._-

1 Y1 32"; —-. L. 1. ' — -— 1 , . _ 1. ,V._._-_H . .'V'VTW-va A»,
:1 I I. . "' .1 7- ‘ - ' . . '2 '_ 1" 1“,. .l." (Wooqylnu'fi- Al..,,,."- 1
.I I' ,1, , 1 'KI' j, , '  ’ ' 351;, ,' '3‘,“ ,I'! '§,.'I' I... “\I‘ :II‘I‘f-I .°_1 1

‘ - ‘ u r”) , '

 

7'
I211" I-;.:_~, ., 1
.1 ,|,'1:I , 1 , 1. [I II II, ,,1",1"1II} I.“ "1'., ,1I'I' ~ uék' '1'.
1 1111.1: ;"'I 1"" "- I 1111 111111111 ~"1-"I- '15-!!! 1‘11"? “11‘- 1‘1'.
11‘ In 1,.1‘} ,1 ., ' 1f .1 1 ..;', 11 , ,.1
:1;I 1,113 1‘1 1 ‘ "“1 j’ . ,1 '2I'i;'1'1‘1!‘I;I."&I1-a"cI-‘
1'1'I ILA “11:!" 1', " I . "1 .1 H, , bf" (1:;00'm,z;{$7|1':&‘," HIIE1‘: '35
"W ‘ " ‘ 1' ' ' "I' 'I ' 1 "' ' I 'l :I' , I I." I” :"'I' ‘6 l
h$1Mw1w1w'~1mWw%MIfigyfit
,1 ”’J'I , 1'I'1'1'I1‘""1 1. {1'1 11,1 .' .1 ‘11‘ ' In,1,.~.'.'I.',H,',"Y,, ".I
1'1, .‘1‘1'11I,1'!1‘1'[II.', ,1 , ‘_ 1, J .1" III §""'1I1’." ' '.I,-',',),,11.1.,'- {AII',W
1,1,_1r,,,n .1". , 1:11“, 1'. , y 1, 217.", ,11'.I1I;"'.- an,»
i “11,2114,“ ,II11.I, 1f, , ,,~ , I1 I,'. , 11 .11' 11, :I’ ,I (19' I1 '1 1"! 11-1I1'31I:.,I1
1' I"I|' ‘ .1 ’.. 1 .I ‘ I. 1' "" '.
11..' I.I.'I1'-.I'|‘.' 1" III II' ' "" . ' '1' 1I1"" III!" -'I' :1"-"I'.dp wiI
III '- 1 1' I 1I 1 " "I I." 1 ' ' 1.11 I ' ' 'I ~'I"~' ' III'I
r | III .1 ‘I ,' .'IIII1'1' . | ". 'I" I" "'1' . :Il',: '!'| 'l |" 1 '9'“.
1": ' I, . . I"' ':I, 1 11.2.1.1 11 I 1"' II ' I" '1']:
,- 1, 1 ' . 'I,l, 11'!.,1;;1,. I.“ . '1wa.5" "'"'I"" "I"; I ,,!!-._1 1' 11.111 III.
1l 1 ' ' '! ."' 1' Iu,-1' ' " ' I"'... 1 I,' *" I " "1"Il" ‘11"".I "fl
. I' I ' "I'II "II:' ‘I '1'1""I 1'" -"' '1 ' ' ' ' '1 'N'<'IL- II I 2"”;
. 1,. ‘ I I' III II I“, II, ‘ 1 ,,,' 1.1I ‘ “H ',I 1“: .u 1" 1,13,?" . 3"
" III" . II1"':" ' I, I 1' IL" " '9'” II " "" l-"'l' "' |
11 'QII' 1,! 1 1 .‘II"III1II, 1| , I‘.’ 1111' ',' 'n'III' , I"‘ "1' F121»
' " I111 1"III11. 1 1- 111:1 -. 1 1‘ 1 1~ .11 1. I .1
" I [11: 1 'I1 15:11! 1:11 11"1'1.11I '.u.:.'."1’!'! ,1 '1, '! ‘13} 1111'
' 'II' ‘ 1') I '11I'1'1'1I“ 12,,I I"‘II'” .' II 11 ' 11' “Wk . ‘ 1' '-
I "3:11: 1 "W131: 1:. .r' 1.1 '1 1" III 11I‘11. ,1 1,1",1111II"11' '--1‘11I1
' ' " ."Iu .'I " I I"‘”. '1 .1' -".' W (WW1
' I I II' "'1I" 1 I
‘ I ‘ ' I I. 1* I. I II I 1
1, ! , ‘

  
 
  
  
  
   

  

TEL". 1 o -
.‘ ., ,1. ug- -'
1 ‘ . I. 1, k0, , ,
11 11:1? , ., 111'?” 1...~1,£h . 1
I,:I."':; 9- fl 7“. -
I, 1" j::l.' 5:; ' , ' "t '
II“ “Iii, .- ”Ski“ !('\‘|'l1k*',4 "'::-'?"-h ,I

W
i" _ a
‘ 1' 51"“, . 1::‘f'1OIII. W;:‘fi

. .‘! [WI ' "4' erII'I' ‘ git“: .'1' Yflfp “II? 2:)” ~11”: :’?:,.€
'I 1' " 11;,11 I.‘ . I 213"“:“1'1I': \. é?" ‘ {-511
.1, ;1.13111111111151111‘3’5‘ 1 ‘1 ””3113"?

'k' ”WNJfikvgt §1%
. W... “I, (111‘, 1,111,911 1v

   

   

      

I' ,II" I‘ 'II' HI;"# I" 3' '-
11,,“ In 1.1. I It I III R 121*, fig-
‘1' IV 1:11,]: §¥§,:,I.J,fi ,I ,l i: “1‘93 . .- ," 53‘4” ,, 3%.?an

  
  
  
 

 
 

: 'i'iT1f 1,1,". I11- 1_ ,
I'. 1 ‘1!"

‘73:}?

 
 

\' A's ~r.-
1125,“;

III aw:

       

     

Ihk EV.

'I'II "ii: {'93,
I.‘ I I I
1‘ I 0‘
”Murmuw
I'.I..““1 1- ’1' ' I -~ 1

11...,1 111.11 21:11.11

 
     
 
    
 
      

an“ 'I' ”"' :'"'I:. “III

 
 
 
 

    
    
  
 
 
  

 
   

   

  

;, ’ 1'1... .1 .. 11'1“
,1,‘,'”,I,.I ,, .‘ 1‘ ' MI I'I"II',1 III- .,I
1 1 .111 11.111111111111111.

I1 I
1" 1 ' ‘ "I'4 1 41’
"U“ 'f' "‘ ' 'J'JI" 1I. "I. ‘I In" 'IVJIII 1"." ‘1" III‘II’ IIIIL' '1\' I

 

”55‘s in. , .xa "an“; 3
9-1". 2___ (1.....2...
.._, :1

‘u .~ .-..’ :; 1' I. q, «22 g-g‘,’

o 9
. 3".‘1-‘A?f-£1
Lang 6 if! 1335,

 

 

 

This is to certify that the

dissertation entitled
Genetic Analysis of Cooking time, Nutritional, and Culinary

Quality in Dry Beans (Phaseolus vulgaris L. )

presented by

Nassratullah Naimatullah Nassimi

has been accepted towards fulfillment
of the requirements for

Ph .0. degree in Pl ant breeding

 

/ Major prote’ssor/

 

A 11 3, 1985
Date pr

 

MSU is an Affirmative Action/Equal Opportunity Institution 0-12771

 

 

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.

 

 

 

 

GENETIC ANALYSES OF COOKING TIME, NUTRITIONAL,
AND CULINARY QUALITY IN DRY BEANS
(Phaseolus vulgaris L.)

BY

NASSRATULLAH NAIMATULLAH WASSIMI

A DISSERTATION

SUBMITTED TO
MICHIGAN STATE UNIVERSITY
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY

DEPARTMENT OF CROP AND SOIL SCIENCES

1985

L)

ABSTRACT
GENETIC ANALYSES OF COOKING TIME, NUTRITIONAL,
AND CULINARY QUALITY IN DRY BEANS
(Phaseolus vulgaris L.) -

BY

Nassratullah N. Wassimi

The importance of dry beans as a protein complement to
cereal diets has long been recognized; however, tmm
presence of antinutritional factors and prolonged soaking
and cooking have caused people in lesser developed countries
of the world to turn away from eating beans. In order to
increase the consumption of beans it is necessary to develop
high yielding cultivars that are resistant to pests and have
good food quality. It will be difficult to incorporate
selection for nutritional and culinary quality into bean
breeding programs that historically selected for increased
and stabilized yield. Modern technology'has provided the
means by which nutritional and culinary quality study may be
conducted on small amounts of seed. The present study was
undertaken to determine the inheritance of cooking time and
uniformity, tannin and protein content, and the culinary
quality of a diverse population of beans. Eight strains
were crossed in diallel and the eight parents and F2 and F3
progenies grown at two locations for evaluation. Highly

significant differences were observed among entries for

cooking time, uniformity of cooking, tannin and protein

C)
C)
:3
('1
(I)

'II
l1)
'1
('f
H
1"

pattern

Nassratullah N. Wassimi

content and for eight of nine culinary quality traits.
Partitioning of variability into GCA, SCA, and reciprocal
causes revealed highly significant GCA mean squares for the
traits. In some cases SCA variance was also significant;
however, when significant, SCA variance components were
always smaller in magnitude than GCA components. Reciprocal
differences were detected for a few traits but no consistent

pattern over location or generation was noted.

Quick cooking characteristics of parental strains were
transmitted to progenies. Crosses of low x low and high x
high protein parents had progenies that were also low and
high, in protein content, respectively. Highly significant
correlation between GCA effects in the F2 and F3 and between
parental values and the GCA effects of parents was observed.
Selection aimed at improving the cooking time and cooking
uniformity, soakability, and palatability of beans can be
practiced in generations when plants are more heterozygous
after the initial cross. Selection for low tannin and high
protein among progeny from a cross should result in the

stabilization of these traits in a single cultivar.

Vt; \-:‘

Note: This dissertation is presented as a series of three
papers written in the style and format required by
Crop Science and, the Journal of the American
Society for Horticultural Science.

—.l,
d'

IOm

 

To my father, and all other members of my family.

.5

53
mini

 

,
J.
0..

FPO?
ms, 21
Mar=cx

guidance a
o

mat is s:
that

~11. '5

2
Q

ACKNOWLEDGMENT

As a major professor, Dr. George L. Hosfield provided
many challenges and unlimited stimuli to my professional
growth and understanding. Above all, he provided friendship
that is sincerely appreciated. His strong professional
guidance and criticism have equalled by support and
encouragement in every aspect of research and academic
advancement. It is with humbleness and sincere gratitude
that I say thank you to Dr. G. L. Hosfield for the
opportunity to have been his student.

Appreciation and gratitude is extended to Dr. M. v.
Adams, Dr. J. D. Kelly, Dr. M. A. Uebersax, and Dr. P.
Markakis members of my guidance committee for their
constructive criticism when reviewing this manuscript.

Special thanks to Dr. T.(L Isleib, Dr.(L Cress, and
Dr.1L Ghaderi for their help in the statistical analysis
and its interpretation.

I am also grateful to Dr. F. A. Bliss, Mr. L. Hudson,
and Mr. L. Telek for their help and collaboration when using
their laboratory facilities. Gratitude is expressed to Mr.
Jerry Taylor, Mr. Jeff Redoutey, Ms. Mary Saam, and Ms.
Sallie Nellso for their field work assistance and sample
preparation.

The friendship, encouragement and support of Catalina
Samper, Joe Tohme, Khushal and Homyra Habibi, and all

members of the "bean program" is sincerely appreciated.

  
   
  
 
 

A st;
szudent a;
‘mcoaplair.
‘mrmal‘ i:

c.'

 

A study program such as this is not borne by the
student alone. Family members offer support and tolerate
uncomplainingly many inconveniences and disruptions to a
'normal' family life. Waranga, Wagma, and Atal are such a
family and have greately contributed to the successful

completion of this study program.

~I‘ER l

t

:mocucr

PITA
“‘91

3'
d

A

TABLE OF CONTENTS
PAGE

LIST OF TABLES O O O O O O 0 O O O O O O O O O O O O 0 O O O O O O O O O O O O O O O Vii
LIST OF FIGURES O O O O O O O O O O O O O O O O O O O O O O O O O O O O O I O O O O Xiii
INTRODUCTION 0 O O O O O O O O O O O O O O O O O O O I O O O O O O O O O O O O O O O 0 1

CHAPTER 1. GENETIC CONTROL OF COOKING TIME
' AND UNIFORMITY OF DRY EDIBLE BEANS
(Phaseolus vulgaris L.)

Abstract ............................. 5
Introduction ......................... 6
Materials and Methods ................ 8
Results .............................. 13
Discussion ........................... 27
References ........................... 29

CHAPTER 2. GENETIC CONTROL OF TANNIN CONTENT
AND PERCENTAGE PROTEIN OF DRY AND COOKED
DRY BEANS (Phaseolus vulgaris L.)

Abstract ............................. 30
Introduction ......................... 32
Literature Review .................... 34
Materials and Methods ................ 40
Results .....:........................ 45
Discussion ........................... 76
References ........................... 79

CHAPTER 3. GENETIC CONTROL OF PHYSICO-CHEMICAL
CHARACTERISTICS RELATED TO CULINARY
QUALITY OF DRY EDIBLE BEANS
(Phaseolus vulgaris L.)

Abstract ............................. 85
Introduction ......................... 87
Literature Review .................... 90
Materials and Methods ................ 94
Results .............................. 101
Discussion ........................... 148
References ........................... 154

INTERPRETIVE SUMMARY OOOOOOOOOOOOOOOOOOO0.0...... 158

APPENDICES
Appendix A O O O O I O O O O O O O O O O O O O O O O O O O O O O 163
Appendix B O O I O O O O O O O I O O O O O O I O O O O I O O O O 167

BIBLIOGRAPHY 0.0.0.0....0.0.0.000...OOOOOOOOOOOOCO 175

vi

CHAPIE’ l

l.

\M
O

 

 

LIST OF TABLES

CHAPTER 1 Page

1. Characteristics of parents used in a diallel

crOSSOOOOOOCOOO0.0.0.0...OOOOOOOOOOOOOOOOOOO 17

2. Mean squares of analyses of variance for cooking
time and percent hard seed in the F2 and F3
generationSOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO 18

3. Mean squares form analyses of variance of
(Wr-V ) for cooking time and % hard seed to
test the adequacy of additive-dominance model.. 19

4. Analyses of variance and estimates of variance
components and their standard error for general
and specific combining ability and reciprocal
effects for cooking time and percent hard seed

measured on F2 and F3 progeny Of an 8X8
diallel cross grown at two locations in 1982.. 20

5. Estimates of general combining ability effects
for cooking time and S hard seed measured on
the F2 and F3 generation mean of an 8-parent
diallel cross............................... 21

6. Mean cooking time (min.) of parents and crosses

involving that parent in the F2 and F
generation grown at two locations in I982... 22
CHAPTER 2

1. Seed coat color and tannin content and percentage
of seed protein of 8 dry beans strains used as
parents used in diallel cross............... 51

2. Analysis of variance for tannin content of F
seed coats in dry bean parents and progeny c

an 8x8 diallel (31.038.000.000...ooooooooooooo 52

3a Analyses of variance and estimates of
variance component and their standard
error for general and specific combining
ability and reciprocal effects for tannin
content of seed coats of parents and F

Egogeny of an 8x8 diallel cross grown at East
nsing and Saginaw, Michigan in 1982....... 55

vii

()\

—A

C)
.

Loci

am

an;
th
tar

.
q .-

H‘,‘

Mi:

Ana
prc
an

V

Cor
rel
con

(1::
10C

CO;
per
Pro
8P0

1n

COm
0f
and
of
gen.

N

61‘0‘
in

CHAPTER 2 (continued)

4.

10.

11.

General combining ability effects for tannin
content of seed coats of each parent used in
in a diallel cross of dry beans and grown

at East Lansing and Saginaw, Michigan in 1982.

Mean square from analyses of variances and
combining ability analyses for tannin content
measured on parents and F2 rogen from

an 8-parent diallel cross 0% dry beans and
grown at E. Lansing and Saginaw, Michigan

in1982.00.00.00...0.00000000000000IOOOOOOOO

Mean percent tannin content of seed coats

of arents and F2 progeny Of an 8-parent

dia lel cross in dry beans grown at E. Lansing
and Saginaw, Michigan in 1982...............

Analyses of variances of (“r-Vr) for testing
the adequacy of additive-dominance model for
tannin content of F2 pro any of an 8x8
diallel cross grown at: . Lansing and Saginaw,
Michigan, in 1982...........................

Analyses of variances for raw and cooked bean

protein of parents and F2 and F3 rogeny of
an 8x8 diallel cross grown at E. Eansing and

Saginaw, Michigan in 1982....................

Correlation coefficient indicating the
relationship between F and F3 for protein
content of raw and cooRed beans grown at
East Lansing (upper triangle) and Saginaw
(lower triangle), Michigan and between
locations (diagonal). .....................

Analyses of variance for general, specific
combining ability and reciprocal effects for
percent protein of parents and F2 and F3
progeny of an 8x8 diallel cross of dry eans
grown at East Lansing and Saginaw, Michigan

in19820.00...0.000.000.0000.000.......00...

Component of variance and standard deviation
of general and specific combining ability
and reciprocal effects for percent protein
of raw and cooked bean seed of F and F
generation mean of an B-parent d allel cross
grown at E. Lansing and Saginaw, Michigan

in 1982.00.00...0.00.0.0...OOOOOOOOCOOOOOOOO.

viii

Page

54

55

56

57

58

59

60

61

n-"‘

-...b
\ N

14.

 

 

CHAPTER 2 (Continued)

12.

13.

14.

150

16.

17.

Estimates of general combining ability
effects for percent protein content of raw (R)
and cooked (C) bean seeds measured on F2

and F3 progeny of an 8x8 diallel cross and
grown at E. Lansing and Saginaw, Michigan in

19820....OOOOOOOOOOOOOOOOOOOIOC0.00.0.0...O.

Mean squares for locations, entries, generations
and general, and specific combining ability,
reciprocal, maternal and non-maternal reciprocal
effects and their interaction with location
for protein content of raw and cooked bean
measured on combined data of parents and F2

and F progeny of an 8x8 diallel cross grown at
East Eansing and Saginaw, Michigan in 1982...

Estimates of variance components of GCA, SCA,
and reciprocal effects and their standard
deviations for protein and tannin content
measured on parents and progeny of an 8-parent
diallel cross grown at E. Lansing and Saginaw,
MiChiganin1982.00.00.00000000COO-OOOOOOOOOOO

General combining ability effects and their
interactions with locations for protein and
tannin content for each parent of an 8x8
diallel cross grown at East Lansing and
Saginaw, Michigan in 1982....«..............

Parental values for seed protein and tannin

content of raw beans (Yr) and variance SPECific
combining ability ($2) of each parent from

data measured on the F2 and F3 progeny
an 8x8 diallel cross averaged over locations

in1982.00.00.000000000000000000000000000000

Analysis of variance of ("r-Vr) for raw
and cooked bean seed protein in the F2 and

generation of an 8-parent diallel cross
to test the adequacy of additive-dominance

ix

Page

62

65

64

65

66

CHAPTER 2 (continued)

18.

19.

Mean protein content of raw and cooked bean
seeds of parents and F2 and F3 progeny

of an 8-parent diallel cross grown at East
Lansing, Michigan in 1982...................

Mean protein content of raw and cooked bean
seeds of parents and F2 and F progeny

of an 8-parent diallel cross grown at Saginaw,
M1Chigan in 19820000000000.0.000000000000000

CHAPTER 3

1.

Mean squares from analysis of variance and
combining ability analysis and their
interaction with location for 9 culinary
quality traits measured on parents and

F? and F grogeny of an 8-parent diallel
0 dry edi le beans grown in 1982...........

General combining ability effects (81)
average over locations and their interaction
with location (81x e) for 9 culinary

quality traits measured on F2 progeny Of
an 8-parent diallel cross in dry edible beans
grownin1982.0.00.00.00.00.00000000IOOOOOO.

General combining ability effectS'(gi)
average over locations and their interaction
with location (81x e) for 9 culinary

quality traits measured on F3PP°890Y Of
an 8-parent diallel cross in dry edible beans

grownin1982.00.00.00000000000000000000000C

Estimates of variance components and their
interaction with location for 9 culinary
quality traits measured on parents and F2
and p progeny of an 8-parent diallel cross
in dry edible beans grown in 1982...........

Parenta1 values (yr), for texture and washed
drained weight and variance of specific

combining ability effects (32) of a dry
bean genotypes evaluated with their F2 and

F progeny from an 8-parent diallel cross
ggown at East Lansing and Saginaw in 1982..

Correlation coefficient (r) for general
combining ability effects between F2 and

F progeny of an 8x8 diallel cross and
pgrental values vs. GCA effects in the two
generations of dry edible beans grown at

E. Lansing and Saginaw, Michigan in 1982....

X

Page

68

69

114

116

118

120

121

122

(I)

11.

12.

f '1 C')
0 Fr (0 ()
5—1"

.4",
I3 2 m

Appendix

Pr:
ati

Appendix

1.

Meg
0f

dr:
get
81‘:

Me;
of

d?!

gei
81‘:

5.31
68‘
6f:
01’

V‘!‘

CHAPTER 3 (continued)

7.

10.

11.

12.

Correlation coefficient (r) indicating the
relationship between pairs of culinary quality
traits Of 55 F2 (upper triangle) and F

(lower triangle) progeny of an 8x8 diaIlel
cross grown at E. Lansing, Michigan in 1982.

Correlation coefficient (r) indicating the
relationship between pairs of culinary quality
traits of 56 F2 (upper triangle) and F

(lower triangle) progeny of an 8x8 diaIlel
cross grown at Saginaw, Michigan in 1982....

Mean of texture and processing traits for two
generations of an 8-parent diallel cross grown
at East Lansing, Michigan in 1982...........

Means for soaking and mass ratio traits for

F2 and F3 progeny of an 8-parent diallel
cross grown a East Lansing, Michigan in 1982.

Means of texture and processing traits for two
generations of an 8-parent diallel cross grown
at Saginaw, Michigan in 1982................

Means for soaking and mass ratio traits for F2
and F3 progeny of an 8-parent diallel cross
grown a Saginaw, Michigan in 1982..........

Appendix A

Procedures of tannin extraction and determin-

ation COOOOOOOOOOOOOOOOOOOOOOOO0.0.0.0....O.

Appendix B

1.

Mean squares from combining ability analyses
of variance of 9 culinary quality traits of
dry beans measured on the F2 and F

generation means of an 8-parent di llel cross
grown at E. Lansing, MI, 1982...............

Mean squares from combining ability analyses
of variance of 9 culinary quality traits of
dry beans measured on the F2 and F

generation means of an 8-parent diallel cross
grown at Saginaw, MI, 1982..................

Estimates of variance components and standard
deviations of GCA, SCA, reciprocal and maternal
effects measured on 9 culinary quality traits
of the F2 and F generation mean of an

8-parent dialleI cross in dry beans grown at

E. Lansing, MI, 1982........................

xi

Page

123

124

125

126

128

129

163

167

168

169

lganf‘T
Afrut 9‘

c‘e‘
e1":
cf

5-;

See

5. Est
for

F
H

3.18

E.

loc.

 

APPENDIX B (Continued)

4.

Estimates of variance components and standard
deviations of GCA, SCA, reciprocal and maternal
effects measured on 9 culinary quality traits
0f the F2 and F generation mean of an
8-parent dialleI cross in dry beans grown at
Saginaw, MI, 1982...........................

Estimates of general combining ability effects

for 9 culinary quality traits measured on the

g; and F generation means of an 8-parent
allel gross in dry edible beans grown at

E. Lansing, MI, 1982........................

Estimates of general combining ability effects
for 9 culinary quality traits measured on the

F2 and F2 generation means of an 8-parent
d allel ross in dry edible beans grown at

&ginaW’ MI,1982.00.00.00000000000000000000

Analyses of variance of (wr-vr) for culinary

quality traits to test the adequacy of additive-

dominance model.............................

Correlation coefficients indicating relationships

between F2 and p generation means for
culinary quality traits measured on 56 crosses
of an 8-parent diallel cross grown in two
locations in 1982..00.0000000000000000000000

xii

Page

170

171

172

173

174

Val
ti.
La
re.
reg
unc

'v'ai
tin
Ric
re;
reg
unc

Var
see

snc
pa:
rel
pci

Poi

LIST OF FIGURES

CHAPTER 1

1.

Variance (vp)-covariance (Np) graph for COOKinS

time data 0f the F2 generation grown at East
Lansing and showing the position of points
representing the 7 parental arrays and their

regression line relative to a limiting parabola
un er WhiCh all pOintS MUSt 1180 00000000000

Variance (V )-covariance (Wr) graph for 000K138
time data 0% the F3 eneration grown at Puerto
Rim: and showing The position of points
representing the 8 parental arrays and their
regression line relative to a limiting parabola
under which all points must lie.............

Variance (Vr)-covariance (Wr) graph for 5 hard
339d data 04‘: the F2 grown at E. Lansing and
showing the position of points representing the 8
parental arrays and their regression line
relative to a limiting parabola under which all
points must lie.............................

Variance (Vr)-covariance (Np) graph for 1 hard
3399 data 0f the F grown at Puerto Rico and
showing the position of points representing the 8
parental arrays and their regression line
relative to a limiting parabola under which all
paints must 11e000000000000.000000000000000.

CHAPTER 2

1.

Variance (Vr)-covariance (Hp) graph for % tannin
content data of the F2 seed coat grown at East
Lansing and showing the position of points

representing the 8 parental arrays and their
regression line relative to a limiting parabola
under which all points must lie. ...........

Variance (Vr)-covariance (Np) graph for S tannin
content data of the F 869d coat grown at Saginaw
and showing the position of points representing
the 8 parental arrays and their regression line
relative to a limiting parabola under which all
points must lie.............................

Variance (Vr)-covariance (Hr) graph for protein

content data of the F 81“”!m at E. Lansing
and showing the posit on of points representing

the 8 parental arrays and their regression line
relative to a limiting parabola under which all
paints must lieoo00.00000000000IOOOOOOOOO...

xiii

Page

23

24

25

26

7O

71

72

fiv' {1 HT:
“lure-'3

'0'} ('10) 0“.

V1
Cl
a:
ti
{‘6

pc

Va
cc
ar.
tr
re

pc

CHAPTER

1.

CHAPTER 2 (continued) Page

4.

Variance (Vr)-covariance (Np) graph f0? protein
content data of the F grown at E. Lansing

and showing the position of points representing
the 8 parental arrays and their regression line
relative to a limiting parabola under which all
POintsmust118............................. 73

Variance (Vr)-covariance (Wr) graph for protein
Content data or the Ff grown at Saginaw
on

and showing the posit of points representing
the 8 parental arrays and their regression line
relative to a limiting parabola under which all
points must lie............................. 74

Variance (Vp)-covariance (Wr) graph for protein
content data of the F? grown at Saginaw
on

and showing the posit of points representing
the 8 parental arrays and their regression line
relative to a limiting parabola under which all
points must lie............................. 75

CHAPTER 3

. 1.

Variance (Vr)-covariance (Hr) graph for clump
data of the F generation grown at E. Lansing

and showing t e position of points representing
the 7- parental arrays and their regression line
relative to a limiting parabola under which all
points must lie. ........................... 131

Variance (Vr)-covariance (We) graph for clump
data of the F generation grown at E. Lansing

and showing t e position of points representing

the 6 parental arrays and their regression line
relative to a limiting parabola under which all
points must lie............................. 132

Variance (Vr)-covariance (Np) gra h for clump
data of the F2 generation grown a Saginaw and

showing the position of points representing

the 8 parental arrays and their regression line
relative to a limiting parabola under which all
points must lie............................. 133

Variance (Vr)-covariance (wr) graph for clump

data of the F generation grown at Saginaw and
showing the position of points representing the

6 parental arrays and their regressionline

relative to a limiting parabola under which all
points must lie............................. 134

xiv

CHAP’I

5.

10.

11.

ER}

Var
as

and
the
rel‘
poi

Var
dat
and
tn

rel
po;

Var
cat
sh:
8 1

F9.

. ~n'l

CUflWQ<

Cryonine:

CHAPTER 3 (continued) Page

5.

10.

11.

Variance (Vp)-covariance (WP) graph for splits
data or the Fag generation rown at E. Lansing
and showing t e position 0 points representing

the 8 parental arrays and their regression line
relative to a limiting parabola under which all
pOints must lie.IO..0OOOOOOOOOOOOO’OOOOOOOOOO 135

Variance (Vr)-covariance (WP) graph for splits

data 01‘: the Fag generation grown at E. Lansing

and showing t e position 0 points represent ng

the 8 parental arrays-and their regression line
relative to a limiting parabola under which all
points must lie 136

Variance (Vp)-covariance (Np) graph for splits
data 0f the F2 eneration grown at Saginaw and
showing the pos tion of points representing the

8 parental arrays and their regression line
relative to a limiting parabola under which all
paints must lie. OOOOOIOOOOOOOOOOOOOOO0.00.. 137

Variance (Vr%- covariance (Np) graph for splits
data 0f the generation grown at Saginaw and
showing the posfi tion of points representing the

8 parental arrays and their regression line
relative to a limiting parabola under which all
p°1nts muSt lie. 00.0.0.0....OOOOOOOOOOOOOOO 158

Variance w(Pd-covariance (w ) graph for washed
drained ght data or the £2 generation pawn at

E. Lansing and showing the2 pgosition 0 points
representing the 8 parental arrays and their
regression line relative to a limiting parabola
under which all points must lieuunnu.u.. 139

Variance (Vr)-covariance (Hp) graph for washed
drained weight data of the F3 generation grown

at E. Lansing and showing the position of points
representing the 8 parental arrays and their
regression line relative to a limiting parabola
under which all points must lie..n.u.u.u 140

Variance (V )-covariance (Hr) graph for washed
drained wefght data of the F2 eneration grown
at Saginaw and showing the posTtion of points
representing the 8 parental arrays and their
regression line relative to a limiting parabola

under which all points must lie............. 141

XV

CHAPTER 5

12. Var

-_'

13. Var

14. Va:

n4
re‘.i

1:. Var

dai
an:

 

16.

17.

 

CHAPTER 3 (continued)

12.

13.

14.

15.

16.

17.

Variance (Vr)-covariance (Wr) graph for washed

drained weight data of the F3 generation grown

at Saginaw and showing the position of points
representingthe8 parental arraysandtheir
regression line relative to a limiting parabola
under which all points must lie.".u.u.u.

Variance (Vp)-covariance (Np) gra h for texture
data of the FR generation grown a E. Lanaing
e

and showing t position of points representing
the8parentalarraysand their regression line
relative to a limiting parabola under which all
points must lie. ...

Variance (Vr)-covariance (Hr) gra h for texture
data of the EH generation grown a E. Lansing
e

and showing t position of points representing
the 6parental arrays and their regression line
relative to a limiting parabola under which all
points must lie. .

Variance (Vp)-covariance (Hr) gra h for texture
data of the F generation grown a E. Lansing

and showing tge position of points representing
the 8 parental arrays and their regression line
relative to a limiting parabola under which all

DOints must lie. ......OOOIOOOOO0.0.0.0...

Variance (Vr)-covariance (WP) graph for texture
data of the F6 generation grown at Saginaw
e

and showing t position of points representing
the 8 parental arrays and their regression line
relative to a limiting parabola under which all
points must lie.

Variance (Vr)-covariance (WP) gra h for texture
data of the F generation grown a Saginaw
and showing tge position of points representing
the 8 parental arrays and their regression line
relative to a limiting parabola under which all
points must lie. .........................

xvi

Page

142

143

144

145

146

147

alleVia‘
one of ti
and can

to consur

851

CbJeCtive

 

INTRODUCTION

The high cost and general non-availability of animal
protein in diets of people in poorer countries of the world
necessitates seeking of alternative sources of protein to
alleviate protein malnutrition. Edible legumes constitute
one of the most important sources of plant protein the world
and can go a long way in supplying this and other nutrients
to consumers in lesser developed countries of the world.

Usually an increased and stabilized yield is the major
objective of most legume breeding programs. However, since
the primary use of edible legumes is a human food,
nutritional, culinary and sensory qualities that are of
direct concern to the consumer must be at an acceptable
level (Hawtin, et al. 1977).

Dry beans (Phaseolus vulgaris L.) constitute 32% of
the total world legume production (FAO 1982), and can
provide significant amounts of protein, calories, minerals
and vitamins to human diets. Despite the nutritional
benefits, bean consumption on a world-wide basis is
decreasing. The causes of the reduced consumption of beans,
beside their low yields and high prices, are prolonged
soaking and cooking times, tannins, and heat labile anti-
nutritional factors.

Low sugar, fat, and high dietary fiber content of
beans will not predispose consumers to diabetes and coronary

heart disease (Walker, 1982). Leeds (1982) reported that

:iets C

advanta

may be

CF88 C93

content

 

 

diets containing large quantity of beans may have special
advantages for diabetic patients, as carbohydrate from beans
may be digested and absorbed more slowly from beans than
from other foods. Simpson et al. (1981) reported that many
indices of diabetic control were improved, but in particular
blood glucose levels after meals were lower in patients
treated with high-carbohydrate high-leguminous dietary fiber
content. Bressani et al. (1963) reported that long ( >30,
min. at 16 lb pressure and at 49° C ) cooking times
decreased the nutritive value of bean protein but the
addition of 0.2% methionine to the bean diet of rats
significantly improved weight gain, protein efficiency and
biological value. Bressani and Elias (1977) observed greater
intraspecies and interspecies variability for heat labile
anti-physiological factors responsible for low protein
digestibility in beans.

Tannins are phenolic compounds that have the
characteristic ability to precipitate protein (Gustavson,
1956). Their molecular weights usually range from 500 to
3,000MW. Tannins are generally classified as either
hydrolyzable <nr condensed (non-hydrolyzable) tannins based
on their structural type and their reactivity toward
hydrolytic agents, particularly acids (Haslam, 1966).

The preparation of beans for consumption usually
involves soaking in water to raise moisture content ranging
between 53-57%, followed by cooking. The soaking and cooking

characteristics of dry beans are important in any effort to

1T;FO 'v"
unifor
facili
toxic

proteas

reporte

 

improve nutrient delivery through increased utilization.
uniform hydration and appropriate cooking of beans
facilitates denaturation of protein and inactivation of
toxic protein such as trypsin inhibitors and other
proteases, and make them palatable. Bressani, et al. (1963)
reported a reduction in protein efficiency ration from 1.08
to 0.63 of beans cooked for 30 and 180 minutes,
respectively. One of the major limitations in the
utilization of dry beans is the increase in cooking time
that occurs when dry beans are stored at high temperatures
and humidity (Morris, 1956 ; Muneta, 1964).

The culinary quality of dry edible beans is an
aggregate of properties perceived by consumers and
processors that have bearing on their preferences and
requirements for dry and cooked grains (Hosfield, et al.
1984). There is more awareness on the part of the consumers
in regard to characteristics relating to been color and
appearance, ease of preparation, wholesomeness , texture and
digestibility.

The recognition of the nutritional importance of dry
beans in the diets of a large proportion of consumers in
the poorer countries of the world has led to breeding
programs aimed at improving nutritional quality and ease of
cooking. Realization of these goals could lead to an
increased utilization ofbeans on a world basis.

This research investigation was part of an initiative

to help improve nutrient availability and utilization of

beans in
nutritio
develop
objectiv

inner; ta:

cooked b

culinary

 

 

beans in human diets. A paucity of genetic information on
nutritional and culinary traits in beans has hampered the
development of efficient breeding strategies. The
objectives of this research were to: (1) to study the
inheritance of cooking time and the uniformity of cooking of
individual grains, (2) ascertain the inheritance of tannin
content in the seed coat and protein content of raw and
cooked bean seed and.(3) determine the genetic control of

culinary quality traits.

plant

Cookir

CHAPTER 1

GENETIC CONTROL OF COOKING TIME AND UNIFORMITY
OF COOKING IN DRY EDIBLE BEANS
(Phaseolus vulgarisL.)

 

Abstract

Prolonged cooking is one constraint that limits beans
in diets of people in poor countries of the world. The-
presence of genetic variability for this trait would allow
plant breeders to develop bean cultivars that are fast
cookingandin which the individual grains cook to eating
softness uniformly. The objective of this study was to
investigate the genetic of cooking time to eating softness
and uniformity in cooking of individual seeds. Eight parents
were crossed in diallel fashion and genetic analyses of the
F2 and F3 generations were made. Highly significant GCA and
SCA were observed for cooking time but only GCA was
significant for cooking uniformity. Examination of GCA
effects of parents for cooking time revealed that fast
cooking parents produced progenies that were also fast
cooking. Highly favorable GCA effects for cooking time
persisted in the F3 generation indicated that selection
could be practiced in the F2. The vr,wr graphic analysis
indicated that cooking time and uniformity in the parents
were controlled by a system of genes that exhibited complete
dominance.

Additional index words: diallel cross, general combining

ability, specific combining ability, complete dominance,

selection strategy.

baked

the cot

factors
Ex
aworld
that th
input ‘t
cocking
beans ir
COOK bea
are fopc
from eat
Al
needed
C°0king
baans p;
PPOpot-ti
comprise
Al

beans m0!

INTRODUCTION

Dry edible beans (Phaseolus vulgaris LJ are generally

 

soaked and must be cooked before eating. Cooking tenderizes
the cotyledons and increases palatability, renders protein
and starch digestible, and inactivates endogenous toxic
factors that can markedly limit final nutritional value.

Energy for cooking is becoming a limiting resource on
a world wide basis. Some dry beans have a special problem in
that they require a long cooking time with a high energy
input to render them soft enough for eating. Prolonged
cooking is one constraint that limits the utilization of
beans in diets. The increased cost of energy required to
_cook beans, and the scarcity of fuels needed for preparation
are forcing many consumers in poorer countries to turn away
from eating beans. .

Although beans with a relative short cooking time are
needed in lesser developed countries, rapid and uniform
cooking cultivars are also important characteristics for
beans produced in the U.S.A. This is because a large
proportion of - U.S. grown beans, especially those that
comprise colored seed coated market classes, are exported.

Although technological means are available to cook
beans more quickly, these are energy intensive and require
the addition of a large amount of sodium or phosphate salts
to the cooking medium (Rockland, 1972). Increased salt

consumption by swine, rats, and humans over a period of time

has been a
(Guotkin :
Grommet,
The
uniform
objectiv
increase
the app
Sciatini
Precipi4
Primarily
Tr

throngh

has been associated with detrimental physiological effects
(Gwotkin and Plummer, 1946; Hartroft, 1967; Altschul and
Grommet, 1980).

The development of beans with a relatively short and
uniform cooking time would be a useful plant breeding
objective. It would save fuel, thus reducing costs and
increase utilization through shorter preparation time. Since
the application of heat to beans causes starch to
gelatinize, protein to denature, pectic substances to
precipitate, and cellular deformation, cooking time is
probably a complex trait.

The present investigation was undertaken to ascertain
through genetic analysis of an 8-parent diallel cross, the
type of gene action involved in cooking time of beans. A.
further objective was to determine whether dry bean strains
cooked uniformly and if this characteristic shows genetic

variability.

tn

produc
color,

resists
parents

1-CM-M-

(I)

(81" ).:

comprise

MATERIALS AND METHODS

Eight strains of dry beans adapted to the bean
production areas of Michigan and differing in seed coat
color, nutritional and culinary quality, growth habit and
resistance to pests and environmental stresses were used as
parents in the study. The strains Brasil-Z (B-2), FF 16-15-
1-CM-M-M (FF), 15-R-148 (15-R), A-30, Black Turtle Soup
(BTS), Sanilac (SAN), San Fernando (SF) , and Nep-2 (N-Z)
comprised a broad genetic base (Table 1) and were crossed in
all possible combination in the winter of 1980-1981. The 8
parents and 56 F1 hybrids were space planted in an
unreplicated plot in a nursery the following summer. The
seeds from the 8 parents and the F1 plants from each cross
were harvested in late September and bulked. A sample of
seed from each parent and F2 progeny was sent to the
University of Puerto Rico, and planted in a winter nursery
at the experimental farm at Isabela. In the summer of 1982,
the F3 and parental seed produced in Puerto Rico and remnant
F2 and parental seed produced in Michigan in the previous
year were tested for time and uniformity of cooking at
Pullman, Washington. The seeds of the 8 parents and 56 F2
and parents and 56 F3 progenies were divided into triplicate
samples which served as laboratory replication and<evaluated
using the Washington State University modification (Hudson
1982) of the experimental bean cooker described by Mattson

and Morris (1946). Since the experimental cooker operated on

a force in
an devic:
(PEA). The
a10 x 1C

bean. Pr

a force induced puncture principle brought about by gravity,
the device will be referred to as the pin fall apparatus
(PFA). The PFA contained 100 aluminum test cells arranged in
a 10 x 10 grid. Each cell was filled with an individual
bean. Prior to placing each bean in a cell, it was
scarified by nicking a corner on the side opposite of the
hilum, soaked for 12 hours in tap water at 270 C, and
blotted dry. Each bean was placed in a cell of the PFA so it
lay in a tangential plane. A small piece of seed coat on the
upper and exposed bean surface was carefully peeled away to
reveal a cotyledon. The 100 hollow stainless steel rods of
the PFA each with a 1.47 mm flat tip and previously filled
with 90.610.2 g of No. 8 bird shot were rested on the
exposed cotyledon of each seed. The PFA with 100 seeds and
100 weighted rods (plungers) resting on each seed was.
lowered by a winch into a water bath at 93.30 C and cooked
for 60 min. The time it took each been to soften and the
weighted plunger to penetrate was recorded. Ten seed of each
strain were cooked simultaneously and ten strains were
evaluated together in any replication. Any seed of an entry
that did not soften during cooking (plunger did not
penetrate) was considered as uncooked and hard seed. After
60 minutes, data were taken on each entry for average
cooking time and percentage of hard (uncooked) seeds.

Statistical Analysis. All data were analyzed on an entry
mean basis using the analysis appropriate to a randomized

complete block design. Because the variation associated with

entries
combinir
Griffing
extended
into mate
(1963).

and the .

infereno

ihere:

bi is the

C. I.

8.

UK is th
L .
‘19th f5

J 18 the

Ute“

10

entries was significant for the traits, analyses of
combining ability were performed using the formulas of
Griffing's (1956) model I, method I. These analyses were
extended to include the partitioning of reciprocal variation
into maternal and nonmaternal sources according to Cockerham
(1963). In Griffing's (1956) model I all effects are fixed
and the experimental material is the population about which
inferences are made. Components of genetic effects for
combining ability and reciprocal variations were computed
using least square estimates according to Cockerham (1980).

The effects model appropriate to this study were:

Y13k=u + bi + 33 + sk + 331: + m3 - ms; + rjk + e131:
= “a + bi + A3 + A}; + 633033 + ajknjk + okkak
+ m3 - mg + rjk + e133 '-
E(23)a E L 1/2bp (Y.J. + Y..j) - 3 J
- (AJ- X.) + 633(D33-‘533.) + 63k(53- - 5..)
“3310- E [1/2b ( Y-jk + Ln; ) - 3- 2;) - 2x]
= 63K(Djk - bj.-B.k + 5..)
Where:

pols the population mean

b1 18 the 1’5” block effect, 2b1'0
83 is the GCA effect associated with the 3th parent,
, 0
I?83' .
333 is the SCA effect associated with the cross between
D
the 3““ female and the kth male parents, Sjk'SKJ: £33130

m3 is the maternal effect associated with the 3th parent

‘ p
when it is used as a female, sz'O
I

F3; is

be tween

9
2

K1
eijx 1‘
in the
53 is
gamete
X. is
5.. is

Popula

033 is

11

rjk is the reciprocal effect associated with the cross
between the jth female and kth male parent, pjk=-rkJo

9.:
Kim-0
913k is the error term associated with the 3th and kth parent
in the ith block, Ein~ND(o,3)
A3 is the sum of the additive effects for genes in a
gamete from 3th parent.
I. is the mean additive effect.
5.. is the mean dominance effect of the hybrid diallel

population.

D33 is the sum of the dominance effects associated with

the jjth parent.

933. is the mean of the dominance effects associated with
the 3th parent

53. is the mean of the dominance effects associated with
progeny of the jth parent

Djk is the sum of the dominance effects for genes from
mating of 3th parent with kth parent

533 is the expected proportion of loci homozygous for the
allele derived from the 3th parent, 533:1/2F

65k is the expected proportion of heterozygous loci,

5jk=1'F'

Analyses of covariance between the offspring of each
parent (array) and the nonrecurrent parent minus the variance
of their offspring in each parental array (wr'Vr) were
performed according to Hayman's (1954a, b) analyses of

diallel experiments to ascertain whether gene expression was

appropriate to an additive dominance model or wmether

the

theo:

12

appropriate to an additive dominance model or whether
epistasis was a general feature of the system. A significant
F test for the arrays indicate a failure of one or more of
the seven assumptions underlying Hayman's diallel cross
theory (1954a,bL.

Genetic relationships among the parents were studied
with the variance-covariance (Vr,VhJ graphical analysis
technique developed by Jinks (1954) and Hayman (1954a, b).
The Vr, wt. graph was drawn for the means of three
replications. If the regression coefficient (b) of a graph
was not significantly different from zero,a disturbance in
the assumptions underlying the dial lel analyses of Hayman
(1954 a, b) was implied. To continue the analysis in the
event of a disturbance, it was necessary to determine whether
epistasis was a general feature of the genetic system, or
whether it was due mainly to the influence of one or few
parents. The procedure followed was suggested by Hayman
(1954) and involved the removal of parental arrays until a
subset of the diallel table is found that meets all the

assumptions.

RESULTS

The analyses of variance of cooking time and % hard
seed indicated that significant differences existed in the
F2 and F5 among parents and crosses for these traits (Table
2). This permitted the partitioning of among-entry variation
into general and specific combining ability'(GCA‘and SCA;
respectively) and reciprocal effects. The analyses of
variance 0f (“r-VF) exhibited no significant differences
among the arrays (Table 3) indicating that the additive
dominance model was adequate to describe the gene action
involved in the expression of the traits. The F tests of GCA
and SCA were highly significant for cooking time in both F2
and F3 generations, and the GCA mean square was highly
significant for percent hard seed in both generations
(Table 4). No differences were noted for SCA for the hard
seed character. Significant reciprocal effects were present
and were due to both maternal and nonmaternal causes (Table
4).

Estimates of variance components computed from the
formulas outlined by Griffing (1956) revealed that variances
of GCA effects predominated in the F2 for percent hard seed
while SCA.effects predominated for cooking time (Table 4).
In the F3 generation these estimates were about equal for
both traits (Table 4). General combining ability effects
calculated for each parent indicated that B-2 and 15-R

transmitted more genes for a longer and nonuniform cooking

13

14

time to thEiP F2 progenies than did the other parents (Table
5). San Fernando in addition to B-2 and 15-R, transmitted

genes for prolonged cooking time and hardseededness to F3
progeny (Table 5). A similar trend was observed when the
mean of the parents was compared with the mean of crosses
for cooking time in each generation (Table 6). Parents with
longer cooking time generally produced progenies, that
required a longer time to cook whereas the quicker cooking
parents produced quicker cooking progenies (Table 6).
Examination of the Vr, "r graphs for cooking time and
percentage hard seed provide specific information on the
type of gene action involved in trait expression for each
parent and the relationship among parents for dominance or
recessiveness of the.trait (Fig.1,2,3 and 4). The regres-
sion coefficient of "r on VP for cooking time of the 8
parental arrays in the F2 at East Lansing was significantly
different from unity but not significantly different from
zero. This indicated a failure in one or more of the dial lel
analysis assumptions (Hayman 1954 a, b) and prompted the
elimination of A-50 (parent 4) from the analysis because it
had an exceedingly large covariance. The subsequent analysis
(A-50 eliminated) showed the regression coefficient to be
significantly different from zero but not from unity (Fig.
1). Figure 1 showed that the regression line intersected the
WP axis slightly below the origin and was not significantly
different from zero indicating that complete dominance

governed the genes for cooking time. The position of the

array po
on cue V
carried
time. in
enes i:
genie i

relatiVe

15

array point for B-Z, SAN, and N-2 was distal to the origin
on the VP, wr graph (Fig. 1) indicating that these strains
carried a preponderance of recessive alleles for cooking
time. The strains BTS and SF carried predominantly dominant
genes for cooking time. The strains FF and 15-R exhibited
genic interaction because these showed a high variance
relative to covariance for cooking time (Fig. 1). The vrvwr
graph for cooking time of F3 seeds (produced in Puerto Rico)
showed that the regression coefficient was significantly
different from zero but not significantly different from a
unit slope (Fig. 2). The intercept of the regression was
above the origin and was not significantly different from
zero indicating that complete dominance controlled the
expression of this trait. Strains 15-R, A-SO, SAN and N--2
clustered together close to the origin suggesting that these
arrays carried an excess of dominant genes for cooking time.
BTS and SF carried more dominant than recessive alleles
while B-2, and FF exhibited a preponderance of recessive
genes. The change in position of array points from the F2
(seed produced in East Lansing) and F5 (seed grown in Puerto
Rico) indicated that generation and environmental effects
both may have influenced the expression of cooking time.

The Vr,wr graph for percent hard seed in the F2 (FigeB)
showed that the regression coefficient was significantly
different from zero but not significantly different from
unity..The interceptLof the regression line was above the
origin and not significantly different from zero indicating

that among the 8 parents overall, completely dominant genes

16

controlled this trait. Within the set of parents, A-BO, BTS
and SF had mostly dominant genes for percent hard seed while
B-2, SAN, FF and N-2 had both dominant and recessive genes
for this trait. Strain 15-R had mostly recessive genes for
percentage of hard seed.

Figure 4 showed that the percentage of hard seed in the
F3 generation was controlled by genes that exhibited
complete dominance. The distribution of points representing
the parental arrays showed that B-2 and BTS had a
preponderance of recessive genes while the remaining 6
strains carried mostly dominant alleles for this trait (Fig.
4). The position of the points corresponding to the
parental arrays near the limiting parabola ( Fig. 4)
indicated that these parents carried completely recessive
alleles for percentage of hard seed. The proximity of points
to each other suggested that these parents had a similar
genetic make up for hard seeded trait and that differences
among them were probably due to difference in background

genes.'

hble 1.1

Strain

Brasil-Z
FF 16-15
15-H-145
HO
Black Tu
Soup
Sanilac
San Fern
Nep-2
\
*= Incr
sqakingi:
altar 72
Y = Batza

17

Table 1. Characteristics of 8 genotypes used in a diallel cross.

 

 

Wt.gain*
after Cooking Hard
Testa soaking time Texturez seed
Strain .Source color (g) (min.) (Kg/100g) (%)
Brasil-Z CIAT Beige 11.5 30.1 469 3.6
FF 16-15-1 MSU Red 11.1 27.9 992 5.9
15-R-148 UW Red 9.5 40.3 870 24.6
A-30 CIAT Yellow 5.1 35.1 1053 56.4
Black Turtle MSU Black 11.6 30.7 665 0.0
Soup
Sanilac MSU White 11.4 26.3 605 0.0
San Fernando MSU Black y 36.4 y y
Nap-2 MSU White y 28.4 y y

 

* = Increasein weight of 10 gram of seed after48hourof
soakingin deionized water,z 2 Kg force per 100 gram of sample
after 72 hour of soaking and 30 minutes of cooking at 93°C.

y = Data not taken.

Iasle 2.

Source of
variation

‘

Z=Th A‘
EF‘S
Puerto t

 

 

18

Table 2. Mean squares of analyses of variances for cooking time
and percent hard seed in F2 and F3 generation.

 

 

 

Source of Cooking time2 Hard seedz
variation d.f. (min) (%)

. F2 F5 F2 F3
Entry 63 51.54** 61.52** 63.72** 258.20**
Error 126 7.93 2.74 29.72 134.69

 

** 2 Significant at 1% probability level.

2 a The F2 data are from E. Lansing and the F3 data are from
Puerto Rico.

Table 3

 

Source
of varia

19

Table 3. Mean squares from analyses of variances of(W -V ) for
cooking time and percent hard seed to tes the
adequacy of additive-dominance model.

 

 

Source Cooking time2 Hard seed2
of variation d.f. (min.) (5)

F2 F3 F2 F3
Array 7 35.0 34.8 173.1 833.2
Error 63 21.6 29.4 122.4 2399.4

 

z a The F2 data are from E.Lansing and the F5 data are from
Puerto Rico,respectively.

Table 4.

‘
lean squa;
and compo:
OI varian:

20

Table 4. Analyses of variance andestimates of variance
components and their standard errors for general and
specific combining ability and reciprocal effects
for cooking time and percent hard seed measured on

 

 

 

 

 

 

 

F2 and F3 progeny of an 8x8 diallel cross
grown at two locations in 1981.
Cooking time (min.) Hard seed (i)
Mean squares
and components
of variance d.f. FZY F52 FZY F52
Mean squares
Crosses 63 15.30** 20.39** 21.15** 87.20**
GCA 7 52.63** 107.03** 77.50** 278.78**
SCA 28 12.57** 9.65** 11.30 58.63
Reciprocal 28 8.71** 9.46** 16.90* 67.86
Maternal(MR) 7 5.55 15.67** 14.28 42.30
Nonmaternal 21 9.76* 7.39 17.77* 76.38
Reciprocal(NMR)
Error 63 2.64 4.25 9.91 44.90
Components of
variance
GCA 7 3.1211.55 6.42;3.15 4.22:2.29 14.621 8.23
SCA 28 9.9313.28 5.4012.60 1.3913.39 13.73i17.06
Reciprocal 28 3.03;1.19 2.6111.32 3.4912.42 11.48; 9.91
MR 7 0.1810.17 0.7110.46 0.27:0.44 0.0
NMR 21 3.5612.91 1.5712.30 3.9312.76 15.73111.93
y,z = Grown at East Lansing and Puerto Rico, respectively.
*,** significant at 5 and 1 % probability level, respectively.

rents

2
u

P

0 «$08

9 t 9

Du O .n P ..w

2 . 212w it! 2 ...L 6 )al.
. . . .m .n. . .VlRU : 1.1.. ..Yi...

. :w «\b ZFP. Q» .4

SD all 11: AH ..du «\v .\V I.“

21

Table 5. Estimates of general combining ability effects for
cooking time and percent hard seed measured on F2
and F3 generation means of an 8-parent diallel
cross grown at two locations in 1981.

 

Cooking time2 Hard seedz

 

 

 

Parents F2 F3 F2 F3

B’Z 1074 2008 1009 5032
FF -1068 -5026 0006 -5010
15-R 3.53 2.04 4.85 5.44
A-30 -1.51 -1.81. -2.05 -2.38
BTS -O.56' 0.34 -1.62 1.77
SAN -1.41 0.33 -1.20 -2.60
SF 0016 2066 -0015 4026
N-Z -0027 -0038 -0099 -4070
7 0.38 0.48 0.74 1.57
$5 0.58 0.73 1.11 2.37

 

z=The F data are from E.Lansing and the F5 data arefrom
Puerto co.

5?, S5 standard error and standard errore of the difference
between two means, espectively.

fable 6. Me'
im
at

22

Table 6. Mean cooking time (min) of parents and crosses

involving the parents in F2 and F5 generation grown
at two locations in 1981.

 

 

 

 

F2’ F32
Strain Parents Crosses Parents Crosses
8-2 5001 5608 5000 4202
FF 27.9 33.2 29.0 36.8
15‘R 4003 3704 4007 4305
A‘Bo 3501 3204 5702 3906
BTS 30.7 34.1 43.0 41.2
SAN 26.3 33.7 43.0 41.1
SF 36.4 34.1 47.3 43.8
N‘Z 2807 3407 4402 4002

 

 

y,z a Grown at East Lansing and Puerto Rico, respectively.

40

 

30

Wr 20

1O

 

 

23

Brasll-z 0 6. Sanilac

o 1.
o 2. FF16-15-1' o 7. San Fernando
o 3. 15-8-148 - e 8. Nap-2

. 5.

  

 

40' BTS
30
WVr 20
10
b =0.76 20.19"”r
a = -1.22 2 2.46
04
/ 10 20 30 40
Vr

Figure 1.

Variance (7r)-covariance (Hr) graph for cooking
time data of the pa generation grown at East
Lansing and showing the position of points
representing the 7 parental arrays and their
regression line relative to a limiting parabola
under which all points must lie. '* = Significant
at the 11 level of probability.

(A,
(I l

C)

Fiau

. 24.

0 1. BrasH-Z o 5,315

o 2.FF16-15-1'- o 6. Sanilac
45 * 0 3, 15-3-143 o 7. San Fernando
° 4. A-3o o 8. Nap-2

  

355»

:25

Wr

‘15,

-o.9: 0.16”
a = 4.622 2.16

 

s 15 25 35 45
Vr

Figure 2. Variance (VA-covariance (Hr) graph for cooking
time data or the F3 generation grown at Puerto
Rims and showing the position of points
representing the 8 parental arrays and their
regression line relative to a limiting parabola
under which all points must lie. " = Significant
at the 1% level of probability.

45

Wr

 

25

O 1. BrasH-z O 5. BTS
o 2.FF16-15-1 o B. Sanilac
O 3. 15-R-148 0 7. San Fernando

 

. 4. A-SO O 8. Nap-2
455»
35 .
.3
25 _ s
0
cl
.6
'15,
2 e
4 O b: 0.772 0.24
a= 5.83 23.56
5’, .7
0 L A 1 1 _}
g s 15 25 35 45
Vr
Fisure 3. Variance (Vr)-covariance (Hr) graph for 3 hard

d
data of the 5 grown at E. Lanisng an
gagging the position 6f points representing the 8
parental arrays and their regression line
relative to a limiting parabola under which all
points must lie. " = Significant at the 1: level

of probability.-

H!

1‘ ‘1‘ r

 

 

455.

35.

 

26'

O 1. Brasil-z O 5. BTS

o 2.FF16-15-1 o B. SanHac
I. 3, 15-52-143 0 7. San Fernando

0 4'. A-ao e 8. Nap-2

  

0.87: 0.16**
'-5. 89.. 8.91

15 25 35 ‘2‘5
Vr

Variance (Vr)-covariance.(Wr) graph for hard
seed «data of the F3 grown at uerto 1Rico and
showing the position of points representing the 8
parental arrays and their regression line
relative to a limiting parabola under which all
points must lie. *' = Significant at the 1S level
of probability.

Hen

DISCUSSION

The distribution of array points representing the
parents on the Vr, WI. graphs (Fig. 1,2,3 and 4) showed that
both dominant and recessive genes controlled the expression
of cooking time and percentage hard seed in this population.
Parental arrays differed for dominance and recessiveness of
trait expression depending on the generation. Differences
between the F2 and F3 may have been due to location effects
(East Lansing and Puerto Rico) or to the change in dominance
relationship. This point needs further experimentation to
resolve because the effects of location and generation were
confounded in the present experiment. The combining ability
mean squares indicated highly significant GCA for cooking
time and percent hard seed in both F2 and. F3 generations but
the SCA mean square estimates were highly significant only
for cooking time (Table 3). The magnitude of GCA was larger
than SCA suggesting that the type of genetic variance in the
reference population was mostly additive but that
non-additive variance was also present and should notlbe
Overlooked. The significant GCA mean square for percent
hard seed suggested that selection against hard seed defect
will not be difficult. Superior performing progeny can be
Produced by crossing parents with high general combining
ability. Parents with large SCA effects could also be
Crossed and recurrent selection practiced to maximize the
use of fixable non-additive genetic variances. In the present

sttidy, parents with large negative effects for cooking time

27

and

FF i
Alt:

Dr

the

fa

871'

28

and percentage of hard seed after cooking would be the
superior parents to use in crosses. For example, the use of
FF in crosses should lead to an improvement in both traits.
Although variance of reciprocal effects were sizable (Table
4) they were of mostly non-maternal origin. Non-maternal
reciprocal effects are complex interactions that are
difficult to interpret biologically. The use of such
effects in plant breeding scheme would be largely
unpredictable.

The reduction of cooking time in dry beans would be an
important step to increasing the consumption of dry beans in
human diets. This is especially true in lesser developed
countries where beans are a staple in the diet but the
shortage of fuel has forced people to turn away from eating
beans. Uniformity in cooking, i.e., reduced percentage of
hard seed after cooking, would improve the palatability of
cooked beans. This characteristic is important for bean
cultivars sold to been processors in the UALA. and would be
favored by both domestic and foreign consumers. The signifi-
cance of this investigation to plant breeders is that it
provided information about the genetics of the traits and
their consistency in two different generations despite the
fact that each generation was grown in a different

environment.

\N
O

11.

REFERENCES

1. Altschul,A., and J. K. Grommet. 1980. Sodium intake and
sodium sensitivity. Nutrition review 38(12):393-402.

2. Griffing, B. 1956. Concept of general and specific
combinig ability in relation to diallel crossing system.

3. Gwotkin,R., and P. J. G. Plummer. 1946. Toxicity of
certain salts of sodium and potassium for swine.
Canad. J. Comp. Med. 10:183-190.

4. Hartroft, W. S. 1967. Nutrition, pharmacology, and
hypertension. Nutrition reviews. 25(6): 161-164.

5. Hayman, B. I. , 1954 a. The analysis of variance of
diallel tables. Biometric 10:235-244.

6. , 1954 b. The theory and analyis of
diallel crosses. Genetics 39:789-809.

7. , 1958. Theory and analysis of diallel
crosses. II. GEnetics 43: 155-172.

8. .Hosfield,G‘.L., and M.A.Uebersax.1980. Variability in
physico-chemical properties and nutritional components
of tropical and domistic dry bean germplasm. J. Amer.
Soc. of Hort. Sci. 105: 246-252. .

9. Hudson, L. 1983. Personal communication-
10. Jinks, J} in, 1954. The analysis of continuous variation

in a diallel cross of Nicotiana rustica varieties.
Genetics 39:767-788.

 

11. , and B. J. Hayman, 1953. The analysis of
diallel crosses. Maize Genetics Coop.News letter 27:48-54.

12. Mattson, S. 1946. The cookability of yellow'peas. Acta
Agr. Suecana II 2, 185.

13. Rockland, L. B. 1972. Quick cooking process inmproves
bean image. Canner packer 141(9):8.

14. Wassimi, N. N., G. L. Hosfield, M. A. Uebersax, 1981.
Effect of seed coat color on soaking characteristics
of dry beans. Michigan Dry Bean Digest 6(4): 2-3.

29

nef
Ad fi

beans
he

..b

BU

0

DGPCEJ
of ta
to-pr
e.erz
nigh

CHAPTER 2

GENETIC CONTROL OF TANNIN CONTENT AND
PERCENTAGE PROTEIN OF DRY AND COOKED
DRY BEANS (Phaseolus vulgaris L.)

ABSTRACT

The presence of tannins in the seed coat of colored
beans has been associated with an impaired nutritional
benefit from eating beans. Current thinking among breeders
favors the strategy of breeding beans with a low tannin
content and maintaining or increasing seed protein
percentage. The present study investigated the inheritance
of tannin and protein content in beans with varying seed
coat colors. Eight parents were crossed in diallel fashion
to produce an F2 and F3 generation on which to evaluate
combining ability variance and effects. Tannin content was
determined by the vanillin-HCl method and protein percentage
was determined using an infrared reflectance analyzer.
Parents and progeny differed among themselves for tannin and
protein content. Both GCA and SCA mean squares were
significant for tannin and protein percentage. Maternal
effects for the traits were revealed. Strains that were
high and low in tannin and protein content produced
progenies that were also high and low in tannin and protein
content. The correlations between parental values and their

GCA effects for tannin and protein content in the F2 were P

30

F8118
CORD

Addi

poly;

21

=- O.88** and r = O.94**, respectively. Graphical analyses
revealed that both tannin and protein content were
controlled by dominant genes.

Additional Index words: Diallel cross, combining ability,

polyphenols, nutritional quality, gene action.

SOUP!

1ij

the:
am}
DOS
qua

Ad

4|]

9P0
sol

dUe

FPO

INTRODUCTION

Grain legumes constitute one of the most important
sources of food nutrients to the people in many countries of
the world. They are good source of protein, carbohydrates,
oil, minerals, and vitamins. In dry beans (Phaseolgs
vulgaris.LJ crude protein percentage varies from 17 to 35 S
(Meiners and Litzenberger 1975; Evans and Gridely 1979 ).
Although bean protein is not complete, because of a
deficiency in sulfur bearing amino acids, it has a high
lysine content. High protein and high lysine in beans makes
them an ideal complement to cereal based diets (Bressani,
1983)-

The presence of variation in the protein content and
amino acid profiles of beans indicates that it should be
possible to improve this nutrient both qualitatively and
quantitatively. Adams (1975) suggested the possibility of
increasing seed protein percentage and sulfur amino acid
content of dry beans by genetically manipulating regulator
genes that control specific protein fractions. Osborne
(1894) and Buchbinder (1980) recognized that bean seed
protein could be separated into fractions based on
solubilityg'The most abundant fraction had the properties of
a globulin and was designated "phaseolin".

Enhanced seed protein‘percentage has been shown to be
due to a increased amount of phaseolin and nonphaseolin
protein (Sullivan and Bliss 19830. Breeding lines selected

for either high or low seed protein percentage have been

33

reported to show corresponding increase or decrease in
phaseolin content (Ma and Bliss 1979). A 60 % loss in
protein content during cooking have been reported in dry
beans (Watt, 1963). These losses might be associated with
losses of soluble solids during cooking.

In addition to an imbalance in the amino acid
complement of beans, antinutritional factors and poor
digestibility limits bean nutritional quality. The poor
protein digestibility of legumes, and certain cereal grains
is in part thought to be caused by complexing of the protein
with condensed tannins ( Eliase, et al., 1979; Price and
Butler 1980 ). Tannins are known constituents of many
cereals and legumes. The type, amount and distribution of
tannins within plants are characteristic for each species.
Within a particular species, the variation depends on stage
of growth, physiological condition of plantq time of year
and the genotype ( Johanson, 1940; Maxson, etual. 1972 ).
Tannin content has been shown to be under genetic control in
several plant species including dry beans (Parada 1975; Ma
1978; Croft et al. 1980; Woodruff et al. 1982).

The objectives of this study were: 1) to look at the
inheritance of tannin in the seed coat of diverse colors of
dry bean seed using the diallel cross. 2) To study the

inheritance of crude protein.

REVIEW OF THE LITERATURE

Protein. Sullivan and Bliss (1983 ) reported that seed
protein percentage can be increased by raising the amount of
accumulated protein in relation to a constant amount of non-
protein dry matter or, conversely, by reducing the amount of
accumulated non-protein dry matter while the total protein
fractions remain unchanged. Hence, a change in protein
percentage among several lines may be similar but the‘basis
for the change may be different and may reflect differences
in the genetic control of synthesis and accumulation of the
constituent fractions. Mutschler et al. (1980) found that
patterns of accumulation of phaseolin, nonphaseolin seed
protein, and non-protein dry matter showed genetic
variation,'and that the accumulation patterns affected seed
protein percentage. Leleji, et al. (1972) reported negative
correlations ( r=-0.45 ) between yield and percentage crude
protein in F2 and F3 plant. Similar low and negative
correlation (rs-0.30) between seed yield and percentage
protein has been reported by Kelly and Bliss ( 1975 ). The
same authors also showed that seed yield and percentage
available methionine were uncorrelated, but a positive
correlation (r=Oe33) existed between percentage protein and
percentage available methionine.

Definition of Tannins. Tannins are any naturally occurring
water-soluble compounds of a high molecular weight (between

500 and 3,000 MW) and containing a sufficiently large number

34

35

of phenolic hydroxyl or other suitable groups to enable it
to form effective cross-links between protein and other
macromolecules ( Jones, 1971 ). Tannins of biological
significance are condensed molecules and are often called
condensed tannins, condensed polyphenols, and procyanidins.
Localization of tannins. In pulses, the seed coat is tissue
of maternal origin, and when the plant senescences,
monomeric phenolic compounds are transformed into immobile
polymers (condensed tannins and phlobaphenes; Freudenberg,
1962). Since condensed tannins are largely formed after
physiological maturity, they might be expected to be found
in high concentration in seed coats ( Bakshy, et al. 1978 ).
, In sorghum (Sorghum bicolor L.) , cultivars with pigmented
seed coat and dark pericarp are always higher in tannin
content than the light colored cultivars (Freudenberg,
1962). The epicarp, hypocarp, and pericarp layers and tests
in sorghum contain tannin and other pigments (Bate-smith,
1969). The presence of these layers is under genetic control
( Quinby, 1954 ). Three major genes, B1,32 and s, have been
shown to control pericarp color, and the pigmented tests
can significantly affect tannin content in sorghum seeds
(Maxson, 1972 ). The presence or absence of pigments in the
seed coat of lima bean (Phaseolus lunatus L0 is determined
by‘a single locus (Allard, 1953L.Only'in the presence of
recessive conditions, is pigment production restricted and a
white seed coat results. In the presence of the dominant

tallele pigment is produced, but the particular pigment color

36

depends on genes at several other loci. In faba beans (ligia
_f__a_b_a L.) the palisade and hourglass cells are primarily
responsible for seed coat thickness, and it is in the cell
walls of the palisade cells and lumina of the hourglass
cells that condensed tannins are deposited (Erith 1930 ).

Inheritance g_f_‘ Tannin. In faba been the white flowers are
determined by a single gene and produced a tannin-free seed
coat (Croft, et al. 1980 ). The white flowers have a number
of pleiotropic effects resulting in blockage of the
production of certain phenolic compounds in stems, stipules,
flowers, and testa. In faba beans, several sources of a gene
for tannin-free seed coat have been found among cultivars
with white flowers and white (or buff) seeds. There are at
least two different complimentary genes that confer the
tannin-free characteristic, suggesting that the synthesis of
condensed tannins can be blocked in at least two different
stages (Picard, 1976). Among the tannin-containing cultivars
of faba beans there is a wide range of tannin concentration,
but a considerable amount of the variability is due to
environmental variation ( Marquardt et al. 1978 ). On the
other hand, the analysis of four F2 populations of dry beans
resulting from crosses between lines differing in tannin
content revealed both a continuous and discontinuous pattern
of inheritance (Ma, 1978 ). One F2 population did not
deviate significantly from an expected model of one
incomplete dominant gene for tannin content. The other

three F2 populations showed a continuous pattern of inheri-

tance (Ma 1978). Moreover, low tannin was dominant to high

37

tannin and broad sense heritabilities were Ov84 to 0.97.
These results indicated that a few genes are probably respo-
nsible for genetic differences in tannin content. In sorghum
tannin content has been shown to be controlled by few genes
(Woodruff, et al. 1982 ) and low tannin was dominant to high
tannin in one case Parada (1975)1and high tannin to low in

another case (Woodruff, et al. 1982). In birdsfoot trefoil

 

(£9325 corniculatus L.) tannin content was controlled by a
single gene and high tannin was dominant to low (Dalrymple,
et al. 1984). In dry beans, the F2 segregation for tannin
content was not associated with a particular seed coat
(Rannenkamp, 1977). This implies that tannin content and
seed coat color are independently inherited.

Biological effects 3; tannins. Studies on tannins in
different crops have indicated that tannin may affect the
nutritional quality of and/or disease resistance ( Lindgren
1975; Schaffert 1974; Staller 1970). Butler (1978) ascribed
low nutritional value of high-tannin sorghum to the interac-
tion of tannins with either dietary protein, lowering its
digestibility, or with proteins from the digestive tract,
diminishing the effectiveness of the digestive processes. A
high tannin sorghum contains enough tannin to precipitate
under optimal conditions more than twice as much protein as
is in the grain. Feeding trials of rats (Butler, 1978,1982;
Handler 1944; Joslyn 1969; Lease 1940 ) and poultry (
Ringrose 1940; Rockland 1972 ) have shown that a small

percentage of tannic acid in the diet resulted in lower

38

weight gains of experimental animals. When different sorghum
cultivars were used to supply 50% of the diet of chicks, the
cultivars with high tannin content produced.a slow growth
and similar to that produced when the same level of tannic
acid was added to the diet ( Chang 1964 ). The in xitgg
digestibility of faba bean seed coats by ruminant bacteria
was three times greater when they contained no tannin ( Bond
1976 ). It seemed likely that tannin inhibits enzyme systems
responsible for fiber digestion either by inhibiting cellu-
lases or other carbohydrases, or by inducing nitrogen defi-
ciency in the in gitgg systems due to protein binding, or
both (Schaffert'1974 ). In a rat feeding trial using dry
beans, a reduction in protein efficiency and weight loss was
found with increased tannin content ( Rannenkamp 1977 ). In
poultry feeding trials with tannin-free and tannin-
containing faba beans, nutrient retention increased with
tannin-free cultivars ( Marquardt 1979 ). In the same
experiment, autoclaving of tannin-containing faba beans
affected their use by presumably destroying condensed
tannins which accounted for approximately one-half of the
total growth depression.

However, not all workers agree that condensed tannin
is an anti-nutritional factor. Feeding trials in rats,
poultry, and pigs failed to show differences in feed intake
and improved digestibility when faba beans with and without
the testa and raw and autoclaved were compared ( Muller
1953; Sjodin 1973; Wilson 1972 )-

It has been shown that condensed tannins retarded

39

preharvest seed germination in sorghum ( Harris 1970 ),
wheat ( Triticum aestivum L.) Stoy (1976 ) and barley
(Hordeum vulgare In) Jacobson (1977). Tannins act as gibbe-
rellin antagonists in germinating wheat and barley seed, and
the growth of cucumber (Cucumis sativus Lu) and pea (Eisgm
sativum In) seedlings (Corcoran 1972 ). High concentrations
of phenolics and their derivatives haveibeen.found around
points of infection in plants. In faba beans Levin (1971)
reported an inhibitor of polyphenoloxidase. When this inhi-
bitor was destroyed, the polyphenols were oxidized into
compounds which inhibited the growth of pathogenic fungi.
In certain herbaceous legumes, tannins complexing with
proteins can be advantageous in preventing bloat in
,ruminants ( Jones 1971 ). 0n the other hand, the disad-
vantage of tannin-containing plants is their astringency and
low palatability that discouraged animal feeding ( Abdalla
1976; Donnelly 1969 ).

0t)

MATERIAL AND METHODS

Genetic material. Twenty strains of dry beans were grown

in a replicated nursery during the summer of 1980 at the
Montcalm Research Farm near Stanton, Michigan. These strains
comprised a broad genetic base and differed in seed-coat
color, size, and shape. They varied in growth habit and
plant morphololgy, reaction to heat and drought stress, seed
filling characteristics, maturity, and yield.

After harvesting and threshing plants from one repli-
cation in the fall, seeds of each strain were evaluated for
protein percentage, tannin content, and soaking characteris-
tics including the percentage of hard seed after 48 hours
soaking time (Wassimi, 1981). Based on the characteristics
of dry and soaked seed, 8 strains (Table 1) were selected
for genetic analysis of tannin content, and percentage crude
protein. All possible F1 hybrids were made in the greenhouse
during the winter of 1980-81. The 56 F1 crosses and 8
parents were grown in a nursery in East Lansing during the
summer of 1981 to produce F2 seed.

Field plot procedure. The F2 seed from each cross was
harvested in bulk in September of 1981, and a sample from
each parent and each cross was grown in a winter nursery in
Puerto Rico in the winter of 1981. Parental and F3
generation seed was harvested in March of 1982 bulked and
returned to East Lansing, Michigan. A random sample of 800

seeds from each cross of the Puerto Rico produced F3

generation and parental strains and the remnant F2 and 8

4O

41

parents grown in Michigan the previous year were planted in
an111 x 12 rectangular lattice with three replications at
the Saginaw Valley Bean and Sugar Beet Research farm near
Saginaw and at East Lansing, Michigan in May and June of
1982. Seeds were precision drilled with a tractor mounted
air planter (Taylor, 1975) into two row plots guarded on
each side by the cultivar ‘Seafarer'. Rows were 4.9 meter
long and 50.8 cm apart. Herbicide and fertilizer applica-
tions were made per seasonal recommendations.

Mature plants were harvested and threshed from a 6
meter row length of each plot in mid-to-late September.
Clean seeds were used to determine percentage protein and
tannin content.

Sample preparation

1. Crude protein. About 30 g of raw bean seed and 30 g from
the dry cooked beans from another experiment of both F2 and
F3 generations and grown at Saginaw and East Lansing were
ground to a 40 micron particle size with a Udy- cyclone
mill. The percent crude protein was determined on the dry
and cooked bean flour with a Neotec GQA (model 31) near
infrared reflectance (NIR) grain quality analyzer.

2. Tannin determination. For tannin determination only F3
‘seed grown in two replications at each location were used. A
random sample of 200 seed from each cross and parents was
taken. The seeds from the crosses were separated according
to testa color because of segregation for color in the

progeny from variable testa colored parental material used

42

in the dial lel cross. Based on the number of seeds for each
color obtained from a cross, the proportion of each color
that could comprise a 20 seed sample was determined and a
composite of 20 plump seed prepared. Each individual seed
was split in half and Soaked in about 10-15 ml of n-Heptane
for 12 hours. This facilitated removal of the seed coat by
dissolving waxes. The heptane was decanted, and the seed
coat was carefully removed using a fine pointed scalpel. The
excised seed coat was dried for 12 hours at 60° C, and
ground with a Wiley mill and passed through a no. 40.
screen.
Methods of tannin determination. The vanillin hydrochloric
acid method of Burns (1971) as modified by Telek (1983) was
used to estimate tannin content. Since in the tannin proce-
dure catechin equivalent is used as'a standard, tannin
content was estimated as % catechin equivalent. However,
for purpose of this paper % catechin equivalent will herein
after be referred to as % tannin. A detailed description of
3 crude protein extraction and determination of % tannin
content are given in appendix A.
Statistical procedures. Separate analyses of variance were
performed on F2 data for tannin content and F2 and F3 for %
protein at each location to test for variability among
entries. Tests of homogeniety of error variance were also
made. Error variances were found to be homogeneous by
Bartletfifls procedure described in Steel and Torrie (1980).
Analyses of variance of combined data were computed. Data

were subjected to combining ability analyses of model I

43

method I of the diallel cross of Griffing (1956). The fixed
effects model of Cockerham (1980) was applied to the data and
solved by unweighted least square procedure. The model was
fitted to include general and specific combining ability

effects and is as follows:

Yijk= # + bi + 33 + gk + Sjk + m3 - mg + rjk + eijk
= no + bi + A3 + Ag + 633033 + 5jijk f ékkak
+m3-mk+l‘jk+eijk I
EKEJ): E [ 1/2bp (Y.j. + Ynj) - 3]
= (A3- A.) + éjj(D,j.j" D33.) + 531.1(133- " 5")
E(gjk)= E [1/2b ( Y-jk + Y.kj ) - a ‘ gj ' gk]
3 63k(D,jk - BU" 5.x + 1.3..)
Where:

u is the p0pulation mean

b1 13 the 1th block effect,‘2b1=0
. a p
33 is the GCA effect associated with the 3th parent, 733:0
Sjk is the SCA effect associated with the cross between the
9.
3th female and the kth male parents, sjk=8kja Esjk‘o
m3 is the maternal effect associated with the 3th parent
P
when it is used as a female, zmjgo
l
ij is the reciprocal effect associated with the cross
between the 3““ female and kth male parent, ija-rkj,
9-1 0
kgjrak
eijk is the error term associated with the 3th and kth parent
2
in the ith block, E13k~ND(0.°)
A3 is the sum of the additive effects for genes in a gamete

from 3th parent.

44

A. is the mean additive effect.

5.. is the mean dominance effect of the hybrid diallel
population. .

933 is the sum of the dominance effects associated with the
33th parent

533. is the mean of the dominance effects associated with
the jth parent

53. is the mean of the dominance effects associated with
progeny of the 3th parent

933 is the sum of the dominance effects for genes from
mating of 3th parent with kth parent

633 is the expected proportion of loci homozygous for the
allele derived from the 3th parent, 633:1/2F

63k is the expected proportion of heterozygous loci,
djkz1-F.

Analyses of covariance between the offspring of each
parent (array) and the nonrecurrent parent minus the variance
of their offspring in each parental array (”r-Vr) were
performed according to Hayman%3(1954a, b) analyses of dial-
lel experiments to ascertain whether or not gene expression
was appropriate to an additive dominance model or epistasis
was a general feature of the system.

Genetic relationships among the parents were studied
with the variance-covariance (Vr. Wr) graphical analysis
technique developed by Jinks (1954) and Hayman (1954a, b).
The VP, Wr graph was drawn for the means of two

replications.

RESULTS

Tannin content. The individual analyses of variance showed
highly significant differences among entries for percent
catechin equivalent.(Table 2). Combining ability analyses of
variance of separate location data for S tannin revealed
highly significant GCA and maternal effects present in this
population (Table 3). The SCA and nonmaternal reciprocal
effect variances.were significant only at Saginaw'(Table
3). Estimation of the variance components for GCA, SCA, and
reciprocal effects mean squares showed that GCA variance
predominated in trait expression (Table 3). The GCA effects
for % tannin calculated for each parent revealed that A-30,
SAN and N-2 transmitted genes that reduced the tannin
content of their progenies at both locations while the other
five strains imparted a high x tannin content to their
progenies (Table 4).

The combined analyses of variances revealed that inte-
raction of combining ability and reciprocal variation with
location was absent (Table 5). Reciprocal effects determined
for each parent indicated that significant maternal
variation was probably due to large maternal effects asso-
ciated with FF (0.32), 15-3 (-0.32), and SF (0.45) (data
not given). The GCA effects in the combined analyses were
consistent with those found in the separate analyses (Table
13). The tannin content of parental strains was reflected in
their progenies (Tables 1 and 6). Parents with none or a low

content of tannin produced progenies that also had a low

45

46

content of tannin in their seed coat. The analyses of
variance of (wr-vr Table 7) showed that the F test for the
arrays was not significant at either location indicating
thatLan additive dominance model was adequate to describe
gene action in the population.‘The regression of WI. on Vr
for % tannin at East Lansing and Saginaw showed that the
regression coefficients at East Lansing (b=0.92 3: 0.03**) -
and at Saginaw (b=0.98 i 0.05**) were significantly
different from zero but not significantly different from
unity (Fig. 1 and 2). The intersect of the Wr 3X18 at East
Lansing (a= 0.1310.01**) and at Saginaw (a=0.15_t 0.01**) are
above the origin and are significantly different from zero .
This indicates that a partially dominant gene system con-
trolled tannin content (Fig. 1 and 2). The distribution of
the parental array points under the limiting parabola along
the regression line showed that the genes controlling tannin
'content for SAN and N-2 were predominantly recessive while
the remaining parents had a relatively high proportion of
dominant factors. The distribution of array points represen-
ting the parents under the limiting parabola along the
regression line clustered consistently at each location for
those entries with mostly recessive genes and those with
mostly dominant genes (Fig. 1 and 2).

Variance of SCA calculated for each parent showed few
non-additive effects for B-2, FF, 15-R, and A-30. However,
it was indicated that the remaining parents.(BTS, SAN, SF,
and N-2) had high SCA variances. These results indicated

47

that progenies resulting from crosses of these parents would
produce progeny with relatively'more or less tannin than
would be expected on the average (Table 16).
Gauge protein. Analyses of the combined data indicated that
significant differences existed among entries for percentage
protein in raw and cooked beans (Table 13). This permitted
the partitioning of the variation due to entries into GCA,
SCA, and reciprocal effects. Variation from reciprocal
differences among entries was further broken down into
maternal and non-maternal causes. Mean squares for the
various main effects and their interaction with locations
are summarized in table 13. All F tests for GCA main
effects for protein content were highly significant and the
F tests for SCA main effects were significant or highly
significant. The combined analyses reflected the signifi-
cant GCA and SCA variation detected in the individual ana—
‘lyses except that the SCA mean square was nonsignificant at
Saginaw for the F3 generation (Table 10). Significant reci-
procal effects were inconsistent in the various tests and
were due to both maternal and non-maternal causes. Except
for a significant non-maternal reciprocal effect in the F2
for protein content of raw beans, reciprocal variation disa-
ppeared from the combined analyses (Table 13). A closer
examination of reciprocal variation (Table 10) revealed
that the significant non-maternal reciprocal variation for
raw bean protein in the F2 could have been caused by crosses
of A-30 x 15—R, BTS x A-30, BTS x N-2, SAN x 15-R, and SAN x
N-2. Interactions of the combining ability effect with

48

locations were essentially interaction of general combining
ability effects with locations (Table 12). There were no
significant interactions of SCA x locations. There was
significant variation among generations for raw bean protein
indicating that entries performed differently from the F2 to
F3 generations. Variation among entries within a generation
was also inconsistent from location to location; however,
when the four tests (two locations and two generations) were
combined together no entry x location interactions were
noted (Table 13).

Estimates of variance components revealed that with the
exception of the F2 generation at Saginaw, GCA effects
predominated for cooked beans (Table 14). The relative
magnitude of these components indicated the relative impor-
tance of the sources of variation. In the combined ana-
lyses, the GCA to SCA ratio ranged from 3.7 : 1 for protein
content of raw beans in the F3 to 5.3 : 1 for raw bean
protein in the F2 (Table 14). These results indicated that
additive effects of the parents were more important than
non-additive effects in determining performance in crosses.
Variance components for GCA and SCA x locations were always
smaller than the main effect components (Table 14).

There was considerable variation in the general combi-
ning ability (GCA effects) contribution of each parent to
protein content (Table 15). Specific rankings of the parent
that increased or decreased the protein in raw and cooked

beans were not exactly the same for each generation and

49

location. However, the same lines were generally in the
same positive or negative grouping. Shia.few cases a line
had a negative GCA effect at one location and positive GCA
effect at the other. For example, BTS had a -0.10 value for
protein content in raw beans at East Lansing and a CL29 GCA
effect value at Saginaw.

When means of the diallel progeny were examined indivi-
dually, crosses of low x low and high x high parents tended
to produce progenies that were low and high in protein
content, respectively (Tables 18 and 19). Some parents had
high SCA variances in one or both generations (Table 16)
which indicated that progeny from these crosses could have a
higher or lower protein content than what is expected on the
average.

The analyses of variance of (Wr-vr) were non-signifi-
cant among arrays for protein content at both locations in
the F2 and F3 generation, indicating that an additive
dominande model was appropriate for describing gene action
among crosses (Table 17). The regression of Wr on Vr (the
covariance and variances of parental arrays) showed that the
regression coefficients for percent protein in the F2
(b=1.0210.16** Fig. 3) and in the F3 (b= 10810.32“) at E.
Lansing was significantly different from zero but not signi-
ficantly different from unity. Similarly, the regression
coefficients for the F2 (be-1.091 0.10** Fig. 5) and for the
F3 (b=1.2510.30** Fig. 6) at Saginaw were significantly

different from zero but not significantly different from

unity. The intercepts on the WP 8X18 were Significantly

50

different from zero in the F2 at E. Lansing (aao.6o_+_o.14**)
and at Saginaw (a=0.791 0.08**) indicating that genes con-
trolling protein content were partially dominant. In the F3
generation the intercepts on the Wr axis at E. Lansing
(a=0.0930.32) and at Saginaw (a=0.40_+_0.34) were not signifi-
cantly different from zero, indicating that the genes con-
trolling protein content were completely dominant. Distribu-
tion of array points representing the parental arrays along
the regression line under the limiting parabola showed con-
sistency for some parents from the F2 to F3 generation
across locations but not for other parents (Fig. 3.4.5 and
6L.This indicated that some entries were more stable than

others.

51

Table 1. Seed coat color and tannin content and percentage of
seed protein of 8 dry beans strains used as parents
in a diallel cross.

 

 

Seed Tannin Protein

coat contentz content
Strain color (5) (5)
Brasil-Z Beige 9.6 22.5
FF 16-15-CM-M-M Red 8.6 26.2
15-R-148 Red 8.1 27.9
A-30 Yellow 3.4 22.4
Black Turtle Soup Black 4.7 23.9
Sanilac White 0.0 23.0.
San Fernando Black 4.4 25.0
Nep-2 White 0.0 24.8

 

z = Tannin content is expressed as % catechin equivalent.

52

Table 2. Analysis of variance for tannin content of F
seed coats in dry bean parents and progeny 0% an
8x8 diallel cross.

 

I

Mean squares
Source of

 

 

variation . d.f. E.Lansing Saginaw
Entries 63 9.3* 12.3*
Error 63 1.1 2.0

 

* 2 Significant at 1% probability level.

53

Table 3. Analysesof variance and estimates of variance
components and their standard errors for general
and specific combining ability and reciprocal
effects for tannin content of seed coat of
parents and of F2 progeny 1»er dry beans
parents and progeny of an 8x8 diallel cross
grown at East Lansing and Saginaw, Michigan

 

 

 

 

in 1982.
Mean square
Source d.f. “*East Lansing Saginaw
Entry 63 4.6** 10.5**
GCA 7 35.3** 26.5**
SCA 28 1.0 6.3**
Reciprocal 28 0.6 10.6**
Maternal (M) 7 1.4* 36.3**
Nonmaternal 21 0.3 2.0*
Reciprocal(NMR)
Error 63 0.6 1.0
Component of
variance
GCA 7 2.1711.04 1.60:0.78
SCA 21 0.3910.26 5.32:1.63
Reciprocal 28 0.00 » 4.8111.42
M 7 0.05:0.04 2.2111.07
NMR 21 0.00 0.53:0.62

 

*,** . Significant at 5 and 1% probability level, respectively.

54

Table 4. General combining ability effects for tannin content
of seed coats of each parent used in a diallel
cross of dry edible beans and grown atEh Lansing
and Saginaw, Michigan in 1982.

 

 

 

Parent . East Lansing Saginaw
Brasil-2 1.58 0.99

FF 16-15-CM-M-M 1.29 . 0.64
15-R-148 0.54 0.94
A-30 -O.34 -0.21
Black Turtle Soup 0.28 1.39
Sanilac -2.62 -2.21
San Fernando 1.10 0.02
Nep-Z -1.73 —1.58

S? 0.18 0.23

SD 0.27 0.35
SYoSD = Standard error of a mean and standard error of a

mean difference, respectively.

55

Table 5. Mean squares from analyses of variances and combining
ability analyses for tannin content measured on
parents and F2 progeny from 8-parent diallel cross
of dry edible beans grown in 1982.

 

 

Source
of
variation d.f. Tannin contentz
Location(L) 1 29.8
Reps/L 2 4.9
Entry(E) 63 19.1 **
GCA 7 145.5 **
SCA 28 4.7 **
Maternal(M) 7 4.8 *
Nonmaternal
reciprocal(NMR) 21 1.0
E x L 63 1.6
GCA x L 7 1.6
SCA x L 28 2.0
M x L 7 2.3
NMR x L 21 0.9
Error 126 1.6

 

z z Tannin content is expressed as % catechin equivalent.
*,** 2 Significant at 5%and1%levelofprobability,
' ’ respectively.

56

Table 6. Mean tannin content of seed coats of parents and F2
progeny of an 8x8 diallel cross in dry beans grown at
East Lansing and Saginaw, Michigan in 1982.

 

 

 

Parents .Percentage tannin content
and

crosses East Lansing Saginaw
Farents

3.2 (1) 909 906
FF (2) 8.2 8.6
15-R(3) 6.2 8.1
A-30(4) 3.9 3.4
BTS (5) 5.6 4.7
SAN (6) 0.0 0.0
SF (7) 6.5 4.4
N-2 (8) 0.0 0.0
Crosses .
1 x 2 7.6 7.6
1 x 3 7.8 6.6
1 x 4 6.7 6.1
1 x 5 6.6 8.8
1 x 6 4.1 4.1
1 x 7 7.6 9.0
1 x 8 6.1 . 6.0
2 x 3 7.6 7.5
2 x 4 6.7 6.2
2 x 5 6.3 9.0
2 x 6 3.9 4.6
2 x 7 7.5 10.5
2 x 8 6.2 5.9
3 x 4 5.8 6.0
3 x 5 5.8 7.2
3 x 6 3.6 3.8
3 x 7 6.7 7.7
3 x 8 4.4 5.4
4 x 5 5.7 7.0
4 x 6 1.9 3.3
4 x 7 6.4 8.0
4 x 8 2.9 3.2
5 x 6 3.7 5.3
5 x 7 7.2 8.3
5 x 8 4.9 5.7
6 x 7 5.9 6.0
6 x 8' 0.0 0.0
7 x 8 5.2 6.0
Mean (x) 5.5 6.2
$5 0.8 1.0

 

SE = Standard error of the difference between two means.

57

Table 7. Analyses of variances of(Wr-Vr) for testing the
adequacy of additive-dominance model for tannin

content of F progeny of an 8x8 diallel cross
grown at E. Eansing and Saginaw, Michigan in 1982.

 

 

 

Source Mean squares
of

Variation d.f. E. Lansing Saginaw
Array 7 0.2 0.7

Error 15 0.1 1.1

 

58

Table 8. Analyses of variance for raw and cooked bean seed
protein of parents and F2 and F progeny of
an 8x8 diallel cross grown at East Lansing and
Saginaw, Michigan in 1982.

 

Mean squares

 

 

 

 

East Lansing Saginaw
F F
Source 2 3
of Raw Cooked Raw Cooked
variation d.f. beans beans beans beans
F2 63 2.8** 5.5** 3.53** 7.44**
F3 63 2.7** 6.5** 7.49* 8.50**

 

*,** s Significant at 5% and 1% probability level,respectively.

59

 

 

 

Table 9. Correlation coefficient (r) indicating the relationship
between F and F3 .for protein content in raw (R) and
cooked (C7 beans3 grown at East Lansing (upper triangle)
and Saginaw (lower triangle) and between locations
(diagonal).

Protein content (%)
Generation FZR F2C F3R F3C
F2 R 0.67* 0.77* 0.75* 0.71*
F2 C 0.75* 0.69* 0.71* 0.70*
F3R 0.72* 0.88* 0.78* 0.85*

 

* = Significantat 1% probability level.

60

 

 

 

 

 

 

 

Table 10. Analyses of variances for general, specific combining
ability and reciprocal effects for percent protein of
parents and F2 and F3 progeny of an 8x8 diallel cross
of dry beans and grown at East Lansing and Saginaw,
Michigan in 1982.

Mean squares
East Lansing Saginaw

Source Protein (%) *Protein (3)

of

variation d.f. Raw CSoked Raw Cooked

F2

Entry 63 1.45** 2.66** 1.44** 3.75**

GCA 7 9.44** 16.05** 7.63** 27.73**
SCA 28 0.47* 0.79 0.98** 1.08**

Reciprocal 28 0.44* 1.18* 0.36** 0.42

Maternal (M) 7 0.45 0.90 0.44** 0.41
Nonmaternal 21 0.43* 1.35* 0.33** 0.43
Reciprocal (NMR)

Error 63 0.24 ‘ 0.64 0.01 0.24

F3

Entry 63 1.39** 3.22** 1.78 4.24**

GCA 7 8.72** 21.35** 13.13** 32.36**
SCA 28 0.61* 1.19* 0.52 0.90**

Reciprocal 28 0.35 0.88** 0.20 0.54**

M 7 0.37 1.21* 0.19 0.66**
NMR 21 0.34 0.57 0.20 0.50**
Error 63 0.31 0.63 1.58 0.14

 

*,** Significant at 5 and 1 % probability level, respectively.‘

61

Table 11. Components of variance and their standard errors for
general and specific combining ability and reciprocal

effects for percent protein in raw and cooked bean

seed of F2 and generation means of an 8x8 diallel

F
cross grown at East Lansing and Saginaw, Michigan in

 

 

 

 

1982.
Components
F F
Location 2 3
and Raw Cooked Raw Cooked
source beans beans beans beans
‘E; Lansing
GCA 0.5810.28 0.96:0.47 0.5310.26 1.30:0.63
SCA 0.2310.13 O.1510.23 0.3010.17 0.5610.33
Reciprocal 0.1010.06 0.2710.17 0.0210.05 0.1310.13
Maternal (MR) 0.0110.01 0.0210.03 0.00 0.0410.04
Nonmaternal 0.1910.07 0.7110.21 0.0310.06 0.00
reciprocal (NMR)
Saginaw .
GCA 0.4810.25 1.7210.82 0.7210.39 2.0110.95
SCA 0.9710.25 0.8410.28 0.00 0.7610.23
Reciprocal 0.1710.05 0.0910.06 0.00 0.2010.07
MR 0.0310.01 0.0110.01 0.00 0.0310.02
NMR 0.3210.05 0.19:0.07 0.00 0.36:0.07

 

62

Table 12. Estimates of general combining ability effectsof
parents for protein content of raw (R) and cooked (C)
bean seeds measured (n1 F2 and F‘ progeny of an 8x8
diallel cross and grown azt East iansing and Saginaw,
Michigan in 1982.

 

 

 

 

 

 

East Lansing Saginaw
F2 F3 . F2 F3
Protein(%) Protein(%)
Parents R C R C R C R C

B-2 ~ -0.92 '-0.99 -0.82 -0.76 -0.91 -1.11 -1.03 -1.35
FF 0.50 0.29 0.40 0.63 0.27 0.19 0.08 0.07
15-R 1.34 2.06 1.45 2.46 1.23 2.79 2.03 3.19
A'30 -0095 -0089 -0085 -1042 -0090 -1057 -0069 -1038
BTS -0.10 -0.07 -0.04 -0.10 0.29 0.23 -0.03 -0.12
SAN -0.39 -0.90 -0.21 -0.28 0.07 0.36 0.02 0.02
SF 0.12 0.29 -0.18 -0.44 0.03 -0.28 -0.29 -0.39
N-2 0.39 0.20 0.25 -0.09 -0.08 -0.62 -0.08 -0.22
. SY 0.06 0.15 0.07 0.15 0.002 0.06 0.37 0.03
SD 0.09 0.23 0.11 0.23 0.004 0.09 0.56 0.05

 

5’, Sfi= Standard error of a mean and standard error of a
d fference between two means, respectively.

63

Table 13. Mean squares for locations, entries, generations, and
general, and specific combining ability, reciprocal,
maternal, and nonmateranl recirpocal effects and their
interaction with locations for protein content of raw
and.cooked bean measured on combined data of
parents and F2 and F3 progeny of an 8x8 diallel
cross grown at East Lansing and Saginaw, Michigan

 

 

 

 

in 1982.
Mean squares
Source Degrees Protein (%)
of of _ _f,
variation freedom Raw beans Cooked beans
Location(L) 1 617** 1147**
Reps/L 2 2.2 8.3
Entries(E) 127 5.7** 12.4**
Generations(G) 1 7.6** 5.0
G x L 1 0.4 0.5
Entry/G 126 5.7** 12.5**
Amon F2
GC 7 41.5** 81.0**
SCA 28 1.1* 2.6*
Maternal(M) 7 1.0 2.0
Nonmaternal
reciprocal(NMR) 21 .1.3* 1.9
Among F5
GCA 7 43.1** 107**
SCA 28 1.4** 2.8**
M 7 0.8 2.7
NMR 21 0.6 1.6
E x L 127 0.6 1.5
F2X L
GCA x L 7 2.3** 7.4**
SCA x L 28 0.6 1.2
M x L 7 0.7 0.6
NMR x L 21 0.3 1.5
F3): L
GCA x L 7 1.4* 2.9*
SCA x L 28 0.6 1.2
M x L 7 0.3 1.0
NMR x L 21 0.4 0.3
Error 254 0.5 1.2

 

*,** = Significant at 5% and

respectively.

level of probability ,

64

Table 14. Estimates of variance components of GCA, SCA, and
reciprocal effects and their standard deviations for
protein and tannin content measured on parents and
progeny of an 8-parent diallel cross grown at East
Lansing and Saginaw, Michigan in 1982.

 

Rrotein content

 

 

Tannin

Component Raw beans Cooked beans contentz
Among F2

GCA 0.6410.31 1.2410.59 2.2511.07

SCA 0.12:0.08 0.2910.18 0.76:0.31

Maternal(M) 0.01:0.01 0.01:0.02 0.0510.04

Nonmaternal

reciproca1(NMR) 0.1610.10 0.1010.15 0.0
Amon F3

00 0.6610.32 1.6510.79 y

SCA 0.18:0.09 0.3310.19 y

M 0.00310.006 0.0210.02 y

NMR 0.0 0.04:0.14 y
szL .

GCA x L 0.0610.03 0.1910.11 0.0

SCA x.L 0.0310.08 0.0 0.1810.29

M x L 0.01:0.01 0.0 0.02+0.03

NMR x L 0.0 0.1310.25 0.0
FBXL

GCA x L 0.03:0.02 0.0510.04 y

SCA x L 0.0510.09 0.0 y

M x L 0.0 0.0 y

NMR x L 0.0 0.0 y

 

z = Tannin content expressed as catechin equivalent.
y a No data taken.

Table 15. General combining ability effects and their

65

interaction with locations for protein and
tannin content for each parent of an 8x8
diallel cross grown at East Lansing and

Saginaw,

Michigan in 1982.

 

F2

F3

 

Protein content

 

 

Protein content

 

 

Tannin _
Raw Cooked content Raw Cooked
Parents beans beans (%) beans beans
3.2 -00965* -10034* 1030* “00909* -10076*
FF 0.338* 0.287* 1.34* 0.225* 0.344*
15-R 1.6* 2.459* 0.49* 1.768* 2.87*
A-50 -1 0939* -1 o 226* ’0 o 58* -0 0805* -1 o 405
BTS 0.035 0.06 0.58* -0.097 -0.12
SAN ‘00204 ”00352.“. -2057* -00062 “0.048
SF 0.027 0001 1025* “00206* ‘00383*
N-2 0.108 -0.224 -1.82* 0.087 -0.181
Sy 0.209 0.274 0.29 0.181 0.271'
SD 0.125 0.197 0.23 0.125 0.197
'B-2 x L 0.025 0.045 0.19 0.124 0.278
FF x L 0.158 0.068 -0.04 0.136 0.267
15-R x L -0.275* -0.337* 0.06 -0.273* -0.332
Ab30 x L 0.085 0.328 0.13 -0.075 -0.031
BTS x L -0.163 -0.186 -0.29 -0.008 -0.006
SAN x L -0.189 -0.572* -0.03 -0.104 -0.247
SF x L -0.085 0.279 -0.13 0.058 0.012
"-2 X L 00273.} 00373“ 0010 00142 00048
5? 0.178 0.326 0.29 0.178 0.326
Sfi 0.128 0.197 0.23 0.128 0.197

 

* 2 Significant at 5% level of probability.
= Standard error of a mean and standard error of

S' s’
diffege

nce between two means, respectively.

66

Table 16. Parental values for seed protein and tannin content of

raw beans (Yr ) and variance of specific combining

ability (S2)r of each parent from data measured on F2

and F? progeny of an 8x8 diallel cross averaged over
0

 

 

 

 

content (%)

A

 

content (%)

locat ns in 1982.
Generation
F2 F3-
Protein Tannin Protein

content (%)

 

 

 

Parents Yr 3‘ Yr 3‘ Yr 32
8-2 2104 0012* 906 0014 2206 0007
FF 24.9 0.0 8.6 0.05 24.9 0.10
15"R 2707 0005 801 000 2708 0001
A-30 2203 0003 304 0006 2206 0016**
BTS 23.8 0.02 4.7 0.47** 22.9 0.14*
SAN 22.6 0.18** 0.0 0.26* 25.4 0.13*
SF 24.0 0. 0 4.4 1.34** 24.0 0.01
N-2 23.6 0. 02* 0.0 0.83** 24.0 0.14*
s? 0.25 1.28 0.25

35 0.36 1.28 0.36

 

Sb": Standard error of a mean and standard error of a
dffference between two means, respectively.

67

 

 

 

 

Table 17. Analyses of variances of ( Wr-Vr) for raw and cooked
bean seed protein in the F2 and F5 generation Of an
8x8 cross to test the adequacy of additive-dominance
model.

Mean squares
East Lansing Saginaw

Source Raw Cooked Raw Cooked

of beans beans beans beans

variation d.f (%) (%) (%) (%)

F2 7 0.09 0.18 0.02 0.17

F3 7 0.09 0.17 1.07 0.40

 

68

Table 18. Mean protein content of raw and cooked bean seeds of
parents and F2 and F3 progeny of an 8x8 diallel cross
grown at East Lansing, Michigan in 1982.

 

Protein content (%75

__

 

 

 

 

F2 F3

Parent Raw Cooked Raw Cooked
and cross beans beans beans beans
Parents

FF (2) 26.2 24.6 26.3 25.9
A-30 (4) 22.4 23.1 23.1 20.9
BTS (5) 23.9 22.4 24.0 22.9
SAN (6) 23.0 20.5 24.6 22.8
SF (7) 25.0 24.8 25.4 23.3
Crosses

1 x 2 24.3 23.2 25.4 24.1
1 x 3 26.2 25.2 25.3 25.7
1 x 4 23.4 21.7 23.6 21.6
1 x 5 24.5 23.2 24.8 23.4
1 x 6 23.1 '21.0 23.7 22.0
1 x 7 24.6 23.9 24.3 23.5
1 x 8 24.8 22.2 23.9 22.0
2 x 3 27.1 26.3 27.3 27.4
2 x 4 24.9 22.5 24.9 22.5
2 x 5 25.4 24.0 24.9 23.0
2 x 6 25.5 23.4 26.0 25.2
2 x 7 25.6 24.5 25.0 24.8
2 x 8 25.8 23.2 25.5 23.5
3 x 4 25.0 24.3 25.0 25.5
3 x 5 26.5 25.5 27.4 27.0
3 x 6 26.1 24.9 27.0 26.2
3 x 7 26.5 26.2 26.8 24.6
3 x 8 26.3 25.7 26.9 26.5
4 x 5 23.7 23.1 24.4 21.6
4 x 6 24.1 22.2 24.6 22.9
4 x 7 24.4 22.1 23.9 21.4
4 x 8 25.3 23.4 25.8 23.5
5 x 6 25.1 23.0 24.8 23.8
5 x 7 25.0 23.0 25.7 24.5
5 x 8 25.9 24.6 25.8 24.3
6 x 7 25.2 24.2 24.3 22.5
6 x 8 25.7 24.6 24.3 22.5
7 x 8 25.4 24.1 25.1 23.2
55 0.59 0.98 0.68 0.98

 

SD=Standard error of a difference between two means.

69

Table 19. Mean protein content of raw and cooked bean seeds

of parents and F2 and F3 progeny of an 8x8 diallel
cross grown at Saginaw, Michigan in 1982.

 

Protein content (i7

 

 

 

 

F2 F3
Parents -—Raw Cooked Raw Cooked
and cross beans beans beans beans
Parents
FF (2) 23.7 21.9 23.7 21.6
15-R (3) 27.5 26.6 27.7 28.1
A-30 (4) 21.4 17.8 22.8 18.5
BTS (5) 23.8 21.2 23.8 20.4
SAN (6) 22.3 21.8 22.2 20.6
SF (7) 23.0 19.8 22.7 20.5
N-2 (8) 22.4 18.6 22.0 19.6
Crosses
1 x 2 23.0 21.4 22.3 19.5
1 x 3 23.9 23.0 24.1 22.7
1 x 4 20.0 17.6 21.4 18.5
1 x 5 22.3 20.3 23.3 20.0
1.x 6 21.1 19.0 21.6 19.0
1 x 7 21.9 19.5 21.9 19.2
1 x 8 22.1 19.2 22.0 18.9
2 x 3 24.8 23.3 25.5 24.5
2 x 4 21.8 18.9 21.8 18.7
2 x 5 22.8 21.3 22.5 19.9
2 x 6 23.1 20.9 24.0 21.5
2 x 7 22.7 20.7 23.0 20.9
2 x 8 22.4 19.0 23.0 20.3
3 x 4 23.6 21.5 24.0 21.3
3 x 5 24.3 22.4 24.8 23.5
3 x 6 24.8 24.3 25.1 24.4
3 x 7 24.5 23.6 24.8 23.7
3 x 8 24.6 23.5 25.4 24.1
4 x 5 .21.9 19.5 21.9 19.6
4 x 6 21.8 18.9 22.9 19.1
4 x 7 22.0 19.1 21.9 19.1
4 x 8 22.0 20.0 23.0 21.0
5 x 6 23.2 21.4 22.7 21.2
5 x 7 22.9 20.8 23.1 20.4
5 x 8 23.0 20.6 23.1 20.9
6 x 7 23.3 21.2 23.3 20.9
6 x 8 22.8 21.1 23.5 21.2
7 x 8 22.0 18.8 22.1 19.1
35 0.13 0.60 1.54 0.46

 

Standard error of a difference between two means.

to
U
M

 

FiSure 1.

  

70

0 1. Brasll-Z O 5. BTS

O 2.FF16-15-1 O B. Sanilac
0 3 ‘15-3-143 0 7. San Fernando

. 4. A-so O 8. Nap—2

b =o.92= 0.03”
a=0.13: 0.01”

Variance (Vr)-covariance (Hr) graph for S tannin
content data of the F2 seed coat grown at East
Lansing and showing the position of points
representing the 8 parental arrays and their
regression line relative to a limiting parabola
under which all points must lie. '* = Significant
at the 11 level of probability.

0.8 ..

0.6

 

Figure 2.

71

0 1. BrasH-Z
o'2.FF16-15-1
O 3. 15-R-148
0 4. A-3O

0 5. BTS
. 6. Sanllac
0 7. San Fernando

o 8. Nap-2

 

b =0.93 20.05“
a=0.15 2 0.01 **

1 J 1 1 L 1 1 —J

0.6 0.3
Vr

Variance (Vr)-covariance (Hr) graph for S tannin
content data of the F seed coat grown at Saginaw
and showing the position of points representing
the 8 parental arrays and their regression line
relative to a limiting parabola under which all
points must lie. '* = Significant at the 15
level of probability.

72

O 1. Brasil-z O 5. BTS

0 2.FF16-15-1 0 6. Santlac
0 3, 15.-3-143 0 7. San Fernando

0 4. A-so 0 8. Nap-2

 

 

 

0.5 1.0 1.5 2,0 2.5
Vr

Figure 3. Variance (Vr )-covariance (Hr ) graph for protein
content data of the F 870““ at E- Lansing
and showing the position of points representing
the 8 parental arrays and their regression line
relative to a limiting parabola under which all
points must lie. " : Significant at the 11
level of probability.

734. .

0 1. Brash-2 0 5. BTS

0 2.FF.16-15-1 0 6. Sanilac
O 3, 15.-3-143 0 7. San Fernando

2'5 0 4. A430 0 8. Nap-2

2.0

1.5

1.0

0.5

 

  

*

b =1.08 2 0.32
3 =0.09 20.32

 

Figure 4.

0.5 1.0 1.5 2.0 2.5
Vr

Variance (Vr)-covariance (Hr) graph for protein
content data of the F grown at E. Lansing

and showing the posit on of points representing
the 8 parental arrays and their regression line
relative to a limiting parabola under which all
points must lie. * = Significant at the 53
level of probability.

 

74

- I 1. BrasH-z O 5. BTS

0 2.FF16-15-1 0 6. Sanuac
0 3. 15-3-143 0 7. San Fernando

. 4. A-3o O 8. Nap-2

  
 

b =1.09 20.10”
a =0.79 : 0.08“

 

Figure 5.

Vr

Variance (Vr)-covariance (Hr) graph for protein
content data of the F grown at Saginaw

and showing the position of points representing
the 8 parental arrays and their regression line
relative to a limiting parabola under which all
points must lie. ** = Significant at the 11
level of probability.

75

O 1. Brasil-z O 5. BTS
0' 2.FF16-15-1 0 B. Sanuac

4 0 3, 1543-443 0 7. San Fernando
0 4, A-30 0 8. Nap-2

  

1.25: 0.30“
0.40 2 0.34

 

 

Figure 6. Variance (Vr)-covariance (Hr) graph for protein
content data of the F grown at Saginaw
and showing the position of points representing
the 8 parental arrays and their regression line
relative to a limiting parabola under which all
points must lie. *5 = Significant at the 1%
level of probability.

DISCUSSION

The presence of genetic variation was demonstrated by
the significant F test of entries for tannin content and %
crude protein (Tables 2 and 8). The entry by location
component was not significant. For percent protein the
location effect was significant, leading to a significant
GCA by location interaction (Table 12). The significant GCA
mean square for tannin and crude protein % in the combined
analyses using Griffing's model I method 1 (1956) indicated
that the differences among progeny for these traits were due
to genes with primarily additive effects. Although of
lesser magnitude than GCA, significant SCA was involved in
tannin and protein content in the F2 generation at both
locations. This suggests that non-additive genetic effects
are also influencing traits expreSsion in this generation.
For the percent protein trait in the F3, SCA variance dimi—
nished. This was possibly due to the reduction in heterozy-
gosity accompanying selfing. The significance of maternal
effects for % tannin could have arisen because seed coats
are maternal tissue. Strength of this argument comes from
the fact that parents with a high % tannin content
contributed the most to the maternal effects mean square.
These parents also had significant and positive GCA effects
for this trait» For 5 protein the non-maternal reciprocal
mean square was significant, primarily because of the con-
tribution of A-30 x 15-R (0.58), BTS x A-30 (0.52), BTS x N-
2 (0.62), SAN x 15-R (-0.67) and SAN x N-2 (-0.55). However

76

77

the number of crosses that contributed to non-maternal
reciprocal effects decreased from 5 to 2 ( FFxN-2=-0.53,
BTSxFF=0.49) in the F3 generation. The comparison of
parental and progeny means revealed that genotypes with low
tannin content produced progenies that were also low in
tannin content. The GCA picture for tannin and protein
content indicated that it should be possible to develop
lines with low tannin.and high protein content. This con-
clusion is supported by the high correlation between
parental value and GCA effects for 75 tannin in the F2
(r=0.88**) and 75 crude protein (r=0.97** and r=0.94** ) in
the F2 and F3 generation, respectively. Similarly
significant correlations were obtained for GCA effects
between F2 and F3 generation for raw ('r=0.91**) and cooked
bean protein (r=0.98**).

Graphical analysis of % tannin revealed that this
trait is controlled by a gene system of complete dominance.
Sanilac and Nep-2 with white seed coat and zero 95 catechin
equivalent are the parents that carry a preponderance of
recessive genes and the rest of the parents carry mostly
dominant genes for this trait (Fig. 1 and 2). This finding
is in agreement with the work of Croft et a1. (1980) that
the white or buff seed coat associated with white flower in
1_i__<_:_i_a faba). caused by a single recessive gene (Rowlands,
1962) had no tannin. White seed coat in Nep-2 has been
reported to be caused by change of'a single locus from the

dominant to recessive (Mob, 1971). This indicated that the

78

color gene has a pleiotropic effect on substrate production
leading to the sysnthesis of tannin .

Graphical analyses of % crude protein demonstrated
that on the average a gene system with partial to full
dominance controlled this trait. The position and distribu-
tion of array points representing the parents for % protein
was relatively consistent for some of the genotypes within
each generation and location, while consistency was lacking
for other genotypes. This indicated the presence of

genotype by environment interactions.

8.

10.

11.

References

Abdalla, M. M., M. M. Morad, and M. Roushdi. 1976.
Some quality characteristic of selections of Vicia
faba L. and their bearing upon field bean breeding.

Z. Pflanzenzuchtg. 77: 72-99.

 

Adams, M. W. 1973. On the quest for quality in field
bean. P. 143-149. In: M. Milner (ed.). Nutritional
improvement of food legumes by breeding. ( Proceedings
of a symposium sponsored by PAG, Rome, July, 1972).
Protein Advisory Group, United Nations, New York.

Allard, R. W. 1953. Inheritance of some seed coat
color and patterns in lima beans. Hilgardia 22: 167-
177.

Aykroyd, W.R., J. Doughty, and A.F. Walker. 1982.
Legumes in human nutrition. 2nd edn. FAO Nutritional
studies No. 19. FAQ, Rome.

Bakshy, A. K., E. T. Mertz, and J. D. Axtel. 1978.
Effect of dehulling on tannin content, protein
distribution and quality of high and low tannin
sorghum. J. Agric. Food Chem. 26: 679-683.

Bate-Smith, E. C., and V. Rasper. -1969. Tannins of
grain-sorghum: Lutrafurol (Leucoluteolinidin)
34,45,7- pentahydroxyflavin. J. Food Sci. 34: 203-
209.

Bliss, F. A., and J. W. S. Brown. 1983. Breeding
common bean for improved quantity and quality of seed
protein. Plant Breeding Rev. 1: 59-102.

Bond, D. A. 1976. £2 Vitro digestibility of testa in
tannin-free field bean (Vicis faba Lu) J. Agric. Sci.
Comb. 86: 561-566.

Bressani, R. 1983. Nutritional quality of dry beans.
Bean improvement cooperative and nationdal dry bean
council biennial conference. Minneapolis, Minnesota,
November 8-10.

Buchbinder, B. U. 1980. Polysomal and messenger RNA
directed in vitor synthesis and processing of
phaseolin, the G1 seed storage protein of Phaseolus
vulgaris L. PhD Thesis, University of Wisconsin,
Madison.

 

 

Burn, R.E. 1963. Method of tannin analysis for forage
crop evaluation. Tech. Bull. N.S. 32, Georgia Agric.
Exp. Stn. Athens.

79

80

12.Butler, L. G. 1978. Tannins in sorghum grain:

13.

14.

15.

16.

17.

.18.

19.

20.

21.

22.

123.

241.

Problems, solution, and opportunities, 33rd annual
corn and sorghum research conference. PP: 190-201.

1982. Proceedings, International

 

symposium on‘Grain Quality, ICRISAT. 28-31 Oct. 1981,
Hyderabad, India. p: 294-310.

Chang, S. I., and H. L. Fuller. 1964. Effect of
tannins content of grain sorghum on their feeding value
for growing chicks. Poul. Sci. 43: 30-36.

Cockerham, C. C. 1963. Estimationof genetic variances. P.
53-94. In W. D. Hanson and H. F. Robinson (ed.)
Statistical genetics and plant breeding. Nat. Acad. Sci.
Nat. Res. Coun. Pub. 982.

1980. Random and fixed effect in
plant genetics. Theor. Appl. Genet. 56: 119-131.

 

Corcoran, M. R., T. A. Geisman, and B. 0. Phinney.
1972. Tannins as gibberellin antagonists. Plant
Physiol. 49: 323-330.

Crofts, H. J., L. E. Evans, and P. B. E. McVetty.
1980. Inheritance, characterization and selection of
tannin-free faba beans (Vicia faba 1..) Can. J. of Plt.
Sci. 60(4): 1135- 1140.

Dalrymple, E. J., B. P. Goplen, and R. E. Howarth.
1984. Inheritance of tannin in birdsfoot trefoil. Crop
Sci. 24: 920-923.

Donnelly, E.D., andW.B. Anthony. 1969. Relationship
of tannin, dry matter digetibility and crude protein
in Sericea lespedeza. Crop Sci. 9: 361-362.

Eliase, L. G., D. G. DeFernandez, and R. Bressani. ,.

1979. Possible effects of seed cost polyphenolics on

the nutritional quality of bean protein. J. Food Sci.
44 : 524-527.

Erith, A. C. 1930. The inheritance of color, size,
form of seeds and flower color in Vicia faba L.
Genetica 13: 477-510.

Evans, A. M., and H. E. Gridley. 1979. Prospects for
the improvement of protein and yield in food legumes.
Current Adv. In plant Sci. 32: 1-17.

FAO 1982. FAO Production Year Book. Vol. 36. FAQ,
Rome.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

81

Freudenberg, K., and K. Weinges. "Catechin and Flavanoid
Tannins", In the chemistry of flavanoid compounds, ed.
T. A. Geissman, New York: MacMillan C0. 1962. pp. 212-
214. '

Griffing, B. 1956. Concept of general and specific
combinig ability in relation to dial lel crossing system.
Austral. J. Biol. Sci. 9:463-493.

Handler, P., and R. D. Baker. 1944. Toxicity of orally
administered tannic acid. Science 99: 393-394.

Harris, H., B. and R. E. Burns. 1970. Influence of
tannin content on preharvest seed germination in
sorghum. Agron. J. 62: 835-836.

Hayman, B. I. , 1954 a. The analysis of variance of
diallel tables. Biometric 10:235-244.

: 1954 b. The theory and analyis of
diallel crosses. Genetics 39:789-809.

 

, 1958. Theory and analysis of dial lel

 

crosses. II: Genetics 43: 155-172.

Ivan, M., and J. P. Bowland. 1976. Digestion of
nutrients in small intestine of pig's diet containing
raw and autoclaved faba beans. Can. J. Animal Sci. 59:
129-1330 .

Jacobson, A., and M. R. Corcoran. 1977. Tannins as
giberrellin anatagonists in the synthesis of alfa-
amylase and acid phosphatase by barley seeds. Plant
physiology 59: 129-133.

, and B. J. Hayman, 1953. The analysis of

 

diallel crosses. Maize Genetics Coop.News letter 27:48-54.

Johansen, D. A. 1940. Plant MicroteghiniQue. New York
and London: McGraw Hill Book Co. p.193.

Jones, W. T., and J. W. Lyttleton. 1971. Bloat in
cattle. XXXIV. A survey of legumes forages that do
and do not produce bloat. N. Z. J. Agric. Res. 14:
101-107.

Joslyn, M. A., and Z. Click. 1969. Comparative effect
of gallotannic acid and related phenolics on the growth
of rats. J. Nutr. 98: 119-126.

Kelly, J. D., and F. A.,Bliss. 1975. Heritability
estimates of percentage seed protein and availables
methionine and correlation with yield in dry beans.

Crop Sci. 15: 753-757.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

5CL

51 .

82

. 1975. Quality factors

 

affecting the nutritive value of bean seed proteins.
Crop Sci. 15: 757-760.

Lease, E. J., and J. H. Mitchell. 1940. A study on the
tannins of Lespedeza sericea. S. Car. Expt. Sta. Ann.
Rep. 53: 710.

Leleji, 0. I., M. H. Dickson, L. V. Crowder, and J. B.
Bourke. 1972. Inheritance of crude protein percentage
and its correlation with seed yield in beans, Phaseolus
vulgaris L. Crop Sci. 12: 168-171.

Levin, D. A. 1971. Plant phenolics: An ecological
perspective. Amer. Nat. 105: 157-181.

Lindgren E. 1975. The nutritive value of peas and
field beans. Swedish J. Agric. Res. 5: 159-161.

Ma, Y., and F. A. Bliss. 1978. Seed proteins of common
bean. Crop Sci. 18: 431-437.

. 1978. Tannin content and
inheritance in common beans. Crop Sci. 18: 201-204.

 

Maxson, E. D., L. E. Clark, and L. W. Johanson. 1972.
Factors affecting the tannin content of sorghum grains
as determined by two methods of tannin analysis. Crop

Marquardt, R. R., A. T. Ward, and L. E. Evans. 1978.
Comparative properties of tannin-free and tannin-
containin cultivars of faba beans (Vicia faba LQL
Can. J. Plt. Sci. 58: 753-760.

Marquardt, R. R., A. T. Ward. 1979. Chick performance
as affected by autoclave treatment of tannin-containing

and tannin-free cultivars of faba beans. Can. J. Anim.

Meiners, J. P., and S. C. Litzenberger. 1973. Breeding
for nutritional improvement. P. 131-141. In: M.
Milner(ed.) Nutritional improvement of food legumes by
breeding. (Proceedings of a symposium sponsored by PAG,
Rome, July, 1972). Protein Advisory Group, United
Nations, New York.

Mutschler, M. A., F. A. Bliss, and T. C. Hall. 1980.
Variation in the accumulation of seed storage protein

among genotypes of Phaseolus vulgaris L. Plant
physiology. 65: 627-630:

Osborne, T. B. 1894. The protein of kidney bean. J.
Amer. Chem. Soc. 16: 703-712.

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

83

Parada, R. S., M. L. Saini, and S. K. Arora. 1975.
Inheritance of tannin content in Eu-sorghum. Z.
Pflanzenuchtg. 74: 251-256.

Picard, .L. 1976. Apercu sur 1heredite du caractere
absence de tanins dans les graines de feverole (Vicia
faba L0)0 Anno Amelior0 plante80 26 0 101-1060

Price, M. L., and L. G. Butler. Tannin and nutrition.
Station bulletin 272, March 1980, Dept. Biochem. Agric.
Expt. Sta. Purdue Univ., Indians.

Quinby, J. R., and J. H. Martin. 1954. Sorghum
improvement. Advances in Agronomy. Academic press, New
York and London. 6: 305:359.

Rannenkamp, R. R. 1977. The effect of tannins on
nutritional quality of dry beans Phaseolus vulgaris L.
Ph.D. Thesis Purdue University.

Ringrose,R. C., and C. L. Morgan. 1940. The nutritive
value of lespedeza for poultry feeding. South Carolina
Eptc Stn0 AflflU0 Rep0 53: 91-920

Schaffert, R. E., V. L. Lechtenberg, D. L. Oswalt, J. D.
Axtell,R. C. Pickett, and C. L. Rhykerd. 1974. Effect
of tannin on in vitro dry matter and protein
disappearance in sorghum grain. Crop Sci. 14: 640-643.

Sjodin, J. 1973. Breeding for improved nutritional
quality in field beans (Vicia faba LAL.First FAO/SIDA
seminar on improvement and production of field crops.

Statler, G. D. 1970. Resistance of bean plants to
Fusarium solani f} phaseoli. Plant Dis. Rep. 54 (4):
698.6990

Steel,RsG., and J.H. Torrie. 1980. Principles and
procedures of statistics 2nd ed. McGraw-Hill, New York.

Stoy, V.,and K. Sundin. 1976. Effect of growth
regulating substance in cereal seed germination. Cereal
Res. Comm. 4(2): 157-163.

Sullivan,J.G., and F. A. Bliss. 1983. Expression of
enhanced protein content.in inbred backcross lines of
common bean. J. Amer. Soc. Hort. Sci. 108(5) :787-791.

1983. Recurrent mass selection for
increased seed yield and seed protein percentage in the
common bean (Phaseolus vulgaris L.) using a selection
index. J. Amer. Soc. Hort. Sci. 108(1): 42-46.

 

 

Telek, L. 1983. Tannin determination in dry beans.
Personal communication.

66m

67}

68.

69’.

70.

71.

84

Taylor, J. 1975. A modified air painter ("plantair") for
use in experimental plots. Proc. Bien. Conf. of the Bean
Improv. 000p., Nov. 5-7, East Lansing, Mich. P.36.

Walker, A.F. 1982. Physiological effects of legumes in
the human diet. J. of Plant Foods 4: 5-14.

Wassimi,N., G.L. Hosfield, M.A. Uebersax. 1981. The
effect of seed coat color on soaking characteristics of
dry edible beans. Michigan Dry Bean Digest 5(4): 12.

Watt,B.K., and AsL. Merrill. 1963. Composition of
Foods, Agriculture Handbook No. 8., U.S. Department of
Agriculture.

Wilson, B.J., J.M. Mcnab,and H.Bently. 1972. The
effect of autoclaving and methionine supplemnetation on
the growth of chicks given diets containing field beans
(Vicia faba L.). Brt. Poult. Sci. 13: 67-73.

Woodruff, B. J., R. P. Cantrell, J. D. Axtell, and L. G.
Butler. 1982. Inheritance of tannin quantity in
sorghum. The J. of heredity 73: 214-218.

CHAPTER 3

THE GENETIC CONTROL OF PHYSICO-CHEMICAL SEED
CHARACTERISTICS RELATED TO CULINARY QUALITY
IN DRY BEANS (Phaseolus vulgaris L.)

ABSTRACT

Dry edible beans (Phaseolus vulgaris L“) are important
as a world food crop because they can provide significant
amount of energy and nutrients to consuming populations.
Middle and low income families in many poor countries of the
world rely on beans for calories and protein. In order to
encourage an increased production of dry beans in lesser
developed countries, an economic incentive must be present.
This could best be achieved by the development of high
yielding cultivars that are resistant to diseases and
insects and have acceptable culinary quality. Researchers
have concentrated their efforts on the inheritance of yield
components and factors causing yield reductions. Little
information is available concerning the inheritance of
culinary quality traits in dry beans. The type of gene
action involved and the degree to which the expression of
genes are influenced by environmental variation are
determinant factors in the development of breeding
strategies. Culinary quality is a complex trait and
previous work has shown that several seed characteristics of
a physico-chemical nature are related to culinary quality.
The present study was undertaken to obtain information from

an 8-parent diallel cross on the genetic control of culinary

85

86

quality traits in beans. The parents and 56 F2 and 53
progenies were grown at two locations and data were taken on
nine traits.

Highly significant differences were observed among
genotypes in both generations for the traits. Except for
clumps, soaked bean weightLand soaked bean moisture, there
were no differences between generations. The variances due
to GCA were'highly significant in both generation for most
characters. Significant SCA variances were noted for some
traits but the SCA mean squares were always smaller than the
GCA ones. Variance estimates of SCA effects for the washed
drained weight and texture traits indicated that progeny
resulting from crosses of 'A-30', 'SAN', 'SFH, and 'N-2'
will be better or poorer than the average expected on the
basis of GCA. Genetic relationships among the parents were
studied with the variance-covariance (Vrvwr) graphical
analysis technique and showed that most of the traits were
controlled by dominant or partially dominant genes. Highly
significant correlations were found between GCA effects in
the F2 and F3 generations. The conclusion that genes with
mostly additive effects controlled culinary quality traits
in this population of genotypes was supported by the highly
significant correlations between parental performance and
their respective GCA effects.

Additional Index Words: Diallel analysis, Combining

ability, Texture, Processing traits, texture.

INTRODUCTION

Legumes are an important source of protein, in human
diets and especially in poorer countries of the world. In
addition, legumes can make a significant contribution to the
carbohydrate, mineral, and vitamin fraction of the diet
(Hawtin et a1. 1977). Because of the low sugar, low fat,
high protein and high dietary fiber content of dry beans,
their contribution to Western diets as a replacement for
animal products in terms of dietary fiber and complex
carbohydrates would bring diets in developed countries
closer to recommended dietary goals (Walker 1982).

Legumes are most frequently considered in terms of
their complementary nutritional value and particularly in
relation to amino acids, to cereal diets for people living
under marginal nutritional conditions. Legumes are low in
the sulfur-containing amino acids but high in lysine (except
for groundnut), Aykroyd et al. (1982). Consumers of
food legumes have sensory and palatability requirements for
the dry and cooked seeds. Seed characteristics not meeting
consumer expectation may render a cultivar unacceptable
regardless of how agronomically superior it is. Color,
flavor, soakability, cookability, and degree of tenderness
after cooking are important aspects of legume seeds that
influence consumers preference and these seed characteris-
tics are referred to as culinary quality. The genetic
composition of legume seeds is a major determinant of culi-

nary quality; however, other factors such as environmental

87

88

influences, handling and storage, and processing procedures
play important roles (Ghaderi, et al. 1984).

Dry edible beans (Phaseolus vulgaris LJ are among the

 

most important food legumes in the diets of humans in Latin
America, parts of Africa, and India (Jaffe, 1972). Beans
comprise 32 5 of the world total of edible legumes. Eight
percent of the world production of legumes is in the U.S.A.
and domestic production of beans comprise 82 % of the total
edible legumes grown (FAO, 1982). .

Dry beans are generally soaked and must be cooked prior
to eating to render seed palatable and the proteins
nutritionally available. There are a number of tests that
- allow a food scientist to differentiate the quality of
different samples” However, the utility of these techniques
in bean breeding depends on the extent of genetic variation
for the measured traits. Moreover, the type of gene action
involved and the degree to which the expression of genes are
influenced by environmental variations determine which
breeding strategies would be most successful for trait
improvement. 1

A diallel.crossing system can be uSed to systematically
evaluate a strain's performance. These techniques also
provide an analytical tool to estimate the combining ability
for traits under selection and to evaluated the selection
potential of individual crosses. Superior performing
parents may produce superior performing progeny in

generations subsequent to hybridization.

89

There is relatively meager information concerning the
inheritance of culinary quality traits in dry beans. The
present investigation was undertaken to obtain genetic
information in a population of dry beans and ascertain the
effect of genotype x environmental interactions in trait
expression. The implication of the work will be useful for

bean breeding programs.

LITERATURE REVIEW

Good cooking quality in peas (Pisum sativum L.) was

 

reported to be controlled by two recessive genes, although,
there was some indication of genotype by environment
interaction (Gfeller, 1967). Halstead (1964) studied the
cooking quality of two cultivars of field peas and found
improved culinary characteristics in one cultivar by adding
P and K to the soil, the cooking quality of the other
cultivar remained unchanged. The same author reported that
cooking quality varied among cultivars with ash and phytin P
content (Halstead 1964). Wassimi et al. (1978) reported
that cooking quality of lentils (Le_r_1_s_ culinaris Med.) grown
in a pot experiment was significantly influenced by major
and trace elements, particularly by high levels of Na and K.
A significant genotype and genotype by season interaction
and heritabilities of 37% and 63% were found for whole and
dehulled pigeon peas (Cajanus M L.) seeds, respectively
(Singh, et al. 1973). The addition of one percent sodium
citrate solution, to the soak water greatly reduced the
hardness of whole peas (Muller, 1967). Rizley and Sistrunk
(1979) reported that black eyed peas soaked in 1%
Pyrophosphate or 1% NaHC03 solution at pH 8.5 each produced
softer texture peas. The bicarbonate treatments produced
peas with less desirable color but with softer texture and
better flavor as compared to the other treatments (Rizley
and Sistrunk 1979). Fast cooking legumes were produced by
Bjarkavist (1972) by soaking them in a solution of calcium

91

sequestering salts until they absorbed.an amountLOf water
equal to their dry weight and were dried to a 3 percent
moisture content . A decrease in cooking time for beans

(Phaseolgg vulgaris LJtcorresponded to an increased

 

leaching of organic phosphate at cooking temperature of 90°
C. Storage of cowpeas (Viggg unguiculata Lu) at 85 percent
relative humidity and 29° C for one year caused a decrease
in cooking rate and formation of hard cooking seed. This
was associated with an incomplete breakdown of the middle
lamella (Sefa-Dedeh, et a1. 1979). Bhatty et al. (1983)
reported that location had greater influence on cooking
quality of lentils than did the cultivar: The location and
season effect was clearly demonstrated in the microstructure
of cooked and undercooked lentils (Bhatty, et al., 1984).
The microstructure of the lentils showing the degree of
cellular breakdown closely correlated with the shear force
values (texture) obtained with a Kramer shear press.

High temperatures, high moisture content, and long
storage times contributed to impaired cookability in dry
beans and lima beans (Phaseolus lunatus L.) (Burr, et a1.
1968). A few-fold increase in cooking time was observed in
beans held under conditions that may be often encountered in
storage. For example, one year keeping time of 21° C and a
moisture content below 18% (Burr, et al. 1968). Morris, et
al. (1956) reported that beans held at 13 % moisture at 24°
C deteriorated significantly in texture as well as flavor

after six months while beans held at less than 10 % moisture

92

content maintained their cooking quality for two years
almost as well as control samples stored at minus 10 F.
Hosfield and Uebersax (1980) reported significant
variability in important nutritional and processing traits
among tropical and domestic dry beans. Large variability
for five water-soluble vitamins and nine minerals was
observed between different commercial classes of dry beans
(Augustin, et al., 1981) with between class variability
larger than within class. Increased firmness of dry beans
has been observed with an increased calcium concentration in
soak and brine liquid. A correlation of 0.97 to 0.99 between
Kramer shear press values and calcium concentration was
reported in a storage time study (Uebersax, 1980). Nordstrom
and Sistrunk (1979) showed that been type, blanch treatment,
initial moisture and storage time all had a significant
effect on shear press values and splitting of beans after
cooking. Rockland and Metzler (1967) proposed a method for
quick-cooking;dry beans which consisted of loosening seed
coats by vacuum infiltration of a salt solution, soaking the
beans in the same salt solution and rinsing and drying. The
resulting product cooked in 15 minutes (Rockland and Metzler
1967). Varriano-Marston (1979) reported that sodium salts
affected the mineral content as well as the amount of pectic
substances solubilizing from beans during soaking and
cooking periods. X-ray microanalysis suggested that
mechanisms of ion exchange and chelation were operative in
the dissolution of the intercellular cement and the

subsequent cell separation. Ghaderi, et al. (1984) observed

93

significant cultivar, and cultivar by location differences
for culinary quality traits in navy and pinto classes of dry
beans. Hosfield, et al. (1984) reported that the expression
of soaked and cooked bean traits was strongly influenced by
genotype x season interactions. The estimates of season and
genotype x season variance components were larger than the
genotype component. They also reported that strains differ
in their genetic potential to respond to varying

environments.

MATERIAL AND METHODS

Genetic material. Twenty strains of dry beans were grown

in a replicated nursery during the summer of 1980 at the
Montcalm Research Farm near Stanton, Michigan. These strains
comprised a broad genetic base and differed in seed-coat
color, size, and shape. They varied in growth habit and
plant morphology, reaction to heat and drought stress, seed

filling characteristics, maturity, and yield.

After harvesting and threshing plants from one repli-
cation in the fall,seeds of each strain were evaluated for
protein percentage, tannin content, and soaking characteris-
tics including the percentage of hard seed after 48 hours
soaking time (Wassimi, 1981). Based on the characteristics
of the dry and soaked seed, 8 strains were selected for
genetic analysis of culinary quality characteristics. All
possible F1 hybrids were made in the green house during the
winter of 1980-81. The 56 F1 crosses and 8 parents were
grown in a nursery in East Lansing during the summer of 1981
to produce F2 seed.

Eielg‘gigg procedure. The F2 seed from each cross was harve-
sted in bulk in September of 1981, and a sample from each
parent and each cross was grown in a winter nursery in
Puerto Rico in the winter of 1981. Parental and F3 genera-
tion seed was harvested in March of 1982 bulked and returned
to East Lansing, Michigan. A random sample of 800 seeds from

each cross of the Puerto Rican produced F3 generation and

parental strains and the remnant F2 and 8 parents grown in

94

95

Michigan the previous year were planted in an 11 x 12 recta-
ngular lattice with three replications at the Saginaw Valley
Bean and Sugar Beet Research farm near Saginaw and at East
Lansing, Michigan in May and June of 1982. Seeds were
precision drilled with a tractor mounted air planter
(Taylor, 1975) into two row plots guarded on each side by
the cultivar 'Seafarer'. Rows were 4.9 meter long and 50.8
cm apart. Herbicide and fertilizer applications were made
per seasonal recommendations.
Mature plants were harvested and threshed from a 6
meter row length of each plot in mid-to-late September.
Quality evaluation. Single samples from each plot were eva-
luated for culinary quality traits of dry seeds and at
soaked and cooked stages of processing. Prior to their being
processed, bean moisture was determined and based on that ~
moisture a 100 g equivalent weight of total solids was
determined and weighed. The bean sample was placed in nylon
mesh bags and soaked. The soaking treatment used in this
study was a 2-stage procedure that has been shown to
maximize differences between genotypes for water uptake,
cotyledonary hydration, and the degree of cotyledonary sof-
tening during cooking. The initial soak was for 30 min in
21°C water to facilitate seed-coat softening and expansion.
Immediately after the cold soak,.beans were transferred to
water maintained at 88° C in a stainless-steel kettle for an
additional 30 min. All soaking was done in tap water con-

taining about.50 ppm calcium. The soaking procedure Just

96

described yields an end product that has minimum bean damage
and is similar to beans soaked continuously in the high-
temperature systems common throughout the LLB. canning indu-
stry. After soaking, beans were momentarily cooled under
cold tap water and drained for 2 min. The weight gained
through water imbibition during bean soaking was used to
calculate:

a) Soaked bean weight which was measured in grams and b) the
hydration coefficient which was calculated by dividing the
weight of soaked beans by the fresh weight of the dry beans.
The percentage water content of soaked beans was determined
by the formula (Soaked bean weight (g)- Initial weight of
beans (g)/ Soaked bean weight) x 100.

After weighing, beans were filled into 303 x 406 cans
and covered with boiling brine prepared by adding 142.0 g of
sucrose and 113.4 g of salt to 9.1 kg of tap water
containing 50 ppm calcium. Cans were sealed and processed
in a retort without agitation for 45 min at 116° C. After
thermal processing, cans were uniformly cooled to 38° C
under cold tap water and stored for 2 weeks at room
temperature before evaluation. The storage period after
processing permits canned beans to completely equilibrate
with water in the canning medium.

After the cans were opened, the washed drained weight
of processed beans was determined by decanting the can
contents on a number 8 mesh sieve, rinsing them in 21° 0 tap

water to remove adhering brine, draining for 2 min on the

97

sieve positioned at a 15° angle, and weighing (g). Texture
was determined by using a Kramer Shear Press fitted with a
standard multiblade shear compression cell (Food Technology
Corp., Reston, VA”). .A 100-g sample of washed processed
beans was placed in the compression cell and force was
applied until blades passed through the bean sample» The
water content of the canned beans (final moisture
percentage) was determined from the 100-g texture samples.
These were oven dried at 81° C until the weight remained
constant.

Subjective bean-quality evaluations were made on
contents of all processed cans while beans were drained on
the mesh screen. The degree of clumping (packing in can)
and splitting were scored on a 1-5 scale (5-point range) to
represent the minimum and maximum expression of the traits,
respectively.

Statistical procedures. All data were subjected to an

 

analysis of variance appropriate to a randomized complete
block design. Separate analyses of variance were performed
on F2 and F3 data at each location. Tests of homogeniety of
error variance were made using Bartlett's procedure
described in Steel and Torrie (1980). Analyses of variance
of combined data were computed. In the mathematical model
used for the combined analyses, replications and locations
were considered to be a random sample of the population of
replications and locations. Since the parental lines used
were selected for their quality characteristics, the genetic

effects of the strains were of primary interest.

98

Analyses of combining ability were performed using the
program of the theory for combining ability formulas of
Griffing (1956) model 1 method 1. Since the parental strains
used in the diallel cross were selected,they does not
represent a random sample of the p0pulation of dry beans
genotypes. Therefore, the fixed effect model of Cockerham
(1980) was applied to the data and solved by unweighted
least square procedure. The model was fitted to include
general and specific combining ability effects and is as

follows:

Yidka u + bi + g3 + gk + Sjk + m3 - mg + ij + eijk
al‘g + bi + A3 + Ak + ijDjj + djijk + okkak
+ m3 - mg + rjk + 913k
, A
E(§3)I E [ 1/2bp (Y.j. + Y..j) - M)
' (A3- A.) + 533(033- 533.) + 633(53. - D..)
A . A A A
E(Sjk)= E L1/2b ( v.3k + Y.k3 ) - u - gj - 33]
= djk(Djk " Bj0'B0k ‘1' .500)
Where:

p is the population mean
b1 is the 1“! block effect, 2131-0

p ,
83 is the GCA effect associated with the 3th parent,z;gj-0

1
83k is the SCA effect associated with the cross between the

[0
3th female and the kth male parents, SJKsSKJ, Esjkao
m3 is the maternal effect associated with the 3th parent
. p '
when it is used as a female,‘zm3=o
I

lfijk is the reciprocal effect associated with the cross

between the 31:11 female and ktn male parent, rjka—rkj,

.5335"

eijk is the error term associated with the 3th and Kt“ parent
in the 1th block, Eijk~ND(o,3)

A3 is the sum of the additive effects for genes in a gamete
from.,jth parent.

A. is the mean additive effect.

5:. is the mean dominance effect of the hybrid diallel
population.

033 is the sum of the dominance effects associated with the

33th parent

033. is the mean of the dominance effects associated with.
the 3th parent

55. is the mean of the dominance effects associated with
progeny of the 3th parent

Djk is the sum of the dominance effects for genes from

mating of 3th parent with kth parent
533 is the expected proportion of loci homozygous for the
allele derived from the 3th parent, 533:1/2F

65k is the expected proportion of heterozygous loci,
65k=1-F.

Analyses of covariance between the offspring of each
parent (array) and the nonrecurrent parent minus the variance
of their offspring in each parental array (wr-Vr) were
performed according to Haymanfls (1954a, b) analyses of dial-
lel experiments to ascertain whether or not gene expression

was appropriate to an additive dominance model or epistasis

was a general feature of the system.

1OO

Genetic relationships among the parents were studied
with the variance-covariance (Vr, W1.) graphical analysis
technique developed by Jinks (1954) and Hayman (1954a, b).

The Vr9 Wr graph was drawn for the means of three

replications.

RESULTS

Analysis of variance of individual generation and
location data revealed highly significant differences among
entries. Combined data analyses of variance were computed
after determining that error variance from the separate
tests were homogeneous by Bartlett's procedure (Snedecor
1964). The test for hydration ratio, washed drained weight,
and cooked bean moisture were not homogeneous. Except for
hydration ratio, significant and highly significant differe-
nces were detected for all traits (Table 1). There were
significant differences between generations for clumps,
soaked bean weight and soaked bean moisture. The differences
among entries within generation were highly significant for
all the traits except hydration ratio (Table 1). The) inte-
raction of generation and entry by location were not signi-
ficant for any of the traits investigated. The computation
of general combining ability (GCA) mean squares in each
generation over locations revealed (Table 1) that GCA mean
square was highly significantzfor all the traits in the F2
generation while specific combining ability (SCA) mean
square was significant for clumps, washed drained weight,
and texture (Table 1). Cooked bean moisture and hydration
ratio were the trait for which the maternal and non-maternal
reciprocal mean squares were significant, respectively. In
the F3 generation, the GCA mean squares were significant or
highly significant for all traits except hydration ratio and

the SCA mean squares were highly significant for clumps,

102

splits, washed drained weight, and texture (Table 1). Only
washed drained weight showed significant maternal reciprocal
effect mean squares and none of the traits showed any signi-
ficance for the non-maternal reciprocals. .

The traits splits, hydration ratio, and texture showed
significant GCA x location and SCA x location mean squares
in the F2 generation (Table 1) but in the F3 generation only
the GCA x location interaction was significant for splits
and texture. The interaction of location with maternal
effects was significant for cooked bean moisture and the
interaction of non-maternal reciprocal effects was signifi-
cant:for clumps, hydration ratio,«cooked bean moisture and
texture in the F2 generation, but in the F3 generation no
significant reciprocal difference were detected (Table 1).
The GCA effects and their interactions with environments
were computed for the eight parents in each generation
(Tables 2 and 3) which revealed a significant GCA variance
for some parent for each traits. Some inconsistencies in the
ranking of strains between generations were observed for GCA
effects. Texture and washed drained weight were the traits
that showed the most consistent GCA effects in both genera-
tions. For example, 15-R, A-30, BTS, SAN, and N-2
transmitted significantly large and negative effects to
their F2 and F3 progeny for texture, but they transmitted
significantly large positive effects to their progenies for
washed drained weight (Tables 2 and 3). The interaction of

GCA effects by location showed no consistency from genera-

103

tion to generation for most of the traits except texture.
Parental strains B-2, FF, A-30, BTS, SAN in the F2 and FF:
A-30. BTS. SAN. SF in the F3 generation had either signifi-
cantly positive or negative GCA interaction effects with
locations.

The estimates of variance components for GCA, SCA,
maternal and non-maternal reciprocal effects showed (Table
4) that variance of GCA was significant for all the traits
in the F2 and all but hydration ratio in the F3 generation.
The SCA component of variance was significant for clumps,
splits, hydration ratio, washed drained weight, and texture
in the F2 generation. Variance component for these traits in
the F3 generation were significant except for hydration
ratio. The interactions between combining ability and
maternal and non-maternal reciprocal effects variance
components and locations generally followed similar trends
as the main effects (Table 4). '

The correlation of general combining ability effects
between F2 and F3 generation for the measured traits were
large and highly significant (Table 6). Strains A-30, BTS,
SAN, SF, and N-2 had large and highly significant SCA effect
variances in the F2 for texture but in the F3 strains B-2,
A-30, SAN, and N-2 had large and significant SCA effect
variances (Table 5).

Separate combining ability analyses for F2 and F3
generations at both locations showed that the GCA mean
squares were highly significant for most of the traits

(Tables 1 and 2, appendix 8). Similarly, the SCA mean

104

square estimates were either significant or highly signifi-
cant for all of the traits except clumps and hydration ratio
at both locations and cooked bean moisture at East Lansing
in the F2 generation. In the F3 generation the SCA mean
square for splits, texture, and washed drained weight at
East Lansing, and texture, hydration ratio, washed drained
weight, and washed drained ratio at Saginaw were either
significant or highly significant. Washed drained weight was
the only trait that showed significant mean squares for
maternal and non-maternal reciprocal effects across loca-
tions in both generations (Tables 1 and 2 appendix B). The
magnitude of GCA mean squares was larger than SCA.in most
cases. The estimates of variance components of GCA and SCA
in the F2 and F3 generation at each location are significant
with SCA estimates being larger in magnitude than GCA
(Tables 3 and 4 Appendix B). The estimates of variance
components of maternal and non-maternal reciprocal effects
were generally zero. In cases where these estimates were
positive, they were not significant except for washed
drained weight (Tables 3 and 4 Appendix B).

The adequacy of the additive dominance model was
tested.by the analysis of variance for covariance between
the parents and their progenies in each array minus the
variance 03: each array (Wr-Vr) for each trait according to
Hayman (1954a,b). The mean squares for the traits measured

on beans in the F2 and F3 generations grown at Saginaw were

nonsignificant but the entries grown at East Lansing showed

105

significant mean squares for the traits soaked bean moisture
in the F3 and texture in the F2 and F5 generation (Table 5
Appendix B). The non-significant differences of mean squares
of ("r-Vr) for arrays is indicative of the adequacy of
additive dominance model for describing the variation in
these traits. However, the significant mean squares for some
of the traits indicated that the additive-dominance model is
not appropriate to describe the variation in the expression
of these traits. This suggests that several assumptions
underlying the diallel analysis are invalid (Hayman 1954a,
b).

Correlation between pair of soaking traits in the F2 at
East Lansing were positive and significant (Table 7).
However, when the soaking traits were correlated with washed
drained ratio, they were significant but negative. No other
trait showed any significant correlation except texture with
soaked bean weight (r=-0.29*). A similar trend was observed
in the F5 generation at East Lansing except that clumps and
washed drained weight were significantly correlated with
soaked bean weight (r=0m28* and r=-0.28* , respectively).
The clumping trait was significantly and negatively corre-
lated with splits (r=-0.44** and r--0.38**) and texture (r:-
0.42** and r=-0.46) in the F2 and F3 generations, respecti-
vely. A negative correlation (rs-0.50") between clumps and
washed drained weight was observed in the F3 generation at
East Lansing (Table 7). The trend and the sign of the corre-
lation coefficient was similar in both generations but the

levels of significance of correlation varied from generation

106

to generation. Texture was negatively correlated with cooked
bean moisture. Similar results were obtained in data from
Saginaw with respect to water uptake traits in both soaked
and cooked beans (Table 8). A negative correlation of
texture with clumps, splits, washed drained weight, and
cooked bean moisture was observed. Significant positive
correlation coefficients were obtained for the measured
traits between F2 and F3 generation.

Graphical analyses of traits identified parents with
dominant and recessive factors influencing trait expression.
The Vr, Wr graphs presented in this study are for clumps,
splits, washed drained weight and texture, traits of primary
importance to consumers and processors. Figure 1 showed
that the regression coefficient (ba0.90:_ 0.22**) for clumps
in the F2 generation after parent 3 which exhibited high
variance was eliminated, was significantly different from
zero but not significantly different from unity . The regre-
ssion line intersected the WP axis Slightly below the origin
(as-0.00110.045) but not significantly different from zero,
.indicating complete dominance of genes controlling the
trait. Sanilac appeared to contain a preponderance of rece-
ssive genes for clumps in the'FZ generation at E. Lansing
while the remaining parents carry mostly dominant genes for
this trait. At Saginaw in the F2 generation (Fig- 3) the
regression coefficient (bx 1.2410.37*) was significantly
different from zero but not significantly different from

unity. The intercept of the Wr axis was below the origin

107

(as-0.1610.11) but not significantly different from zero,
suggesting that complete dominance of genes governs this
trait. Here again Sanilac contained mostly recessive genes,
while FF'16-15-1-CM-M-M.and A-30 appeared to contain overdo-
minant genes, and BTS had mostly dominant genes for this
trait. The remaining parents were midway along the regres-

sion line indicating equal proportion of dominant and reces-

sive genes. Figure 2 showed regression of WP on V, for
clumps in the F3 at E. Lansing. This indicated that the
regression coefficient (ba0.5310.18) was significantly
different from zero but not significantly different from
unity after elimination of parents 1 and 4 which showed
relatively high covariances. The regression line intersected
the hr axis above the point of origin (a=0.031 0.04) but it
is not significantly different from zero indicating complete
dominance. The distribution of array points representing the
parents changed. Parents BTS and N-2 appeared to carry
predominantly recessive genes while parents FF and 15-R seem
to have dominant genes. Parents 6 and 7 were also near the
origin. The graph of the F5 generation at Saginaw for clumps
(Fig. 4) after elimination of strain FF and SAN indicated
that the regression coefficient is significantly different
from zero but not significantly different form unity
(b=0.9910.19**). The intercept of W, axis was above the
origin (a=0.04;0.05) but not significantly different form

zero suggesting full dominance. The distribution of points

108

representing parents showed that N-2 is carrying predominan-
tly recessive genes. Strains B-2, 15-R, A-30, and BTS
contained dominant gene, and strain N-2 had equal proportion
of dominant and recessive genes. Figures 1, 2, 3, and 4
clearly revealed that the pattern of genetic influence for
the clumping trait can not be easily be interpreted because
of either genie interaction or errors in the way the trait
was measured.

The Vrv Wr graph of splits in the F2 generation at
East Lansing is shown in figure 5. The regression coeffi-
cient:(b=1.00:.0.12**) was highly significantly different
from zero but not significantly different from unity. The
intercept was (a=0.06_+_0.06) not significantly different from
zero which indicated that complete dominance of genes con-
trolled splits in these genotypes. The distribution of array
points indicated that strain 15-R seemed to carry mostly
recessive genes, strains BTS and SAN predominantly dominant,
strains A-30 and N-2 had an intermediate balance of dominant
and recessive genes for splits while parent SF shows overdo-
minance for this trait. A similar picture appeared for
splits in the F3 generation at East Lansing (Fig. 6). The
regression coefficient (ba1.07;0.13**) was significantly
different from zero but not significantly different from
unity. The intersect on the WP axis was through the origin
indicating full dominance for splits in the F3. The distri-
bution of array points is similar to the F2 except that
strain SF showed interallelic interaction and strain BTS

complete dominance. Strain FF was an outlier and A-30, SAN,

109

and N-2 had an intermediate balance of dominant and
recessive genes. Strains B-2 and 15-R carried a
preponderance of recessive genes. In the F2 generation at
Saginaw the Vf,Wr graph showed that the regression coeffi-
cient (b=0.94; 0.13**) was significantly different from zero
but not significantly different from unity (Fig. 7). The
intercept 015 wt. axis was a=0.24_1_-_0.06* which was significan-
tly above the origin. This response suggested that the gene
system controlling this trait was partially dominant. Since
the array points representing BTS and 15-R- are closest to
the point of intersection between the regression line and
limiting parabola, it indicated that BTS and 15-R carried a
preponderance of dominant genes. B-2 and A-30 carried mos-
tly recessive genes. Strains SAN and SF seemed to carry a
balance of dominant and recessive genes while FF and N-2
seemed to carry dominant genes in excess of recessive genes.
The over all picture of splits in the F3 generation at
Saginaw (Fig. 8) was the same. The regression coefficient
was significantly different from zero but not significantly
different from unity (b=0.94;0.12**) and the intercept of
Wr axis was significantly above the origin (a=0.2310.05*)
indicating that genes with partial dominance effect
controlled splits in the F3. The distribution of array
points indicated that recessive genes seemed to prevail in
strains B-2 and SF while strain N-2 carried mostly dominant
genes. The remaining parents had a balance of both dominant

and recessive genes.

110

Figure 9 represented the Vr,wr graph for washed drained

weight in the F2 generation at East Lansing. The regression
coefficient (b=0.95 i 0.15**) was significantly different
from zero but not significantly different from unity. The
intercept of the regression line was not significantly
different from origin (as-0.021 0.03) indicating that
complete dominance of genes controlled thisitraitn .Array
point distribution showed that strains B-2 and SAN carried
predominantly recessive genes, strain 15-R carried a prepon-
derance of dominant genes, while N-2 showed a balance of
dominant and recessive genes. Strains FF, A-30, BTS, and SF
displayed interallelic.interaction. The graph of washed
drained weight in the F3 generation at East Lansing was the
same as that of the F2 generation with respect to overall.
dominance (Fig. 10) but the distribution of array points
representing the parents was different. The slope of the
regression line was significantly different from zero but
not significantly different from unity (b=1.321 0.27**).
The intersect of the regression was not significantly
different from zero (as -&L0210.02) suggesting full
dominance of genes for this trait. Parents FF and BTS
carried mostly dominant genes but 15-R and SF carried mostly
recessive genes. Parents A-30 and N-2 had a balance of
dominant and recessive genes, while parents B-2 and SAN
carried an excess of dominant genes for this trait. The
graph of washed drained weight in F2 in Saginaw "33 the same
as the F3 in E. Lansing (Fig. 11). In the F3 generation at
Saginaw the vr, wr graph for washed drained weight (Fig. 12)

111

showed that the slope of the regression line was
significantly different from zero but not significantly
different from unity (b=1.3010.26*). The intersect of WI.
axis was slightly below the origin but not significantly
different from zero indicating that the gene system
controlling this traitiexhibited complete dominance. The
distribution of array points representing the parents indi-
cated that parents B-2, FF, 15-R, and BTS carried a prepon-
derance of dominant genes, and parents A-30, SAN, and N-2
had mostly recessive genes. Parent SF had a balance of
dominant and recessive genes for this trait.

The Vr, Wr graph of texture of the F2 generation at
East Lansing indicated that the regresSion coefficient
(b=0.41-+CL21*) was significantly different from one but
not significantly different from zero (Fig. 13). This sug-
gested that the assumption of no genic interaction was
invalid in this case. This disturbance in the anticipated
slope was caused by parents SF and N-2 which were
interacting in this case in opposite direction. EliminatiOn
of these parents and recalculation of the regression of WP

on Vr (Fig. 14 ) showed that the slope of the regression
line became significantly different from zero but not signi-

ficantly different from from unity (b=1.08 _+_ 0.06**). The
intercept of the ”r axis was significantly different from
the origin (a=8.18i1.03**) indicating that partially
dominant genes controlled this trait. This changed the dis-

tribution of array points along the regression line. Strain

112

15-R appeared to carry a preponderance of recessive genes,
and FF and SAN seem to carried mostly dominant genes. B-2,
A-30 and BTS exhibited a balance of dominant and recessive
genes. Figure 15 showed the Vr'wr graph for texture in the
F3 generation at East Lansing. The regression coefficient
(b=0.85;0.19** ) was significantly different from zero but
not significantly different from unity. The intersect on
the WI. axis was significantly above the origin
(a=18.6617.69**) suggesting that a partially dominance
system of genes controlled this trait. Array points distri-
bution indicated that strains FF, 15-R, BTS, and SF carried
preponderance of dominant genes and strains SAN, and N-2 a
preponderance of recessive genes. Parents B-2 and A-30 had
an intermediate balance of dominant and recessive genes.
The slope of the regression line for texture in the F2
generation at Saginaw (Fig. 16) was significantly different
from zero but not significantly different from unity
(b=0.84:0.10**). The intercept of the WI. axis was signifi-
cantly different from zero (a=15.0513.0**). This suggested
that a partially dominant system of genes controlled texture
in the F2. The distribution of array points representing
the parents along the regression line indicated that SAN and
N-2 carried a preponderance of recessive genes while BTS and
SF carried mostly dominant genes for this trait. The remai-
ning parents had a balance of both dominant and recessive
genes for texture. Figure 17 showed the Vr, Wr graph for
texture in the F5 generation at Saginaw, and revealed that

the regression coefficient was significantly different from

113

zero but not significantly different from unity (b=1&)3;
0.16**). The intersect of Wr axis was significantly above
the origin indicating that a partial dominance of genes
controlled this trait. The distribution of array points
representing the parents indicated that A-30, SAN, and N-2
carried.a preponderance of recessive gene and 15-R, BTS,
and SF carried a preponderance of dominant genes while B-2

and FF had a balance of both dominant and recessive genes.

114

 

 

Table 1. Mean square from analyses of varaince and combining
ability analyses and their interaction,wiht location
for 9 culinary quality traits measure on parents and
F2 and F progeny of an 8x8 diallel cross in
dry bean; and grown at E. Lansing and Saginaw,
Michigan in 1982.

Source Soaked Soaked Hydra-

of beans beans tion

variation d.f. Clumps Splits weight moisture ratio
(scale) (scale) (s) (%)

Toatal 768

Location (L) 1 2.9 42.7 2094 79.7 1.28

Reps/L 4 9.6 3.8 4063 153 0.37

Entry 127 1.24** 3.1** 229** 9.2* 0.75

Generation (G) 1 1.09* 0.03 948* 30.0** 0.99

Entry/G 126 1.24** 3.07** 223** 9.0** 0.75

G x L 1 0.001 0.16 340 11.2 0.98

Among F2

GCA 7 4.91** 24.1** 951** 41.3** 2.31**
SCA 28 1.10** 0.93 185 7.3 1.11
Maternal (M) 7 0.50 0.42 62 2.2 0.87
Nonmaternal
reciprocal (NMR) 21 0.45 0.54 76 3.2 1.92**
Among F3
GCA 7 6.07** 18.8** 1015** 38. 6* 0.012
SCA 28 0.99** 1.17** 164 6.4 0.01
M 7 ' 0.32 0.23 111 4.6 0.002
NMR 21 0.26 O;56 87 3.5 0.004
Entry x Loaction 127 0.50 0.60 140 5.8 0.71
F x L
2GCA 7 0.48 1.29** 351 15.0 1.50*
SCA 28 0.58 0.99** 206 8.5 1.17*
M 7 0.34 0.25 62 2.2 0.80
NMR 21 0.67* 0.31 78 3.1 1.89*

F ,

BGCAL 7 0.77 1.17** 170 7.1 0.007“
SCA 28 0.45 0.44 132 5.4 0.005
M 7 0.33 0.26 79 3.3 0.004
NMR 21 0.32 0.39 84 3.8 0.003

Error 508 0.41 0.42 119 4.8 0.72

Experiment mean 2.14 2.88 229 56.2 1.86

C.V. (%) 29.9 22.5 4.8 3.9 45.5

 

*,** = Significant at 5% and 1% probability level, respectiely.

115

 

 

Table 1. (cont'd.).
Source Washed Washed Cooked
of drained drained beans Texture
variation d.f. weight ratio moisture (Kg/100 g)
(8) (%)
Toatal 768
Location (L) 1 80.7 0.11 0.5 2506
Reps/L 4 1720 0.07 8.1 196
Entry .127 331** 0.02* 2.7** 245**
Generation (G) 1 483 0.006 2.8 2.1
Entry/G 126 329** 0.02** 2.7** 247**
G x L 1 54 0.005 1.5 24
Among F2
GCA 7 1207** 0.10** 8.4** 1189**
SCA 28 245* 0.008 2.3 140**
Maternal (M) 7 16 0.003 5.0** 25
Nonmaternal
reciprocal (NMR) 21 91 0.004 2.4 23
Among F3
GCA 7 1327** O.10** 7.1** 1898**
SCA 28 332** 0.009 1.6 137**
M 7 325* 0.004 0.9 31
NMR 21 159 0.005 1.3 41
Entry x Loaction 127 137 0.01 1.7 39
F2XL '
GCA 7 186 0.01 1.1 147**
SCA 28 94 0.009 123 38*
M 7 73 0.004 3.6** 21
NMR 21 95 0.003 2.6* 31*
FBXL
GCA 7 248 0.006 2.9 156**
SCA 28 107 0.005 1.0 22
M 7 271 0.002 0.6 28
NMR 21 205 0.004 1.7 11
Error 508 95 0.005 1.7 17
Experiment mean 311 1.36 64.5 44.8
C.V. (%) 3.1 5.2 2.0 9.2

 

*,** 3 Significant at 5% and 1% probability level, respectiely.

116

Table 2. General combining ability effects averaged over

location and their interaction with location for

culinary quality traits measured on Fg progeny of an
e

8-parent diallel cross in dry edible ans in 1982.

 

 

Soak bean Soak bean

Clump Split weight moisture
Parent (scale) (scale) (g) (%)
Among F2
B-2 -0.109 -0.104 1.672* 0.351
FF 0.203* -0.677* 2.649* 0.461*
15-R 00286* -00823* -1 0066 -0027?
A-Bo -00276* 00208.”, -70019 ‘1 0470
BTS 0.161* 0.313* 1.094 0.263
SAN 0.161* 0.156* 1.738* 0.415
N-2 -0.172* 0.552* -1.027 -0.158
57 0.131 0.144 1.636 0.452
55 0.102 0.112 1.707 0.348
F2 x Location
B-2 x L -0.052 0.104 0.299 3 0.027
15-R x L 0.031 0.010 -2.774* -0.594*
A-BO X L 00010 00104 “10084 -00261
BTS X L ’00094 -00104 20321”. 00465*
SAN x L -0.052 -0.114 0.084 0.018
SF x L -0.031 -0.104 2.853* 0.594
N-2 x L 0.114 -0.073 0.056 0.060
57 0.116 0.116 2.037 0.409
SD 0.092 0.094 1.573 0.316

 

S; and S’ a standard error of means and standard error
0

a dif erence between two means,respectively.

* = Significant at 5% probability level.

Table 2.

(cont'd.).

117

 

Wash drain Wash

cooked

 

 

Hydration weight drain bean moist. Texture
Parent ratio (g) ratio (%) Kg/100g
Among F2
FF “00089 -00458 ’00017* -00455* 40199*
A-30 -0.164* 5.628* 0.071* 0.011 -0.478
BTS 0.287* ‘-2.121 -0.017* 0.129 -1.846*
SAN 00117 40420... 00006 00383" -60077*
SF 0.115 -5.198* -0.036* -0.086 4.504*
N-2 -0.077 -0.166 0.004 0.365* -1.690*
3? 0.158 .195 0.014 270 175
S5 0.121 .689 0.014 185 904
F2 x Location
B-2 x L 0.079 0.802 0.002 -0.021 0.815*
FF x L 0.078 -0.219 0.008 -0.098 -1.651*
A-3O x L 0.093 1.591 0.016* 0.071 -2.261*
SAN X L -0008? -1 0093 -00005 “00148 00914*
SF x L -0.076 1.390 -0.012 0.033 0.595
S? 0 158 1.824 0.013 0.243 0.772
85 0 122 1.408 0.01 0.188 0.966
S' and S‘ = standard error of mean and standard error
otia dif erence between two mean, respectively.

* a Significant at 5% probability level.

118

Table 3. General combining ability effects averaged over
location and their interaction with location for
culinary quality traits measured on F progeny of an
8-parent diallel cross in dry edible Beans in 1982.

 

 

Soak bean Soak bean

Clump Split weight moisture
Parent (scale) (scale) (g) (%)
Among F3
B-2 -0.018 -0.044 0.669 0.156
FF -0.080 -0.523 1.289 0.127
15“R 00232“. “00773 1 0599 00265
A-30 -0.258* 0.247* -7.859* -1.542*
BTS 0.242* 0.070 1.333 0.278
SAN 0.357* 0.122 1.095 0.283
SF -0.331* 0.372* 2.126 0.413
N-2 -0.143 0.529* -0.253 0.022
3? 0.132 0.145 2.214 0.453
S5 0.102 0.112 1.707 0.348
F3 x Location
B-2 x L 0.070 0.039 1.072 0.189
FF x L 0.049 0.206* -2.335* -0.497*
15“R x L 00049 00018 '1 0703 “00338
BTS x L -0.190* -0.075 0.577 0.129
SAN x L -0.013 -0.023 0.151 0.019
SF x L 0.070 -0.169* 0.926 0.235
N-2 x L 0.028 -0.055 0.018 0.041
st 0.119 0.121 2.044 0.409
S5 0.092 0.044 1.573 0.316

 

Sf and S5 = standard error and standard error of difference
respectively. * 2 Significant at 5% probability level.

119

Table 3.(cont%L).

 

 

 

 

Wash drain Wash cooked

Hydration weight drain bean moist. Texture
Parent ratio (g) ratio (%) Kg/100g
Among F3
B-2 -0.002 -2.930* -0.017* -0.302* 2.291*
FF -0.016 -2.478* -0.015* -O.443* 5.417*
15-R 0.007 1.251 -0.003 0.012 -0.820
A-30 -0007? 50164* 00072* -00015 -0053?
BTS 0.027 -3.119* -0.022* 0.137 -2.217*
SAN 0.034 2.775* 0.003 0.344* -7.362*
SF 00019 -40584* -00033* -00039 60005*
N-2 0.007 3.921* 0.016* 0.306* -2.776*
S? 0.136 2.196 0.014 0.242 1.177
S5 0.121 1.689 0.014 0.185 0.904
F3 x Location
8-2 X L -000004 00746 -00003 00206 -00445
FF x L -0.011 0.011 0.015* 0.054 -1.444*
15-R x L .-0.012 -2.673* -0.0008 -0.206 0.412
A-30 x L 0.012 2.698* 0.005 0.069 -1.883*
BTS x L -0.001 1.006 -0.00001 0.086 1.270*
SAN x L 0.005 0.197 -0.0001 -0.316* 1.114*
SF x L 0.007 -1.208 -0.012 -0.012 1.497*
N-Z X L 000009 -0077? -00004 00119 -0052}
Sf 0.158 1.823 0.013 0.242 0.775
SB 0.122 1.408 0.01 0.188 0.966
S? and $5 a standard error and standard error of—HTTTerence

respectively. * sSignificant at 5% probability level.

120

.0000> 00000000 00000 au0> 00 00000000 I III

 

 

 

III III III 0.000 0.0n III III III III III «:2
00.00-.0 III III 00.00 n0.n III III III III III 3
~0.~0nn.0 III III 0.000 00.0 III 00.000~.0 00.00000.0 00.00 00.0 000.00000.0 000
nn.0000.~ n0.00n0.0 III 0n.~0 0~.n III 90.00n0.0 «0.0 050.0 00.00 00.0 000.00000.0 n 000
. u 0 00000000
00.n000.0 n~.000n.0 III 00.000 «0.0 00.00 on.0 III III no.0usa0.0 0x:
00.0000.0 00.0000.0 III III 000.000000.0 I III III III x
on.n0n0.0 III 0000.00n000.0 III 000.00 «0.0 05.00n~.0 00.00000.0u 00.00 00.0 nn0.00~no.0 00m
00.0000.“ III «0000.000000.0 00.00 00.0 000.00 000.0 n0.000~.0 sc.n «00.0 00.00 «0.0 n00.00000.0 N 000

. n 0 00000000
mn.~0«~.0 III III 0~.00nn.n III III III III III 0::
III III III 00.00n0.0 III III III III III a
n0.00~n.00 III III 0.000 0.0m III «n.0000.0 00.0 000.n 00.00 00.0 n00.00000.0 0cm
~n.000n.00 00.0000.0 n000.0«0000.0 n.o 0 0.00 III o0.000n.0 00.0 000.0 00.00 00.0 0N0.000n0.0 n «“0
b 0000
III 2 .22 .0 III III 23.3 «2 .0 III III III III 300 088009:
0000000000:
III «00.00n0.0 III III 000.00 «00.0 . III III III III “:0 0000000:
00.0000.00 00.0000.0 III «.000 0.50 «n0.00 ~00.0 0n.000~.0 00.0 000.5 000.00n00.0 000.00 00.0 000
00.0000.00 00.0000.0 n000.000000.0 0.n 0 0.00 000.00 000.0 00.000n.0 00.0 000.0 000.0uncu.0 «00.00000.0 0 0w0
u 0000
I~u000 Auv ~00 an. an» Hmwumww a000o00 00000000:
\00000 0:» 00000000 00000 00000: 00000 00000000 00000: 000000 000000 00
0000009 00000000 0000000 0000000 0000000»: 000a 0000 0000000000
000003 000003 001000 000000
.0000 00 0000000: .3000000 000 0000000 0003 00 0300. 00000 000000 500 00 00000 0000000 000000I. 00 00 >000000
an 000 «n .0000000 00 00000000 000000 5000000 50000000 0 000 000000000 0003 00000000000 00000 000 0000000000 00000000 00 000000000 .0 0000a

.>00>00000000 .0000000 5000000 000000000 0000000 000 00000 N 0003000 0000000000 00 00000 00000000 I 0m

.h00>00000000 .000000 >000000 000000000 0000000 000 0000 00 00000 00000000 I mm

 

121

 

0.0.0 0.0 0.0 0.0 mm

0.0 0.0 0.0 0.0 mm

.0a.00 0.000 0000.00 0.00 0In.e~ 0.00m 0.0.00 0.0a 0.2
..0.00 0.000 «I00.c0 0.00 00.0 0.000 «I~.n0 .0.on um
00.: «.000 IInc.nH «.00 00.00 0.000 0I0.¢0 0.00 z<m
00.00 «.000 00.0 0.00 000.0 0.000 I0¢.n 0.00 «an
00.0 0.000 0.00.00 0.00 000.00 0.000 .«ne.o 0.00 cnI<
.mn.00 0.000 00.0 0.00 0.: o.0Hn 00.0 0.00 0In.
00.0 0.00m 0.0 0.00 0.0 0.000 0.0 0.00 00
0.0 0.000 00.0 0.00 .0.00 0.000 0.0 0.00 ~In
0000 a» 0000 0» 0000 a» ‘ 0000 I 0» acouam

 

 

 

Maw 00u003 0000000 000003 Ad000x 00000.0Wv 0000009 any 000003 0000000 000003 ~0000x00000IuMW0000009

 

 

0000000000

.~000 00 3000000 000 0000000 000m 00 00000
n N 0
00000 0000000 000000I0 00 0000 a000000 m 000 a 00000 0003 000000000 000000000 0000 00w 0 «0 a «00000000
0000000 000000000 00000000 00 00000000> 000 000003 0000000 00000: 000 0000000 000 .A 00 00000» 00000000 .0 0000s

122

 

 

 

 

 

 

Table 6. Correlation coefficient (r) for general combining
ability effects between F and F3 progeny of an 8x8
diallel cross and parentazl values vs. GCA effects
in the two generations of dry edible beans grown at
East Lansng and Saginaw, Michigan in 1982.

Correlation coefficient (rfz
Parental value vs.
GCA effects
GCA effects between
Traits F2 and F3 F2 F3
Soaked beans
Weight (g) O.92** O.88** 0.72*
Moisture (%) 0.92** O.96** O.93**
Hydration ratio 0.72* 0.62 0.59
Cooked beans
Clumps(scale) O.83** 0.77* 0.82**
Splits(scale) O.98** 0.98** O.87**
Moisture (%) O.99** 0.76* 0.72*
Texture(kg force/100g) O.99** 0.92** O.94**
Washed drained weight (g) O.86** O.90** O.91**
Washed drained ratio O.98** 0.82** 0.76*

 

*,** afi§ignificant at§% and 1% probability level,

respectively. 2 a Data averaged over 2 locations.

123

.5~o>fiuomqmmc .Ho>m~ >u-flnmnocq me com an on» no ucmoamacwam mom
vm.0 now 00.0 con» copoocm no on #0300 acmdoawmmoo Ho m3~m> musfiomn<

 

 

 

 

*
.3oc map a“ moon» on noncommecoo cezaoo on» :H cones: muamca u N
::: Nv.0 mn.0 0n.0: 55.0 «9.0: Nw.0 ww.0 nw.0: .0
No.0: ::: no.0 vn.0: «0.0: 50.0 00.0: 55.0: «5.0: .0
nw.0 NN.0 ::: 00.0: 0v.0 00.0: 5N.0: 59.0: 0N.0: .5
00.0: 50.0: 55.0: ::: 50.0:. Nv.0: 05.0 00.0: 5e.0 .0
er.0 v0.0: nm.0 00.0: ::: v¢.0: 00.0 00.0 no.0 .0
05.0: no.0 v0.0 0v.0: 0m.0: ::: v0.0 0w.0 0N.0 .v
e0.0 00.0: 5P.0 Pe.0 5F.0 50.0 ::: 05.0 00.0 .n
«v.0 00.0: ne.0 Fe.0 00.0 no.0: m0.0 ::: m0.0 .N
nm.0: 55.0: FN.0: 0N.0: 09.0 we.0 00.0 «0.0 ::: .w
80 80 A50 30 RV . 20 30 30 AS
monumfioa owpmc usmwm: muzuxma muwdam mqezao .umfioe oflumn unmamz
momma oocqmnn occamco momma cow» momma
vmxooo cosmos cocoa: omxmom :00051 ooxmom
*Acv acmaoawmooo coHumHmucoo nmufimcfi
000.. 5 20033: 050:3 .0 no mczocm mmoco H333 0.0
cm Ho mammona Aoamcmficu cmzoav 0 com Amamcmacp uoaozv
mm 00 Ho madman huaamac acacaaso no magma cmmspon
aficmcoaumamc on» mcwumoaocw Acv powwowuumoo coapmdmecoo .5 magma

124

.>~m>waomammo .~m>m~ hpwafiamaooo up com an on» um uomoHHHoMHm mom

vm.0 com 0N.0 omou umummaw co on Hmocm nomaoamumoo no mo~m> muoaoma<

 

 

 

 

*
.3oo mop oH mmoop on moooammoooo oeoaoo moo oH omaeoo muamoa u N
::: 00.0 0N.0 55.0: mw.0 rm.0 mp.0 5v.0 0w.0 .m
F5.0 ::: 0F.0 v0.0: PN.0 mm.0: 00.0: 00.0: 09.0: .0
09.0 0N.0 ::: 0r.0: «0.0 00.0 0N.0: 0n.0: NN.0 .5
N5.0: 05.0: 0N.0: ::: Mr.0: 50.0: FN.0: 0v.0: 0P.0: .0
mm.0 00.0 mn.0 00.0: ::: mm.0: 0m.0: No.0 Pv.0: .0
No.0 we.0 0N.0: 09.0: N¢.0: ::: 00.0 0n.0 00.0 .v
nm.0 00.0: Nn.0: mr.0: 0F.0: 0P.0: ::: N0.0 N5.0 .n
00.0 0n.0: No.0 00.0: 50.0: 0F.0: nv.0 ::: 00.0 .N
05.0: 5v.0: v0.0 v9.0: mn.0 0N.0: 0v.0 00.0 ::: .P
Amv A00 A50 A00 Amv. Aev Amv Amy A90
monumfioa oHomo unwams moouxme maaqom moan~0 .umHoa oHumo “swam:
momma omoamao omoqmoo momma ooHu momma
omxooo omommz omommz omxmom :mo053 cmxmom
*Aeo memaoamomoo eoeemameeoa unnameb
.000? o“ ommaoowz .3moH0m0 um zoom mmooo Hmaamao 0x0
om mo zommooq Amamomqop omzoqv .0 com Amamomaoo cmooov
Wm 00 Ho muamou zuaamoc ammoaaoo Ho moamo ommsuma
ofiomoo«omamu moo moaamofioow on oomwofimumoo oowomamoooo .0 mHame

125

Table 9. Means of texture and processing traits for two
generations of an 8-parent diallel cross grown in
East Lansing, MI (1982).

 

 

 

 

 

Generations

F2 F3

Traits Traits
Parents t -
and CIumps Splits Texture Clumps Splits Texture
crosses (scale) (scale) (Kg/100g) (scale) (scale) (Kf/100g)
Earents
B-2 (1) 108 203 5301 200 207 5605
FF (2) 2.7 1.0 52.4 2.3 1.3 56.0
15-R (3) 203 1.0 4907 207 107 4609
A-30 (4) 1.0 4.0 38.6 1.7 3.7 34.7
BTS (5) 2.7 3.0 39.0 3.0 3.0 39.5
SAN (6) 3.3 2.3 37.6 3.0 3.7 30.4
SF (7) 1.7 3.3 50.8 1.7 3.7 60.1
N'Z (8) 200 307 3303 107 400 3308
Crosses ~
1 x 2 2.5 2.8 50.8 2.0 '2.8 49.3
1 x 3 2.0 1.8 53.4 2.5 1.8 50.3
1 x 4 1.7 3.5 45.4 2.0 3.8 45.7
1 x 5 2.0 3.5 49.3 2.0 3.5 47.8
1 x 6 2.0 3.2 43.8 3.0 2.8 - 40.4
1 x 7 1.3 "3.8 42.9 1.7 3.8 53.8
1 x 8 1.8 4.0 48.3 2.2 3.2 44.8
2 x 3 2.5 1.2 50.1 2.5 1.6 49.8
2 x 4 2.0 3.0 48.5 1.3 3.5 48.5
2 x 5 2.2 3.0 48.1 1.8 3.0 51.3
2 x 6 2.3 2.8 44.8 2.5 3.2 44.7
2 x 7 2.0 3.3 52.7 2.0 2.8 58.2
2 x 8 2.3 3.8 45.0 2.2 4.0 47.9
3 x 4 2.0 2.7 42.8 1.8 2.0 46.6
3 x 5 2.3 3.0 43.1 2.7 2.8 44.8
3 x 6 2.3 3.0 40.3 2.3 2.7 42.1
3 x 7 3.0 2.3 44.8 2.3 3.3 43.3
3 x 8 2.3 3.5 46.3 2.3 3.3 43.3
4 x 5 2.0 3.8 40.5 2.0 3.7, 42.8
4 x 6 2.0 3.7 43.0 1.7 3.2 38.3
4 x 7 1.7 4.0 49.3 1.7 3.3 52.2
4 x 8 2.2 3.3 46.7 2.0 3.7 48.3
5 x 6 2.0 3.7 42.1 2.3 2.8 40.9
5 x 7 2.0 3.2 49.8 2.2 2.8 51.7
5 x 8 1.7 3.5 51.8 1.3 3.2 47.6
6 x 7 1.2 3.7 54.7 1.8 3.5 57.5
'6 x 8 2.3 3.5 29.3 2.8 3.3 30.3
7 x 8 1.2 3.5 59.6 1.5 3.7 53.6
SB 004 004 207 004 004 205

 

55: Standard error of a difference between two means.

126

Table 10. Means for soaking and mass ratio traits for F2

and F progeny from an 8-parent diallel cross
grown at East Lansing in 1982.

 

 

 

 

Generations
F2 F3
Soaked Soaked Hydration soaked Soaked Hydration

Parents bean bean ratio been been ratio
and wei ht moisture weight moisture
crosses (3% (%) (g (%)
Parents

B-2 (1) 231.8 46.8 1.9 235.3 44.5 1.9
FF (2) 230.4 43.8 1.8 232.2 44.1 1.8
15-R (3) 214.4 41.8 1.7 221.5 44.7 1.8
BTS (5) 232.6 47.3 1.9 228.4 46.7 1.9
SAN (6) 230.1 46.9 1.9 221.3 46.3 1.9
SF (7) 231.6 46.4 1.9 225.4‘ 47.6 1.9
Crosses

1 x 2 237.1 44.4 1.8 223.2 43.5 1.8
1 x 3 225.9 43.4 1.8 231.7 44.9 1.8
1 x 4 216.9 40.1 1.7 222.6 42.4 1.7
1 x 5 234.8 46.9 1.9 232.7 45.8 1.8
1 x 6 221.2 45.0 1.8 226.8 45.6 1.8
1 x 7 233.6 - 45.7 1.8 232.4 45.5 1.8
1 x 8 228.4 45.9 1.8 230.3 44.8 1.8
2 x 3 218.3 43.0 1.8 216.5 42.3 1.7
2 x 4 206.5 40.8 1.7 214.5 41.9 1.8
2 x 5 229.3 45.5 1.8 230.6 44.5 1.8
2 x 6 234.9 46.6 1.9 233.9 46.4 1.9
2 x 7 234.7 47.1 1.9 233.1 45.2 1.8
2 x 8 230.0 45.7 1.8 228.8 44.6 1.7
3 x 4 209.4 39.9 1.7 217.2 40.8 1.8
3 x 5 226.7 45.6 1.8 231.1 46.5 .1.9
3 x 6 231.4 45.8 1.9 240.9 48.2 1.9
3 x 7 236.7 46.9 1.9 233.4 45.4 1.9
3 x 8 220.6 43.9 1.8 228.0 44.9 1.7
4 x 5 231.8 44.9 1.8 222.2 42.7 1.8
4 x 6 229.4 45.6 1.8 226.6 45.8 1.8
4 x 7 224.7 44.2 1.8 227.7 43.1 1.8
4 x 8 220.8 45.3 1.8 221.2 43.9 1.8
5 x 6 227.6 46.3 1.9 229.4 46.3 1.9
5 x 7 231.0 46.5 1.9 232.6 46.9 1.9
5 x 8 227.6 46.4 1.9 229.4 46.3 1.9
6 x 7 231.3 46.1 1.9 229.6 46.3 1.9
6 x 8 222.7 44.8 1.8 222.5 44.7 1.8
7 x 8 228.9 46.4 1.9 231.3 46.0 1.8
S5 5.9 1.2 0.05 6.4 1.3 0.1

 

0‘ Standard error of a difference between two means.

127

Table 10. (cont'd.).

 

 

 

 

 

 

Generations

Wash Cooked Wash Wash Cooked Wash
Parents drain bean drain drain bean drain
and weight moisture ratio weight moisture ratio
crosses (g) (%) (g) . (%)
Parents
B-2 (1) 373.3 63.4 1.3 376.1 63.2 1.3
FF (2) 379.3 64.2 1.2 376.1 64.1 1.3
15-R (3) 594.1 65.2 1.3 388.1 64.9 1.3
A-3O (4) 421.9 65.5 1.5 398.3 65.2 1.5
BTS (5) 386.8 64.1 1.3 376.1 64.0 1.3
SAN (6) 401.1 66.6 1.4 393.7 66.0 1.4
SF (7) 371.8 64.3 1.3 367.6 64.8 1.3
Crosses
1 x 2 317.0 62.5 1.3 306.4 64.3 1.4
1 x 3 306.4 63.5 1.4 307.1 ' 63.1 1.3
1 x 4 325.9 63.8 1.5 321.2 64.7 1.4
1 x 5 30909 6405 103 31000 6402 103
1 x 6 305.9 63.8 1.4 310.0 64.1 1.4
1 x 7 308.2 64.9 1.3 304.6 65.1 1.3
1 x 8 307.0 66.0 1.3 318.8 66.1 1.4
2 x 3 303.5 63.8 1.4 310.0 63.6 1.4
2 x 4 315.8 64.6 1.5 310.0 64.2 1.4
2 x 5 310.0 64.1 1.4 314.1 64.9 1.4
2 x 6 310.0 64.1 1.4 317.6 64.0 1.4
2 x 7 304.1 63.6 1.3 305.2 63.5 1.3
2 x 8 312.9 64.2 1.4 307.6 63.7 1.3
3 x 4 312.9 64.2 1.5 325.4 64.4 1.5
3 x 5 308.2 65.9 1.4 310.5 64.2 1.3
3 x 6 304.7 64.2 1.3 318.2 65.2 1.3
3 x 8 310.5 64.3 1.4 310.0 64.4 1.4
4 x 5 317.0 64.0 1.4 312.3 64.6 1.4
4 x 6 323.0 64.5 1.4 322.9 64.3 1.4
4 x 7 312.3 64.4 1.4 314.7 64.5 1.4
4 x 8 310.6 64.7 1.4 324.8 64.8 1.4
‘5 x 6 311.2 65.5 1.4 312.9 65.2 1.4
5 x 7 302.3 64.4 1.3 312.3 64.6 1.3
‘5 x 8 299.4 65.4 1.3 302.9 65.1 1.3
6 x 7 308.8 64.3 1.3 305.9 63.4 1.3
6 x 8 328.9 65.7 1.5 322.3 74.7 1.4
7 x 8 301.6 69.7 1.3 315.3 64.3 1.4
$5 4.3 1.2 5.3 0.7 0.04
'33: Standard error of differences.

128

Table 11. Means of texture and processing traits for two
generations of an 8-parent diallel cross grow
Saginaw, MI (1982).

 

 

 

 

 

 

gfienerations

F2 F3

Traits Traits
'Parents __ V¥_, _ __
and Clumps Splits Texture Clumps Splits Texture
crosses (scale) (scale) (Kg/1003) (scale) (scale) (Kr/100
Parents
8.2 (1) 103 200 4501 203 200 4902
FF (2) 1.7 1.0 55.8 1.3 1.0 56.0
15'R (3) 203 100 4103 205 100 4202
A-3O (4) 1.7 3.0 37.0 2.3 3.3 34.5
BTS (5) 2-3 3-7 37-6 3.3 3.3 35.4
SAN (6) 3-7 3-7 24.5 3.3 3.3 21.3
SF (7) 1.3 3.7 51.3 1.3 3.7 50.8
N-2 (8) 1.0 3.3 34.9 1.0 3.7 29.7
Crosses
1 x 2 2.7 1.0 54.0 2.3 1.8 55.7
1 x 3 2.2 1.5 45.6 2.3 1.5 45.7
1 x 4 1.7 2.3 47.8 2.0 3.0 43.9
1 x 5 2.8 3.0 39.4 2.2 2.8 40.1
1 x 6 2.5 2.7 39.2 2.7 2.7 37.1
1 x 7 1.8 3.5 45.5 1.2 3.8 48.2
1 x 8 1.8 3.3 40.4 2.2 3.0 44.3
2 x 3 2.2 1.5 - 45.5 1.8 1.3 45.1
2 x 4 2.0 1.7 52.2 1.3 2.0 55.2
2 x 5 2.5 2.3 45.7 2.7 2.2 43.4
2 x 6 2.0 2.2 40.7 2.5 1.8 41.4
2 x 7 2.5 2.2 51.0 2.3 2.2 50.5
2 x 8 2.8 2.3 48.2 2.5 3.2 50.0
3 x 4 2.5 1.5 43.4 2.5 1.8 45.1.
3 x 5 2.8 2.8 39.4 2.8 2.2 39.0‘
3 x 6 2.8 1.8 34.2 2.8 2.5 34.2
3 x 7 2.2 1.8 45.2 2.5 2.0 40.6
3 x 8 2.3 2.5 40.6 2.2 2.7 40.5
4 x 5 2.2 3.2 40.8 2.5 2.8 41.3
4 x 6 1.3 3.3 41.4 1.8 3.0 40.8
4 x 7 2.0 3.5 49.9 1.7 3.7 52.0
4 x 8 1.7 3.3 47.7 2.2 3.2 45.6
5 x 6 2.5 2.8 36.7 2.8 2.7 33.8
5 x 7 2.2 3.2 42.0 2.3 3.5 42.4
5 x 8 2.0 3.5 43.3 2.7 3.0 39.0
6 x 7 2.2 3.0 45.1 2.2 3.0 45.0
6 x 8 2.0 3.7 28.1 2.7 3.5 26.2
7 x 8 1.3 4.0 47.4 1.2 3.8 48.9
$5 0.4 0.3 2.0 0.4 0.4 2.1

 

Sfia Standard error of a differences between two means.

129

Table 12. Means for soaking and mass ratio traits for F2 and F
progeny of an 8-parent diallel cross grown at Saginaa

 

 

 

 

 

in 1982.
Cifierations
F2 F3
‘Soaked ’Soaked Soaked' Scaked
Parents bean bean Hydration bean bean Hydration
and wei ht moisture ratio weight moisture ratio
crosses (a? (i) (g) (5)
'Farents
B-2 (1) 23107 4500 108 23106 4506 108
FF (2) 255.9 47.8 1.9 246.6 48.7 1.8
15-8 (3) 24208 4603 109 24105 4802 109
BTS (5) 233.2 46.4 1.9 226.9 46.2 1.9
SAN (6) 234.1 47.3 1.9 232.6 46.3 1.9
SF (7) 229.2 47.7 1.9 234.0 47.3 1.9
Crosses .
1 x 2 235.4 44.9 1.8 234.8 44.8 1.8
1 x 3 234.8 45.0 1.8 226.9 43.6 1.8
1 x 4 229.1 42.0 1.7 220.8 43.0 1.7
1 x 5 228.4 45.1 1.8 238.3 47.2 1.9
1 x 6 225.6 45.7 1.8 231.6 45.8 1.8
1 x 7 233.7 46.0 1.8 230.2 45.5 1.8
1 x 8 230.2 45.5 1.8 235.0 45.8 1.8
2 x 3 241.0 45.0 1.8 239.6 45.6 1.8
2 x 4 227.6 42.5 1.7 221.0 40.5 1.7
2 x 5 226.9 44.5 1.8 237.8 45.5 1.8
2 x 6 232.0 45.5 1.8 238.5 46.4 1.9
2 x 7 218.2 42.0 1.7 238.4 45.0 1.8
2 x 8 228.7 44.1 1.8 230.8 44.4 1.8
3 x 4 222.4 42.1 1.7 234.4 44.1 1.8
3 x 5 224.9 44.9 1.8 228.7 45.6 1.8
3 x 6 231.8 45.6 1.8 237.7 47.4 1.9
3 x 7 232.8 44.6 1.8 245.9 47.2 1.9'
3 x 8 221.2 43.1 1.8 229.7, 44.7 1.8
4 x 5 220.7 41.2 1.7 223.7 43.2 1.8
4 x 6 229.2 43.2 1.8 226.8 44.8 1.8
4 x 7 215.1 41.0 1.7 211.9 40.0 1.7
4 X 8 21500 4002 107 22509 4301 108
5 x 6 228.9 45.5 1.8 237.1 47.0 1.9
5 x 7 230.8 46.7 1.9 233.5 47.3 1.9
5 x 8 226.4 45.7 1.8 238.4 47.2 1.9
6 x 7 229.2 45.8 1.8 235.2 45.9 1.8
6 x 8 232.5 45.3 1.8 226.6 46.2 1.9
7 x 8 233.9 46.5 1.9 233.4 46.2 1.9
$5 503 101 0003 607 103 0015

 

021

Standard error of differences.

130
Table 12. (cont'd.).

 

 

 

 

CEnerations
F2 F3

Hash Cooked Wash Wash Cooked Wash
Parents drain bean drain drain bean drain
and weight moisture ratio weight moisture ratio
crosses (g) (5) (g) (%)
Parents
8-2 (1) 373.3 63.4 1.3 376.1 63.2 1.3
FF (2) 379.3 64.2 1.2 376.1 64.1 1.3
15-R (3) 394.1 65.2 1.3 388.1 64.9 1.3
A-30 (4) 421.9 65.5 1.5 398.3 65.2 1.4
BTS (5) 386.7 64.1 1.3 376.1 64.0 1.3
SAN (6) 401.1 66.6 1.4 383.7 66.0 1.4
SF (7) 371.8 64.3 1.3 367.6 64.8 1.3
N-2 (8) 388.1 65.5 1.3 404.3 65.1 1.4
Crosses
1 x 2 305.3 63.9 1.3 302.4 63.2 1.3
1 x 3 302.3 63.2 1.4 308.3 63.6 1.4
1 x 4 311.8 64.1 1.4 312.4 64.3 1.3
1 x 5 308.8 64.1 1.4 307.6 64.5 1.3
1 x 6 307.6 64.7 1.4 309.8 64.8 1.4
1 x 7 307.6 64.3 1.3 304.1 64.6 1.3
1 X 8 30501 6406 103 30701 6403 103
2 x 3 318.8 64.4 1.3 317.7 64.5 1.3
2 x 4 309.4 63.8 1.4- 307.1 64.0 1.4
2 x 5 308.8 64.1 1.4 312.4 64.2 1.3
2 x 6 315.3 64.2 1.4 313.6 64.5 1.3
2 x 7 304.7 64.8 1.4 303.5 64.0 1.3
2 x 8 308.8 63.9 1.3 307.6 64.3 1.3
3 x 4 312.9 64.3 1.4 324.1 64.6 1.4
3 x 5 307.6 64.7 1.4 312.4 65.2 1.4
3 x 6 31707 6408 104 31905 6507 103
3 x 7 30701 6401 103 31509 6409 103
4 x 5 311.2 64.5 1.4 311.7 64.5 1.4
4 x 6 317.1 64.2 1.4 317.1 65.0 1.4
4 x 7 299.4 64.1 1.4 305.3 64.0 1.4
4 x 8 309.4 64.1 1.4 316.8 64.5 1.4
5 x 6 311.8 64.8 1.4 309.4 65.2 1.3
5 x 7 300.0 64.8 1.3 301.7 64.9 1.3
5 x 8 303-5 64-3 1-3 304.7 64.7 1.3
6 x 8 323.6 65.7 1.4 320.0 66.0 1.4
7 x 8 301.1 64.3 1.3 337.2 64.7 1.4
$5 4.3 0.4 0.03 7.7 0.4 0.05

 

F

U)
CI

= Standard error of differences.

131

01. Brasll-z
02. FF 18-15

    

_1 04. A-30
8x10‘ 05.8TS
06. Sanilac
07. San Fcrnando
08. N0p-2
6
4

b -..- 0.90 e 0.22 H
a =-0.00120.045

 

 

2
O J 3.... _1
4 a 8:10

-1

Figure 1. Variance (V,)-covariance (Hr) graph for clump
data of the 1" generation grown at E. Lansing
and showing t e position of points representing
the 7 parental arrays and their regression line
relative to a limiting parabola under which all
points must lie. ** = Significant at the 1% level
of probability.

132

 

 

 

2.FF 16-15
3.15-8-148
- 5. 8113
8X1? ‘ 6. Sanilac
7. San Fernando
8. Hep-2
6
Wr
4
0.18"
2 0.04
-1
0 8x10
Figure 2.

Variance (Vr)-covariance (Hr) graph for clump
data of the F generation grown at E. Lansing
and showing t e position of points representing
the 6 parental arrays and their regression line
relative to a limiting parabola under which all

points must lie. * = Significant at the 51 level
of probability.

8x10

 

-2

Figure 3.

133

1. Brasll-Z
2. FF 16-15
4. A-SO

5. BTS

6. Sanilac

7. San Fernando
8. Hep-2

   

b = 1.24: 0.37"
a {0.16: 0.11

 

-1
6 8x10

Variance (Vr)-covariance (Hr) graph for clump
data of the F2 generation grown at Saginaw and
showing the position of points representing

the 8 parental arrays and their regression line
relative to a limiting parabola under which all

points must lie. * = Significant at the 51 level
of probability.

8x10

 

Figure 4.

 

134

Brash-2
15-3-148
A-SO

.BTS
. San F0rnando
.N0p-2

1.
3
4.
5
7
8

30.19H
30.05

'1
I 4 6 8x10

Variance (Vr)-covariance (Hr) graph for clump
data of the F generation grown at Saginaw and
showing the position of points representing the
6 parental arrays and their regressionline
relative to a limiting parabola under which all

points must lie. " = Significant at the 11 level
of probability.

135

01. BrasH-z- 05. BTS

 

 

 

11x10 02- FF 16-15 06. Sanilac
03. 15-R-148 07. San Fernando
04. A-SO 08. Hep-2
1O 2
O
8 3
1
‘C
6
Wr H
0 b =1.oo:0.12
4 4 a =0.06."'.0.06
1.
5
2 O s
0 ___ _1
2 4 6 a 10x10
'1 .3 V1'

Figure 5. Variance wry-covariance (Hr) graph for slpits
data of the F generation grown at E. Lansing
and showing t e position of points representing
the 8 parental arrays and their regression line
relative to a limiting parabola under which all

points must lie. *' = Significant at the 11 level
of probability.

136

01. Brash-2 05. BTS
02- FF 16-15 06. Sanilac
03. 15-8-148 07. San Fernando

 

 

10x15" 04. A-SO 0a. Hep-2
F
21.
8
6.
Wr .'3 1 “I.
b:=1.07'=(3.13
4 a =0.00 .'!: 0.05
1
. 0 4
6 C
2
.,5
0 . 8 ..1
2 4 6 3 1OX1O

Vr

Figure 6. Variance (Vr)-covariance (Hr) graph for splits
data of the F generation grown at E. Lansing
and showing t e position of points representing
the 8 parental arrays and their regression line
relative to a limiting parabola under which all
points must lie. '* : Significant at the 1% level
of probability.

137

12x15 02- FF 16-15 06. Sanilac
I 03. 15-R-148 07. San Fernando
04. A-30 08. Nap-2

 

‘10

4‘

.y.
b = 0.94 20.13
a =o.24 2 0.06 "

 

 

2 4 3 ’1
Vl’ 8 10x10

Figure 7. Variance (Vr)-covariance (Hr) graph for splits
data of the F2 generation grown at Saginaw and
showing the position of points representing the
8 parental arrays and their regression line
relative to a limiting parabola under which all
points must lie. *_' = Significant at the 11 level

of probability.

138

10x161 0:0.94: 0.12M
a 20.23 2 0,051”

‘VVF 01-8rasH-2
5. 5 0 2' FF 16-15
0 3. 15-R-148
0 4. A-30
.8 05.8TS
06.SanHac
07. San Fernando
2 08. Hep-2

 

 

-1
2 4 6' 8 10x10
Vr

Figure 8. Variance (Vr)-covariance (Hr) graph for splits
data of the F generation grown at Saginaw and
showing the position of points representing.the
8 parental arrays and their regression line
relative to a limiting parabola under which all
points must lie. '5 = Significant at the 1% level
of probability.

139

01.8rasu-2
0 2oFF 16-15

 

 

.3I 03015-8-148
18x10 04. A-3O
05. BTS
15 05. Sanilac
07. San Fernando
14 08. Nep-z
12 fl
6
1o 0
Wr
8
5 08
b: o.95:0.15**
4 a: -0.02:0.03
2
6
0 '7 0.1 '4 3
2 4 6 8 10 12 14 15 18x15”

Vr

Figure 9. Variance (Vr)-covariance (Hr) graph for washed
drained weight data of the F generation grown at
E. Lansing and showing the position of'points
representing the 8 parental arrays and their
regression line relative to a limiting parabola
under which all points must lie. ** = Significant
at the 15 level of probability.

140

01. BrasH-z 05. BTS
1 02- FF 16-15 06. Sanilac
.3. 15-R-148 .7. San Fernando
o4. A-30 08. Nap-2

18x10

  

b=1.32 20.27**
a = -0.02:0.02

 

 

4 5 8101214

-3
16 18x10
05 Vr

Figure 10. Variance (Vr)-covariance (Hr) graph for washed

drained weight data of the F3 generation grown
at E. Lansing and showing the position of points
representing the 8 parental arrays and their
regression line relativerto a limiting parabola
under which all points must lie.** = Significant
at the 11 level of probability.

141

O1. BrasH-Z 05. BTS
02- FF 16-15 06. Sanilac
08. 15-3-148 07. San Fernando
18x10 '
04. A-3O .8. Nap-2

   

b: 1.32 2 0,27 **
a: -o.o2 : 0.02

 

 

6 3 1° 12 14- 16 18x10-3
5 Vr

Figure 11. Variance (Vr)-covariance (Hr) graph for washed

drained weight data of the F2 generation grown
at Saginaw and showing the position of points
representing the 8 parental arrays and their
regression line relative to a limiting parabola
under which all points must lie. '* = Significant
at 15 level of probability.

18x10

 

 

142.

o1-BrasH-2

o 2. FF 16-15
0 3o15-R-148
o4. A-3O
05.8TS

06. Sanilac

07. San Fernando
08. Nep-z

b=1.30 1': 0.26“”
=-0.01 1' 0.01

 

Figure 12.

2 4 5 8 10

~~3
12 14 16 18x10
Vr

Variance (Vr)-covariance (Hr) graph for washed
drained weight data of the 1“ generation grown
at Saginaw and showing the pogition of points
representingthe8 parental arraysandtheir
regression line relative to a limiting parabola
under which all points must lie. '9 = Significant
at the 1% level of probability.

143

oi-BrasH-Z
o 2- FF 16-15

 

 

10° 0 3.15-3—143
_ o 4. A-3O
05.8TS
80 ’ .6. Sanilac
07. San Fernando
L 08. Nep-2
6C).
Wr
1 0
4C)? . 3
6
03 .
2°~ 3'1 b=0.41 20.21
0
. ‘ a=8J828J0
I 5 '
o A A 1 #4 4 . L 4 1 .
20 40 60 30 100
Vr.
o
1

Figure 13. Variance (V, )-covariance (Hr ) graph for texture
data of ther F generation grown at E. Lansing
and showing tge position of points representing
the 8 parental arrays and their regression line

'relative to a limiting parabola under which all
points must lie.

o 1. Brasfl-z o 4
o 2. FF161-15-1 . 5, BTS
. 3. 1542-148 . 6

 

144

. A-SO

. Sanilac

    

 

410.
:30
{
2°, b=1.oezo.oe**
a = 8.63 21.03“
10
0 f i . .4
10 20 30 4O
Vr
Figure 14. Variance (Vr)—covariance (Hr) graph for texture

data of the F generation grown at E. Lansing
and showing the position of points representing
the 6parental arrays and their regression line
relative to a limiting parabola under which all
points must lie. '5 = Significant at the 11
level of probability.

 

14S

o1-BrasH-2
o 2-FF 16-15
0 3-15-R-148

 

o4. A-3O
05.BTS
10‘” o6. Sanilac
07. San Fernando
.8. N3p‘2 6
80 .
r
60 . g
C .4
Wr ’ 1
l
‘30, ' w *4;
2 b: 0.85 20.19
P .3 6:18.662 7.69‘
O
20 b- S .
1
o r _i J— 1 L 4 1 1 4 n
20 4o 60 so 100
Vr
Figure 15. Variance (Vr)-covariance (Hr) graph for texture

data of the F generation grown at E. Lansing
and showing the position of points representing
the 8 parental arrays and their regression line
relative to a limiting parabola under which all
points must lie...H = Significant at the SS
and 1% level of probability, respectively.

146

I1. Brash-2 05. BTS
.2. FF 16-15 06. Sanilac
100 .3. 15-3-143 .7. San Fernando

o4. A-ao oa. NeP'2

   

8C).

 

 

6C).
Wr r
1
4C),

b=os4zono**

2° a'=15.06:3.o**
’7

O_ 1 4 L i L L ,
20 4o 60 30 100
Vr

Figure 16. Variance (Vr)-covariance (Hr) graph for texture
data of the F generation grown at Saginaw
and showing tge position of points representing
the 8 parental arrays and their regression line
relative to a limiting parabola under which all
points must lie. ** = Significant at the 15
level of probability.

147

   
  

 

   

 

10C)
P .8
so.
01. BrssIl-z
.2. FF 16-15
so. .0 3.15-3-148
o4. A-3O
05.8TS
Wr as. Sanilac
. 07. San Fernando
4O+ 08.Nop-2
20. b=1.o:o.16**
e1 “127126.97“
o f . . . . . 4 L .
2° 40 60 so 100
Vr
Figure 17. Variance (Vr)-covariance (Hr) graph for texture

data of
and showing t

the F generation grown at Saginaw
3e position of points representing

the 8 parental arrays and their regression line
relative to a limiting parabola under which all

points must lie.'," -.-

Significant at the SS

and 11 level of probability, respectively.

DISCUSSION

Significant.F tests of culinary quality traits indi-
cated that genetic variability existed among the parents and
progeny for these traits (Table 1). Significant genetic
variability indicated that these traits would lend
themselves to improvement through selection. Clumps, splits,
washed drained weight, and texture, are traits important to
consumers and processors and were significantly different
within the F2 and F3 generations at both locations. The
generation mean square was not significant except for
clumps, soaked bean weight, and soaked bean moiéture. The
nonsignificant generations mean square indicated that the
expression of culinary quality traits was similar from one
generation to the next. Location effects for the traits were
nonsignificant. This was surprising because location and
seasonal effects are generally significant for culinary
quality traits in dry beans. These results could be due to
the larger F value required to declare significance because
of a few number of degrees of freedom for testing locations
or the fact that we evaluated segregating populations that
were more broadly adapted. The interaction of generations x
locations and entries x locations was not significant. This
suggested that the performance among crosses was similar at
each location.

The combining ability effects can be measured on popu-

lations at any level of inbreeding but its estimates depends

on the generation tested and on the other hand additive

149

variance remain constant for a given population through a
generation of inbreeding provided no selection occurs but
the dominance variance decreases with inbreeding. The GCA
effect of a parent comprises its additive effect and the
average dominance interaction associated with that parent in
hybrid combination with all other parents and with itself.
With inbreeding, GCA includes a genetic component associated
with the set of dominance interactions within the homozygous
loci of the parent itself. The contribution of this compo-
nent increases and the contribution of average dominance
decreases in proportion to the level of inbreeding.

Significant GCA mean squares were observed for all
traits in the F2 and 8 out of 9 traits in the F3 generation,
suggesting that genes with primarily additive effects con-
trolled trait expression. Significant SCA mean squares for
splits, clumps, washed drained weight and texture (Table 1)
showed that nonadditive genetic variance was also important
in trait expression. In most cases the GCA mean square was
larger than SCA.mean square» The type of genetic variance
in a reference population indicates the type of breeding
scheme that maximizes trait improvement" In this case the
preponderance of additive variation for traits and the
presence of significant SCA suggested that reciprocal
recurrent selection would be useful strategy for utilizing
both additive and the fixable component of non-additive
genetic variance in trait improvements Parents with large

SCA effects could be crossed and reciprocal recurrent

150

selection practiced to maximize the use of both types of

genetic variance in advanced generations.

The lack of consistent reciprocal and maternal effects
indicated that it makes little difference as to the choice
one makes for use of a plant as either a pollen of seed
parent. Reciprocal and maternal effects are generally
absent in plant species. However, they have been shown in
some crops, for example, onions (Aligm 3223 L”) in which
cytoplasmic male sterility was used to control pollen.
These effects arose probably because of relic heterozygosity
present in an essentially homozygous seed (nonrecurrent)
parent after its development by back crossing. This
situation does.not occur in been breeding because a useful
male-sterile is not available thus limiting the production
of hybrids in favor of pure lines. The interaction of GCA
and SCA with locations for several traits (Table 1)
suggested that effects changed from location to location.

GCA effects and their interaction with locations indi-
cated the contribution that a parent made to its progeny and
the uniformity of this performance from site to site (Tables
2 and 3). The GCA changed in direction and magnitude for
some of the traits depending on the location, Texture and
washed drained weight were two traits in which parents
showed a consistent GCA in both the Fg'and F3 generations.
Strains A-SO, 15-R, BTS, SAN, and N-2 (Tables 2 and 3)
transmitted significantly large and negative GCA effects for

texture and significantly large and positive GCA effects for

washed drained weight to their progeny. Up to a point,

151

negative GCA effects for texture are desirable because beans
with firm texture (positive effects) may be discriminated
against by consumers. Because texture affects the perceived
stimulus for chewing, it influences to a large degree a
consumer's acceptance of a1 food product. Textural
properties of processed beans must fall within prescribed
acceptability limits (Adams and Bedford, 1973). Beans may
be unacceptable if they are too firm "tough beans", or too
soft, "mushy beans" , after cooking.

High values for the washed drained weight trait in
beans are desirable to consumers because the washed drained
weight indicates the amount of total solids available for
consumption. It has been shown that washed drained weight
and texture are negatively correlated in beans (Hosfield and
Uebersax, 1980, 1984; Nordstrom and Sistrunk 1977).

Variances of specific combining ability of parents for
washed drained weight and texture (Table 5) showed that some
highly significant non-additive effects were present in some
parents while other parents had no significant SCA variance.
. The highly significant correlations of general
combining ability effects between F2 and F3 generation
(Table 6) indicated that the mode of gene action did not
change from generation to generation. This suggested that
recurrent selection could be useful in fixing additive
genetic variance in early generations following hybridiza-
tion. Moreover, it should be possible to identify progeny

with desirable gene combination even though they are hetero-

152

zygous. This point of view is held by Shebeski (1967) and
supported by our data (Table 6) showing a highly significant

correlation between parental value for texture (r=0.92** and

O.94**) and the washed drained weight (r=0.90** and O.91**)

in the F2 and F3 generation, respectively; The correlations
between pairs of the 9 culinary quality traits in the F2 and
F3 generations at each location indicated that the soaking
characters were significantly and positively correlated
among themselves and negatively correlated with the washed
drained ratio. The washed drained ratio and washed drained
weight were significantly and positively correlated (Tables
7 and 8). Tbxture was negatively correlated with the washed
drained weight . The pattern of correlation between the
soaking traits and texture was negative in both F2 and F3 at
Saginaw but inconsistent in direction and magnitude at East
Lansing. Since correlation between pairs of traits varied
from generation to generation and locations, they would be
unreliable in a plant breeding program.

Graphical analyses developed by Jink.(1954) and Hayman
(1954) of the covariance between the offspring of each
parental array and the nonrecurrent parent (up) and the
variance of their offspring in each parental array (VP) for
clumps, splits, washed drained weight and texture in both
generation at each location revealed that the regression
coefficient was significantly different from zero but not

significantly different from unity (Figures 1.u17). The

intersects 0f the Wr axis passed through the origin for most

of the traits except texture, where it intersected the axis

153

above the origin. These data indicated that genes
controlling the expression of clumps,splits, and the washed
drained weight were completely dominant, but that genes with
partially dominant effects controlled the expression of
texture.

Comparison of array point distribution and GCA effects
for texture in the p2 and F5 generation at East Lansing and
Saginaw revealed that the SAN, and N-2 had a preponderance
of recessive alleles for texture and also a significantly
large and positive GCA effects. It is tempting to speculate

that soft texture is a recessive trait while firmness is

under the control of dominant alleles. With respect to the
washed drained weight trait, it was observed that strains
with a reduced washed drained weight had significant and
negative GCA effect values and those with a high washed
drained weight value, had significant and positive GCA
effect values. The distribution of array points for this
trait showed similar results for SAN and N-Zh This suggested
that for these two genotypes, washed drained weight is
controlled by recessive alleles but controlled by dominant

alleles in the other strains.

10.

11.

LIST OF REFERENCES

Adams,M.W., and C.L. Bedford. 1975. Breeding food
legumes for improved processing and consumer acceptance
properties. Pages 299-309. In: 1L. Milner, ed.
Nutritional improvement of food legumes by breeding.
Proceedings of a symposium, sponsored by PAG, Rome,
July, 1972. Protein Advisory Group (PAG), United
Nations, New York.

Augustin, J., C. 8. Beck, G. Kalbfleish, L. C. Habel

and R. H. Mathews. 1981. Variation in the vitamin and
mineral content of raw and cooked commercial Phaseolus
vulgaris L. classes. J. Food Sci. 46: 1701-1706?

 

Bhatty, R. s., M. A. Nielson, and A. E. Slinkard. 1985.
Comparison of cooking quality of Laird and commercial

Chilean lentils grown in the Canadian prairies. Can.

Inst. Food Sci. Tech. J. 16: 104-110.

. 1984. Cooking quality of lentils'
grwon in Canadian prairies. Can. J. Plant Sci. 64: 17-
24. .

 

Bajarkavist, R.V., V. Karlson, and O. Bengtson. 1972.
Method for producing quick-cooking pulses. Swedish
patent application 341777.

Bourne, M.C. 1980. Texture evaluation of
horticultural crops. Hortscience. (15:58-59.

Burr, H. K., S. Kon, EL .L. Morris. 1968. Cooking
rates of dry beans as infleunced by moisture content

and temperature and time of storage. Food techs 22(3):
88-90 0

Chernick,A.,and B. A. Chernick. 1963. Studies of:
factors affecting cooking quality of yellow peas. Can.

Cockerham,C. C. 1980. Random and fixed effect in plant
genetics. Theor. Appl. Genet. 56:119-131.

Gfeller, F. and H. L. Halstead. 1967. Selection for
gook6ing quality of field peas. Can. J. of Plt. Sci. 47:
31- 340

Ghaderi, A., G. L. Hosfield, M. W. Adams, and M. A.

Uebersax. 1984. Variability in culinary quality,
component interrelationships, and breeding implications

in navy and pinto beans. J. Amer. Soc. Hort. Sci.
109(1): 85-90.

12.

15.

14.

15.

16.

17.

18.

19.

20.

21.

22.

25.

24.

155

Griffing, B. 1956. Concept of general and specific
combinig ability in relation to diallel crossing system.
Austral. J. Biol. Sci. 9:463-493.

Halstead, R. L., and F. Gfeller. 1964. The cooking
quality of peas. Can. J. of Plt. Sci. 44: 221-228.

Hayman, B. I. , 1954 a. The analysis of variance of
diallel tables. Biometric 10:235-244.

Hayman, 8.1. 1954 b. The theory and analyis of dial lel
crosses. Genetics 59:789-809.

Hosfield,G.L., and M.A.Uebersax.1980. Variability in
physico-chemical properties and nutritional components
of tropical and domistic dry bean germplasm. J. Amer.
Soc. of Hort. Sci. 105: 246-252.

Hosfield, G.L., A. Ghaderi, and M. A. Uebersax. 1984.
Factor analysis of yield and sensory and physico-
chemical data from tests used to measure culinary
quality in dry edible beans. Can. J. Pl t. Sci. 64:
285-293.

Jaffe, w. G. 1973. Factors affecting the nutritional
value of beans. P. 43-49. In. M. Milner (ed.).
Nutritional improvement of food legumes by breeding
(proceeding of a symposium sponsored by PAG, Rome,
July, 1972). Protein Advisory Group United Nations,
New York.

Jinks, J} Lu, 1954. The analysis of continuous variation
in a diallel cross of Nicotiana rustica varieties.
Genetics 39:767-788.

Kon, S. 1979. Effect of soaking temperatures on cooking
and nutritional quality of beans. J. Food Sci. 44(5) 3

Mattson, S. 1946. The cooking of yellow peas. A
colloid-chemical and biochemical study. Acta. Agr.
Suecana, 2: 185-231.

, E. E. Akerberg, E. Ericksson, E. Koutler-
Andersson, and K. Vahtras. 1950. Factors determining
the composistion and cookability of peas. Acta. Agr.
Scand. 1: 40-61.

 

Morris, H. J., E. R. Wood. 1956. Influence of
moisture content on keeping quality of dry beans. Food
Technol. 10: 225.

Muller, F. M. 1967. Cooking quality of pulses. J. of
the Sci. of Food and Agric. 18: 272-275.

25.

26.

27.

28.

29.

30.

31.

32.

55.

34c

35.

36.

156

Nordstrom, C. L., and W. A. Sistrunk. 1977. Effect of
types of beans, soak time, canning media, and storage
time on quality atributes and nutritional value of
canned dry beans. J. Food Sci. 42: 795- 798.

. 1979. Effect of type of

 

beans, moisture level, blanch treatment and storage
time on quality atributes and nutrient content of
canned dry beans. J. of Food Sci. 44: 392-395.

Rizley,N.F., and W. A. Sistrunk. 1979. Effects of
maturity, soaking treatment.and cooking method on the
quality and mineral content of southern peas. J. Food
Sci. 44(1): 220-221.

Rockland, L. R. 1972 (Abstract SI 735) U.S.A. Dept. of
Agric. U.S.A. Patent 3635.728.

Rockland,LuB., and A.E.Metzler.1967. Quick
cooking lime and other dry beans. Food Technol. 21:
345.

Sefa-dedeh, S., D. W. Stanley, and P. W. Voi-sey. 1979.
Effect of storage time and conditions on the hard- to-
cook defect in cowpeas (Vigna unguiculata) J. of Food

Sci. 44: 790- 796

Shebeski, luH. 1976. Wheat and breeding. In: K.F.
Nelson (ed.)..Proceedings of the Canadian Centenial
Wheat Symposium, P.253. Modern Press, Saskatoon.

Silva, C. A. B., R. P. Bates, and J. C. Deng. 1981.
Influence of soaking and cooking upon the softening and
eating of black beans (Phaseolus vulgaris Ed. J. Food
Sci. 46: 1716-1720.

Singh, L., D. Shaema, A. D. Daodhar, and Y. K. Sharma.
1973. Variation in protein, methionin, tryptophan, and

cooking period in pigeon peas (Cajanus cajan). Indian
J. of Agric. Sci. 45(8): 795- 79

Uebersax, M.A., and C. L. Bedford. 1980. Navy bean
processing: Effect of storage and soaking method on
quality of canned beans. Research report No. 410 from
the Michigan State University Agric. Expt. Station East
Lansing.

Varriano-Marston, EL, and E. DeOmana. 1979. Effect of
sodium salt solution on the chemical composition and

morphology of black beans (Phaseolus vulgaris L). J.
Food Sci. 44(2): 531-536.

Wassimi, N., S. Abu-Shakra, R. Tannous, and A. H.
Hallab. 1978. Effect of mineral nutrition and cooking

quality of lentils. Can. J. Plant Sci. 58: 165-168.

157

37. Wassimi, N., G.L. Hosfield, M.A. Uebersax. 1981. The
effect of seed coat color on soaking characteristics of
dry edible beans. Michigan Dry Bean Digest 5(4):12.

INTERPRETIVE SUMMARY

The studies presented in this thesis have dealt with
the genetic of cooking time and uniformity of cooking of
individual seed, tannin and protein content, and culinary
quality traits of dry beans (Phaseolus vulgaris L.). Both
GCA and SCA effects were significant for determining cooking
time and uniformity of cooking in parents and progeny of an
8x8 diallel cross. The magnitude of GCA was larger than SCA
for both traits. This suggested that the genetic variance
in this population was primarily additive in nature but
non-additive effects.also influenced trait;expression. It
was noted that quick cooking parents produced progenies that
were also easy to cook. Thus it should be possible to
select superior cooking progenies from crosses involving
quick and uniform cooking parents because of the
preponderance of additive genetic variance. Graphical
analyses for cooking time and 5 hard seed revealed that
these traits were governed in the parents by both dominant
and recessive alleles. It was observed that B-2, SAN, and N-
2 carried a preponderance of recessive genes for cooking
time while BTS and SF had predominantly dominant genes. The
GCA.effects.of each parent revealed that FF, A-30 and N-2
contributed genes for quick cooking and‘a reduced percentage
of hard seed while B-2 and 15-R contributed genes for longer
cooking time and a higher percentage of hard seed.

The partitioning of the among entry source of variance

for the 5 tannin trait into GCA, SCA and reciprocal effects

158

159

indicated highly significant GCA and SCA mean squares.
Estimates of variance components revealed that GCA variance
predominated in tannin content expression. The GCA effect
of each parent indicated that A-30, SAN and N-2 reduced the
% tannin content in their progenies at both locations under
test while the other strains contributed to a higher tannin
content in their progenies. Reciprocal effects determined
for each parent indicated that the significant maternal
effect present for % tannin content was probably due to the
large maternal effects of FT‘(0.32), 15-R (-0.32) and SF
(CL45). It was also observed that parents with non or low
tannin content produced progenies that also had low tannin
content in their seed coats.

Graphical analyses revealed that tannin content was
controlled by partially dominant systems of genes. It also
indicated that SAN and N-2 had a preponderance of
recessive genes and the remaining parents had a
relatively high proportion of dominant factors. The
data indicated that in this population the difference
.among progenies for tannin content was due to genes
with primarily additive effects but nonadditive effect
had also influenced trait expression.

Significant differences among entries were present for
percentage protein in raw and cooked beans. .All F tests for
GCA main effect were highly significant and the F tests for
SCA were either Significant or highly significant for both

raw and cooked bean protein content. No consistency in the

160

reciprocal variation was observed. Estimates of variance

components revealed that GCA effects were more important
than SCA in both cases suggesting that additive effects of
parents were more important than non-additive effects in
determining protein content in crosses. Comparison of
parental and progeny means indicated that crosses of low x
low and high x high parents tended to produce progenies that
were low and high in their protein content, respectively.
The crude protein content of uncooked parents and progenies
was reflected in similar fashion in the cooked bean samples.
Graphical analyses of raw bean seed protein content revealed
that this trait was controlled by a partial to complete
dominance system of genes. The presence of genetic
variability suggested that selection in this population for
low tannin and high protein content would be possible.

The genetic analysis of culinary quality of dry beans
revealed significant and highly significant differences
among entries for all traits except the hydration ratio. It
was found that GCA mean squares were highly significant for
all traits in the F2 while SCA mean squares were significant
for clumps, washed drained weight and texture. In the F3
generation the GCA mean square was significant for all the
traits except the hydration ratio and the SCA mean square
was significant for clumps, splits, washed drained weight
and texture. The GCA effect for each parent'in each
generation was significant for texture and washed drained

weight. Estimates of variance components showed that GCA

variance was significant for all the traits in the F2 and

161

all but hydration ratio in the F5 generation. The SCA
components of variance was significant for clumps, splits,
hydration ratio, washed drained weightg and texture in the
F2,generations SCA variance component for these traits in

the F3 was significant except for hydration ratio. The
interactions between combining ability effects and locations
generally followed a similar trend as the main effects.
Variances of SCA for A-30, BTS, SAN, SF and N-2 were
Significant in the F2 for texture but in the F3 parents B-2,
A-30, SAN and SF had significant SCA variances. This might
have led to the significant SCA variability for texture.
The results also indicated that certain crosses among
parents would produce progenies that would be either firmer
or softer than expected on the average. Highly significant
correlation between parental value and GCA effects, and GCA
effects between F2 and F3 generation indicated that the mode
of gene action did not change from generation to generation.
The following conclusions are drawn on the basis of
this research:
1) Significant GCA and SCA mean squares for cooking time and
% hard seed suggested that both additive and nonadditive
variance influenced trait expression. However, additive
effects were of greater influence. Reciprocal recurrent
selection would be appropriate to improve traits in this
population by utilizing both additive and fixable non-

additive variance.

2) High tannin content was dominant to low tannin content.

162

3) Strains with white or beige seed coats had no or low

tannin content and should produce progenies that also have
low tannin. Similarly, it would be possible to select
progenies that would have low tannin content and high
protein content.

4) This research indicated that protein losses during
cooking are not as severe as one might expect.

5) Both dominant and recessive genes influenced culinary
quality traits. Strains B-2, SAN, and N-2 had mostly
recessive genes for texture and the washed drained weight
and crosses among these parents should produce progenies
that have softer texture and a higher washed drained weight.
6) The high correlation coefficients between parental value
and GCA effects and between the GCA effects of the F2 and F3
suggested that;the mode of gene action did not change from
generation to generation.

The genetic control of cooking time and cooking
uniformity of individual grains, tannin and protein content,
and culinary quality traits in dry beans provided
information to develop efficient breeding strategies to

improve bean cultivars.

Appendix A

Appendix A

The methods of protein and 76 catechin equivalentare

described step by step as it follows:

A.

933g; protein determination.

Prior to the analysis of raw and processed bean flour for
protein content, a sample of 20 strains of dry beans
representing a wide range in color, growth habit, and
protein content was analyzed for percentage protein by
the Kjeldahl method of nitrogen determination. The
nitrogen content of each sample was multiplied by 6m25 to
obtain the total percent crude protein. These samples
were tested.for percent crude protein by the NIR.method
and the results compared. A correlation coefficient of
(LL98) was obtained. The samples from crosses and parental
material was then tested by the NIR method for protein
content and the results were checked every 20 samples

with an internal standard.

B.Sample extraction for tannin content.

The sample preparation for extraction and phenolics
determination was done according to Telek (1983). The
procedure described step by step as follows:
1. A 0.15-0.20 gram sample of ground testa is carefully
weighed and transferred into a 100 ml medicine bottle.
2. An acidic methanol solution is made by thoroughly
mixing 80 ml absolute methanol:19.5 ml distilled
320: 0.5 ml concentrated hydrochloric acid, (V/V/V).

163

164

3. Take 35 ml of the acidic 80% methanol solution and add
to the bottle containing the ground testa.

4. Extract in a shaker bath at 7000 for 30 minutes.

5. Decant the extract over a porcelain filter crucible
lined with glass microfiber filter (GF/D whatman, 2.5
cm) into a 100 ml volumetric flask.

6. Take the residue from the filter and repeat step # 5
two additional times. Combine all extracts,and make
up the volume with 80% acidic methanol solution.

7. Carefully pipette 5 ml of the extract into a 25 ml
volumetric flask and bring up to volume with a 30%
sulfuric acid solution.

8. From this 25 ml volume, carefully pipette a 3 ml
sample into each of three 10 ml volumetric flasks.

9. Add 3 ml of a 0.5% vanillin solution to two of the 10
ml sulfuric acid solution.

10. Add only the sulfuric acid solution to the third 10
ml flask.

11. Let all 3 flasks stand for 20 minutes. Read the absor-
bance of each flask at 500 um. i

12. While the flasks are Standing, prepare 2 the vanillin
blanks, by pipetting 3 ml of 0.5% vanillin solution
intoa 10 ml volumetric flask and bring up to volume
with a 30 % sulfuric acid solution.

C. Preparation o_f_ 3&9. catechin standard

1. A 0.05 gram sample of catechin is carefully weighed

and transferredintoa50 ml volumetricflask. Itis

dissolved in 1-2 ml absolute methanol and brought to

D.

1.

165

volume with distilled water.
A 5 ml sample of the catechin solution is pipetted
into a 2001nlvolumetric flask.and brought to volume
witka 30 % sulfuric acid solution. 1.
A 3 ml sample is pipetted from this 200 ml into a 10
ml volumetricflask in duplicate,a 3 ml of a 0.5%
vanillin solution is added to each of the flasks
and broughtto volume with a 30% sulfuric acid solution.
Prepare the 0.5% vanillin and catechin solutions
fresh eachday and just prior to pipetting the seed
coat extract into the 10 ml volumetric flasks.
Reading the absorbance
Set the spectrophotometer to zero witha vanil lin
blank by putting the blank in both sample and refere-
nce cuvette.
The vanillin blank is left in the reference cell and
the catechin standard is read at 500 nm against a
vanillin blank.
The sample blank is placed in both the reference and
sample cell and read, then the sample cuvette is
rinsed and the actual sample is poured into the

cuvette and read against the sample blank.

Determining the catechin equivalent

A.

Day factor
Day factor=( wt. of catechin/ 0.D of catechin) x

(dilution factor of sample / dilution factor of

catechin) x 100

166

B. % catechin equivalent =(0.D. of sample/wt. of sample)

x Day factor

 

Appendix B

167

0000:00 00:00: a 03 I

 

 

 

 

r
.0—o>_.cs::st .ps>o0 000000.0653 N0 ass N0 00 00300000000 I .

00 0

000.00 000.00 000.0 000.0 000.0 000.0 000.0 000.0 000.0 00 totem
00000.00 000.00 000.0 000.0 000.0 000.0 0000.00 000.0 000.0 00 0acuounacoz
000.0 000.00 000.0 000.0 000.0 000.0 000.00 000.0 000.0 0 0n=uounz

0000.00 000.00 000.0 000.0 000.0 000.0 00000.00 000.0 000.0 00 0000000000
0000.00 000.00 000.0 000.0 000.0 000.0 00000.00 00000.0 000.0 00 <00
00000.000 0000.000 00000.0 00000.0 000.0 0000.0 00000.000 00000.0 00000.0 0 <00

00000.00 000.00 00000.0 00000.0 000.0 000.0 0000.00 00000.0 0000.0 00 000:0

00

000.00 00.00 0000.0 000.0 000.0 000.0 000.0 000.0 000.0 00 souuu
00000.00 000.00 000.0 000.0 000.. 000.0 0000.00 000.0 000.0 00 0a=uouqieoz
00000.00 000.00 000.0 000.0 000.0 000.0 000.00 000.0 000.0 0 0asuouuz

00000.00 000.00 000.0 000.0 000.0 000.0 0000.00 000.0 000.0 00 0auoum0u90
00000.00 000.00 00000.0 000.0 000.0 0000.0 00000.00 00000.0 000.0 00 <00
00000.00 00000.000 00000.0 0000.0 000.0 00000.00 00000.000 00000.0 00000.0 0 <00

00000.00 00000.00 00000.0 00000.0 000.0 00000.0 00000.00 00000.0 00000.0 00, 000:0

0..

~00 so0ua0u<>
00000: 000 000003 600.» 600-» any assua0os 000 ousuo0oe 00000x0xv 000.000 000.000 .0.0 06 «canon

00: econ uoxuom +03 :O0uaw00: 0000 000660 :00: amazon museums 000000 003300 0:: eo0uauo=ou

 

000aub

 

.0000000
00000

.0 00 03°00 0:000 000000 000 00 00000 0300000 00000010 :0 me 0:00: so0uauosou
000000 0st no ou0suu 0000.09 Asse00so 0 0: ates—Le> 0c sea>0sss 000000. 0:0c0nloo noun notusee sec: .0

.m000 .00000003
0 so sousoeoa

h ‘0».

000.0

 

.168.

00:0000 09:00: I 03

0

.00u>0uuoaauu .0a>a0 000000.00u0 N0 00. 00 0. 00000000000 I .

 

 

 

00 0

000.00 000.00 0000.0 0000.0 000.0 000.0 000.0 000.0 000.0 00 00.00
00000.000 000.00 0000.0 000.0 000.0 000.0 000.0 000.0 000.0 00 0.:uuuaacoz
00000.000 000.00 0000.0 000.0 000.0 000.0 0000.00 000.0 000.0 0 0.000000

00000.000 000.00 000.0 000.0 000.0 000.0 000.0 000.0 000.0 00 0mucua0090
00000.000 000.00 00000.0 00000.0 000.0 000.0 00000.00 000.0 000.0 00 <00
00000.000 0000.000 00000.0 00000.0 00000.0 00000.0 00000.000 00000.0 00000.0 0 <00

00000.000 000.00 00063.9 000cc.c 00000.: 300.0 0:00.00 00000.0 00000.9 00 000:0

00

000.00 000.00 000.0 000.0 000.0 000.0 000.0 000.0 000.0 00 uouum
00000.00 000.00 000.0 00000.0 000.0 000.0 0000.0 000.0 000.0 00 0.:uou.l:oz
00000.00 000.00 000.0 00000.0 000.0 000.0 000.0 000.0 000.0 0 0.00000:

00000.00 000.00 000.0 00000.0 000.0 000.0 0000.0 000.0 000.0 00 0auoua0umz
00000.00 00000.00 00000.0 000.0 00000.0 0000.0 00000.00 00000.0 000.0 00 <00
00000.000 00000.000 00000.0 00000.0 00000.0 00000.0 00000.000 00000.0 00000.0 0 <00

00000.00 00000.00 00000.0 00000.0 00000.0 00000.0 00000.00 00000.0 00000.0 00 II. 0mm00

«0

000 co0ua0ua>
00000) 000 0000.3 000.0 00u.u 000 00300003 000 ou3u00ol 000000000 0.0.000 0.0.000 .0.0 no 000300

003 0.00 0.0.00 +03 =00u.u000 0.00 000000 0000 003.00 0.30000 .u0000 003300 0:. 0000.00000

 

000.09

.3.:00.m a. 03000 00.00 00000. Ana :0 .0000 0900000 acou.0n0 0. no .00.: :00uau0000

0

.m000 .00000002
0 0:. a :0 00030005
0:..0 00000. 0H0 00 .00.uu 0000.30 00.00030 0 0o ouca0ua> 00 00000000 0000000 000000300 loam 090.300 0.92

000.0

 

.169P

coca-av 6.51.3 I a) +

.uonaaz o>nuuuon «malt hue» an o>uunnon gnu-ouscn I III

 

 

 

 

 

 

IIII « u «a N IIII IIII IIII IIII an .caon .o IIII III «:30qu
s.o—«c.nu °.o «n.q IIII IIII n~.o«cn.c IJII ~.¢ «o.n IIII IIII d-uuouqlcoz
a.ndflo.- I... I..- sun- “H.o«mc.c I--- a.” «m.n nan. I..- guuouauuuu
o.~ «c.nn ~.o «n.s IIII IIII o~.c«o—.o sa.o«~c.o «.o «a.nu ac.ouon.o 0°.ouno.c <um
n.n «n.0n «.n «N.q noc.c«nco.o «co.c«~oo.c nc.o«no.c a~.o«su.o u.n «u.su so.c«cu.c no.9«no.c <00

nu
n.d an.” IIII IIII IIII oc.cuco.c IIII en.c«n~.c IIII do.ouno.c dunk-an:
~.o «a.nu «.5 «0.. IIII IIII um.c«ou.c IIII ~.c «o.u IIII oo.c«~o.c Hash-unleoz
s.~ wa.¢~ n.o «c.n IIII IIII su.c«ou.c IIII o.~ «s.u IIII nc.o«~c.c uncounnoox
~.c~«~.cm ~.n~«a.un doc.ouuoo.o noc.c«noc.o IIII «n.c«oc.n c.0uwn.~n au.¢«~u.o «a.nuaa.o <um
o.u «h.c o.n «n.0u Ho¢.o«~oo.o aco.c«ncc.o oo.o«no.c on.o«Nn.c n.n «c.9n «c.0«cn.c nc.efico.c <00
-
Anv usuuca usuqoa cuunu owuuu Auvouauuuol Auvouauswcl AaOOuxuuv Annie-v Annie-v neocoaloo
+93 anon voxaom couuquvhz coon vuxooo coon voxnom ouauumm ouuaam onlaao an. dawucuoaoo
cud-uh
.Noau .cnuwgoq: .ucnocaa .u u. crown
canon Canaan sue a« noouu «ounces uaouqalo :1 no Gaol couunuocou at can an uo nuunuu mama-ac unusuaao a

no nous-Io! nun-um. accused! van unuounuoou .<om .<oo uo noun. enact-a. as. cue-sonlou cos-«uc> no non-luau»

.n Inauh

 

 

17$!

VOQHIH‘ ‘05:: I a

 

 

 

 

+
.uoaaac o>uuwuoa anal. huo> no o>wuuuoc uuunowvan I III
~.o we.ca III: III: III: III: III: nn.o«nc.o III: III: aacuouqx
n.n~«o.cn IIII IIII IIII IIII IIII H.~ ua.° IIII IIII Hunk-unacoz
c.ao«u.n~ IIII IIII IIII IIII IIII .H.H «o.n IIII IIII uaoouauuox
H.on«m.ns a.naum.» noo.onao.c Hoc.o««cc.c oc.c«~c.o no.cu~n.o u.c «~.HN oc.ouco.c co.cuou.o <um
o.m us.o~ o.h «a.na ~o°.c«noo.c Ho¢.onaoc.o no.9«nu.o n«.c«~c.c n.~a«o.n~ ma.cwo~.o «c.0«cc.c due
nu
o.~ “a.“ IIII IIII IIII IIII IIII n~.c«nc.a IIII IIII aucuouqx
«.0 an.HH n.o «n.n IIII eco.o«oo.o IIII nn. anon. c c. a «a. c no.0«uo.° oo.o«no.o Honk-unicoz
«.0 «o.~u e.n «o.n IIII nco.o«coc.o IIII III N. o «a. o sc.ou~o.c ~c.c«~o.c announwoom
s.n «m.a~ s.sn«s.sn noo.c«~oo.o «co.o«soo.o mo.ouuu.o «n.9nmn.u H. n «e. ma oo.o«~H.o oc.c«an.c <um
n.n «m.om o.e an.» noo.c«~co.o noc.oumoo.c «o.o«sc.o ha.onou.o a. s «n. on om.°«nn.c nc.onno.o <06
«a
Any uguuou mamas: own-u owuau any ouauauol Auvousuaaoa Auoonmav Acnaoov Aounu-v uaoaonlou
+93 6.0: noxaom *9: nouuuuvaa coon voxoou anon vaxuom ouauxoa magnum uaaano van couucuoaou

 

aufiuua

 

cu uuouu Honnauv acouIAIa an ac anon nouunuocom mm van

.NmmH .cowusuu: .ucuuaua .w an abouu canon oHAHvo hut
a no uuuauu huuaosw huocwuno o co vouaoaoa ouuouuo
Hanuouul ecu ancuouqalcoa uncouauoou .<om .<uo uo uouuo uuuvcauu can ouaoconloo oocuuua> no counlwuuu

.c Qua-H

171.

venuauv vegan: I a:

 

 

 

 

+
.hdofinuuon—oou .flocfluouufi—u up: no Houuo cunt-gun van Hound Quflvcuum I ma .mm
a.“ n.~ oHc.c odo.o o~.o s..o a.o «H.° «H.o mm a
n.~ d.“ nac.o o~a.o He.¢ nq.o o.“ «H.o ca.c m .m
~.H a.~ gflc.a HHc.c nfi.c an.o e.o «9.: oc.° Mm »
o.H e.H eco.c o~o.c -.o a~.o 9.: ac.c ac.o m , um
o.H e.~- «ac.° acc.o ~e.c oo.o a.nI h¢.o -.ou Mm.
~.e- «.ou ncc.c n~¢.o ~n.o c~.o- n.~- oe.o go.ou m uuaoz
5.“- n.~ ncc.cu ouo.o no.o- no.c a.» o~.c o~.ou mu
o.Ha n.¢ ~«c.c- «no.9 nc.c- Ho.” ~.n n~.o a~.ou .. u ouamwu.m gum
o.~ e.o coo.o amo.o no.o on.o H.o- cH.c nnuc my
o.~ n." ~eo.o ono.c e~.c nq.c «.ms nc.o H~.o u unauqam
n.c- ~.n «No.cu “No.c n~.c He.o o.H- Ho.os no.o Mu . “aom
a.nn a.~ «no.ou oqc.a -.c a~.c H.H: -.o so.o u oduuaa xu.~n
c.o c..- fisc.c noo.cu no.o an.H- ~.~- Hn.c ~n.o- Mm
... a.nI huc.o “so.o- ao.o- n5.H- ~.~- on.ou -.cn u anu<
~.H- c.~- nco.c- nco.o- mH.o- 5°.o- «.o- o~.cu m~.o mm
n.o- n..- - nfio.o anc.ou H~.o- ~m.oa ~.cu ~o.oa «n.o m a¢HI~ua~
0.? can 3...? 29?. 2.? 8.? «... “as- 3.? M. .116
n.~- n.Hu acc.o- -c.cn an.o- nfi.o a.“ cm.cu a~.o a» -HInH-o~.-
o.a- n.n c~c.cu ~oo.oa od.ou en.c H." ~o.o- n¢.o ~m
o.o- a.~ odo.en ooc.cn en.cn on.o c.n ~c.ou og.ou m a «a.nua
_ .d: .
usage) any usage) caucu can.» Raw ouauauoa any ouauawas Aaconxuxv “Iain-v Audio-v cowuduoaoo vacuum
+93 0603 VOXIOm +93 flauuflu‘ha €603 vow—000 Gama fiflxflom our—vulva. Duddam Inn—5H0
canopy
.Nmmn .auuusuaz .ucwncua .u as cabuu acuua canavo hue :« cacao uaounnno :0 ac anal nouuuuunou
mm as. n no eons-nos cud-nu huwaaav huucuusu a now uuuoumo huaanan acu:HAIou uauucou no nouuanu-u .n canny

172

vacuuuv cos-a: I a:

+
a h

.hno>«uooaoou .oucououufiv o:u no uouuo unavaauu wen nouns eunucqum I In .Im

 

 

~.~ e a an.o n~.o nH.o qe.o «H.o n~.o Mm a
n.~ a a auc.c unc.c ma.o on.o «H.o eo.c m um
c.~ o H ooo.o coc.o oa.o a~.c ac.o oo.o Mm a
c.~ n u soc.o “co.o ao.c n~.o oo.o co.o m r In
n.n nuc.c soo.c s~.c nc.on on.o -.cn mm
H.c eco.o m~o.o -.c -.cu no.o a~.os u ~Inoz
e.~- -o.cu «fia.c «o.ou nfl.o en.o oe.ou Mu
~.nu «Nc.cu noc.o -.cn a~.ou oc.o n«.cs n ova-anon cam
a.~ nco.c «ao.o ...: on.c ¢~.o fin.c. Mm
~.q -o.c nHo.o an.c oe.o -.c -.c u o-agn-m
~.n- nuo.cn «no.c «o.o «H.o «H.e ne.o my mac»
n.~u goo.OI ouc.cs ec.c o~.on «<.c u~.c m «Hausa Joann
n.n coo.o auc.ou c~.cu 50.“: oH.c a~.cu Ma
«.Q ono.o ooc.oa oo.cu H~.~u o~.o a~.¢u a onn<
o.« «oo.ou sao.c -.o sn.o o~.on oH.o Mm
~.~ ~oo.o «Hc.c nc.o ~n.c no.oa o~.o u quIunna
.....- 8...? "8.? :5- an... 2.? 2... Mm 316
n.on owo.ou Huc.o on.cu h5.c no.on n~.c nu nanndno~I~u
n.oa e~c.cn ~oc.o oe.ou oc.o- nc.ou ac.ou «a
a.nu nuc.cu cuc.cu on.cn ma.c Hu.on oo.on h ~.~«-qun
:3. .
ugmuoa any usmwoa owunu aqunu may ousuuuol ARV ousuuuol Amocaxuxv Aoanouv Ananuuv caducnoaoo vacuum
+93 coca voxaom +93 couuauvam anoa voxooo :aoa voxoom Iguana unflauu

 

caucus

 

.Ncou .aauuzowx..aiuuunm an sauna canon «Hawvo hut an uuouo acounalo an no anon nouuuuoaou
a co vans-col ouchu huaaoav huucuaau a you mucouuo huuauaa ucwcwalou nauocou no noun-«aux .9 ounce

 

173

Table 7. Analyses of variance of (Wr-Vr) for culinary
quality traits to test the adequacy of additive-
dominance model.

 

 

 

 

 

 

East Lansing Saginaw
Traits F2 F3 F2 F3
Soaked Beans
Weight 3765.8 4505.2 1647.11 5350.5
Hydration ratio 6.34 8.54* 3.61 7.13
Moisture (%) 0.0 0.0 0.0 0.0
Cooked beans
'Eiumps(sca1e) 0.04 0.024 0.05 0.0227
Splits(scale) 0.03 0.013 0.03 0.0093
Washed drained wt. 1004.3 1796.2 631.40 55564.9
Washed drained ratio 0.0 0.0 0.0 0.0
Texture kg/100g 1546.1* 421.4* 125.1 219.5
Moisture (%) 3.93 1.08 0.053 0.107

 

*,** = Significant at 5% and 1% probability level,respectively.

174

Table 8. Correlation coefficients indicating relationships
between F2 and F generation means for culinary
quality tarits measured on 56 crosses of an 8-parent
diallel grown in two locations in 1982.

 

Correlation coefficient (r)*

 

 

 

 

 

 

Traits East Lansing Saginaw
Soaked bean
1. Weight (g) O.56** 0.44**
2. Moisture (%) 0.56** 0.44**
3. Hydration ratio 0.70** 0.23
Cooked bean
4. Clumps (scale) 0.42** 0.49**
5. Splits (scale) 0.67** 0.77**
6. Texture (Kg/100g) 0.76** 0.87**
7. Wash drain Wt.(g) 0.39** 0.64**
8. Wash drain ratio 0.71** 0.36**
9. Moisture (%) 0.10 0.57**

 

** a Significant at 1iprobability level .

General ‘Bibliography

2.

10.

BIBLIOGRAPHY

Abdalla, M” M., M. M. Morad, and M. Roushdi. 1976.
Some quality characteristic of selections of Vicia
faba L. and their bearing upon field bean breeain .

Z. Pflanzenzuchtg. 77: 72-99.

Adams, M. W. 1973. On the quest for quality in field
bean. P. 143-149. In: M. Milner (ed.)‘. Nutritional
improvement of food legumes by breeding. ( Proceedings
of a symposium sponsored by PAG, Rome, July, 1972).
Protein Advisory Group, United Nations, New York.

Adams,M.W., and C.L. Bedford. 1973. Breeding food
legumes for improved processing and consumer acceptance
properties. Pages 299-309. In: M. Milner, ed.
Nutritional improvement of food legumes by breeding.
Proceedings of a symposium, sponsored by PAG, Rome,
July, 1972. Protein Advisory Group (PAG), United
Nations, New York.

Allard, R. W. 1953. Inheritance of some seed coat
color and patterns in lima beans. Hilgardia 22: 167-
177.

Altschul, A., and J. K. Grommet. 1980. Sodium intake and
sodium sensitivity. Nutrition review 38(12):393-402.

Augustin, J., C. B. Beck, G. Kalbfleish, L. C. Habel

and R. H. Mathews. 1981. Variation in the vitamin and
mineral content of raw-and cooked commercial Phaseolus
vulgaris L. classes. J. Food Sci. 46: 1701-1706.

Aykroyd, W.R., J. Doughty, and A.F. Walker. 1982.
Legumes in human nutrition. 2nd edn. FAO Nutritional
studies No. 19. FAQ, Rome.

Bakshy, A. K., E. T. Mertz, and J. D. Axtel. 1978.
Effect of dehulling on tannin content, protein
distribution and quality of high and low tannin
sorghum. J. Agric. Food Chem. 26: 679-683.

Bate-Smith, E. C., and V. Rasper. 1969. Tannins of
grain sorghum: Lutrafurol (Leucoluteolinidin)

3:4, 45,7- pentahydroxyflavin. J. Food Sci.

54: 203- 209.

Bhatty, R. s., M. A. Nielson, and A. E. Slinkard. 1983.
Comparison of cooking quality of Laird and commercial

Chilean lentils grown in the Canadian prairies. Can.

Inst. Food Sci. Tech. J. 16: 104-110.

11.

12.

15.

14.

15.

16.

17.

18.

19.

20.

21.

22.

176

. 1984. Cooking quality of lentils

 

grwon in Canadian prairies. Can. J. Plant Sci. 64: 17-
24.

Bajarkavist, R.V., V. Karlson, and 0. Bengtson. 1972.
Method for producing quick-cooking pulses. Swedish
patent application 341777.

Bliss, F. A., and J. W. S. Brown. 1983. Breeding
common bean for improved quantity and quality of seed
protein. Plant Breeding Rev. 1: 59-102.

Bond, D. A. 1976. In Vitro digestibility of testa in
tannin-free field bean ZVicis faba in) J. Agric. Sci.

Bourne, M.C. 1980. Texture evaluation of
horticultural crops. Hortscience. 15:58-59.

Bressani,R., LAG. Elias, and Ana Teresa Valiente.
1963. Effect of cooking and amino acid supplementation
on the nutritive value of black beans (Phaseolus
vulgaris L.). Brit. J. Nutr. 17:69-78.

Bressani,R., and LuG. Elias.1977. Tentative
nutritional objectives in major crops for plant
breeder. Pages 51-60. In: J.H. Hulse, K50. Rachie,
and luW. Billingsley,.eds. Nutritional standards and
methods of evaluation for food legume breeders. Int.
Dev. Res. Cent. Ottowa, Canada. IDRC-Ts7e.

Bressani, R. 1983. Nutritional quality of dry beans.
Bean improvement cooperative and nationdal dry bean
council biennial conference. Minneapolis, Minnesota,
November 8-10.

Buchbinder, B. U. 1980. Polysomal and messenger RNA
directed ixl'vitor synthesis and processing of
phaseolin, the G1 seed storage protein of Phaseolus
vulgaris L. PhD Thesis, University of WisconSin,
Madison.

 

Burn, R.E. 1963. Method of tannin analysis for forage
crop evaluation. Tech. Bull. N.S. 32, Georgia Agric.
Exp. Stn. Athens.

Burr, H. K., S. Kon, H. J. Morris. 1968. Cooking
rates of dry beans as infleunced by moisture content
and temperature and time of storage. Food tech. 22(3):

88-90 0

Butler, L. G. 1978. Tannins in sorghum grain:
Problems, solution, and opportunities, 33rd annual
corn and sorghum research conference. PP: 190-201.

 

23.

24.

25.

26.

27.

28.

29.

50.

51.

52.

55.

54.

55.

177

1982. Proceedings, International
Symposium on'GEain Quality, ICRISAT. 28-31 Oct. 1981,
Hyderabad, India. p: 294-310.

 

Chang, S. I., and-H. L. Fuller. 1964. Effect of
tannins content of grain sorghum on their feeding value
for growing chicks. Poul. Sci. 43: 30-36.

Chernick, A., and B. A. Chernick. 1963. Studies of
factors affecting cooking quality of yellow peas. Can.
J. Plant. Sci. 43: 174-183.

Cockerham, C.C. 1980. Random and fixed effect in
plant genetics. Theor. Appl. Genet. 56: 119-131.

Corcoran, M. R., T. A. Geisman, and B. 0. Phinney.
1972. Tannins as gibberellin.antagonists. Plant
Physiol. 49: 323-550.

Crofts, H. J., L. E. Evans, and P. B. E. McVetty.
1980. Inheritance, characterization and selection of
tannin-free faba beans (Vicia faba I") Can. J. of Plt.
Sci. 60(4): 1135-1140.

 

Dalrymple, E. J., B. P. Goplen, and R. E. Howarth.
1984. Inheritance of tannin in birdsfoot trefoil. Crop

Donnelly, E.D., andW.B. Anthony. 1969. Relationship
of tannin, dry matter digetibility and crude protein
in Sericea lespedeza. Crop Sci. 9: 361-362.

Eliase, L. G., D. G. DeFernandez, and R. Bressani.
1979. Possible effects of seed coat polyphenolics on
the nutritional quality of been protein. J. Food Sci.
44 : 524-527.

Erith, A. C. 1930. The inheritance of color, size,
form of seeds and flower color in Vicia faba L.
Genetica 13: 477-510.

Evans, A. M., and H. E. Gridley. 1979. Prospects for
the improvement of protein and yield in food legumes.
Current Adv. In plant Sci. 32: 1-17.

FAO 1982. FAO Production Year Book. Vol. 36. FAQ,
Rome. '

Freudenberg, Km, and K. Weinges. "Catechin and Flavanoid
Tannins", In the chemistry of flavanoid compounds, ed.
T. A. Geissman, New York: MacMillan Co. 1962. pp. 212-
214.

178

56. Gfeller, E., and R. L. Halstead. 1967. Selection for
cooking quality of field peas. Can. J; of P1 t. Sci. 47:

651-654.

37. Ghaderi, A., G. L. Hosfield, M. W. Adams, and M. A.
Uebersax. 1984. Variability in culinary quality,
component interrelationships, and breeding implications

in navy and pinto beans. J. Amer. Soc. Hort. Sci.
109(1): 85-90.

38. Griffing, B. 1956. Concept of general and specific
combinig ability in relation to dial lel crossing system.
Austral. J. Biol. Sci. 9:463-493.

39. Gwotkin, R., and P. J. G. Plummer. 1946. Toxicity of
certain salts of sodium and potassium for swine.
Canad. J. Comp. Med. 10:185-190.

40. Gustavson, K.H. 1956. "The chemistry of tanning
process." Academic press, New York.

41. Halstead, R. L., and F. Gfeller. 1964. The cooking
quality of peas. Can. J. of Plt. Sci. 44: 221-228.

42. Handler, P., and R. D. Baker. 1944. Toxicity of orally
administered tannic acid. Science 99: 393-394.

43. Harris, H. B., and R. E. Burns. 1970. Influence of
tannin content on preharvest seed germination in
sorghum. Agron. J. 62:.835-836.

44. Hartroft, W. S. 1967. Nutrition, pharmacology, and
hypertension. Nutrition reviews. 25(6): 161-164.

45. Haslam, E. 1966. "Chemistry of vegetable tannins."
Academic press, New York.

46. Hayman, B. I. , 1954 a. The analysis of variance of _
diallel tables. Biometric 10:235-244.

47. , 1954 b. The theory and analyis 0f
diallel crosses. Genetics 39:789-809.

48. Hawtin, G.C., K.0. Rachie, and J.M. Green. 1977.
Breeding strategy for the nutritional improvement of
pulses. Pages 43-51. In: J.H. Hulse, K.0. Rachie, and
L.W. Billingsley, eds. Nutritional standards and
methods of evaluation for food legume breeders. Int.
Dev. Res. Cent., Ottawa, Canada. IDRC-Ts7e.

49. Hosfield,G.L., and M.A.Uebersax.1980. Variability in
physico-chemical properties and nutritional components
of tropical and domistic dry bean germplasm. J. Amer.
Soc. of Hort. Sci. 105: 246-252.

50.

51.

52.
55.

54.

55.

56.

57.

58.

60.

61.

179

Hosfield, G.L., A. Ghaderi, and M. A. Uebersax. 1984.
Factor analysis of yield and sensory and physico-
chemical data from tests used to measure culinary
quality in dry edible beans. Can. J. Plt. Sci. 64:
285-295.

Hosfield, G.L., M.A. Uebersax, and T.G. Isleib. 1984.
Seasonal and genotypic effects on yield and physico-
chemical seed characteristics related to food quality
in dry edible beans. J. Amer. Soc. Hort. Sci.
109(2): 182-189.

Hudson, L. 1983. Personal communication.

Ivan, M., and J. P. Bowland. 1976. Digestion of
nutrients in small intestine of pig's diet containing
raw and autoclaved faba beans. Can. J. Animal Sci. 59:

Jacobson, A., and M. R. Corcoran. 1977. Tannins as
giberrellin anatagonists in the synthesis of alfa-
amylase and acid phosphatase by barley seeds. Plant
physiology 59: 129-133.

Jaffe, W. G. 1973. Factors affecting the nutritional
value of beans. P. 43-49. In. M. Milner (ed.).
Nutritional improvement of food legumes by breeding
(proceeding of a symposium sponsored by PAG, Rome,
July, 1972). Protein Advisory Group United Nations,
New York.

Johansen, D. A. 1940. Plant Microtechinique. New York
and London: McGraw Hill Book Co. p.195.

Jones, W. T., and J. W. Lyttleton. 1971. Bloat in
cattle. XXXIV. A survey of legumes forages that do
and do not produce bloat. N. Z. J. Agric. Res. 14:
101-107. -

Joslyn, M. A”, and Z. Glick. 1969. Comparative effect
of gallotannic acid and related phenolics on the growth

Kelly, J. D., and F. A. Bliss. 1975. Heritability
estimates of percentage seed protein and availables

methionine and correlation with yield in dry beans.
Crop Sci. 15: 753-757.

. 1975. Quality factors
affecting the nutritive value of bean seed proteins.
Cr0p Sci. 15: 757-760.

 

Kon, S. 1979. Effect of soaking temperatures on cooking
and nutritional quality of beans. J. Food Sci. 44(5) :

62.

65.

54.

65.

66.

67.

68.

’69.

70.

71.

72.

75.

74.

180

Lease, E. J., and J. H. Mitchell. 1940. A study on the
tannins of Lespedeza sericea. S. Car. Expt. Sta. Ann.
Rep. 53: 710.

Leeds, A.R. 1982. Legumes and gastrointestinal
function in relation to diets for diabetics. J. Plant
Foods. 4: 23-27.

Leleji, 0. I., M. H. Dickson, L. V. Crowder, and J. B.

Bourke. 1972. Inheritance of crude protein percentage
and its correlation with seed yield in beans, Phaseolus
vulgaris L. Crop Sci. 12: 168-171.

Levin, D. A. 1971. Plant phenolics: An ecological
perspective. Amer. Nat. 105: 157-181.

Lindgren E. 1975. The nutritive value of peas and
field beans. Swedish J. Agric. Res. 5: 159-161.

Ma, Y., and F. A. Bliss. 1978. Seed proteins of common
bean. Crop Sci. 18: 431-457-

. 1978. Tannin content and

 

inheritance in common beans. Crop Sci. 18: 201-204.

Mattson, S. 1946. The cookability of yellow peas. Acta
Agr. Suecana II 2, 185.

Mattson,S., E. E. Akerberg, E. Ericksson, E. Koutler-
Andersson, and K. Vahtras. 1950. Factors determining
the composistion and cookability of peas. Acta. Agr.
Scand. 1: 40-61.

Marquardt, R. R., A. T. Ward, and L. E. Evans. 1978.
Comparative properties of tannin-free and tannin-
containin cultivars of faba beans (Vicia faba LJ.
Can. J. Plt. Sci. 58: 755-760.

Marquardt, R. R., A. T. Ward. 1979. Chick performance
as affected by autoclave treatment of tannin-containing
and tannin-free cultivars of faba beans. Can. J. Anim.
Sci. 59: 781-789.

Maxson, E. D., L. E. Clark, and L. W. Johanson. 1972.
Factors affecting the tannin content of sorghum grains
as determined by two methods of tannin analysis. Crop
Sci. 12: 255-255.

Morris, H. J., E. R. Wood. 1956. Influence of
moisture content on keeping quality of dry beans. Food

Technol. 10: 225.

75.

76.

77.

78.

79.

80.

81.

82.

83.

84.

85.

86.

87.

181

Meiners, J. P., and S. C. Litzenberger. 1975. Breeding
for nutritional improvement. P. 131-141. In: M.
Milner(ed.) Nutritional improvement of food legumes by
breeding. (Proceedings of a symposium sponsored by PAG,
Rome, July, 1972). Protein Advisory Group, United
Nations, New York.

Muller, F. M. 1967. Cooking quality of pulses. J. of
the Sci. of Food and Agric. 18: 272-275-

Muneta, P. 1964. Cooking time of dry bean after
extended storage. Food Tech. 18: 130-151.

Mutschler, M. A., F. A. Bliss, and T. C. Hall. 1980.
Variation in the accumulation of seed storage protein
among genotypes of Phaseolus vulgaris L. Plant
physiology. 65: 627-350.

 

Nordstrom,C.L., and W.A. Sistrunk. 1977. Effect of
types of beans, soak time, canning media, and storage
time on quality atributes and nutritional value of
canned dry beans. J. Food Sci. 42:795-798.

. 1979. ~Effect of type of
beans, moisture level, blanch treatment and storage
time on quality atributes and nutrient content'of
canned dry beans. J. of Food Sci. 44: 592-595-

 

0sborne, T. B. 1894. The protein of kidney bean. J.
Amer. Chem. Soc. 16: 703-712.

Parada, R. S., M. L. Saini, and S. K. Arora. 1975.
Inheritance of tannin content in Eu-sorghum. Z.
Pflanzenuchtg. 74: 251-256.

Picard, J. 1976. Apercu sur 1heredite du caractere
absence de tanins dans les graines de.fleverole (Vicia
faba L.). Ann. Amelior. plantes. 26 : 101-106. .

Price, M. L., and L. G. Butler. Tannin and nutrition.
Station bulletin 272, March 1980, Dept. Biochem. Agric.
Expt. Sta. Purdue Univ., Indians.

Quinby, J. R., and J. H. Martin. 1954. Sorghum N
improvement. Advances in ronom . Academic press, ew
York and London. 6: 3053355.

Rannenkamp, R. R. 1977. The effect of tannins on
nutritional quality of dry beans Phaseolus lglgagig L.
Ph.D. Thesis Purdue University.

Ringrose,R. C., and C. L. Morgan. 1940. The nutritive
value of lespedeza for poultry feeding. South Carolina
Ept. Stn. Annu. Rep. 53: 91-92.

88.

89.

90.

91.

92.

95.

94c

95.

96.

97.

98.

99.

100.

182

Rizley, . F., and W. A. Sistrunk. 1979. Effects of
matur ty, soaking treatment;and cooking method on the

quality and mineral content of southern peas. J. Food
Sci. 44(1): 220-221.

Rockland,LuB., and A. E. Metzler. 1967. Quick
cooking lime and other dry beans. Food Technol. 21:
545.

Rockland, L. R. 1972 (Abstract SI 735) U.S.A. Dept. of
Agric. U.S.A. Patent 3635,728.

Rockland, L. B. 1972. Quick cooking process inmproves
bean image. Canner packer 141(9):8.

Schaffert, R. E., V. L. Lechtenberg, D. L. Oswalt, J. D.
Axtell,R. C. Pickett, and C. L. Rhykerd. 1974. Effect
of tannin on in vitro dry matter and protein
disappearance in sorghum grain. Crop Sci. 14: 640-643.

Sefa-dedeh, S., D. W. Stanley, and P. W. Voisey. 1979.
Effect of storage time and conditions on the hard-to-
cook defect in cowpeas (Vigna unguiculata) J. of Food
8010 44: 790-7960

Shebeski, IuH. 1976. Wheat and breeding. In: K.F.
Nelson.(ed~). Proceedings of the Canadian Centenial
Wheat Symposium, P.253. Modern Press, Saskatoon.

Silva, C. A. B., R. P. Bates, and J. C. Deng. 1981.
Influence of soaking and cooking upon the softening and
eating of black beans (Phaseolus vulgaris L.). J. Food
Sci. 46: 1716-1720.

Simpson, H.C.R., R.W. Simpson, S. Lousley, R.D.
Carter, M. Geekie, T.D.R. Hockaday, and J.I. Mann.
1981. A high carbohyrate leguminous fiber diet
improves all aspects of diabetic control. Lancet. 1:
1-50

Singh, L., D. Shaema, A. D. Daodhar, and Y. K. Sharma.
1973. Variation in protein, methionin, tryptophan, and

cooking period in pigeon peas (Caaanus cajan). Indian

Sjodin, J. 1973. Breeding for improved nutritional

quality in.field beans (Vicia faba L.L.First FAO/SIDA
seminar on improvement and production of field crops.

Statler, G. D. 1970. Resistance of been plants to

Fusarium solani f} phaseoli. Plant Dis. Rep. 54 (4): 698-

699 0

Steel, R. G., and J.H. Torrie. 1980. Principles and
procedures of statistics 2nd ed. McGraw-Hill, New York.

101.

102.

105.

104.

105.

106.

107.

108.

109.

110.

111.

112.

115.

185

Stoy, V., and K. Sundin. 1976. Effect of growth
regulating substance in cereal seed germination. Cereal

Sullivan, J.G., and F. A. Bliss. 1983. Expression of
enhanced protein content in inbred backcross lines of
common bean. J. Amer. Soc. Hort. Sci. 108(5) :787-791.

1983. Recurrent mass selection for

 

increased seed yield and seed protein percentage in the
common bean (Phaseolus vulgaris L.) using a selection
index. J. Amer. Soc. Hort.'Sci. 108(1): 42-46.

Telek, L. 1983. Tannin determination in dry beans.
Personal communication.

Taylor, J. 1975. A modified air palnter ("plantair") for
use in experimental plots. Proc. Bien. Conf. of the Bean
Improv. Coop., Nov. 5-7, East Lansing, Mich. P.36.

Truswell, AJS. 1980. Food and nutrition goods in
Australia. Food Tech. Aust. 32: 295-298.

Uebersax, M. A., and C. L. Bedford. 1980. Navy bean
processing: Effect of storage and soaking method on
quality of canned beans. Research report No. 410 from

.the Michigan State University Agric. Expt. Station East

Lansing.

Varriano-Marston, E., and E. De0mana. 1979. Effect of
sodium salt solution on the chemical composition and
morphology of black beans (Phaseolus vulgaris LJ. J.
Food Sci. 44(2): 551-556.

Walker, A.F. 1982. Physiological effects of legumes in
the human diet. J. of Plant Foods 4: 5-14.

Wassimi, N., S. Abu-Shakra, R. Tannous, and A..H.
Hallab. 1978. Effect of mineral nutrition and cooking
quality of lentils. Can. J. Plant Sci. 58: 165-168.

Wassimi,N., G.L.Hosfield, M.A. Uebersax. 1981. The
effect of seed coat color on soaking characteristics of
dry edible beans. Michigan Dry Bean Digest 5(4): 12.

Watt, B.K. and A.L. Merrill. 1963. Composition of
Foods, Agriculture Handbook No. 8., U.S. Department of
Agriculture.

Wilson, B.J., J.M. Mcnab, and H. Bently. 1972. The
effect of autoclaving and methionine supplemnetation on
the growth of chicks given diets containing field beans
(Vicia faba L. ). Brt. Poult. Sci. 13:67-75.

184

114. Woodruff, B. J., R. P. Cantrell, J. D. Axtell, and L. G.
Butler. 1982. Inheritance of tannin quantity in
sorghum. The J. of heredity 73: 214-218.