism by which the plant counters an environmental stress ( 30 L Since there is no 107 108 agreement among workers as to whether or not yield is limited by the size of the source or sink of the plant ( 14 ) further study of the partitioning of assimilates and the interactions between source and sink is warranted. The presence of genotypic variability in physiological components such as seed growth rate and duration and canopy development as wellas their associations with the components of yield might enable one to develop a strategy for improving yield. However, the choice of the appropiate selection strategy for breeding for morphological traits depends on the type of gene action present in the population for the traits and their heritability. If the morphological traits are controlled by genes with additive effects or effects involving additive types of epistasis then it should be possible to improve these characters through recurrent selection or other strategies that tend to increase the percentage of favorable alleles in a population. The objective of this study was to estimate combining ability and genetic effects for morphological components of plant architecture related to yield. Literature Review Combining Ability. General combining ability (GCA) and specific combining ability (SCA) were first defined by Sprague and Tatum ( 28 ) and indicates the nature of a parent's performance for a particular trait when it is in a hybrid combination. General combining ability is the average performance of a parent in hybrid combination while specific combining ability is used to designate those cases in which certain combinations do relatively better or worse than would be expected on the basis of the average performance of the lines involved in the same crosses. Breeders often rely (Ml combining ability estimates in selecting elite parents for breeding programs. Combining ability, can be estimated by several procedures among which the most widely used is theiiiallel mating system ( 16 ). Griffing ( 16 ) defined four experimental diallel methods for estimating combining ability and the method one chooses depends on whether parental inbreds and reciprocal Fl's are included. Griffing ( 16 ) also proposed models when parental lines or experimental material are assumed totxaa random sample from some population about which inferences are to be made or whether they are deliberately chosen for some specific 109 110 reason. Certain assumptions underly the theory of diallel crosses. These are : 1)diploid segregation, 2)no difference between reciprocal crosses 3)independent action of non- allelic genes 4)no multiple allelism 5)homozygous parents and 6)genes independently distributed between parents ( 18 ). Despite its wide spread usage, the diallel cross analysis has been criticized as a method for studying gene action of quantitatively inherited traits. When Baker ( 3 ) reviewed 13m; important issues concerning diallel analysis, he opined that two assumptions are violated in interpreting the results. The assumption of independent distribution of genes seemed to be least acceptable because the failure of this assumption will often result in overestimation of the average level of dominance ( 3 ). The assumption that epistasis is absent may also invalid, since epistasis affects estimates of GCA and SCA variances. The presence of epistasis leads to the overestimation of these variances ( 3 ). Baker ( 3 ) concluded that the best use of diallel cross analysis in plant breeding is to estimate genetic effects of individual parents. Types pf Gene Action Yield and yield comppnents. Dickson ( 10 ) reported that the number of pods per plant, number of seeds per pod and the number of seeds per plant were determined by an additive gene 111 system. Pod number showed incomplete dominance and mostly a recessive behavior in dry beans. Singh and Saini ( 26 ) found significant and high general combining ability effects in dry beans for the number of pods per plant and pod diameter, which indicated a preponderance of additive gene effects. However, seed yield had high specific combining ability effects. Foolad and Bassiri ( 12 ) pointed out that the ratio of GCA to SCA mean squares in dry beans was low for yield, number of pods and seeds per plant and high for 100-seed weight. Significant reciprocal effects were observed for yield and number of seeds per plant. Chung and Goulden ( 6 ) showed that the dominance component of variance in dry beans was higher than the additive one, and overdominance was observed for number of pods and yield. They observed significant differences between additive genetic effects in different locations for lOO-seed weight and between dominance effects for yield. Except for pod number there appeared to be more recessive than dominant genes controlling yield and 100-seed weight. Safari ( 25 ) reported partial or near complete dominance for pods per plant, seeds per pod and lOO-seed weight in dry beans. Neinhuis and Singh ( 21 ) found that GCA was more important than SCA for yield and yield components in both F1 and F2 analysis in dry beans. 112 Soybean and peas. Weber et a1 ( in) ) reported that significant SCA was associated with seed yield in soybean; however, general combining ability estimates were larger than the SCA effects. Krarup and Davis ( 20 ) found that yield and the components of yield in peas were controlled by an additive genetic system. On the other hand, Gritton ( 17 ) showed that both GCA and SCA were important for pods per plant, seeds per pod, seed per plant, seed weight and seed yield. Pods per plant also showed some maternal influence. Gene Action pf Some Architectural Traits Qppmon bean. Ghaderi and Adams ( 14 ) found that plant type and hypocotyl diameter were governed by both additive and dominance gene action in two crosses involving Seafarer, a standard navy type and two architype genotypes. In the first cross, the number of branches, racemes, pods and seeds per plant were influenced by both types of gene action. Only dominant effects were present in nodes below 15 cm and yield, while additive effects were present for number of nodes above 15 cm and pod lenght. In the second cross, the number of nodes above 15 cm and pod length were under both additive and non additive gene effects while number of branches showed neither. Nodes below 15 cm showed additivity and the remaining characters all showed dominance. Coyne ( 7 ) showed that the association between indeterminate growth habit and late maturity in dry beans was 113 due to a genetic linkage between indeterminancy and late flowering. Frazier et al ( 13 ) found that determinate growth habit was controlled primarily by a single recessive gene. However, least three major recessive genes or many minor genes were controlling the upright bush habit. Coyne ( 7 ) concluded that the growth habit was inherited quantitatively because of the continuous nature of the distribution in segregating populations. On the other hand, Bliss ( 4 ) reported that a single gene controlling the sprawling growth habit was completely dominant to the bush type in one cross but epistatic effects were present in another cross. Dickson ( 10 ) reported additive gene action for days to flowering, but overdominance for earliness. Foolad and Bassiri ( 12 ) reported significant GCA for days to flowering. Chung and Goulden ( 6 ) found significant differences between additive genetic effects at different locations for flowering time. Coyne and Matton ( 8 ) reported that blooming period was inherited.quantitatively and that dominance effects for a long flowering period were observed. Davis and Frazier ( 9 ) concluded that the genes conditioning the expression of growth habit, plant height, and number of internodes were on the average recessive. Additive effects predominated for growth habit, internode number and height of pod attachment. Sgypggp. Weber et a1 ( 31 ) reported significant specific [I 114 combining effects for maturity date, plant height and oil content in soybean. Paschal and Wilcox ( 23 ) found that SCA was more important than GCA for maturity and height in soybean but GCA more important than SCA for lodging, plant weight and harvest index. Location.2< GCA effects were significant for maturity, lodging, plant weight and HI and a year x SCA effect interaction was observed for maturity, lodging and harvest index. gppp. Snoad and Arthur ( 27 ) indicated that days to flowering, node position of first flower, number of flowers per node and number of ovules per pod were under p ygenic control. However, mean internode length was addi:.onally controlled by a major gene. In general, additive gene action was more important than dominance and epistasis in controlling the inheritance of these traits. Gritton ( 17 ) reported that general and specific combining ability were important for days to bloom and plant height. Days to bloom showed a significant maternal influence in one experiment. Pandey and Gritton ( 22 ) also found that both SCA and GCA were important for days to bloom and plant height. Both traits also presented significant maternal effects ( 22 ). Watts et a1 ( 29 ) showed that the flowering period was under the control of an additive system. Dominance effects influence late flowering. However, Rowland ( 24 ) indicated 115 that flowering was largely under the control of a polygenic system in which lateness was dominant to earliness and the gene action was mostly additive in peas. Materials and Methods Six genotypes of dry beans were used as parents in this study.They were chosen on the basis of differential growth habit, number of days to maturity, and seed filling characteristics (Table 1 ). All possible F1 hybrids including reciprocals were made among the parents. The 30 F1 hybrids and parents were planted at the Michigan State University Botany Farm at East Lansing and the Saginaw Valley Bean and Sugar Beet Research Farm near Saginaw, Michigan. Production and management practices were those normally followed for dry bean production at these research facilities. Beans were planted during the first week of June at Saginaw and the third week of June at E. Lansing in 1985. Seeds were planted into two-row plots with a hand-held mechanical planting device in a randomized complete block design with two replications in each location. Each row was 8 m long and 47 cm apart. Plant spacing within a row was 15 cm. Each plot was bordered on either side by a single row of Montcalm kidney bean. Each two-row plot was divided into 8 segments each 2 m long where 5 plants of the experimental material was planted. Each segment was bordered on each end by three plants of 'Montcalm'. This was necessary to 116 117 Table 1. Growth habit, architectural form, seed size commercial class and days to maturity of six genotypes grown at two locations in 1985. Archi- Growth tectural 100-seed Commercial Days to Genotypes Habit+ form weight class maturity Seafarer Type I Traditional 17.8 Navy 68 Laker Type I Architype 17.2 Navy 82 T-39 Type II Traditional 19.2 BTS 80 Swan Valley Type II Architype 18.7 Navy 84 Harris Type III Postrate 26.4 Great 70 vine northern Carioca Type III Postrate 26.6 Undefined 82 vine +Growth habit classification used by the International Center for Tropical Agriculture (CIAT), Cali, Colombia. 118 facilitate sampling by separating each sampling unit from another. The five plants in each segment were harvested at weekly intervals for eight consecutive weeks at both locations. In order to determine the reproductive period for each genotype plants in all plots were evaluated every two days after flowering at both locations. Harvesting began after the first pod was formed on each genotype and was accomplished by cutting plants at the soil line. Plants were then separated into stems, leaves, pods and seeds. The parts were dried in a forced air oven at 75 C for 48 hours. Individual mature plants from each plot were harvested and thresehed to measure final seed yield. Individual plants were also used to determine yield components. Statistical analysis. All data were subjected to an analysis of variance appropriate for a randomized complete block design. The analysis of combining ability for yield, yield components and some morphological traits was performed using the formulas presented by Griffing ( 16 ) for Model 1, Method I. This analysis was extented to include the partitioning of reciprocal variation into maternal and non maternal effects according to Cockerham ( 5 ). The components of genetic effects for combining ability, reciprocal (maternal and non maternal effects) were computed using the 119 least squares procedures. In Model I, treatment effects are fixed, therefore, the inferences from this study pertain only to the material analyzed in this work. The effects model appropiate to this study were: I Y(1:!)(k1)m = 9+ bi + gdk} + 9 (1m) + 8(le (1m) + m(jk) - “(1m + r(jk)(1m) + €(jk1m)n Where: H is the population mean b. is the ith block effect g. is the GCA effect associated with the jkth parent sjk is the SCA effect associated with the cross between the jkth female and the lmth male parent m. is the maternal effect associated with the jkth parent when is used as a female parent rjk is the reciprocal effect associated with the cross between the jkth female and the lmth male parent Ejklm is the error term associated with the jkth and lmth parent in the ith block Results The combined analysis of variance indicated that significant differences among genotypes existed for all the traits except seed yield (Table 2 ). The genotypic variation was partitioned into GCA and SCA. General Combining Ability Yield and yield components. The GCA mean squares averaged over growth habit, architectural forms and early and late maturing genotypes were significant for all yield and yield components (Table 2 jLThis indicated that additive genetic effects controlled a large portion of the expression of the traits considered in this study. The GCA mean squares for yield and yield components among the different growth habits was significant for yield components but not for seed yield per plant . This indicated that growth habit of a genotype influence its GCA for yield components. Estimates of GCA effects of the individual genotype are given in Table 3 for yield and components of yield. Type I genotypes transmitted a high number of pods per plant but a low number of seeds per pod and lOO-seed weight to their crosses, Type II genotypes greatly increased the number of pods per plant and seeds per pod but reduced the 100-seed 120 121 Table 2. Analysis of variance for GCA, SCA and Reciprocal effects for components of yield and seed yield for a six parent diallel cross grown at two locations in 1985 Pods/ Seed/ 100-seed Seed yield/ Source+ df plant pod weight plant Total 143 Location 1 1.37 0.13 0.71 1201.08** Replication(L) 1 12.13 0.09 24.59 34.16 Genotypes 35 150.18* 1.26** 79.69** 40.96 GCA 5 361.74** 2.46** 240.13** 161.99* Among 2 461.35** 2.69** 327.99** 129.79 Within 3 275.34* 2.30** 181.55** 183.45* SCA 15 116.55 0.96** 82.07** 18.09 Among 3 50.07 0.90 307.39** 29.73** Within 12 133.16 0.97** 25.74* 15.19 Maternal 5 167.91 1.35** 17.53 56.47 Among 2 293.85* 3.15** 3.57 73.71 Within 3 83.94 .0.12 26.83 44.98 Reciprocal 10 85.98 1.01 25.59 6.98 Among 1 0.62 0.01 0.01 0.58 Within 9 95.46 1.19** 28.43 7.69 Error (L x G) 70 86.74 0.19 10.98 25.35 *,** Significant at 5% and 1% probability level, respectively. + Among and within refer to growth habit. 122 Table 3. Estimates of GCA effects for components of yield and seed yield measured on F1 progeny from a six- parent diallel cross grown at two locations in 1985. Pods/plant Seed/Pod Parental Parental effect Effect effect Effect over between over between Growth growth growth growth growth Parent habit+ habits habits habits habits 1 Seafarer I -0.45 0.05 2 Laker I 3.45 -0.29 1.50 -0.12 3 T-39 II 0.42 0.36 4 S Valley II 1.78 0.02 1.10 0.19 5 Harris III -4.68 -0.19 6 Carioca III -.260 0.05 -2.60 -0.07 100-seed weight Seed yield/plant Parental Parental effect Effect effect Effect over between over between Growth growth growth growth growth Parent habit habits habits habits habits 1 Seafarer I -0.71 —2.54 2 Laker I -0.65 0.65 -0.68 -0.94 3 T-39 II -0.85 -0.26 4 S Valley II -1.98 —0.44 -1.41 —0.35 5 Harris III 4.41 -0.48 6 Carioca 111 -0.23 3.08 2.09 1.29 +Growth habit classification used by the International Center for Tropical Agriculture (CIAT), Cali, Colombia. 123 weight. Type III genotypes increased seed weight in their progeny but decreased number of pod per plant and seeds per pod. There was considerable variation in the GCA contribution of each parent to the various traits analyzed (Table 3). No one parent had all positive or negative GCA effects for yield and yield components. However, there were some parents which could complement each other by combining positive GCA effects for yield and yield components. Based on its GCA effects, Laker transmitted an increased number of pods per plant and seed yield to its progeny but reduced the number of seed per pod and the lOO-seed weight (Table 3 ). However, Harris tended to impart a high 100-seed weight but a low number of Table 4. Analysis of variance for GCA, SCA and reciprocal effects of morphological traits from a six-parent diallel cross grown at two locations in 1985 Bio- Days to Stem dry Pod wall logical 50% Source+ df weight dry weight yield flowering Total 143 Location 1 1.20 225.28** 2225.72** 1002.78** Replication (L) l 8.65 6.32 31.60 11.01 Genotypes 35 24.40** 17.26 119.96 99.73** GCA 5 46.72** 20.89** 310.10** 292.91 Among 2 69.75** 37.05 135.87 263.67** Within 3 31.37** 10.12 426.25** 312.39** SCA 15 12.76 10.35 58.06 58.85** Among 3 17.02 13.41 91.36 76.62** Within 12 11.69 9.58 49.74 53.65** Maternal 5 71.39** 40.11* 215.32* 191.18** Among 2 73.33** 47.06* 99.42 14.76 Within 3 70.09** 35.48* 292.58* 308.80** Reciprocal 10 7.20 14.38 70.04 18.73* Among 1 0.42 3.34 0.07 0.26 Wtihin 9 7.95 15.61 77.82 20.78** Error (LxG) 70 8.32 12.58 75.86 7.68 *,** Significant at 5% and 1% probability level, respectively. + Among and within refer to growth habits. 125 squares within growth habits were larger than the GCA mean squares between types for biological yield and days to 50% flowering. The variation between types was twice as large as within growth habits. Estimates of GCA effects for morphological traits are presented in Table 5. Type I and III genotypes reduced stem dry weight, pod wall dry weight per plant, biological yield and imparted lateness to their progeny while type II genotypes imparted increased stem dry weight, pod wall dry weight per plant, biological yield and lateness to its progeny. Seafarer, Harris and.Carioca tended to reduce the stem dry weight, pod wall dry weight per plant and biological yield of their progenies while Laker, T-39 and Swan Valley increased the expression of these traits in their progenies. Examination of the GCA effects of individual parents indicated that Harris was the best combiner for earliness followed by Seafarer, whereas, the progenies of Swan Valley were among the latest (Table 5). Specific Combining Ability Yield and yield ppmpgpgpps. The SCA mean squares were significant for number of seed per pod and 100-seed weight (Table 6 L The variation among and between growth habits for number of seeds per pod and days to 50% flowering were both large for the expression of these traits which was 126 Table 5. Estimates of GCA effects for morphological traits measured on F1 progeny of a six-parent diallel cross grown at two locations in 1985. Stem dry weight Wall pod dry weight Parental Parental effect Effect effect Effect over between over between Growth growth growth growth growth Parent habit+ habits+ habits habits habits l Seafarer I -0.86 -0.72 2 Laker I 0.68 0.39 -0.09 -0.16 3 T-39 II 1.23 0.63 4 S Valley II 0.56 0.75 0.89 0.69 5 Harris III -1.33 -0.46 6 Carioca III -0.28 -0.59 -0.80 -0.52 Biological yield Days to 50% flower. Parental Parental effect Effect effect Effect over between over between Growth growth growth growth growth Parent habit habits habits habits habits 1 Seafarer I -3.85 -1.97 2 Laker I 2.05 0.13 -0.90 -0.92 3 T-39 II 1.55 1.53 4 S Valley II 1.15 2.30 1.35 1.91 5 Harris III -2.59 -3.91 6 Carioca III -1.70 1.92 —0.45 -0.99 +Growth habit classification used by the International Center for Colombia. Tropical Agriculture (CIAT), Cali, 127 Table 6. Estimates of SCA effects for components of yield and seed yield measured on F1 progeny of a six-parent diallel cross grown at two locations in 1985. Pods/plant Seed/plant Parental Parental effect Effect effect Effect over between over between growth growth growth growth Parent habits+ habits habits habits Seafarer 3.83 1.20 -0.63 -0.12 Laker -3.09 1.20 0.23 0.12 T-39 1.11 0.22 -0.27 0.04 S Valley 0.99 0.22 0.54 0.04 Harris 8.39 1.85 -0.18 0.21 Carioca —4.47 1.85 -0.25 0.21 lOO-seed weight Seed-yield Parental Parental effect Effect effect Effect over between over between growth growth growth growth habits habits habits habits Seafarer -l.76 -1.98 ' -1.78 -l.55 Laker -1.54 -1.98 1.34 1.57 T-39 -2.57 —1.50 0.04 1.52 S Valley -0.70 -1.50 -l.98 -0.50 Harris -9.81 -5.00 -2.16 -l.33 Carioca -3.71 -5.00 1.32 2.15 +Growth habit classification used by the International Center for Tropical Agriculture (CIAT), Cali, Colombia. 128 Table 6. (Continued) Pods/plant Seed/plant Parental Parental effect Effect effect Effect over between over between growth growth growth growth Crosses habits+ habits habits habits 1x2 2.02 1.20 -0.04 -0.12 1x3 0.97 0.22 0.24 0.14 1x4 4.34 0.22 0.20 0.14 1x5 -9.75 -1.41 0.02 -0.03 1x6 -l.41 -1.41 0.21 -0.03 2x3 -1.64 0.22 0.46 0.14 2x4 -2.80 0.22 -0.32 0.14 2x5 2.35 -l.41 0.08 -0.03 2x6 3.16 1.41 -0.42 -0.03 3x4 -0.60 0.22 -0.06 0.04 3x5 0.21 -0.44 -0.11 0.18 3x6 0.05 —0.44 -0.26 0.18 4x5 -2.95 -0.44 -0.45 0.18 4x6 1.02 0.44 0.08 0.18 5x6 1.75 1.85 0.63 0.21 100-seed weight Seed yield Parental Parental effect Effect effect Effect over between over between growth growth growth growth habits habits habits habits 1x2 -2.31 -l.98 -0.24 -0.01 1x3 -0.01 0.76 1.25 0.81 1x4 -2.23 0.76 1.05 0.62 1x5 4.40 2.74 1.83 2.04 1x6 1.89 2.74 -2.12 -l.91 2x3 1.23 0.76 0.42 -0.02 2x4 -2.06 0.76 -0.97 -1.40 2x5 2.51 2.74 -0.63 -0.42 2x6 2.15 2.74 0.08 0.29 3x4 -l.36 -1.50 -1.99 0.51 3x5 1.19 2.26 0.74 -0.30 3x6 1.51 2.26 -0.46 -1.50 4x5 4.95 2.26 1.47 0.43 4x6 1.40 2.26 2.41 1.37 5x6 -3.25 5.00 -l.24 -0.41 +Growth habit classification used by the International Center for Tropical Agriculture (CIAT), Cali, Colombia 129 reflected in their mean square components. However, the SCA mean squares among types for lOO-seed weight was more than 10 times greater than the variation found within the growth habits. The former result was expected because of the presence of large, medium and small seed size representing each bean class. For instance, Type II genotypes had small seed size (Nap-2 and T-39) whereas Type III genotypes had a medium to large seed size (Carioca and Harris, respectively). Specific combinimg ability effects showed considerable variation with regard to the contribution of each parent to the various traits analyzed (Table 6 ). The estimates of SCA effects indicated that among the six parents Harris showed the highest negative SCA effect for 100-seed weight and seed yield. The large negative SCA effects for lOO-seed weight and seed yield indicated that Harris had the ability'to combine better than expected, based on its own performance for these traits. The positive SCA effects for a hybrid suggested that the mean of the hybrid was greater than expected, based on the mean performance of the lines involved. For instance, Swan Valley x Carioca was the only cross which showed a positive GCA effect for number of pods per plant, number of seeds per pod 100-seed weight and seed yield (Table 6L Morphological traits. The SCA mean square was significant only for days to 50% flowering and the variation between and within types were both significant (Table 7). 130 Table 7. Estimates of SCA effects for morphological traits and seed yield measured on F of a six-parent diallel cross grown at two locations in 1985. Stem dry weight Pod wall dry weight Parental Parental effect Effect effect Effect over between over between growth growth growth growth Parent habits+ habits habits+ habits Seafarer -1.4 -0.3 -1.2 -0.2 Laker -l.7 -0.3 -0.5 -0.2 T-39 -1.7 -0.9 -l.5 -0.3 S Valley -0.4 -0.9 -0.1 -0.3 Harris 2.5 0.8 1.2 0.9 Carioca -1.0 0.8 0.1 0.9 Biological yield Days to 50% flower. Parental Parental effect Effect effect Effect over between over between growth growth growth growth habits+ habits habits+ habits Seafarer -4.6 -1.0 -4.3 -2.0 Laker -1.0 -1.0 —4.8 -2.0 T-39 -2.9 -2.7 2.2 0.2 S Valley -2.8 -2.7 0.1 0.2 Harris 2.5 -0.1 5.8 1.0 Carioca 2.4 -0.1 -6.6 1.0 +Growth habit classification used by the International Center for Tropical Agriculture (CIAT), Colombia. Te Q14 01‘ V.‘ V. .c I... 1 4 l l l 1.. l 2 7. 2 2 To. 10. «J. 4.. 4.. .3. i 71 l l l l 2 2 .4 «4 «J Us To .5 .4. :3 \4. 131 Table 7. (Continued) Stem dry weight Pod wall dry weight Parental Parental effect Effect effect Effect over between over between growth growth growth growth Parent habits+ habits habits+ habits 1x2 0.96 -0.29 0.45 -0.19 1x3 1.64 0.85 1.37 0.68 1x4 1.81 0.85 1.89 0.68 1x5 -2.36 -1.10 -2.46 -0.49 1x6 -0.66 -1.10 -0.09 -0.49 2x3 0.36 0.85 -0.35 0.68 2x4 -0.42 0.85 -0.20 0.68 2x5 -0.62 -l.10 0.79 -0.49 2x6 1.42 -1.10 -0.20 -0.49 3x4 -0.70 -0.88 0.17 —0.33 3x5 0.24 1.25 0.79 -0.35 3x6 0.18 1.25 -0.45 -0.35 4x5 -0.05 1.25 —1.41 -0.35 4x6 -0.23 1.25 -0.35 -0.35 5x6 0.25 0.80 1.05 0.85 Biological yield Days to 50% flower. Parental Parental effect Effect effect Effect over between over between growth growth growth growth habits+ habits habits+ habits 1x2 0.86 —0.96 -.58 -1.99 1x3 4.12 1.79 0.31 1.40 1x4 4.49 1.79 1.78 1.40 1x5 -2.54 -0.82 —1.01 0.59 1x6 -2.34 -0.82 2.66 0.59 2x3 0.43 1.79 1.45 1.40 2x4 -1.90 1.79 2.06 1.40 2x5 -0.18 -0.82 -0.86 -0.82 2x6 1.76 -0.82 1.56 -0.82 3x4 -2.47 —2.69 -0.72 0.22 3x5 2.42 0.86 -3.01 -1.62 3x5 -1.65 0.86 -0.22 -1.62 4x5 0.32 0.86 -3.40 -1.62 4x6 2.35 0.86 0.14 -1.62 5x6 -2.57 -0.04 2.47 1.03 +Growth habit classification used by the International Center for Tropical Agriculture (CIAT), Cali, Colombia. 132 Seafarer and Laker had negative SCA effects for stem dry weight, pod wall dry weight, biological yield and days to 50% flowering. However, Harris had a positive SCA effects for all the traits analyzed. For example, Seafarer x Laker, Seafarer x T-39, Seafarer x Swan Valley'had positive SCA'effects for stem dry weight, pod wall dry weight, biological yield and days to 50% flowering, whereas Seafarer x Harris and Seafarer x Carioca had only negative SCA effects for all the traits considered in this study (Table 7L Maternal Effects Yield and yield components. In general maternal effects were detected only for number of seeds per pod and they were small in magnitude compared with GCA and SCA mean squares. Maternal effects variation among growth habits accounted for most of the variation observed for number of seeds per pod (Table 2 ). Maternal effects for number of pods per plant and number of seed per pod were significantly different between growth habits but within growth habits differences were not detected. Significant differences within growth habits were only observed for 100-seed weight (Table 2 ). Maternal effects are generally attributed to differences in heritable extra nuclear factors such as mitochondrial and chloroplast DNA ( 19 ). Morphological traits. Maternal effects were detected for stem 133 dry weight, pod wall dry weight, biological yield and 50% flowering. For stem dry weight and wall pod dry weight the variation among growth habit was more important than the variation within bean types, whereas for biological yield and days to 50% flowering the mean squares within types were larger than among growth habits (Table 4 ). Reciprocal Effects Yield and yield ggmpgggntg. Reciprocal effects were significant for number of seeds per pod and 100-seed weight. The within growth habit mean squares were much larger than among growth habits. This indicated that more variability was found within than among growth habits (Table 2 ). These effects are generally attributed to interactions between nuclear and extranuclear factors ( 19 ). Morphological traits. Reciprocal effect was significant only for days to 50% flowering and the within growth habit variance component was larger than the between growthhabits (Table 4 ). Interaction components In general, the absence of location x GCA, location x SCA, location x maternal and location x reciprocal efects for almost all the traits indicated that the GCA, SCA, maternal and reciprocal effects associated with parents were almost 134 consistent over locations (Table 8).Ckxthe other hand, the main larger magnitude:of GCA, SCA,1naternal and reciprocal effects compared with the interaction mean square values further suggested that the interaction effects may be small for the traits locations and genotypes evaluated. In spite of that, significant location x genotype interaction mean squares for pod wall dry weight per plant and days to 50% flowering were detected (Table 9), which indicated a differential response of the genotypes grown under different environmental conditions and soil types for these specific traits. When the location x genotype interaction was partitioned for stem dry weight per plant and days to 50% flowering showed significant location x GCA interaction. 0n the other hand, weight of 100—seed and days to 50% flowering presented significative location x SCA interaction. Days to 50% flowering also showed significant location x SCA and location x maternal effects and pods wall presented location x maternal and location x reciprocal effects (Table 9 ). Several other traits showed only significant differences when the general effect was partitioned into among and within growth habits effects. This situation occurred because the general effect indicated the average performance of the trait. 135 Table 8. Analysis of variance and estimates of location x combining ability interactions for components of yield and seed yield measured on F1 progeny of a six parent diallel cross grown at two locations in 1985. Pods/ Seeds/ 100-seed Seed Source+ df plant pod weight yield Location x Genotype 35 86.74 0.19 10.98 25.35 Location GCA 5 50.47 0.32 3.75 34.67 Among 2 7.28 0.16 0.47 70.01 Within 3 79.26 0.42 5.95 11.11 Location SCA 15 67.42 0.19 12.70 20.06 Among 3 27.66 0.12 25.24* 38.49 Within 12 77.48 0.22 9.57 15.45 Location x Maternal 5 145.65 0.22 19.52 28.38 Among 2 1.78 0.09 25.32* 29.89 Within 3 241.55* 0.30 15.64 27.37 Location x Reciprocal 10 10.40 0.10 6.99 27.01 Among 1 0.68 0.09 0.01 5.42 Within 9 115.75 0.11 7.77 29.52 Error 70 79.14 0.21 7.35 59.60* *Significant at 1% probability level. +Among and within refer to growth habits. 136 Table 9. Analysis of variance and estimates of location x combining interactions ability for morphological traits of a 6 x 6 diallel cross grown at two locations in 1985. 810- Days to Stem dry Pod wall logical 50% Source+ df weight dry weight yield flowering Location Genotype 35 8.32 12.58* 75.86 7.68** Location GCA 5 19.80 11.34 98.54 14.91** Among 2 10.27 5.54 79.41 15.67* Within 3 26.16** 25.20 119.29 14.40 Location SCA 15 7.35 8.52 55.15 6.56 Among 3 4.46 6.50 59.00 7.70 Within 12 8.07 9.03 54.19 6.28* Location x Maternal 5 8.98 18.08* 115.43 14.52** Among 2 5.18 9.88 86.87 3.76 Within 3 11.52 23.54* 134.60 21.69** Location x Reciprocal 10 3.71 16.53* 75.79 1.50 Among 1 2.81 2.75 0.43 8.76 Within 9 3.81 18.06** 84.16 1.60 Error 70 6.34 6.07 71.48 3.31 * Significant at 1% probability level. + Among and within refer to growth habits. Discussion In general, these results indicated that both additive and non-additive gene action were involved in the expression of yield and morphological characteristics studied. However, the additive gene action predominated. This indicated that recurrent selection should be effective in increasing yield and changing the levels of the expression of yield components and some morphological traits. These results do not agree with Chung and Goulden ( 6 ), Dickson ( 10 ), Fooland and Bassiri ( 12 ) and Singh and Saini ( 26 ) who reported larger specific combining ability mean squares for yield, yield components and for some morphological traits, but agree with Nienhuis and Singh ( 21 ) who found that general combining ability predominated in the expression of yield and yield components in dry beans. The discrepancies between the results in this and previous studies may be due in part to the different genotypes and locations in which these genotypes were evaluated. Moreover, it appeared that the among type variability was more important than within for almost all traits in this sample of genotypes. This suggested that more progress could be made by selecting among growth habits than within the same type. However, this possibility 137 138 has a practical limitation because plants with indeterminate growth habit may not be well adapted to northern latitude temperate environments were beans are grown in monoculture and under rainfall conditions. The significant and positive GCA effects and medium to large SCA effects suggested the possiblilty of selecting lines which would be superior to their parents. GCA effects accounted for most of the variation among hybrids, therefore, one might expect heritabilities of the traits to be high and thus respond rapidly to selection. However, some traits also had significant SCA effects which could be taken into account in planning breeding strategy for selection within segregating populations. The presence of reciprocal effects indicated that the choice of a parent to be used as either male or female will affect the expression of the progeny; hence, care should be taken in selection of parents. It is not known how often this effect might be detected nor how important it might be in breeding programs involving other parents. Although the interactions of years with GCA, SCA and reciprocal effects were statistically significant for some of the traits, these components of variance were small compared with the main effects. This indicated that a F1 hybrid xnanifested a broad adaptation to the sample of locations inhere they were evaluated. 139 In a practical breeding program of a self-pollinated crop, where the commercial cultivars grown are near homozygous lines, parents with high GCA would be more useful that those with high SCA. Gritton ( 17 ) suggested that the most efficient breeding procedure would be one that would permit the combination of desirable genes from the parents followed by the production of homozygous lines. Superior performing progenies can be produced by crossing parents with high GCA. This study showed the importance of GCA for the traits investigated. The genetic variability for each trait also suggested that it should be possible to make significant and rapid changes using breeding systems to concentrate primarily additive genes present in this population. No inferences to the general population of beans can be «irawn from the present experiment, since the number of parents used in this experiment was rather small, and they xwere selected for their differences in the traits studied and they did not represent a random sample of the dry bean germplasm. BIBLIOGRAPHY l.Adams, M. W. 1982. Plant architecture and yield breeding. Iowa State J. Res. 56:225-254. 2.Donald, C.M. 1967. The breeding for crop ideotypes. Euphytica 17:385-403. 3.Baker, R.J. 1978. Issues in diallel analysis. Crop Sci. 18:533-536. 4.81133, E.A. 1971. Inheritance of growth habit and time of flowering in beans, Phaseolus vulgaris L. J. Amer. Hort. Sci. 96:715-717. 5.Cockerham, C.C. 1980. Random and fixed effects in plant genetics. Theor. Appl. Genet. 56:119-131. 6.Chung, J.H., D.S. Goulden. 1971. Yield components of haricot beans(Phaseolus vulgaris L.) grown at different plant densities. N. Z. J. Agric. Res. 14:227-234. 7.Coyne, D.P. 1968. Correlation, heritability and selection of yield components in field beans, Phaseolus vulgaris L. Proc. Amer. Hort. Soc. Sci. 93:388-396. 8.Coyne, D.P., R.H. Mattson. 1964. Inheritance of time and length of blooming in Phaseolus vulgaris L. Proc. Amer. Soc. Hort. Sci. 85:366-373. 9.Davis, D.W. and W.A. Frazier. 1966. Inheritance of some growth habit components in certain types of bush lines of Phaseolus vulgaris L. Proc. Amer. Soc. Hort. Sci. 88:384-392. 10.Dickson, M.R. 1967. Diallel analysis of seven economic characters in snap beans. Crop Sci. 7:121-124. 12.Foolad, M.R. and A. Bassiri. 1983. Estimates of combining ability, reciprocal effects and heterosis for yield and yield components in a common bean diallel cross. J. Agric. Sci., Camb. 100:103-108. 13.Frazier, W.A., J.R. Baggett and W.A. Sistrink. 1958. Transfer of certain Blue Lake Pole bean pod characteristic to 140 141 bush beans. Proc. Amer. Soc. Hort. Sci. 71:416-421. 14.Ghaderi, A., and M.W. Adams. 1981. Preliminary studies on the inheritance of structural components of plant architecture in dry bean (Phaseolus vulgaris L.). Bean Improv. Coop. 24:35-38. 15.Goldsworthy, 1984. Crop growth and development: The reproductive phase, p.163-212. In: P.R. Goldsworthy and N.N. Fisher (Eds.). The physiology of tropic field crops. John Willey & Sons. N.J. 16.Griffing, B. 1956. Concept of general and specific combining hability in relation to diallel crossing systems. 17.Gritton, E.T. 1975. Heterosis and combining ability in a diallel cross of peas. Crop Sci. 15:453-457. 18.Hayman, B.I. 1954. The theory and analysis of diallel crosses. Genetics 39:789-809. 19.Islieb, T.G., J.C. Wynne, C.H. Elkan and T.J. Schneeweis. 1980. Quantitative genetic aspects of nirogen fixation in peanut. (Arachis hypogaea L.). Peanut Sci. 7:101-105. 20.Krarup, A. and D.W. Davis. 1970. Inheritance of seed yield and its components in a six-parent diallel cross in peas. J. Amer. Soc. Hort. Sci. 95:795-797. 21.Nienhuis, J. and S. Singh. 1986. Combining ability analysis and relationships among yield, yield components and archytectural traits in dry beans. crop Sci. 26:21-27. 22.Pandey, S. and E.T. Gritton. 1975. Inheritance of protein and other agronomic traits in a diallel cross of pea. J. Amer. Soc. Hort. Sci. 100:87-100. 23.Paschal, H.H. and J.R. Wilcox. 1975. Heterosis and combining ability in exotic soybean germplasm. Crop Sci. 15:344-349. 24.Row1ands, D.G. 1964. Genetic control of flowering in Pisum sativum L. Genetica 35:75-94. 25.Safari, A. 1978. A yield-component selection experiment involving american and iranian cultivars od the common bean. Crop Sci. 18:5-7. 26.Singh, A.K. and 8.8. Saini. 1983. Heterosis and combining 142 ability studies in french beans. SABRAO J. 15:17-22. 27.Snoad, B. and A.E. Arthur. 1973. Genetical studies of quantitative charactes in peas. 1. A seven-parent diallel cross of cultivars. Euphytica 22:327-337. 28.Sprague, G.F and L.A. Tatum. 1942. General vs specific combining ability in single crosses of corn. JuAmer. Soc. Agr. 34:923-932. 29.Watts, IHE., E. Stevenson and M.J. Crampton. 1970. Inheritance of flowering time in six pea cultivars (Pisum sativum In). Euphytica 19:405-410. 30.Weber, C.R”.InT. Empig and J.C. Thorne. 1970. Heterotic performance and combining ability of two way Fl soybean hybrid. Crop Sci. 10:159-160. 31.Wilson, D. 1981. Breeding for morphological and physiological traits, p.233-290. In: K.J. Frey (Ed.). Plant Breeding II. The Iowa State Univ. Press. Ames, Iowa. CHAPTER 4 COMBINING ABILITY EFFECTS FOR SEED FILLING TRAITS IN A DIALLEL CROSS OF DRY BEANS WITH DIFFERING GROWTH HABITS AND MORPHOLOGI CAL CHARACTERI STICS ABSTRACT Analysis of variance detected significant differences for total and linear filling duration, linear filling rate mean filling rate and predicted maximum seed weight which indicated that genetic variability existed for these traits. The GCA mean squares from the diallel analysis of the F1 showed that the variation for total and linear filling duration, mean filling rate and seed weight (predicted maximum value) were significant, therefore, the additive gene action was mainly controlling the expression of these traits. The growth habit variation (between and within) was not consistent with any trait, indicating that for some traits the variation between growth habit was larger than within types, but the opposite was true for other traits. LFD was the only parameter that showed significant SCA variation. However, when the SCA mean squares were partitioned into among and between growth habit variation, the mean filling rate and predicted maximum dry weight showed highly and 143 144 significant variation between growth habit components, respectively. Maternal and reciprocal effects were only significant for linear filling duration indicating that the selection of parents should be emphasized to improve this trait. In general, the GCA variances were larger than SCA, and reciprocal components which would indicate that genetic improvement is possible using breeding systems that concentrate additive genes in this population. INTRODUCTION The yield of a crop used for it seeds is a function of the rate and duration of accumulation of dry weight in the seed ( 4 ). Studies in soybeans (Glycine max (L) Merril)( 5, 19 ), wheat (Triticum aestivum L.) ( 14, 20 ), corn (Zea mays L.) ( 4, 23 ) and rice (95.2.2.9. satiya L.) ( 10 ), have demonstrated that species and cultivars within species differ in both rate and duration of grain filling, and since these parameters are heritable ( 7, 18, 19 ) it has been suggested that selection for grain filling duration would result in yield improvement ( 4, 9, 19 ). Limited data on seed filling are available in dry beans, Izquierdo and Hosfield ( 9 ) reported that phenotypic variability was present in dry beans for seed filling duration and rate of seeds. They also found that linear filling duration (LFD) was positively correlated with seed yield. They pointed out that a long LFD and a lack of yield components compensation normally observed in dry beans could be an explanation for the higher seed yield of dry bean architypes as compared to the traditional cultivars grown in the Michigan Area ( 19 ). The grain filling period in dry beans begins at anthesis 145 146 and continues until physiological maturity. At this stage the seed has reached its maximum weight. The extension of the seed filling duration results from shortening the period from planting 1x3 flowering (early flowering) or lengthenning the period from flowering to physiological maturity (increase lateness). However, this longer filling duration should be coupled with a high leaf area index (LAI) to provide an excess of photosyntate for seed filling. The objective of this study was to estimate combining ability and genetic effects for several parameters among 6 genotypes of dry beans that differed in growth habit, architecture, and lenght of maturity. Literature review Genetic Variability The duration of seed filling has been found to be positively associated and significantly correlated with seed yield in dry beans (Phaseolus vulgaris L.)( 9 ), soybean (G1ycine1nax(ld Merril)( 5, 19 L peas (Pisum sativum L.)( 12 ), wheat (Tritiggm sp.)( 14, 15, 20 ), barley (nggeum vulgare L )( 13, 18 ), and corn (Zea mays L.)( 1, 4, l6 ). However, in experiments with rice (Orysa sativa L.)( 10 L soybean ( 5 ), and spring wheat (Tritigum aestiygm L.) the data indicated that the rate of seed filling rather than duration was related to seed yield. Heritability of Duration and Rate gf Seed Filling Limited data are available regarding the heritability of grain filling parameters (rate and duration). Results from work of Hillson and Penny (E3), and Reicosky et a1 ( l9 ) indicated that this trait may be heritable. Hallauer and Russel ( 7 ) suggested that heterosis for grain filling traits was present in the material evaluated. Perenzin et a1 ( l7 ) reported in corn highly significant genotypic 147 148 differences and high heritability estimates for duration of grain filling (84;2%L. grain filling rate (97%), seed weight (96:7%) and protein content (72.8%) in corn. In addition, the kernel weight was positively and significantly correlated with the seed filling duration. This finding provided a simple criterion for selecting for seed filling duration. The possibility of improving the duration of seed filling through selection is supported by the trait's high heritability. Rasmusson.et a1 (lii) found that the estimates of heritability in barley were low for duration of the grain filling period based on single plots (parent-progeny method), but relatively high when based on means of replicated plots (variance component method). These authors recomended that more information should be obtained concerning the relationship between the effective grain filling period and the duration of growth periods before breeding work was done. Type 9f Gene Action Cross ( l ) reported that GCA effects for duration and rate of grain filling in corn predominated over the SCA (effects. This suggested that recurrent selection should be effective procedure in improving this trait. Walton ( 24 ) found that both additive and cknninance variances were important in grain filling in wheat anui that overdominance was expressed for a long grain filling period. Singh et a1 ( 20 ) pointed out that additive genetic 149 variance was more important than non additive genetic variance in determining the expression of the grain filling period in wheat. Neuhausen ( 15 ) showed that additive effects were important for duration of grain filling in the three sets of spring wheat genotypes studied. In one cross, dominance gene effects were equally as important as additive ones. She suggested that selection for these parameters in wheat should be on a family mean basis whereas in those cases in which dominance gene effects were of equal importance to additive, selection would not be as effective as when all the gene action was additive because dominance variance is not fixable and would have little use in a pure line wheat breeding program. Poneleit and Egli ( l6 ) showed heterotic responses in F1 corn hybrids relative to the parents lines. This suggested that traditional hybrid programs might benefit from the use of inbreds with long grain filling periods. Reciprocal differences among F1 kernel genotypes were found for both the effective filling period and kernel growth rate. They pointed out that the differences for kernel growth rate can be caused by dosage effects in the kernel endosperm. Either effects indicated that control of pollination is needed for accurate evaluations of effective filling period. Regulation 9: the Seed Filling Period Little is known about the mechanism by which plants 150 regulate the length of seed filling. Duncan ( 3 ) speculated that one possible determinant could be the relationship between.photosynthetic rate and sink capacity: Daynard and Kennenberg ( 4 ) hypothetized that the duration of grain filling may be established primary by kernel volume. Tollenar and Daynard ( 23 ) suggested that kernel sink strength may affect the rate of dry matter accumulation. Genotype x Environment Interaction Wych et al ( 25 ) reported genotype x year interactions for the seed filling period, seed filling rate, seed yield and seed weight in oats. Reicosky et a1 ( l9 ) showed that the reproductive, seed filling periods were highly correlated across years in soybeans. They also suggested that one plant 'would.be an adequate sample of the measure of seed filling period for a homogeneous line when large differences existed among them. Seed Filling Period and Rate as a Selection Criterion Selection for the duration of the seed filling period and rate could be a useful tool for plant breeders if the heritability was high and positively correlated with yield. Smith and Nelson ( 22 ) reported that the seed filling duration could be used as a selection criterion for developing high yielding soybean cultivars. They found a 151 positive association between seed filling durations in the F4, F5, F6 and F7, however they did not find this relationship in the F2. This might indicate that these parameters have a low heritability and selection in later generations would be more advantageous for improving these traits. Smith and Nelson ( 21 ) also showed a positive association among seed filling duration and F4 plants, F5 plants rows and replicated yield plots of F4 derived lines. This evidence suggested that increases in yield may be made by selecting for a long filling duration, but it is not clear whether the increases would be sufficiently large to compensate for the additional effort required inlneasuring these parameters rather than yield per se. Metzger et al ( 13 ) reported that selection for an optimum seed filling duration to improve seed yield does not appear to be a worthwhile objective in the upper mid-western barley types. Materials and Methods Six strains of dry beans were selected as parents in this study on the basis of differential growth habits, days to maturity, and seed filling characteristics. The materials and methods used in this study were explained elsewherel'2 Results The analysis of variance detected significant differences between genotypes for TFD, LFD, LFR, mean filling rate and predicted maximum seed weight. This indicated that genetic variablity existed for these traits. Further, this genetic variation was partitioned into GCA, SCA and maternal components. General Combining Ability. Seed filling duration. The GCA mean squares from the diallel analysis of the F1 (Table 1 ) showed that the variation for TFD and LFD was signicant. This indicated that additive genetic effects governed large portion of the expression of these traits. 1Chapter 2. Paredes, C.M., 1986, M.S. Thesis, Department of Crop and Soil Science, Michigan State Univ.,East Lansing,MI 8824 Chapter 3. Paredes, C.M., 1986, op cit 152 153 Table 1. Analysis of variance for GCA, SCA and reciprocal effects for total filling duration and linear filling duration measured on F progeny of a six- parent diallel cross grown at wo locations in 1985. Filling duration Source+ df Total Linear Total 71 25.34 1.46 Replication 1 0.06 0.07 Genotypes 35 34.52* 2.55** GCA 5 93.70* 5.35** Among 2 59.08 9.25** Within 3 116.77** 2.75** SCA 15 28.34 2.13** Among 3 26.14 1.67** Within 12 28.89 2.24** Maternal 5 39.03 3.73** Among 2 65.36* 6.08** Within 3 21.46 2.17** Reciprocal 10 11.93 1.21* Among 1 11.60 2.08* Within 9 11.97 1.11** Error 35 16.86 0.35 *,** Significant at 5% and 1% probability level, respectively. + Among and within refer to growth habits. 154 The GCA mean squares within growth habit was twice as large as among growth habit suggesting that more variation was found within the bean types rather than among ones. However, the among types GCA mean square three fold larger than within growth habits. Estimates of GCA effects for TFD and LFD are presented in Table 2.]xigeneral, early genotypes within each growth habit (Seafarer, T-39 and Harris) transmitted a reduced TFD and LFD whereas, the late genotypes ( Laker, S. Valley and Carioca) tended to impart a longer TFD and LFD. Seed filling rate and predicted maximum weight. Only the GCA mean squares for the mean filling rate was significant, indicating the additive gene action was important controlling the expression of this trait (Table 3). The among growth habit variation was also larger than the within types in this sample of genotypes. This suggested if one wants to breed for this trait one should select and cross parents from different growth habits. Estimates of GCA effects for LFR, MFR indicated that Seafarer and T-39 transmitted a reduced effect for these characters. However, Harris showed a large positive effect :flor both traits (Table 1). The T-39 genotype tended to reduce the seed weight (based on its predicted maximum valueL xMhereas, Harris transmitted a positive effect for bean seed size (Table 4). 155 Table 2. Estimates of GCA effects for Total filling duration, Linear filling duration, and yield measured on F progeny of a six parent diallel cross grown a two locations in 1985. Seed filling duration Parent Total Linear Yield Seafarer 0.9 -0.6 -2.5 Laker 0.6 -0.2 0.7 T-39 -1.0 0.3 -0.3 Swan Valley 3.4 0.7 -0.4 Harris -2.4 -0.3 -0.5 Carioca 0.3 0.3 3.1 156 Table 3. Analysis of variance for GCA, SCA and reciprocal effects for linear and mean filling rate, and predicted maximum weight measured on F1 progeny of a six-parent diallel cross grown at two locations in 1985. Predicted Filling duration maximum Source+ df Linear Mean weight Total 71 4.76 1.44 2500.26 Replication l 2.49 2.96 4092.11 Genotypes 35 5.34 2.08** 3298.62* GCA 5 12.83 7.95** 10372.70* Among 2 13.08 15.07** 18419.05** Within 3 12.67* 3.20* 5008.46* SCA 15 5.10 1.55 3233.01 Among 3 9.49 4.26** 5310.19* Within 12 4.01 0.87 2713.72 Maternal~ 5 5.54 0.87 541.96 Among 2 6.80 1.37 384.05 Within 3 4.69 0.53 647.24 Reciprocal 10 1.87 0.55 1238.32 Among l 14.30 0.80 681.01 Within 9 0.46 0.53 1300.24 Error 35 4.25 0.75 1656.46 *,** Significant at 5% and 1% probability level, respe ctively. + Among and within refer to growth habits. 157 Table 4. Estimates of GCA effects for Linear filling rate, Mean filling rate and Predicted maximum weight measured on F1 progeny of a six-parent diallel cross grown at two locations in 1985. Predicted Filling rate Maximum Parent Linear Mean weight Seafarer -0.4 -0.1 —12.4 Laker 0.5 -0.4 - 9.8 T-39 -0.2 —0.3 -24.0 Swan Valley -0 9 -0.5 1.0 Harris 1.2 1.1 35.1 Carioca -0.2 0 2 10.1 158 Specific Combinig Ability. Seed filling duratitwn Linear filling duration was the only parameter that showed significant SCA variation (Table 1). This indicates the importance of the dominance gene effect. However, the magnitude of this component was small compared to the additive ones. The within growth habit variance was larger than the among types for this trait. Specific combining ability effects for LFD showed that Seafarer and Carioca had the ability to combine better than expected, based on their own performance. The positive effect of Harris for LDF suggested that the mean of this hybrid was greater than expected, based on the mean performance of the lines involved in the crosses. Seed filling rate and predicted maximum seed weight. The SCA mean squares were not significant for any of the parameters evaluated. However, when the SCA mean squares were partitioned into among and between growth habits variation, xnean filling rate and predicted maximum dry weight showed luighly'and significant variation between the growth habits component, respectively. Maternal Effects Eheed filling duration. Maternal and reciprocal effects were significant only for LFD. The variation among growth habits \uas larger than within bean types for both LFD and TFD. 159 Seed filling rate and predicted maximun seed weight. Linear filling rate , mean filling rate and predicted maximum seed weight did not showed either significant maternal or reciprocal effects. Discussion The significant differences among genotypes detected by the ANOVA suggested that there existed genetic variability in this population. The existance of genetic variability had been reported previously in beans (99 ), corn, (1” 4, 16 L and small cereals ( 14, 15, 20 ). 0n the basis of the ANOVA of F1 hybrids for the combining ability, the additive genetic system was shown to 'be very important in controlling TFD, LFD, mean filling rate and seed size (maximum predicted value). However, LFD showed :3 highly significant SCA variance, indicating that the «iominance gene action was controlling partly the expression of this trait.lklgenera1q the GCA effects were larger than 'the SCA component. This indicated that heritability for these 'traits could be high and selection could be possible using ‘genetic systems that concentrate primarily additive genes present in this population. This hypothesis agrees with :Perenzini et al ( l7 ) who reported high heritability estimates for grain filling duration, grain filling rate and seed weight in corn. The primarily additive genetic control of duration and :rate is similar to those reported by Cross ( 1 ) in corn, 160 161 Walton ( 24 ), Singh et al ( 20 ) and Neuhansen ( 15 ) in wheat who concluded that the GCA effects for duration and rate ofigrain filling was primarily under additive genetic control in corn. Only LFR showed significant maternal and reciprocal effects, indicating that the selection of parents in a breeding program should be emphasized to improve this trait. This results agreed with Poneleit and Egli ( 16 ) who reported reciprocal differences among Fl genotypes for both TFD and kernel growth rate in maize. They attributed this effect to differences in kernel growth rate to dosage effects in the kernel endosperm. It is interesting to note that the genotype Carioca showed a positive effect for LFD and yield per plant. These results agree with Duncan (44 ) who pointed out that there existed genetic potential for higher yields through an extention of the length of the grain filling period. Izquierdo and Hosfield ( 9 ) and Duncan ( 4) reported that linear filling duration was positively correlated with yield in dry beans and corn, respectivelyu In spite of this fact, it seems that not all the parents with long LFD transmitted a positive GCA effects to its progeny therefore, the selection of the parents is crucial in the future improvement of these traits through genetic manipulation. The results agree with Izquierdo and Hosfield ( 9 ) who 162 pointed out that phenotypic variation there existed for the seed filling parameters and increases in yield may be made by selecting for long filling duration but it is not clear whether the increases would be sufficiently to compensate for the additional efforts required in measuring these parameters rather than yield per se ( 21 ). Bibliography 1.Cross, H.Z. 1975. Diallel analysis of duration and rate of grain filling of senen inbred lines of corn. Crop Sci. 15:532-535. 2.Cockerham, C.G. 1980. Ramdom and fixed effects in plants genetics. Theor. Appl. Genet. 56:119—131. 3.Daynard, T28. and L.W. Kennenberg. 1976. Relationships between lenght of the actual and effective grain filling periods and the grain yield of corn. Can. J. PLant Sci. 56:237-242. 4.Duncan, W.G., A.L. Hatfield and J.L. Ragland. 1965. The growth and yield of corn. II. Daily growth of corn kernels. Agron. J. 57:221-223. 5.Egli, D.B.,J.E. Leggett and J.M. Wood. 1978. Influence of soybean seed size and position on the rate and duration of filling. Agron. J. 70:127-130. 6.Griffing, B. 1956. Concept of general and specific combining ability in relation to diallel crossing systems. Aust. J. Biol. Sci. 9:463-493. 7.Hallauer A.P. and W.A. Russel. 1969. Estimates of the maturity and its inheritance in maize. Crop Sci. 2:289-294. 8.Hillson M.T. and L.H. Penny. 1965. Dry matter accumulation and moisture loss during maturation of corn grain. Agron. J. 57:150-153. 9.Izquierdo J.A. and G.L. Hosfield. 1983. The relationship of seed filling to yield among dry beans with differing archytectural forms. J. Amer. Soc. Hort. Sci. 108:106-111. 10.Jones, D.B., M.L. Peterson and S. Geng. 1979. Association between the planting rate and duration and yield components in rice. Crop Sci. 19:641-644. 11.Kaplan, S.L. and H.R. Koller. 1974. Variation among soybean cultivars in seed growth rate during the linear phase of seed growth. Crop Sci. 14:613-614. 163 164 12.Mahon, JxD. and SJLA. Hobbs. 1983. Variability in pod filling characteristics of peas (Pisgm satiygm L.) under field conditions. Can. J. PLant Sci. 63:283-291. l3.Metzger, D.D., S.J. Czaplewski and D.C. Rasmusson. 1984. Grain filling duration and yield in spring barley. Crop Sci. 24:1101-1105. 14.Nass, H.G. and B. Russer. 1975. Grain filling period and yield relationships in spring wheat. Can. J. PLant Sci. 55:673-678. liLNeuhausen, S. 1982. Genetic variation for the duration of grain filling in a sample of the wheat germplasm pool. M.S. Thesis, Michogan State Univ” East Lansing. l6.Poneleit CLG. and D.B. Egli. 1983. Differences between reciprocal crosses of maize for kernel growth characteristics. Crop Sci. 23:271-275. l7.Perenzin, bk, .F. Ferrari and bk Motto. 1980. Heritabilities and relationships among grain-filling period, seed weight and quality in forty italian varieties of corn (Zea mays L.). Can. J. Plant Sci. 60:1101-1107. 18.Rasmusson, D.G. 1. Mc Lean and T.L. Tew. 1979. Vegetative and grain filling periods of growth in barley. Crop Sci. 19:5-9. 19.Reicoskyg DNA”.£LH. Orf and CAL Poneleitu 1982. Soybean germplasm evaluation for length of the seed filling period. Crop Sci. 22:319-322. 20.Singh V.P., B. N. Dahiya and R.K. Chowdhry. 1977. Genetic studies of grain filling period in wheat. In.AnK. Gupta (ed). Genetics and wheat improvement. 11. Genetics of economic traits. Oxford and IBH Publishing Co., New Delhi. 21.Smith, 33H. and R.L. Nelson. 1986. Selection for seed filling period in soybean. Crop Sci. 26:466-469. 22.Smith J.H. and R.L. Nelson. 1986. Relationship between seed-filling period and yield among soybean breeding lines. Crop Sci. 26:469-472. 23.Tollenar, M. and TKB. Daynard. 1978. Kernel growth and development at two position on the ear of maize (Zea mays L.). Can. J. Plant. Sci. 58:189-197. 24.Walton, PJL 1971. The genetics of yield in spring wheat 165 (Triticum aestivum L.). Can. J. Genet. Cytol. 13:110-114. 25.Wych, R.D., R.L. McGraw and D.D. Stuthman. 1982. Genotype x year interaction for legth and rate of grain filling in oats. Crop Sci. 22:1025-1028. SUMMARY AND CONCLUS IONS Selection of superior plants for economic yield, disease resistance, and tolerance to adverse conditions in breeding programs has led to the development most new crops cultivars. However, an increased understanding of growth and development and physiology of plants may result in more efficient approach to yield improvement than to the traditional hybridization and selection approach. The objectives of the present study was to evaluate whether variation in morphological traits among 12 dry bean genotypes was related to yield and estimate the combining ability and genetic effects of these traits. The genotypes were characterized by a Type I, II and 111 growth habit. Type I and II were represented by two different architectures, namely the traditional form and the architype. The genotypes were selected based on their differences in days to maturity, reproductive growth pattern, seed size and yield. During the course of this research dry matter accumulation in different plant parts of the 12 dry bean genotypes was determined. In addition, several seed filling parameters were estimated based on seed dry matter accumula- 166 167 tion using a cubic regression equation. The following conclusions emerge from this study. 1.Determinate genotypes had fewer days to 50% flowering, days to physiological maturity and seed weight compared with the indeterminate ones. 2.Type II genotypes had a higher seed yield, number of pods per m2, seeds per pod and biological yield than Type III. However, Type III genotypes had a higher 100-seed weight, fewer days to 50% flowering and physiological maturity than Type II. 3.Architypes I genotypes had a higher seed yield, number of 2 and number of seed per pods than traditional I pods perxn genotypes but traditional I showed a higher 100-seed weight. Architype II differed in 100-seed weight when it was compared to traditional II. Late Type III genotypes outyielded in seed and biological yield production to earlier Type III genotypes. 4.In general, indeterminate types produced more total dry 'weight than determinate ones. Type II genotypes produced more leaves than Type I and III which could indicate a greater potential for photosynthetic production. 5.The maximum LAI, LAR and RGR was reached at early stage ( 4-12 days after flowering ) of the reproductive phase, thereafter, these parameter declined as dry matter of leaves decreased. 168 ELGrain yield was positively correlated with number of pods per plant, number od seeds per pod, biological yield , days to 50% flowering, physiological maturity and linear filling rate. 75Total dry weight, leaf dry weight and leaf area were positively correlated with yield at 12 days after 50% flowering. LAI was Positively associated with yield at 4, 12, and 20 days after flowering. This could suggest that an early development of a canopy and long duration of a LAI high is important for high yields. 8.In general, Type I had a shorter total filling duration, linear filling duration and linear filling rate compared with Type II and III. Architypes showed a similar or slightly different total filling duration and linear filling duration to traditional forms. Late genotypes tended to have a longer total filling duration and linear filling duration than earlier genotypes but earlier genotypes had a faster linear filling rate than the later ones. 9.Linear filling duration was positively and significant correlated with yield, number of pods per m2, number of seeds per pod, biological yield and physiological maturity. However, total filling duration was only positively correlated with days to 50% flowering and physiological maturity. Mean filling rate was positive and significatively correlated to number of seeds per pod and biological yield. 169 In 1985, six genotypes were selected as a parents based on their growth habits, number of days to physiological maturity, and seed filling characteristics. All posible F1 hybrids including reciprocals were made among the parents. The 30 F1 hybrids were planted at two locations, East Lansing and Saginaw, Michigan. Five plants were harvested at weekly intervals for eight consecutive weeks at both locations. Harvesting began after the first pod was formed on each genotype and continued until physiological maturity. Plants were separated into plant morphological components, dried and weighed at 75 C for 48 h. The seed dry weight was used to estimate the seed filling parameters (TFD, LFD, LFR, MFR). The objective of this study was to estimate the combining ability and genetic effects of yield, yield components, seed filling parameters enui some other morphological traits. The analysis of combining hability was performed using the formulas presented by Griffing for Model I, Method 1. This analysis was extended to include the partitioning of reciprocal variation into maternal and non maternal effects according to Cockerham. The following conclusions can be drawn from this study. 1. The GCA mean squares were significant for yield, yield components, stem dry weight, biological yield, days to 50% flowering, total filling duration, linear filling duration 170 and.mean filling rate.1fin£;suggested.that additive genetic effects controlled a large portion of the expression of these traits. iLThere was a considerable variation in the GCA contribution of the growth habits (among and within) and each parent to the various traits analized. No one parent had all positive or negative GCA effects for all the traits studied. However, there were some parents which could complement each other by combining positive GCA effects for some traits. ELThe large-seeded parent (Harris) had a positive GCA effect for seed weight but a negative ones for seed yield. This could indicate that using this parent in a breeding program would.be difficult to improve yield and seed weight at the same time. 4.The SCA mean squares were significant for number of seeds per pod, 100-seed weight, 50% flowering and linear filling rate. 5.The significant and positive GCA effects and medium to large negative SCA effects suggested that the possibility of selecting lines which would be superior to their parent. ELThe presence of reciprocal and maternal effects indicated that the choice of a parent to be used as either male or female is going to affect the F1 outcome and special care should be taken in selection of parents. 7.Although the interactions of locations with GCA, SCA and 171 reciprocal effects were statistical significant for some of the traits, these components of variance were small compared with the additive part. 8.This study showed the importance of GCA for the traits investigated. The genetic variability for each trait also suggested that it should be possible make rapid changes in these traits using breeding systems to concentrate primarily additive genes present in this population. 172 Table 1. Regression and determination coefficients from the cubic polynomial regression equation for total dry weight on days after 50% flowering of 12 dry bean genotypes grown in 1984. Genotypes 50 B1 52 k3 R2 Seafarer 10.0291 -0.2878 0.0650 -0.0015 0.849 Brazil-2 9.1811 -0.2974 0.0425 0.0008 0.840 Laker 10.3828 -0.5995 0.0650 -0.0011 0.942 C-49 9.8141 -0.4360 0.6454 -0.0013 0.952 T-39 24.0956 —l.3353 0.1265 0.0023 0.935 Nep-2 9.0291 -0.1344 0.0341 0.0005 0.933 C-20 9.3367 0.1758 0.0251 -0.0004 0.955 8 Valley 13.1267 -0.7597 0.0855 -0.0015 0.941 Carioca 3.9449 1.1999 -0.0334 0.0005 0.932 Viva 10.2051 -0.8738 0.0843 0.0017 0.816 Harris 11.9998 -1.2841 0.1406 —0.0031 0.861 Valley 11.3642 -0.7171 0.0805 -0.0014 0.927 APPENDIX A 173 Table 2. Regression and determination coefficients from the quadratic polynomial regression equation for dry beans on days after 50% flowering of 12 dry bean genotypes grown in 1984. Genotypes B0 61 62 R2 Seafarer 4.0286 0.3994 -0.0177 0.874 Brazil-2 4.9032 0.1452 —0.0076 0.917 Laker 3.3499 0.3488 -0.1270 0.742 C-49 5.0118 0.0796 -0.0068 0.871 T-39 4.4244 0.2996 -0.0112 0.744 Nep-2 4.3690 0.1971 0.0067 0.657 C-20 5.3093 0.2261 -0.0093 0.880 S Valley 5.9661 0.1890 -0.0089 0.750 Carioca 3.8060 0.3511 -0.0140 0.864 Viva 3.2191 0.2322 -0.0098 0.534 Harris 4.3809 0.2066 -0.0100 0.571 Valley 3.4476 0.3685 -0.0115 0.708 174 Table 3. Regression and determination coefficients from the cubic polynomial regression equation for pod wall dry weight on days after 50% flowering of 12 dry bean genotypes grown in 1984. Genotypes 50 51 82 +33 R2 Seafarer -0.0571 0.0090 0.0033 -0 0001 0.865 Brazil-2 -0.0657 -0.0650 0.0089 -0.0002 0.876 Laker 0.2645 -0.0985 0.0099 -0.0002 0.851 C-49 0.2586 -0.1040 -0.0128 -0.0003 0.898 T-39 0.3414 -0.1195 0.0109 -0.0002 0.948 Nep-Z 0.2326 -0.0747 0.0071 -0.0001 0.935 C-20 0.1969 -0.0732 0.0084 -0.0002 0.936 S Valley 0.3167 -0.1028 0.0108 -0.0002 0.941 Carioca -0.4519 0.1096 -0.0026 0.0001 0.900 Viva -0.0648 0.0106 0.0028 -0.0001 0.816 Harris -0.0428 0.0039 -0.0001 -0.0001 0.892 Valley 0.5264 -0.1512 0.0127 -0.0002 0.842 BIBLIOGRAPHY luAdams, NLW. 1967. Basis of yield component compensation in crop plants with special reference to the field bean, Phaseolus vulgaris . Crop Sci. 7:505-509. 2.Adams, M.W. 1981. Update: new bean archetypes. Michigan Dry Bean Dig. 5:12-13. 3.Adams, M.W. 1982. Plant archytecture and yield breeding. Iowa State J. Res. 56:225-254. 4.Adams, M.W., J.V. Wierma, and J. Salazar. 1978. Differences in starch accumulation among dry bean cultivars. Crop Sci. 18:155-157. 5.Askel, R. and L.P.V. Jonhson. 1961. Genetic studies in sowing-to heading and heading-to ripenning periods in barley and their relatopnship to yield components. Can. J. Genet. Cytol. 3:242-259. 6.Baker, R.J. 1978. Issues in diallel analysis. Crop Sci. 18:533-536. 7.Beaver, J.S. and R.L. Cooper. 1982. Dry matter accumulation patterns and seed yield components of two indeterminate soybean cultivar. Agron. J. 74:380-383. 8.Beaver, J.S., C.V. Paniagua, D.P. Coyne and G.F. Freytag. 1985. Yield stability of dry bean in the Dominican Republic. Crop Sci. 25:923-926. 9.Binnie, R.C. 1981. Flower and pod production in Phaseolus vulgaris. J. Agric. Sci., Camb. 97:397-402. 10.Bliss, F.A. 1971. Inheritance of growth habit and time of flowering in beans, Phaseolus vulgaris L. J. Amer. Hort. Sci. 96:715-717. 11.Brame1, P.J.,P.N. Hinz, D.E. Green and R.M. Shibles. 1984. Use of principal factor analysis in the study of three stem termination in types of soybean. Euphytica 33:387-400. 12.Boote, K.J. 1981. Response of soybeans in different maturity groups to march plantings in Southern USA. Agr. J. 73:854-859. 175 176 13.Bunting, AWH. 1975. Time phenology and yields of crops. Weather 30:312-325. l4uButtery, B.R. 1970. Effects of variation in leaf area index on growth of maize and soybean. Crop Sci. 10:9—13. 15.Buzzel, R.I. and B.R. Buttery. 1977. Soybean HI in hill plots. Crop Sci. 17:968-970. 16.Cavaleiri, A. J. and 0.8. Smith. 1985. Grain filling and field drying of a set of maize hybrids released from 1930 to 1982. Crop Sci. 25:856-860. 17.Chang, J.F” ILE. Green and R. Shibles. 1982. Yield and agronomic performance of semi-determinate and indeterminate soybean stem type. Crop Sci. 22:97-101. lBuChung, CLH., D.S. Goulden. 1971. Yield components of haricot beans(Phaseolus vulgaris Lu) grown at different plant densities. N. Z. J. Agric. Res. 14:227-234. 19.Cockerham, CJL 1980. Random and fixed effects in plant genetics. Theor. Appl. Genet. 56:119-131. 20.Coyne, D.P. 1968. Correlation, heritability and selection of yield components in field beans, Phaseolus vulgaris L. Proc. Amer. Hort. Soc. Sci. 93:388-396. 21.Coyne, D.P. 1980. Modification of plant archytecture and crop yield by breeding. Hort. Sci. 15:244-247. 22.Coyne, D.Pn ELH. Mattson. 1964. Inheritance of time and length of blooming in Phaseolus vulgaris L. Proc. Amer. Soc. Hort. Sci. 85:366-373. 23.Crookston, R.Kn J. O'Toole and J.L. Ozbun. 1974. Characterization of bean pod as a photosynthetic organ. Crop Sci. 14:708-712. 24uCross, FLZ. 1975. Diallel analysis of duration and rate of grain filling of seven inbred lines of corn. Crop Sci. 15:532-535. 25.Daynard, T.B. and L.W. Kennenberg. 1976. Relationships between lenght of the actual and effective grain filling periods and the grain yield of corn. Can. J. PLant Sci. 56:237-242. 26.Daynard, T.B., J.W. Tanner and W.G. Duncan. 1971. Duration of the grain filling period and its relation to grain yield 177 in corn, Zea mays L. Crop Sci. 11:45-48. 27.Davis, D.W. and W.A. Frazier. 1966. Inheritance of some growth habit components in certain types of bush lines of Phaseolus vulgaris L. Proc. Amer. Soc. Hort. Sci. 88:384-392. 28.Dickson, M.H. 1967. Diallel analysis of seven economic characters in snap beans. Crop Sci. 7:121-124. 29.Donald, C.M. 1967. The breeding for crop ideotypes. Euphytica 17:385-403. 30.Donald, C.M. and J. Hamblin. 1976. The biological yield and harvest index of cereals as agronomic and plant breeding criteria. Adv. Agron. 28:361-405. 31.Duncan, WBG. 1971. Leaf angles, leaf area and canopy phothosynthesis. Crop Sci. 11:482-485. 32.Duncan, WKG., AJL. Hatfield and J.L. Ragland. 1965. The growth and yield of corn. II. Daily growth of corn kernels. Agron. J. 57:221-223. 33.Dunphy, ELJ., J.J..Hanway and D.E. Green. 1979. Soybean yields in relation to days between specific developmental stages. Agron. J. 71:917-920. 34.Eastin, J.A. and E.T. Gritton. 1969. Leaf area development, light interception and the growth of canning peas (Pisum sativum L.) in relation to plant population and spacing.Agron.;L 61:612-615. 35.Eeuwens, C.J. and W.W. Schwabe. 1975. Seed and pod development in Pisum sativum L. in relation to extracted and applied hormones. J. Exp. Bot. 26:1-14. 36.Egli, D.B. and J.E. Leggett. 1973. Dry matter accumulation patterns in determinate and indeterminate soybeans. Crop Sci. 13:220-222. 37.Egli, D.B. and J.E. Leggett. 1976. Rate of dry matter accumulation in soybean seed varying source sink ratios. Agron..L 68:371-374. 38.Egli, D.BHJLE. Leggett and.J.M. Wood. 1978. Influence os soybean seed size and position on the rate and duration of filling. Agron. J. 70:127-130. 39.Egli, D.Bn J. Fraser, J.E. Leggett and CLG. Poneleit. 1981. Control of seed growth in soybeans (Glygine max (L). 178 Merril). Ann. Bot. 48:171-176. 40.Egli, D.E., J.H. Orf and T.W. Pfeiffer. 1984. Genotypic variation for duration of seed fill in soybean. Crop Sci. 24:587-592. 41.Emecz, T.I. 1962. Suggested amendments in growth analysis and potentiality assessment in relation to light. Ann. Bot. 26:517- 527. 42.Fehr, W.R. and C.R. Caviness. 1977. Stages of soybean development. Iowa State Univ. Coop. Ext. Serv. Spec. Rep. 80. 43.Fehr, W.R., B.K. Lawrence and T.A. Thompson. 1981. Critical stages for soybean defoliation. Crop Sci. 21:259- 262. 44.Fisher, N.M. 1984. Crop growth and development: The vegatative phase, p.119-l6l. In: (Eds.). P.R. Goldsworthy and N.M. Fisher. The physiology of tropical field crops. John Williey & Sons. N.J. 45.Foolad, M.R. and A. Bassiri. 1983. Estimates of combining ability, reciprocal effects and heterosis for yield and yield components in a common bean diallel cross. J. Agric. Sci., Camb. 100:103-108. 46.Frank, S.J. and W.R. Fehr. 1981. Associations among pod dimentions and seed weight in soybean. Crop Sci. 21:547-550. 47.Fraser, J., D.B. Egli and J.E. Leggett. 1982. Pod and seed development in soybean cultivars with differences in seed size. Agron. J. 74:81-85. 48.Frazier, W.A., J.R. Baggett and W.A. Sistrink. 1958. Transfer of certain Blue Lake Pole bean pod characteristic to bush beans. Proc. Amer. Soc. Hort. Sci. 71:416-421. 49.Gay, S., D.B. Egli and D.A. Reicosky. 1980. Physiological aspects of yield improvement in soybeans. Agron. J. 72:387- 391. 50.Gbikpsi, P.j. and K. Crookston. 1981. Effect of flowering date on accumulation of dry matter and protein in soybean seeds. Crop Sci. 21:652-655. 51.Gebeyechou, CL,D.R. Knott and R.J. Baker. 1982. Relationships among durations od vegetative and grain filling phases, yield components and grain yield in durum wheat cultivars. Crop Sci. 22:287-290. 179 52.Ghaderi, A., and M.W. Adams. 1981. Preliminary studies on the inheritance of structural components of plant architecture in dry bean (Phaseolus vulgaris L.). Bean Improv. Coop. 24:35-38. 53.Goldsworthy, 1984. Crop growth and development: The reproductive phase, p.163-212. In: P.R. Goldsworthy and N.N. Fisher (Edts.). The physiology of tropic field crops. John Willey & Sons. N.J. 54.Griffing, B. 1956. Concept of general and specific combining hability in relation to diallel crossing systems. Aust. J. Biol. Sci. 9:463-493. 55. Gritton, E. T. 1975. Heterosis and combining ability in a diallel cross of peas. Crop Sci. 15: 453- 457. 56.Hallauer A.P. and W.A. Russel. 1969. Estimates of the maturity and its inheritance in maize. Crop Sci. 2:289-294. 57.Hanway, J.J. and W.A. Russel. 1969. Dry matter accumulation in corn (Zea mays L.) plants: comparisons among single cross hybrids. Agron. J. 61:947-951. 58.Hanway, J.J. and C.R. Weber. 1971. Dry matter accumulation in light soybean varieties. Agron. J. 63:227-230. 59.Hartung, R.C., J.E. Specht and J.H. Williams. 1981. Modification of soybean plant archytecture by genes for stem growth habit and maturity. Crop Sci. 21:51-56. 60.Hayman, B.I. 1954. The theory and analysis of diallel crosses. Genetics 39:789-809. 61.Hedley, C.L. and M.J. Ambrose. 1980. An analysis of seed development in Pisum sativum L. Ann. Bot. 46:89-105. 62.Hicks, D.R. J.W. Pendleton, R.L.Bernard and T.J. Johnston. 1969. Response of soybean plant type to planting patterns. Agr. J. :290-293. 63.Hillson M.T. and L.H. Penny. 1965. Dry matter accumulation and moisture loss during maturation of corn grain. Agron. J. 57:150-153. 64.Hsu, F.C. 1979. A development analysis of seed size in common bean. Crop Sci. 19:226-230. 65.Islieb, T.G., J.C. Wynne, C.H. Elkan and T.J. Schneeweis. 1980. Quantitative genetic aspects of nirogen fixation in 180 peanut. (Arachis hypogaea L.). Peanut Sci. 7:101-105. 66.Izquierdo J.A. and G.L. Hosfield. 1983. The relationship of seed filling to yield among dry beans with differing archytectural forms. J. Amer. Soc. Hort. Sci. 108:106-111. 67.Izquierdo, J.A. 1981. The effect of accumulation and remobilization of carbon assimilate and nitrogen on abscission, seed development, and yield of common bean (Phaseolus vulgaris L.) with differing architectural forms. PhD Thesis, Michigan State Univ., East Lansing. 68.Johnson, D.R. and J.W. Tanner. 1972. Calculation of rate and duration of the grain filling in corn (Zea mays L.) Crop Sci. 12:485-486. 69.Johnson, D.R. and D.J. Major. 1979. Harvest index of soybeans as affected by planting date and maturity rating. Agron. J. 71:538-541. 70.Jones, D.B., M.L. Peterson and S. Geng. 1979. Association between grain filling rate and duration and yield components in rice. Crop Sci. 19:641-644. 71.Kaplan, S.L. and H.R. Koller. 1974. Variation among soybean cultivars in seed growth rate during the linear phase of seed growth. Crop Sci. 14:613-614. 72.Kelly, J.D., G.V. Varner, M.W. Adams. 1986. The effect of different dry bean growth habit on yield stability. Bean Improv. Coop. 29:58-59. 73.Keuneman, E.A. and D.H. Wallace. 1979. Simplified growth analysis of non-climbing dry beans at three spacing in the tropics. Expl. Agric. 15:273-284. 74.Keumeman, E.A., G. Hernandez-Bravo and D.H. Wallace. 1978. Effects of growth habits and competition on yield of dry beans (Phaseolus vulgare ) in the tropics. Expl. Agric. 14:99-104. 75.Krarup, A. and D.W. Davis. 1970. Inheritance of seed yield and its components in a six-parent diallel cross in peas. J. Amer. Soc. Hort. Sci. 95:795-797. 76.Laing, D.R., P.G. Jones and J.H.C. Davis. 1984. Common bean (Phaseolus vulgaris L.) In: P.R. Goldsworsthy and N.M. Fisher. (Eds). The physiology of tropical field crops. John Willey and Sons. N.J. 85.P comb 15:3. 86.p. fill. and- 87.p HQri. SQQd (ggé 88.1% Chara 89.pc 181 77.Mahon, JuD. and SJLA. Hobbs. 1983. Variability in pod filling characteristics of peas (Pisum sativum L.) under field conditions. Can. J. Plant Scifm63?283-291. 78.Matzger, D.D., S.J. Czaplewski and D.C. Rasmusson. 1984. Grain filling duration and yield in spring barley. Crop Sci. 24:1101-1105. 79.Mc Blain, B.A. and D.J. Hume. 1980. Physiological studies of higher yield in new, early-maturiy soybean cultivar. Can. J.P1ant Sci.60:1315-l326. 80.Metzger, D.D., S.J. Czaplewski and D.C. Rasmusson. 1984. Grain filling duration and yield in spring barley. Crop Sci. 24:1101-1105. 81.Nass, H56. and B. Russer. 1975. Grain filling period and yield relationships in spring wheat. Can. J. Plant Sci. 55:673-678. 82.Neuhausen, S. 1982. Genetic variation for the duration of grain filling in a sample of the wheat germplasm pool. MS Thesis, Michigan State Univu, East Lansing. 82.Nienhuis, J. and S. Singh. 1986. Combining ability analysis and relationships among yield, yield components, and architectural traits in dry beans. Crop Sci. 26:21-27. 84.Pandey, S. and E.T. Gritton. 1975. Inheritance of protein and other agronomic traits in a diallel cross of pea. J. Amer. Soc. Hort. Sci. 100:87-100. 85.Paschal, E.H. and J.R. Wilcox. 1975. Heterosis and combining ability in exotic soybean germplasm. Crop Sci. 15:344-349. 86.Peaslee, D.E”.;LL. Ragland and qu. Duncan. 1971. Grain filling period of corn as influeced by phosphorus, potassium and the time of planting. Agron. J. 63:561-563. 87.Perenzin, hm” F. Ferrari and PL Motto. 1980. Heritabilities and relationships among grain-filling period, seed weight and quality in forty italian varieties of corn (Zea mays L.L Can.£L Plant Sci.60:1101-1107. 88.Poneleit, C.G. and D.B. Egli. 1983. Differences between reciprocal crosses of Inaize 1km: kernel growth characteristics. Crop Sci. 23:871-875. 89.Poneleit, CLG., D.B. Egli, P.L. Cornelius and D.A. 182 Reicosky. 1980. Variation and associations of kernel growth characteristics in maize populations. Crop Sci. 20:766-770. 90.Radford, P.J. 1967. Growth analysis formulae-their use and abuse. Crop Sci. 7:171-175. 91.Rasmusson, D.C. 1. Mc Lean and T.L. Tew. 1979. Vegetative and grain filling periods of growth in barley. Crop Sci. 19:5-9. 92.Reicosky, D.A., J.H. Orf and C.H. Poneleit. 1982. Soybean germplasm evaluation for length of the seed filling period. Crop Sci. 22:319-322. 93.Robitaille, H.A. 1978. Dry matter accumulation patterns in indeterminate Phaseolus vulgaris L. cultivars. Crop Sci. 18:740-743. 94.Row1ands, D.G. 1964. Genetic control of flowering in Pisum sativum L. Genetica 35:75-94. 95.Safari, A. 1978. A yield-component selection experiment involving american and iranian cultivars od the common bean. Crop Sci. 18:5-7. 96.Salado-Navarro, L.R., T.R. Sainclair and K. Himson. 1985. Comparisons among effective filling period, reproductive period duration, and R5 to R7 in determinate soybeans. Crop Sci. 85:1050-1054. 97.8chapaugh, Jr.W.T. and J.R. Wilcox. 1980. Relationships between harvest indices and other plant characteristics in soybeans. Crop Sci. 20:529-533. 98.Shibles, R.M. and C.R. Weber. 1965. Leaf area, solar radiation and dry matter production by soybean. Crop Sci. 5:75-577. 99.Singh, S.P. 1982. A key for identification of different growth habits of Phaseolus vulgaris L. Bean Improv. Coop. 25:92-95. 100.Singh, A.K. and 8.8. Saini. 1983. Heterosis and combining ability studies in french beans. SABRAO J. 15:17-22. 101.Singh, V.P., B. N. Dahiya and R.K. Chowdhry. 1977. Genetic studies of grain filling period in wheat. In A.K. Gupta (ed). Genetics and wheat improvement. II. Genetics of economic traits. Oxford and IBH Publishing Co., New Delhi. 183 102.Sinha, SJL, S.C. Bhargava and AnGoel. 1982. Energy as the basis of harvest index. J. Agric. Sci., Camb..99:237-238. 103.Smith, JuH. and R.L. Nelson. 1986. Selection for seed filling period in soybean. Crop Sci. 26:466-469. 104.Smith, J.H. and R.L. Nelson. 1986. Relationship between seed-filling period and yield among soybean breeding lines. Crop Sci. 26:469-472. 105.Snoad, B. and AnE. Arthur. 1973. Genetical studies of quantitative charactes in peas. I. A seven-parent diallel cross of cultivars. Euphytica 22:327-337. 106.Sprague, G.F and LnA.'Tatum. 1942. General vs specific combining ability in single crosses of corn. J.Amer. Soc. Agr.34:923-932. 107.Spaeth, S.C. and T.R. Sainclair. 1985. Linear increase in soybean harvest index during seed filling. Agron. J. 77:207- 211. 108.Spaeth, S.C., H.C. Randall, T.R. Sinclair and J.S. Vendeland. 1984. Stability of soybean harvest index. Agron. J.76:482-486. 109.Taylor, J. 1975. A modified air planter for use in experimental plots. Bean Improv. COOp. 110.Thorne, J.H. 1979. Assimilate redistribution from soybean pod walls during seed development. Agron. J. 71:812- 816. . lll.Tollenar, M. and T.B. Daynard. 1978. Kernel growth and development at two position on the ear of maize (Zea mays L.). Can. J. Plant. Sci. 58:189-197. 112.Wallace, D.H. and H.M. Munger.1966. Studies of the physiological basis for yield differences. 1. Growth analysis of six dry bean varieties. CrOp Sci. 5:343-348. 113.Wallace, DgH. and HgH. Munger. 1966. Studies of the physiological basis for yield differences. II. Variations in dry matter distribution among aerial organs for several dry bean varieties. Crop Sci. 6:503-507. 114.Wallace, ILH., J.L. Ozbun and H.M. Munger. 1972. Physiological genetics of crop yield. Adv. Agron. 24:97-146. 115.Walton, P.D. 1971. The genetics of yield in spring wheat 184 (Triticum aestivum L.). Can. J. Genet. Cytol. 13:110-114. 116.Watson, D.J. 1952. The physiological basis of variation in yield. Adv. Agron. 4:101-145. 117.Watts, L.E., E. Stevenson and M.J. Crampton. 1970. Inheritance of flowering time in six pea cultivars (Pisum sativum In). Euphytica 19:405-410. 118.Weber, C.R., L.T. Empig and J.C. Thorne. 1970. Heterotic performance and combining ability of two way F1 soybean hybrid. Crop Sci. 10:159-160. 119.Wi1cox, J.R. 1980. Comparative performance of semideterminate and indeterminate soybean lines. Crop Sci. 20:277-280. 120.Wilcox, J.R. and T. Sedimaya. 1981. Interrelationships among height, lodging and yield in determinate and indeterminate soybeans. Euphytica 30:323-326. 121.Wilson, D. 1981. Breeding for morphological and physiological traits, p.233-290. InleJL Frey (ed.L Plant Breeding II. The Iowa State Univ. Press. Ames, Iowa. 122.Wych, RJL, R.L. McGraw and D.D. Stuthman. 1982. Genotype x year interaction for legth and rate of grain fillinginoats.CropSci.22:1025-1028.