s a :.
ruwnfiuyg=
.r in ‘It ti).
3:! J” 15...): .

i

1.. .
flufzsiav.
.i i:
v.3:

3!}

is: .2

A.-

. . .v:

.2 . so?

. 31:12.: xr it
.5

 

v ”‘1‘ IO”
”gums?“ 4m
{'1' ‘1'

L

 

. .. 2.4,.
t
11:.

$2.. 1!:
z. 4). 15013355..
. .,\59:I:...l|€.l...um a
I}! haohgtstli

\- a 3g.‘l’ .

, itals‘iuu

\‘r 2‘3

I

 

 

1:

-xnmxrsvnas

~

3.

av is 5%.“...
1.... 9

3.. R123 buslllx

v. 3)::
.7. J42!

 

S.» A
a: i 2...»:
12.110 .\ .
~vv“ . C
p: .1. 11:19‘!‘
it... 3... a

 

In . ‘ y~ . ‘4 u.
3.5,. ' 1.. Liv. . Luna!!!“ .3“. t; 3...... .14. 3|}
‘ I. u- a - IV a I vfi'

1 . , 13.3.:‘Msfi.tfl..tt»uu

. f. (3:.
.r . . ‘ ,

bcwggmxfigpfifififia «Maw? . , ....,..q..m. filflnauéfinu. : . ”.1”... H .U : . 5.... 57.12.12. u... ..

||

 

This is to certify that the

thesis entitled
EFFECTS OF TERMINAL DROUGHT STRESS 0N BLACK BEANS

presented by

Mark Aar on Frahm

has been accepted towards fummnen t
of the requirements for

J's/’41:; 1.?an and Genet-leg.

ve Action/Equal Opportunity Institution

MS U is an Affi'mmi

 

 

' ——~ - v-rfi '

"N

 

 

TZERARY
Michigan State
University

“.1

 

PLACE IN REI‘URN BOX to remove this checkout from your record,
TO AVOID FINES return on or before date due.

MAY BE RECALLED with earlier due date if requested.

 

 

 

 

 

 

DATE DUE DATE DUE DATE DUE
L

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJClRC/DateDue.p65.p. 15

EFFECTS or TERMINAL DROUGHT STRESS ON BLACK BEANS
By

Mark Aaron Frahm

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE
Plant Breeding and Genetics Program - Department of Crop and Soil Sciences

2002

 

ABSTRACT
EFFECTS OF TERMINAL DROUGHT STRESS ON BLACK BEANS
By

Mark Aaron Frahm

Terminal drought stress severely restricts bean production in Honduras during the
dry season, known as “1a Postrera”. Genetic improvement of common bean (Phaseolus
vulgaris L.) provides a means to assist farmers in production areas affected by drought.
The objectives of this study were to i) identify drought resistant genotypes in two black
bean Recombinant Inbred Line (RIL) populations, ii) evaluate bean root characteristics
for their ability to predict yield performance under stress, and iii) CValuate the
effectiveness ofRAPD markers previously associated to drought resistance in pinto bean.

Two black bean populations segregating for drought resistance were evamated for

yield under moisture stress (Yd) and non-stress (Yp) conditions in Zamorano, Honduras.
Sixteen RILs out-yielded all checks and parents and were identified as drought resistant
based on the geometric mean (GM) of the two treatments. Adaptation of resiStam and
susceptible RILs was tested in Michigan. Despite the low drought stress in Michigan in
2001 , GM was moderately correlated between locations, r = O.63*.
Root length and root architecture were calculated using a pouch method and the
erthionM pro gram. Fine roots and fi’actal dimension were negatively correlated to Yd,
r = ‘0- 1 3*, whereas taproots were positiVely correlated to Yp, r = 0.19". Markers, F06970

and 103 explained 5 % of the variation in Yp in both populations. Root traits

l 130’

Combined with markers accounted for more variation than any one trait alone,

 

ACKNOWLEDGMENTS

I wish to convey my gratitude towards Dr. James D. Kelly for serving as my
major advisor. I appreciate his patience and understanding as I completed this thesis. His
knowledge of plant breeding and common bean is astounding and noteworthy. I will
remember him as the epitome of the plant breeder, who can practically apply his depth of
scientific understanding and who is passionate for his Crop-

I would also like to thank Dr. Foster for taking time out to talk amid the busy
schedules of our lives. Special thanks also go to Dr. Iezzoni who provided insight to
plant breeding and genetics in the classroom and for her patience in the conclusion of this
research. AS a part of my committee, both served valuable roles in my academic growth.

To the bean lab, Jerry Taylor, Norm Blakely and Shitaye MOTges, thank you so
much for your help and our time spent together.

I would personally like to thank Dr. Rosas and PIF for their support in Honduras.
Never before have I planted a bean field by hand or had the opportunities to 16am. A
SPecial thanks to Maria Bravo, Luwbia Aranda, Aracely Castro, Aaron Ortiz and Roberto

Romero for their friendship and kindness to the lone gringo in Honduras.
Lastly, I would like to thank God, my friends and my family for their love and
encouragement throughout this season of my life. I am very grateful for you and am

eXCited to discover how our relationships will grow as I begin a new season. Thank you.

iii

TABLE OF CONTENTS

LIST OF TABLES .............................................................................................................. vi
LIST OF FIGURES ............................................................................................................ ix
KEY OF ABBREVIATIONS .............................................................................................. x
INTRODUCTION ............................................................................................................... l
MATERIALS AND METHODS ....................................................................................... 19
Field Study ............................................................................................................. 19
Parents and Pedigrees. ............................................................................... 19
Population Development ............................................................................ 20
Saginaw, MI 2000 ...................................................................................... 20
Zamorano, Honduras 2001 ------------------------------------------------------------------------- 21
Montcalm, M12001 ................................................................................... 23
Root Protocol. ........................................................................................................ 24
Marker Protocol ..................................................................................................... 25
DNA Extraction ......................................................................................... 25
PCR Protocol ............................................................................................. 26
Electrophoresis ........................................................................................... 26
Statistical Analysis ................................................................................................. 26
RESULTS .......................................................................................................................... 29
Field Study ............................................................................................... 29
Root Study ........................................................................................................... :51
Marker Study ...................................................................................................
Multiple Regression Analysis ............................................................................ :2
DISCUSSION .......................................................................................................
Yield """""""" 59
Root Study .......... 59
Root Length ............................................................................................................... 66
Root Architecture ..................................................................................................... 2;
Molecular Marker572
CONCLUSIONS ................................................................................................................ 74
REFERENCES .................................................................................................................. 76

APPENDIX A: DATA TABLES FROM DROUGHT EXPERIMENTS IN MICHIGAN
AND HONDURAS ............................................................................................................ 85

iv

APPENDIX B: INTROGRESSION OF ROOT ROT RESISTANCE F ROM MIDDLE
AMERICAN LANDRACE BEAN TO CULTIVATED ANDEAN BEAN

GENOTYPES .................................................................................................................. 103
INTRODUCTION ........................................................................................................... 104
MATERIALS AND METHODS ..................................................................................... 106
Backcross #1 ........................................................................................................ 106
Backcross #2 ........................................................................................................ 106

DNA Preparation ................................................................................................. 107

PCR Protocol ....................................................................................................... 107
RESULTS AND DISCUSSION ...................................................................................... 108
Backcross Study ................................................................................................... 108
RAPD Analysis .................................................................................................... 109
REFERENCES ................................................................................................................ 114

APPENDIX C: PRESENCE OF SCAR MARKERS LINKED T O RESISTANCE FOR

COMMON BACTERIAL BLIGHT IN POPULATION L91 ......................................... 115
INTRODUCTION ........................................................................................................... 1 16
MATERIALS AND METHODS ..................................................................................... 117
RESULTS AND DISCUSSION ...................................................................................... 118
REFERENCES ................................................................................................................ 121

LIST OF TABLES

Table 1. Analysis of variance for the RILs at the F 3,5 generation in the L88 and L91
populations for yield and 100 seed weight at Saginaw, MI 2000 ...................................... 31

Table 2. Analysis of variance for 83 genotypes of population L88 in stress and non-stress
treatments from Honduras 2001 ......................................................................................... 34

Table 3. Analysis of variance for 71 genotypes of population L91 in stress and non-stress
treatments from Honduras 2001 ......................................................................................... 35

Table 4. Yield (Yd or Yp), Biomass (BM), and 100 seed weight (100 sw) of the sixteen
RILs selected as highest and lowest yielding based on Geometric Mean (GM) in
population L88 grown under moisture stress and non-stress in Honduras 2001 ............... 38

Table 5. Yield (Yd or Yp), Biomass (BM), and 100 seed weight (100 sw) of the fifteen
RILs selected as highest and lowest yielding based on Geometric Mean (GM) in
population L91 grown under moisture stress and non-stress in Honduras 2001 ............... 39

Table 6. Resistant (16) and susceptible (15) RILs with geometric mean (GM), drought
susceptibility index (D81) and harvest index (HI) under stress and non-stress conditions,
days to maturity (DTM) under the stress treatment and height under the stress and non-
stress treatments in Honduras, 2001 .................................................................................. 40

Table 7. Analysis of variance for 36 genotypes grown under stress and non-stress
treatments in Montcalm, MI 2001 ..................................................................................... 41

Table 8. Yield under stress (Yd) and non-stress (Yp) and 100 seed weight for sixteen
genotypes ranked by Geometric Mean (GM) in population L88 grown in Montcalm, MI
2001 under stress and non-stress treatments ..................................................................... 42

Table 9. Yield under stress (Yd) and non-stress (Yp) and 100 seed weight for fifteen
genotypes ranked by Geometric Mean (GM) in population L91 grown in Montcalm, MI
2001 under stress and non—stress treatments ...................................................................... 43

Table 10. Correlation between yield, biomass, 100 seed weight (100 sw), plant stand at
harvest and disease incidence (DI) at 45 and 75 days after planting in the moisture stress
and non-stress treatments for L88 and L91 RILS in Honduras, 2001 ................................ 47

Table 11. Correlations between yield-based traits including 100 seed weight (100 sw) and

harvest index (HI) and agronomic traits including desirability score (DS) in L88 (below
diagonal) and L91 (above diagonal) in Honduras 2001 ..................................................... 48

vi

Table 12. Selected genotypes and means compared for their disease incidence (DI), plant
stand at harvest, yield under stress (Yd), yield under non-stress (Yp), and geometric mean
(GM) of moisture treatments grown in Honduras, 2001 .................................................... 50

Table 13. Analysis of Variance for the 81 RILS in population L88 for Total root length,
Fractal Dimension, and root length according to 10 different diameter widths (A-J) ....... 52

Table 14. Correlation values between root characteristics and yields in the Saginaw 2000
and Honduras (Hon) 2001 experiments for the 81 RILS of population L88 ...................... 53

Table 15. Mean values of total root length, fractal dimension, fine roots (A-C) and
taproots (H-J) of drought resistant and drought susceptible RILS and parents of
population L88 obtained by the root pouch method .......................................................... 54
Table 16. Presence/absence of RAPD markers in six bean genotypes .............................. 56
Table 17. Coefficients of determination (R2) accounting for the variation in yield under
drought (Yd), yield under non-stress (Y p) and geometric mean (GM) for two RAPD
markers ............................................................................................................................... 56
Table 18. Coefficient of determination (R2) selection method for yield under stress in
Honduras evaluating fractal dimension, root classes B (0.5-1.0 mm) and I (4.0-4.5 mm),
and RAPD marker F06970 (F06) ......................................................................................... 58
Table 19. Coefficient of determination (R2) selection method for yield under non-stress in
Honduras evaluating fractal dimension, root classes B (0.5-1.0 mm) and I (40-45 mm),
and RAPD marker F06970 (F 06) ......................................................................................... 58
Table A1. Yield, rank and 100 seed weight for the 150 RILS in Saginaw, MI 2000 ......... 86
Table A2. Field data for 160 genotypes from drought treatment in Honduras 2001 ......... 88
Table A3. Field data for 160 genotypes from non-stress treatment in Honduras 2001 ..... 92
Table A4. Yield under stress (Yd), yield under non-stress (Yp), and geometric mean
(GM) of 160 genotypes adjusted for plant stand by covariate analysis for the Honduras
experiment 2001 ................................................................................................................. 96
Table A5. Field data of 36 genotypes under drought stress at Montcalm, MI 2001 .......... 98
Table A6. Field data of 36 genotypes under non-stress at Montcalm, MI 2001 ................ 99
Table A7. Mean values of Fractal Dimension, total root length, fine roots (A-C) and

taproots (H-J) for the 81 RILS of population L88 ............................................................ 100

vii

Table A8. Presence/absence of RAPDs associated with drought resistance in T-3016 (T),
Sierra (S), B98311 (B), Raven (R), N98122 (N), Huron (H), VAX 5 (V) and TLP 19

(P) ..................................................................................................................................... 102
Table Bl. Morphological characteristics of BC,Fl cranberry and kidney plants ........... 111

Table BZ: Summary of seed increase methods ............................................................... 112
Table B3: Commercial varieties and breeding lines crossed with BC IFl individuals ..... 112

Table B4: Presence or absence of RAPD markers present in Negro San Luis (NSL),
Redhawk and C97407 compared to the check varieties, FR266 and Montcalm ............. 113

Table C1. Presence/absence of the SAP6 and SU91 SCAR markers for 69 RILS in
population L91 ................................................................................................................. 119

Table C2. Marker and agronomic characteristics of eleven genotypes possessing both
SCAR markers along with the parents ............................................................................. 120

viii

 

LIST OF FIGURES

Figure 1. Frequency Distributions for yield using the adjusted means from each
population in Saginaw, M12000. The drought resistant parent, B98311, is indicated by
(B) ...................................................................................................................................... 30

Figure 2. Frequency Distributions for yield under stress (Yd) and non-stress (Y p) using
the adjusted means from each RIL from population L88 in Honduras. Mean yield of the
parents, B98311 (B) and TLP 19 (T) are noted ................................................................. 32

Figure 3. Frequency Distributions for yield under stress (Yd) and non-stress (Y p) using
the adjusted means from each RIL from population L91 in Honduras. Mean yield of the
parents, B98311 (B) and VAX 5 (V) are noted ................................................................. 33
Figure 4. Frequency Distribution of 150 RILs showing selection of resistant (R) and
susceptible (S) lines based on geometric mean. Parents VAX 5 (V), TLP 19 (T), and
B98311 (B) are included .................................................................................................... 36

Figure 5. Regression analysis of Resistant and Susceptible RILs for Geometric Mean
across the Honduran and Montcalm, MI locations.......... .................................................. 44

Figure 6. Regression of Selected RILS for yield under stress among the Honduran and
Montcalm, MI locations ..................................................................................................... 45

Figure 7. Regression of selected RILs for yield under non-stress conditions within the
Honduran and Montcalm, MI locations ............................................................................. 46

Figure 8. Field incidence of Macrophomina phaseolina compared to yield under stress in
160 genotypes grown in Honduras in 2001 ....................................................................... 50

Figure 9. Representations of herringbone (A) and dichotomous (B) topologies ............... 52

Figure B1. Diagram of inheritance of dwarf lethal genes D11 and D12 from the initial cross
to the BC,Fl ...................................................................................................................... 110

Figure B2. Diagram of the inheritance of morphological markers between kidney and
black bean parents ............................................................................................................ 110

ix

100 sw

AN OVA
ASB
BGMV
cb
CBB
CIAT
cM
CRD
CV
dap
DI
DII
DS
DSI
F sp
GCA
GM
HI
LG
LSD
MAS
masl
MI
NSL
PCR

PIF
PM

ppi
QTL

Ra
RAPD
RCBD
RIL
Rp

KEY OF ABBREVIATIONS

Carbon Isotope Discrimination [(Ra/Rp-1)]
100 Seed Weight

Altitude

Arithmetic Mean

Analysis of Variance

Ashy Stem Blight

Bean Golden Mosaic Virus

Centibars

Common Bacterial Blight
International Center for Tropical Agriculture
Centi-Morgan

Completely Randomized Design
Coefficient of Variation

days after planting

Disease Intensity

Drought Intensity Index [l-(Xd/Xp)]
Desirability Score

Drought Susceptibility Index [(1-(Yd/Yp))/DII)]
F usarium solani pv. phaseoli

General Combining Ability
Geometric Mean

Harvest Index

Linkage Group

Least Significant Difference

Marker Assisted Selection

Meters Above Sea Level

Michigan

Negro San Luis

Polymerase Chain Reaction

Exterior Path Length

Programa de Invéstigaciones en Frij 0]
Physiological Maturity

pre-plant incorporation

Quantitative Trait Loci

Pearson Correlation Coefficient
Coefficient of Determination

Ratio of Carbon in the Atmosphere
Randomly Amplified Polymorphic DNA
Randomized Complete Block Design
Recombinant Inbred Line

Ratio of Carbon in the Plant

SCA
SCAR
SSD
Xd

Yd
Yp

Specific Combining Ability

Sequence Characterized Amplified Region
Single Seed Descent

Mean yield under moisture stress conditions
Mean yield under non-stress conditions
Yield under moisture stress conditions
Yield under non-stress conditions

xi

 

INTRODUCTION

Drought is the second major constraint after disease to negatively affect yield of
common bean, Phaseolus vulgaris L. Approximately 60 % of the bean crop in the
developing world is produced under drought stress (Graham and Ranalli, 1997). An
example of bean production under stress occurs in the lowland tropical areas of Central
America. In Honduras, the bimodal pattern of rainfall permits two seasons of crop
production. The first season, la Primera, is known as the rainy season because 54 % of
the annual rainfall occurs (Cotty et al., 2001). Following 1a Primera (May-Aug), less
frequent rainfall and diminishing soil moisture create the terminal drought stress in the
second production season known as, 1a Postrera (Sept-Dec.) The short life-cycle of
common bean makes it an ideal crop to grow at the end of la Primera. More than 60 % of
the area cultivated to bean in Honduras is planted in la Postrera under a relay system after
corn (Zea mays) has reached physiological maturity or after the corn has been harvested
(Rosas et al., 1991). The bean production area in Honduras increases three-fold during 1a
Postrera despite an overall yield reduction of 50 % due to terminal drought (Cotty et al.,
2001). Since adequate irrigation schemes are unrealistic due to socio-economic
constraints, genetic improvement for drought resistance offers a long-term improvement
of bean productivity under drought stress in Honduras.

The genetic improvement of drought resistance in common bean has been
previously documented (Acosta-Gallegos and Shibata, 1989; Acosta-Gallegos and
Adams, 1991; White et al., 1994a; Singh, 1995; Schneider et al., 1997b; Abebe et al.,
1998; Ramirez-Vallejo and Kelly, 1998; Rosales-Sema et al., 2000; Teran and Singh,

1

2002). Drought resistance can be compared to the evolutionary success of plant
adaptation. Plant adaptation is defined as the relative ability of plants to survive and
produce more biomass and progeny (seed) compared with other plants growing in the
same environment (Hall, 1993). Drought resistance is based on relative yield of a
genotype compared with other genotypes subjected to the same drought and where
drought escape is not a major factor (Hall, 1993). Yield in common bean has been
reduced by 58 % due to water stress (Acosta-Gallegos and Adams, 1991). Each yield
component, pods per plant, seeds per pod and seed weight per 100 seeds, has shown
varying negative responses to water stress (Acosta-Gallegos and Shibata, 1989; Acosta-
Gallegos and Adams, 1991; Ramirez-Vallejo and Kelly, 1998). Pods per plant is the one
yield component most affected by water stress (r = 0.56; Acosta-Gallegos and Adams,
1991). Seeds per pod and 100 seed weight are reduced by water stress but to a lesser
extent than pods per plant. Yield is measured under moisture stress (Yd) and non-stress
conditions (Yp) to calculate drought resistance of individual genotypes. Non-stress
conditions maintained by irrigation reveal the yield potential (Yp) of genotypes. Both
yield variables are commonly combined in different equations to identify genotypes
stable across diverse environmental conditions.

The measurement of drought resistance in common bean has been a topic of
discussion among breeders for many years. Yield differential (Yp-Yd) was commonly
used as a selection criterion for drought resistance, yet it was shown to be
counterproductive due to the likelihood of selecting a low yielding genotype with a
relatively small yield differential due to drought (Samper, 1984). Arithmetic mean of
stress and non-stress treatments (AM = (Y p+Yd)/2) was suggested as selection criteria,

2

based on theoretical experiments (Rosielle and Hamblin, 1981). However, selection for
drought resistance based on AM could be confounded due to genotypes with high yield
potential and low yield under stress.

The variation in yield potential between genotypes can be determined by the
drought susceptibility index (DSI) (Fischer and Maurer, 1978). DSI is a dimensionless
slope calculated from the following formula: DSI = (1-(Yd/Yp))/DII where D11 is the
drought intensity index of the experiment. D11 is calculated by DII = 1-(Xd/Xp), Xd and
Xp being the mean yield of the drought and irrigated treatments, respectively. Since the
D11 of the experiment is considered, individual genotypes can be compared across
locations using DSI. Geometric mean, GM = (Y p*Yd)"’, was introduced as a calculation
that takes into account yield data from both treatments and represented an actual yield
measurement of the genotype. Geometric mean differs from arithmetic mean by
moderating inflated or diminished means resulting from extreme values between
treatments.

Four different selection criteria including yield differential, AM, DSI and GM
were compared for their potential to evaluate drought resistance in common bean
genotypes (Samper and Adams, 1985). Twenty-two bean genotypes of diverse origin
were ranked according to each criterion. Genotypes were ranked similarly based on yield
differential and DSI. Rankings based on AM and GM were similar, yet genotypic
rankings based on GM were drastically different than DSI rankings. A possible
explanation is that the D81 identified low-yielding genotypes that could tolerate drought
well, whereas GM better reflected the actual yield potential of the genotype. The most

effective approach in selection for drought resistance in common bean is based on

3

sequential selection for high GM yield, followed by high Yd yield, low to moderate D81
and harvest index (HI) values (Schneider et al., 1997b; Ramirez-Vallejo and Kelly, 1998).

The genetic makeup of populations created for drought resistance is an important
factor to consider. Interracial populations have been suggested as the most effective way
to combine high yield with drought resistance among different races of common bean
(Singh et al., 1991; Singh, 1995; Teran and Singh, 2002). Genotypes from the Durango
race showed higher yields, larger seed weights and earlier maturity than genotypes from
the Jalisco race (Teran and Singh, 2002). Durango genotypes have an indeterminate type
IH growth habit and a life cycle less than 120 days while Jalisco genotypes exhibit a
climbing type IV growth habit and a life cycle greater than 150 days. For these reasons,
the Durango race is preferred by breeders in interracial crosses to the Mesoamerican race
to improve drought resistance.

Interspecific hybridizations between common bean and tepary bean (Phaseolus
acutzfolius A. Gray) have also been suggested to improve drought resistance since tepary
bean has exhibited high levels of drought tolerance (Thomas et al., 1983; Rosas et al.,
1991). Obtaining viable offspring from interspecific crosses is impossible without
embryo rescue. Over 1500 plants were generated in a P. vulgaris x P. acutz'folius
hybridization where embryo rescue was employed (Mejia-Jimenez et al., 1994).
Recurrent and congruity backcrossing was implemented to overcome any incompatibility
barriers. Tepary beans, highly resistant to common bacterial blight, were successfully
introgressed into common bean germplasm (Singh and Munoz, 1999), yet the impact of
interspecific hybridizations to enhance common bean germplasm for drought resistance

has been limited.

 

Breeding for drought resistance is more difficult due to the quantitative nature of
inheritance. Expression of quantitative traits result from independent segregation of
many genes that have small effects and are more affected by environmental variation
(Paterson et al., 1990). Drought resistance exhibits continuous variation and heritability
estimates have generally been low. Reported values for heritability of drought resistance
in common bean range from 0.09 to 0.80 (White et al., 1994a; Singh, 1995; Schneider et
al., 1997b; Ramirez-Vallejo and Kelly, 1998). The wide range of heritabilities results
from differences in genetic variability among populations and different intensities of
stress. General combining ability (GCA) and specific combining ability (SCA) were
calculated for yield under drought stress from a nine bean diallel grown in tropical mid-
elevation regions (altitude 800-1600m) and semi-arid highlands (1700-2400m) (White et
al., 1994a). GCA for yield was consistently significant and larger than SCA in both
environments. These results suggested the importance of additive gene effects for yield
and 100 seed weight of bean grown under stress. Parental genotypes adapted to both
environments were used in the diallel crosses. At the highland location in Durango,
Mexico, parental genotypes adapted to the mid-elevation environment showed negative
GCA values while all highland genotypes were positive. Reciprocal results occurred in
the mid-elevation location where mid-elevation parents showed positive GCA values
while highland parents had negative GCA values. These location effects underscore the
importance of identifying the target environment before choosing parents to improve
drought resistance in common bean.

The expression of drought resistance or the adaptation to stress is more clearly

illustrated when individual genotypes are compared between locations. Two RIL

5

 

populations of the Durango race were evaluated for drought resistance under two locations
in Michigan, two locations in Zacatecas, Mexico and three in Durango, Mexico (Schneider
et al., 1997b). Yield calculations were made using data from all seven locations
(Schneider et al., 1997b) and from the three locations in Durango (Rosales-Sema etal.,
2000). Different RILS ranked in the top five based on GM for each experiment. These
differences can be explained by the limited ability of the Durango race to adapt to different
environments and the evasive nature of drought resistance.

Drought stress occurs in two contrasting moisture environments (intermittent and
terminal) of the semiarid tropics (Ludlow and Muchow, 1990). Intermittent drought is due
to climatic patterns of sporadic rainfall that causes intervals of drought. The nature of this
rainfall is unpredictable and leads to marginal yields in potentially valuable land. This
rainfall pattern is chronic and endemic to the semiarid highlands (1800 masl) of Mexico
(Singh, 1995). Terminal drought occurs when plants suffer from a lack of water only at
later stages of grth or when crops are planted in a dry season. This farming practice
predisposes the crop to a terminal drought in two very important phases of its life cycle;
flowering and pod-fill. Terminal drought characteristically occurs in lowland tropical
areas when the bean crop is planted at the end of the rainy season.

Different growth habits in common bean offer unique adaptive advantages to the
different types of drought. The type 11 growth habit is characterized by an indeterminate,
upright plant structure with reduced branching angle whereas the type III habit is typical of
an indeterminate prostrate sprawling plant structure (Brothers and Kelly, 1993). The
desired grth habit for resistance to intermittent drought in the Mexican highlands is a
type III plant. The prostrate canopy has an opportunistic growth pattern when moisture is

6

 

available which helps retain moisture in the soil by shading whereas erect growth habits
allow soil moisture to be lost during hot and windy days. Type III genotypes can be
planted at lower densities to reduce inter-plant competition since they have a sprawling
superficial root system that is able to utilize soil moisture in a wider zone than deeper,
narrow taproots of type II growth habit. The type 111 growth habit also produces many
root meristcms and basal roots to access soil moisture in a wider superficial zone (Lynch
and van Beem, 1993).

In the terminal drought environment, a deep penetrating root is needed to maintain
relative water content in the bean plant during the ever-intensifying dry period. The ideal
growth habit for this stress would be a type II. A striking feature of this growth habit is its
deep penetrating root system. The root system of the type 11 grth habit has a
herringbone structure which characteristically goes deep into the soil profile to extract
moisture (Lynch and van Beem, 1993). The erect architecture of the type II shoot allows
continued transpiration to be sustained by deep penetrating roots, so that the plant can
deliver an acceptable yield under terminal drought.

Root architecture is associated to shoot architecture. Shoot height was used as a
selection criterion to predict root depth in soybean (Glycine max) (Mayaki et al., 1976).
Water stressed and non-stressed treatments were used to measure differences in shoot
height and root depth in the field. Root depth was correlated to shoot height (R2 = 0.99)
(Mayaki et al., 1976). The root depthzshoot height ratio was 2:1 from six node stage to
pod fill stage in stressed plots. In non-stressed plots, the 2:1 ratio decreased to 1.4:]
during pod initiation. This research offered a quick and non-destructive method of
predicting rooting depth in the field. Although root growth has been correlated to yield

7

 

under drought stress, its use as an efficient screening technique in common bean has been
limited, due to the cumbersome nature of measurements.

The intensity of the moisture stress can be detrimental to most physiological
functions. Nitrogen fixation has been studied under differing degrees of drought stress
(Acosta-Gallegos, 1988; Foster et al., 1995). In an experiment where DII = 0.41, N
partitioning from the leaf to the seed in common bean was not impaired (Foster et al.,
1995). N-remobilization was severely affected by a more severe stress, DII = 0.92, and
has been suggested as an important drought adaptation strategy under moderate or
intermittent moisture deficits (Foster et al., 1995). N partitioned to seed also decreased
with terminal drought in other leguminous species (Chapman and Muchow, 1985).

Above-ground biomass is one physiological trait that correlates well to yield under
water stress, (r = 0.79), despite the severity of moisture stress (Acosta-Gallegos, 1988).
Since it accounts for the total nutrients fixed in vegetative grth and seed production,
increased biomass is often associated with late maturity in common bean. Harvest index
(HI = plot yield/ biomass) accounts for the efficiency of plant partitioning the nutrients to
seed production. HI must be combined with biomass when selecting for performance
under stress.

Three mechanisms that plants use to respond to water stress are escape, avoidance
and tolerance (Ludlow, 1989). Desert annuals and short season, annual crops use the
escape mechanism during water stress. In Honduras, landraces such as Cuarentefio,
Cincuentefio and Chingo that reach maturity within 65 days are planted by farmers to
escape drought (Rosas et al., 1991). Although earliness is popular among farmers, the trait
is negatively correlated to yield. Drought resistance must combine avoidance and/or

8

 

tolerance mechanisms but not escape mechanisms (Fischer and Maurer, 1978). Breeding
lines with improved yield potential under non-stress must be combined with avoidance
and tolerance traits to increase drought resistance.

The mechanisms of avoidance and tolerance are not mutually exclusive in all
drought resistant traits but their definitions are unique. Plants that avoid drought must do
so because they have tissues that are sensitive to dehydration (Ludlow, 1989). These
plants respond to drought stress by maximizing water uptake and minimizing water loss.
Drought tolerant plants are insensitive to dehydration (Ludlow, 1989). They are
characterized as having a high osmotic adjustment. Since different mechanisms operate in
plants, numerous physiological mechanisms have been evaluated as screening techniques
for yield under drought stress.

Drought tolerance mechanisms involving leaf gas exchange affected by water
stress were studied in common bean (F arquhar et al., 1989; Ehleringer et al., 1990; White
et al., 1990; Ehleringer et al., 1991; White et al., 1994b). Carbon isotope discrimination
(A) was used as an indicator of water use efficiency and adaptation to water deficits in
common bean (White et al., 1994b). Carbon isotope discrimination is defined as A =
(Ra/Rp-l) where Ra and Rp are the 13C/ 12C ratios of carbon in the atmosphere and plant,
respectively (White et al., 1990). A is directly proportional to the intercellular CO2
concentration. With this measurement, higher photosynthetic rates can be derived from
higher A values. The A measurement has only been significantly correlated to biomass
and not to yield. Although A is unsuitable as a screening technique for yield under
drought stress, it could be used to identify different adaptation mechanisms present in

common bean (White et al., 1994b).

 

Roots are recognized as playing an important role in drought avoidance in common
bean. Greater root growth supports yield in common bean through drought avoidance by
extracting more soil moisture at greater depths. Roots of drought resistant bean genotypes
reach greater depths than those in non-stress soils and were hypothesized to be an
important drought avoidance mechanism (Sponchiado etal., 1989).

The root and shoot characteristics of common bean genotypes under water stress
were compared for their association to yield under drought stress in fertile and acidic soils
(White and Castillo, 1989; White and Castillo, 1992). Root and shoot genotypes of
drought resistant and susceptible genotypes were combined through grafting. The
resulting plants were transplanted to the field for evaluation under drought conditions.
When the root of drought resistant genotype, BAT 477, was grafted onto the shoots of
BAT 477 and drought susceptible genotype BAT 1224, the plants yielded 600 and 840
kg/ha, respectively under drought stress. In the reciprocal graft using BAT 1224 as the
root genotype, the shoots of BAT 477 and BAT 1224 yielded 160 and 30 kg/ha
respectively, compared with the normal yields (700 and 40 kg/ha) of BAT 477 and BAT
1224 grown under water stress. This data suggests that the bean root genotype is more
important in drought resistance than the shoot genotype. In both, fertile and acidic soils,
the root genotype had a large and significant effect on yield while the shoot genotype had
no effect. The root systems of four food legumes were compared for their response to
drought (Pandey et al., 1984c). Peanut (Arachis hypogaea L.) with the most extensive
root system when compared to the other three legumes, showed greater yield (Pandey et
al., 1984a) and cooler canopy temperatures (Pandey et al., 1984b) under water stress
suggesting the important role that roots play in drought tolerance.

10

Root reaction of plants to water stress affects stomatal response. In wheat and
sunflower, roots that detect the soil drying consequently sent a message to the leaves,
which induces the. stomates to close (Gollan et al., 1986). This signal was reproduced and
shown to be related to the metabolism of cytokinins (Schulze, 1986). In common bean,
roots under moisture-stressed conditions produced a signal that was transported to the
leaves causing a continuous decline in stomatal conductance (Aguirre-Medina et al.,
1998). Shoot responses to moisture stress detected by roots is also observed as
paraheliotropic leaf movements in common bean (Kao et al., 1994). This paraheliotropic
movement of the shoot decreases the incidence of solar radiation and ultimately minimizes
water loss.

Screening techniques for drought resistance are important since improving cr0ps in
tropical environments by selecting solely on grain yield is problematic because of the
variability in amount and annual distribution of rainfall (Ludlow and Muchow, 1990).
Breeding for high yield would be more efficient if traits correlated to yield under water
stress were identified and could be used in selection. Screening techniques for drought
resistance would be valuable to plant breeders to reduce variety development time and
resource expenditures.

Many physiological measurements have been suggested as an indirect screen for
yield in early generations following hybridization. These traits including 100 seed weight,
leaf area of primary leaves, stem and total dry weight as well as hypocotyl diameter have
been significantly correlated to seed yield in bean (Acosta-Diaz, 1998). The response of
leaf angles to sunlight was suggested as a valuable trait for selection in drought
environments due to its correlation in water use efficiency of the plant (Kao et al., 1994).

ll

Although, these physiological measurements indirectly relate to drought tolerance, their
application as a screening technique can be laborious and time-consuming.

The most recent method in which traits are being indirectly selected is based on
molecular markers linked to the trait of interest. As a screening technique, molecular
markers can be used to screen large numbers of individuals in a relatively short amount of
time. Molecular markers have been associated with qualitative and quantitative traits in
common bean (Kelly et al., 2002, in review). Markers linked to single genes for disease
resistance in anthracnose (Young and Kelly, 1996), bean common mosaic virus (Haley et
al., 1994), bean golden mosaic virus (Urrea et al., 1996), and bean rust (Miklas et al.,
1993) have been developed. Markers have been useful in the identification of single genes
masked by epistatic effects and the building of gene pyramids in common bean (Kelly et
al., 1995). Breeders unable to phenotypically screen for disease resistance can use
markers as a selection criterion. Marker Assisted Selection (MAS) allows selection of
traits in early generations. With MAS, breeders can reduce the number of breeding lines
depending on the presence of the marker and few individuals need to be screened to
identify superior genotypes.

Quantitative traits linked to a simply inherited genetic marker were first observed
in common bean with the co-segregation of seed size and seed coat color (Sax, 1923).
More recently markers associated to quantitative trait loci (QTL) in common bean have
been identified for resistance to ashy stem blight (Miklas et al., 1998), common bacterial
blight (Nodari et al., 1993), BGMV (Miklas et al., 1996), web blight (Jung et al., 1996),
white mold (Miklas et al., 2001), root rot (Schneider et al., 2001), and drought (Schneider
et al., 1997a). Success of QTL analysis has centered on the identification of a few major

12

loci controlling quantitative trait expression. The discovery of major QTLs explaining
large percentages of genetic variation of quantitative traits has encouraged the use of
MAS. The effectiveness of MAS for quantitative traits is inversely proportional to the
heritability of the trait being selected (Lande and Thompson, 1990). Markers for QTLs
associated with drought resistance have been detected in common bean (Schneider et al.,
1997a), rice (Oryza sativa) (Champoux et al., 1995), sorghum (Sorghum bicolor L.
Moench) (Kebede et al., 2001), soybean (Specht et al., 2001), maize (Ribaut et al., 1997),
and barley (Hordeum vulgare L.) (Teulat et al., 1998). In soybean, a major QTL
accounted for 33 to 38 % of the phenotypic variation in yield under various irrigation
regimes (Specht et al., 2001).

In common bean, RAPD markers associated with drought resistance were
identified and used in MAS (Schneider et al., 1997a). Seventy polymorphic primers were
screened across two RIL populations. Nine linkage groups were identified in one
population and ten in the other. A linkage group fiom each population was significantly
associated with Yd, Yp and/or GM. One linkage group explained 8-14 % of the genetic
variation combined across all locations while the other explained 10-16 %. These linkage
groups were used in MAS. Yield under stress among genotypes selected by MAS in one
population was improved by 10 g/m2 despite a severe drought stress (DII = 0.76) imposed
in Michigan. When these same genotypes were grown in two Mexican locations,
significant differences were not detected between drought resistant and susceptible
genotypes. Only significant differences were detected for the second population in the
Mexican locations. MAS was more effective than conventional selection in one of the
populations where heritability estimates for yield were lower (Schneider et al., 1997b).

13

Recently, researchers have used molecular markers to identify and characterize
root morphology traits in rice (Champoux et al., 1995; Lilley et al., 1996; Zheng et al.,
2000). Drought resistance is an important trait for rice breeders as 40 % of the area
planted to rice worldwide experiences water stress. Subsistence farmers grow rice without
irrigation in lowland and highland environments. Root traits perform an important
mechanism in avoiding drought in rice. Researchers generated a mapping population to
study drought resistance in rice by crossing a japom'ca cultivar, Moroberekan, to an indica
cultivar, C039 (Champoux et al., 1995). Moroberekan is a drought resistant cultivar
grown in the highlands and is known to possess a deep, thick root system. C039 is a
lowland cultivar susceptible to drought with a shallow root system yet possessing high
dehydration tolerance traits. Root thickness, root/shoot ratio and root dry weight per tiller
were recorded for 203 RILs in 3 different greenhouse experiments. Plant response to
drought stress was recorded visually by the degree of leaf rolling. This drought avoidance
trait was associated with three root traits mapped to various locations on 10 different
chromosomes. Most of the QTL identified for root characteristics clustered around
chromosomal regions conferring drought avoidance. Markers associated with these root
traits would facilitate selection for otherwise hard-to-score root traits. The linkage map
was used as a basis to add additional drought-related traits in subsequent studies (Lilley et
al., 1996). Osmotic adjustment at 70 % relative water content and lethal osmotic potential
which are characterized as drought tolerant traits were added (Lilley et al., 1996). Three
of the five QT L associated with drought tolerance were mapped to the same chromosomal
regions as were the root traits. Since the drought tolerance and drought avoidance traits
were inherited from separate parents the drought tolerant traits were negatively associated

l4

 

with the root traits that aid in drought avoidance. Linked markers should facilitate the
process of breaking the negative linkage between the avoidance and tolerance traits.

Technological advances have improved our understanding of roots. Historically,
the line intersect method provided the first easy way to estimate the total root length in
plants (Newman, 1966). The excised root is placed in an area with randomly spaced lines.
For each root sample only the number (N) of intersections between root and the straight
lines is recorded. The total length of the straight lines (H) and the observing area (A)
remain constant among experiments. The proposed equation of R = 1rNA(2H)", allows a
fast estimation of root length. The line intersect method can be used to measure 3.43
meters of root in 24 minutes with a coefficient of variation of 4.3 % while direct
measurement took 67 minutes (Newman, 1966). This method was later revised to replace
the randomly oriented lines with a grid and a length conversion factor for A and H
(Tennant, 1975).

The next advance was to study the root system in viva. All previous methods
involved excavation of roots from the soil. In Georgia, an underground laboratory called a
Rhizotron was built (Box, 1996). Angled glass acted as the ceiling of this laboratory.
Roots would grow next to the glass and the roots could be monitored as they grew (Taylor
et al., 1970). The initial investment is too high for the data collected using Rhizotron
technology. This level of technology benefitted the development of the understanding of
root physiology. It is unlikely that information generated through the Rhizotron
technology will aid in breeding since large numbers of genotypes need to be evaluated.

Mini-rhizotrons were developed with the aid of miniature cameras. A glass tube
penetrating the ground at an angle intersects the roots. A miniature video camera can

15

 

traverse the length of the tube and record the grth of the roots at different depths. In a
drought stress experiment in com, a mini-rhizotron showed that a short-term drought
resulted in root losses near the soil surface and large increases in deep root growth (Box et
al., 1989). Root grth over time and root morphology characteristics are measured in
viva. Root images are stored in a video format for a computer analysis.

The first computer program used to analyze root images was the DOS-based Delta-
T Scan (Harris and Campbell, 1989). The program was used to measure root length,
projected area and average diameter. Although the commands and print-out are difficult to
understand, the program allowed the recording of more measurements in a shorter time. A
window-based program, WinRhionM, also measures multiple factors and has easy to use
commands and an easy to understand print-out of the analysis. Roots are scanned into the
computer and a digital image of the root is used to measure length based on pixel size.
Resolution to distinguish root parts is very fine and a color analysis can be conducted to
separate root parts based on color differences. Roots discolored by disease infection can
be separated from healthy roots in WinRhionM. Length separated by diameter and root
morphology characteristics such as topological indices and fractal dimension can also be
measured.

Topology is a method of mathematically describing the root system’s branching
structure. Root systems are, in large part, trivalent branching structures meaning that each
node or vertex has three branches or links (Fitter, 1996). The number of links in a system
can be separated into exterior links which end in a meristem and intemal links which join
other links. The magnitude of any individual link is the number of exterior links it serves.
Other measurements such as the length of links, branching angle, distribution of branches

l6

and relative diameter can describe the system in further detail. Plants with equal
magnitude can vary in branching structure from herringbone at one extreme to
dichotomous to the other (Figure 9). These branching patterns can be quantified using two
parameters, altitude (a), and the exterior path length (p6). Altitude is the number of links
in the longest path connecting an exterior link to a base link. Exterior path length is the
sum of the number of links in all such paths. In common bean, a type 11 grth habit
exhibits a herringbone structure while a type I plant tends to have a dichotomous root
structure (Lynch and van Beem, 1993).

Root growth and architecture may be associated with genotypic adaptation to water
stress. Four different grth habits of common bean and three different root parts
(taproot, taproot laterals, and basal roots) were evaluated for grth rates, dry weight and
final root length (Lynch and van Beem, 1993). The taproot lengths were similar in all
genotypes. No genetic differences were observed in specific root length (length/weight).
Significant genotypic differences were observed in root branching patterns. The number
of apical meristems was highest in type III than type I growth habits. Topological indices
differed significantly between type II (heningbone) and type I (dichotomous). This
research supports an association between root and shoot architecture. Topology and
number of meristcms were very descriptive of grth habit and root architecture. Root
length, dry weight and fractal analysis were equal in usefulness. The utility of fractal
dimension as a selection criterion requires further study in bean.

Fractal analysis has been related to plant root systems (Tatsumi et al., 1989).
Various objects in nature such as clouds, mountains, coastlines and trees have been
described by fractal geometry (Mandelbrot, 1977). The intricacy of shape of the root

17

systems is characterized by the slope of each line as an estimate of the fractal dimension,
D. Methods to quantify root morphology, such as topology and fractal dimension have
been developed but have not been widely applied.

Numerous methods have been used to collect bean root data. Field (Y an et al.,
1995), pouch (McMichael et al., 1985; Yabba, 2001), split-root (Snapp et al., 1995;
Aguirre-Medina et al., 1998), hydroponic (Gabelrnan et al., 1986; Checkai et al., 1987)
and soil-filled PVC tube (Yabba, 2001) mediums have been used to collect root samples.
For breeding purposes, a quick and efficient method of collection and analysis is desired.
Soil-less mediums are less laborious and time consuming. The roots are free from soil or
debris so that measurement is fast and efficient.

The objectives of this study was i) to identify drought resistant genotypes from two
black bean RIL populations grown under moisture stress and non-stress conditions in
Central and North America and ii) to evaluate root characteristics and previously reported
RAPD markers associated with drought resistance for their ability to predict yield under

stress in the two RIL populations.

18

MATERIALS AND METHODS

Field Study
Parents and Pedigrees

Three black bean genotypes were crossed to produce two RIL populations
segregating for drought resistance. The drought resistant genotype, B98311, was
originally derived from a cross between drought resistant breeding line, T-3016 and the
Michigan cultivar, Raven. T-3016 is a non-commercial Durango race breeding line
previously identified as the most drought resistant genotype based on GM from a cross of
Sierra/AC1028 (Schneider et al., 1997b). T-3016 was previously evaluated for root length
(Y abba and Foster, 1997) and RAPD markers associated with drought resistance
(Schneider et al., 1997a). Raven is an early-season black bean with resistance to
anthracnose and Bean Common Mosaic Virus (Kelly et al., 1994). During the 1998
drought in Michigan, B98311 was selected as the highest yielding genotype under stress
(Kolkman and Kelly, 1999).

TLP 19 was developed for tolerance to low phosphorous at the International Center
for Tropical Agriculture (CIAT). Phosphorus-efficient bean genotypes respond to
phosphorus stress by developing a shallow root system (Liao et al., 2001). The
contrasting root architecture of B98311 and TLP 19 was considered in parental selection
to create genotypes with different root systems which could aid in drought resistance.
Under terminal drought stress in Mexico, TLP 19 has shown resistance to Macrophomina
phaseolina (Tassi) Goid., the causal fungus of ashy stem blight (ASB), a disease that is
prevalent under water stress conditions (Mayek-Pérez et al., 2001a; Mayek-Pérez et al.,

19

2001b). The third genotype, VAX 5, was developed at CIAT from an interspecific
hybridization of common and tepary bean and selected for resistance to common bacterial
blight (CBB) (Xanthomonas campestris) (Singh and Munoz, 1999). TLP 19 and VAX 5
were selected as parents for their adaptation to lowland-tropical conditions and good
combining ability with B98311, adapted to temperate conditions. Additional traits such as
commercial seed type, growth habit and disease resistance were considered in the
selection of parents in order to hasten the utilization of any beneficial black genotypes

resulting from this work in the Latin American/Caribbean region.

Population Develogment
The original crosses made in 1998 were B9831 l/T LP 19 and B9831 1N AX 5

which generated populations L88 and L91 respectively. In September 1999, single pods
from each F2 plant were harvested in both populations. F3 seed was advanced to the F4
generation using single seed descent (SSD). Single pods were harvested from F 3 plants
and the SSD process was repeated. The last single plant selection was made in the F3
generation so that seed planted in the greenhouse was at the F3.4 generation. Seed from
each F3.4 genotype was harvested in bulk. This F3.5 seed was planted in Saginaw, MI in
2000 to increase the amount of seed and F3.6 seed was shipped to Honduras for testing in
2001. A total of 81 RILs in L88 and 69 RILs in L91 population were produced for

testing.

Saginaw. M12000

20

 

 

A randomimd complete block design of 160 genotypes was planted with
3

replications 0“ Julie 9‘“, 2000 at the Bean and 363‘ Palm in Saginaw Michigan (43 °41'N
8 4°08' W, 183m), The 150 (69 + 81) RILS and ten checks were space-p1amed in single

- cludg
rows of 20 seeds each. The ten checks 1n d he?“ “\cfigan cultivars Black Jack,

\orr
9 a g \‘Ni
th drought resistant breeding lines

. Atpl .
B98311, N98122, T—3016 and V8025 antlng,

BlackhaWk, Jaguar, Phantom and '11:),

280 . . . .
b and 116)“ t kg/ha of fertilizer 27 .7 .0 plus 4 0/0

Mn and 1 % Zn were applied as 3

o the Se
ed. Th -
Mis‘eguay (fi 6 80119136 at the Bean and

' inaw, MI is a He, '
Beet Farm in Sag mIXed (calcer 0‘13), mesic Aerie

contra lled b a .
Endoaquept5)- Weeds were y Dre-plant mCOIPOI'ation (ppr') of 5 L/ha
10 1"“ Emam EP
( TC). Potato leaf~hoppers were controlled

, . on '
by a2.5 L/ha application of Cyg (dunethoate) at 25 and 33 days after planting (dap).

Frontier (dimethenamid) afid

' at
Benlate (benomyl) was apphed a rate of 3.6 kg/ha and ChMp (copper hydroxi de) at 5
mm on 35 and 46 939 to control fungal and bacterial diseases. Plant stand was reco d d
r e

along with seed weight, percent moisture and 100 seed weight,

Zamorano Honduras 2001

 

«t
On January 23 , 2001, 150 RILS, 3 parents and 7 Checks were Planted by hand in
Zamorano, Honduras (14°00‘ N, 87 002' W, 800m) in conaboration with PrOgrama de
Investigaciones en Frijol (PIF). The seven checks included No PIF breeding lines (Tio
Canela—‘75 and BAP 9510-77), two Mexican genotypes (Tacama and V8025) and three
drought resistant genotypes (BAT 477, Rio Tibagi and SEA 5). This experiment was
designed as a completely randomized design (CRD) with three replications per moisture
treatment. Plots were 5.0m long and 0.70m wide. One-hundred seedS were planted in

21

 

each row and were thinned to S 0 plants for uniform Stands. Rows Were 11,71
Bdforfwrow
irrigation. Weeds were control led by hand when needed. The type ofsoil was
a

sandy-103m isohyperthefinlc ”101110 Ustlfluvent. At planting, one application 0173
0

20:0 was applied 25 days
after planting (clap) at the rate 65 kg/ha. Fertilizer 20:20:20 was applied at 300 L /b a
before flowering at 37 dap- TWO applications OfEndosulfan (endosul

. fan) Were applied at
18 and 32 dap at 1-5 LA“ to comm] Whlte fly (BeniSia T abacz‘). The

fungicide Saprol
(triforin) was applied 45 dap at 1.5 L/ha in order to control MaCmphomjna Agfimicm
which indudes Streptomycm and ”pp“ sulfate was applied 55 dap at 0.7 kg/ha to 00““01
CBB. A second insecticide, Basudin With the active ingredient, Diazinon’ was applied at
28 dap at 1 L/ha to control com rootworm beetles (D’f’bmtica Sp) ‘

u

as lanuary through April is the dry season in Honduras. The moi Stur-
e

, . , Stressed plots
received 269 mn of rainfall and overhead 1mgatron and also 3

i t ‘ .
Iona] Warnings by
irrigation along with 7 furrow irrigatlons. Tenslometers Were inStalled to Verhead

furrow irrigation. The non-stressed plots received 261 mm of rain

moisture. Readings above 60 cb signal that the soil is too dry and Plams ag@ 6
damaged by the water loss. Readings around 20 signify good moisture an being
optimal plant growth. Two weeks before flowering, they recorded 53 centibars (
stressed plots and 22 ch in the non—stressed PlOtS- TWO weeks after the fiFSt flowe .
stressed plots were experiencing 68 cb of soil suction while the non-stressed plors

experienced 21 ch. AdeQuate moisture stress was recorded by soil moisture teStS and

differences in yield between treatments-

22

 

 

 

field 110165

desirability score (198)
Phomina Phaseolina incidence at 45 and 75 dap and plant stand. DeSirabiH
Macro

   

Only 30 plants were harvested per row to record yield. Agrghom'
‘ rc

b efore harvest included days to flower, height, lodging,
taken

- 1), SCOI-e
erall rating from one to nlne 0f plant aI‘Chitecture, number of pods, amount
is an CV

of
disease uniformity in maturity and uniforrnity of plants within the plot_ Yield, biomass

Percent moisture and 100 Seed weight were recorded at harvest. Yield data was used to

calculate GM HI, DSI and D11 for the experiment.

Montcalm MI 2001

 

Using the geometric mean yield “the genotypes grow: in Honduras as the
selection criteria, the top and bottom 10 % Of 150 RILS were selected for resting in
Michigan. Although, RILS from population L38 tended to yield higher than L91 in
Honduras. mequal number from 630'“ P0Pulation was represented in the selections. From
population L88 , eleven resistant and five susceptible RILs were ected and from
population L9l , five resistant and ten susceptlble RILS were selected (Tam
cultivars, Phantom and T-39 and the three parents were inCIUded to Complg § 3). Local
6x6 Square Lattice experimental design, WhiCh W88 Planted at the Montcal V a 36 envy,

. ReSe
Station (43 040' N 85 020! W, 244m) on Junel6‘h, 2001. The $0.11 type IS a Mes arch

h d ) W ed ride sandy
. - \fic Fra iort o s . ater stress and n
loam (coarse-loamy, mixed, 1118819 A g on‘stres

sed
plots were irrigated by overhead spIaYeTS- Irrigated 91°“ received 38 mm more water than
stressed plots. An early drought began seven days 3““ 9‘3““ng Where 1essthan 5.1

of rain fell during the next 30 dayS- Herbicides, Treflan (trifluralin) at 2.5 L/ha and Dual
(metOIOChlor) at 5 L/ha, were applied ppi to control weeds. At 27 clap, 1.25 L/ha Reflex

23

 

 

(fomeSafen) Was also applied. Weeds were pulled by hand when 113 eded, A1) a

of 280 K91“ 0f 19119119 fertilizer was applied as a band at planting. An adwti:::afion
kg/ha OfN Was applied 33 dap- CYgon (dimethome) Was applied to control p0tato 1e 4
hoppers (Ernpoasca fabea) at 2 1 dap. Agronomic field notes taken before harvestin 1a f
days to flower, height, lodging and DS. Seed weight 0 Uded

3 Percent moisture and 100 Seed
weight was recorded at harvest.

Root Protoco 1

The pouch method was used to collect root data (Yabba, 2001). Twenty to thirty
seeds were genn inated in the germination chamber-
each genotype Were transferred ‘0 p°“°hes- A POUCII consisted or: a 25 .4 mm 35 .6 cm
clear plastic bag with 2 1 .0 cm x 37.6 cm germination paper film-(1% - The to P of the

gemnnationpap er is folded into a trough and a hole is Cut so that

. 0t 15
. , the grovn ng hypoc y
have aplace to rest while the root adjusts. The Pouch ls stapled to

piece of 14 pl}r cardboard. The pouch is then placed Verticauy in to 111 X .

a $10“
. . . Q
within a growth chamber. Growth chamber condltlons Included a 23 /20 0 Wooden bOX

temperature and a 15 hr photoperiod. Each sample (pouch) received 360\10 day/Inght
Hoagland’s solution throughout the 14 day gron PefiOd- 0 m1 of

At the end of this period the shoot was excised from the root. The r0 0 t w
removed from the pouch and Put in“) a 0'1 g/L staining solution containing Methyl 1.01
Afier a 24 hour period of staining, the root was transferred to a 30 Cm x 20 cm Plexl- lasset.
plate. Root laterals were separated using tweezers in order ‘0 minimize overlapping of

roots. Root samples Were scanned into a digital image usmg \Nh‘RhiZoTM 4.lOb (Rogent

24

 

instruments lire, 2000) N 1 4 days afier trariSplanting, the averag «3

mm2 “If 03 area- A resol ' . ”Sm-3’ was 0.
mm root per 5 a "no“ 0f 300 dpi and the art 03

Olnatic tin-es l
o ' 01d
WinRhionM were used- Usrng the Batch Analysis, all samples in one fbr

. 1 replication We

tea ' . r

measured for the mOTPhOIOg traits and fractal dlmension. ROOtS were star 6
ed in

. - M2 .
pak bags (4 ounce) contammg 5 0 ml water and staining solution ”1‘

R0“ char aCtefiStics such as total root length, root length ace d
0‘ ing to diamet
er and

fractal dimension were recorded. WinRhionM "lea-Sums length a
CCOrdin . . arid
. . g to p1xel srze
area covered. Root length according to root diameter Was det
crinkled b °
y usmg ten

different root diameter rangos (NJ), each differing by 0 5 mm
' - The procedure to

determine fractal dimension for root systems was 311mm . 89)‘
tinged (T . t al., t9 ,
atsurm 6

A large square frame of a side 1 was placed Over e _
objec di‘”
1; then

th ed into
(llr)2 squares of side r. The number N(r) 0f the quare ’ (ied the object
were cOLmted, and log N(r) was plotted against 10 r. Us that intersec

. . . _ g . at
small values of r, a stra1ght11ne wrth negative 310p e, . D _by measurlng N(r)

interpretation is that the object is fractal and D is the {7an % Obtained, the
since

. . <2 ,
dImenSlOIli s D ’ )
log N(r) =-D 10g r +10g K

where K is a constant, whence
N(r) °< r '0

Note that at one extreme, for objects like straight lineS D become
other extreme, for plane-filling curves, D is 2. S

, and at the
Marker Protocol
DNA Extraction
Leaf tissue from each F324 RIL, check and parental genetype was harvested’
lyophilized and ground. Lyophilized and ground tissue was allocated into 100 ml sam 1e
and DNA was attracted following the mini-prep procedure (Afanatd0r et al., 1993). The s

25

 

 

. 8m] - .
DNA GODCentranon of “on p 13 was quantified using a fluoronn eter (H
Defer TKOI 00,

Hoefer Scientific, 531‘ FranClSCo, CA)’ This Stock sample was diluted to a 101)
g/ml

working solution for amphficatl on by the polymerase chain reaction (PCR).

PCR protocol

The modified PCR procedure (Haley 8t al., 1 994) was used to amplify DNA

. be associated with drought resistanc .
aners reported to e (Schneider et al., 19973) were
used. The DNA was mpfified “Slug a Perkin E11113]- Cetus DNA Thermal CYCIer 480
(Perkin Elmer, Cetus, Norwalk, CT) With the followr'ng QYcles: 1 min at 94°C, 1 min at
35°C, 2 min at 72°C for 3 Gyms; ‘0 sec at 94°C, 20 sec at 40°C 2 min at 12°C ‘0‘ 34

cycles; 5 min at ‘72 °C; unlimited time at 4°C,

We
Approximately 20 ul 0f amPlified DNA was separated by e 1e
. Chophoresls on a
l .4°/o agarose gel containing ethidium bromide 002 ug/ml, 40 mM Trj
S~a

. Q
EDTA. DNA was fluoresced by ultra-Violet llght and recorded by phomg} §fate and lmM
s1:212

Statistical Analysis

Analysis ofvariance (ANOV A) was calculated for each experiment.
e 2000

field experiment at S aginaW, data was analyzed as a randomized complete block des'
Ign

(RCBD) using PROC GLM with the number “hemmed plan‘s Per 910‘ as the com-
an

(S AS Institute Inc., 2000). In the Honduras 2001 experiment, the StI‘ess and non‘stl‘ess

treatments were analyzed as mo CRDs. AN OVA was calculated for each treatrnem With

26

 

 

   

 

harvested plants per plot as the covanant. Each population was anaIyZed
separate 1%

Means, LSD values and CV values were ealelflated afier being adjusted

RIL. for the co""1”th
Yield means for individual 8 of the stress treatment were used With th .

tr 3 e COITeSPOnding
yield means of the non-stress e tment to calculate GM and DSL DH was calenlat
ed

using the overal 1 mean yields 0 f each treminent,

In the Montcalm 2001 experiment, data from 36 genotYDes Were analyzed '
usmg a

6x6 square lattice deSign' ANOVA was calcmated for eaCh treatmenr M - 1d LSD
' can YIC 9
and CV values were calculated for each treatment. Even thOllgh the DH 1 w GM was
was 0 a
calculated among genOtypeS- Regression analysis W as Condua ed to compare yield trends

. 2 s
between locations for the 31 SeleCtEd RILS. The R Values Within corresponding figure

were calculated by the regression function Within Micros”? EXcel ations were

CortCl
made using PROC CORR (SAS Institute Inc., 2000) betWeen yield biomass, \00 seed

wastinrmstane disease incidence (DI) f“ Macrophominaph Q38 1. t 45 and 75
0 ma 3

clap. Correlations were also made among yield, biomass, 100 see

‘ eight
° DS . (13 8 i0
flowering ,height, lodging, days to maturity, and PM. Y

Root measurements of RILS in POPUIation L88 were analyZed usin
P
for mean, LSD.and CV values. Total root length, length according to (H311‘ ROC GLM

tel. C
. . . . 1
fractal dimension were correlated to Yd, Yp and GM usmg 81mple hnear re ass and
rate 111C

2000). Both simple and multiple linear regression were used to associate molecul

(PROC REG) and correlation (PROC CORR) methOdS 0f analysis (SAS Insn.

marker values to yield-based traits. The degree of aSSOCia‘ion between traits was r 0
rted
by the Pearson coefficient values (I) and the Coeffic‘em OfDetennination (R2). Multiple

regression analysis Was used to determine the best model of root and molecular marker

27

 

 

 

 

at that exua'med t“ '
C h‘ghest
amoun
t of variation for '
yield und
er'
Str
685 an
0’ I1
017-5?!
658

conditi 0115'

28

RESULTS
Field Study
Three field CXPefiments Were conducted over W0 years and three locations to
study me genetics of drought resiStance in two black bean populations L88 and L91. 30th
Populations showed marked differences 1n the firm field test in Saginaw, MI in 2000. A
seed increase was needed to meet the requirement of having Sufficient seed for stress and
non-stress treatIIlentS in Honduras. Sp ace-planting allowed each plant to grow without
competition. Therefore the yield results 5' om Saginaw may be inflated, but evaluations of
yield potential and Comparisons between pOpulations Wer e performed,
Significant genotypic differences existed in bath p op” Iations for yide and 100
seed weight (100 5w) grown in sagmaw (Table 1) Mean yield in L88 was “We”
hi gher ($.05) man the mean yield in L91, suggesting that L88 has a greater View
potential. Yiekl for individual RILS in L88 ranged from 2257 to <1 926 kg/i1a (Figure 1).
The range oi yield for L91 RILS represented lower yield potenti a] and ranged from 1868 {C
4323 kglha. Mean 100 sw for L88 was 23.0 g Whereas 100 sw Was 4-9g i I
1). Overall, the RILS in L88 produced a larger number of seeds Whereas q‘ ‘7 L91 (Tab 6
had larger seed size. This relationship is supported by differences in seed

‘S RILS in L9]
1263 1)
parents as VAX 5 is larger than TLP 19.

Yield potential (Yp) and the ability to yield under moisture stress (Yd) were

. tested
in non-stress and stress treatments at Zamorano, Honduras 111 2001. The DH f0r the
Honduras experiment was 0.82 and the D81 Of individual genotypes ranged from 0-52 to

1.20 Conditions W1 thin the tropical climate, SUCh as hlgh temperatures, short day length

29

 

8 RILS

Frequency Distribution of L8
Saginaw, MI

 

 

 

Frequency

 

 

 

 

 

2600 3000 3400 3800 4200 4600 5000
Yield (kglha)

 

 

 

Frequency Distribution of L91 RILs

 

 

 

 

 

5 Saginaw , MI
2
20 1
g 15 i 9
a ‘f
a. '-.
e -
IL .. ggi
'L ‘ i’”
i ' .. . : n i

 

 

 

 

 

 

 

 

 

 

 

2000 2400 2800 3200 3600 4000 4400
Yield (kglha)

 

 

 

v

Figure 1 . Frequency Distributions for yield using the adjusted me

in Saginaw, MI 2000. The drought resistant parent, B98311, is ind-1° ftorn each POPUIatiOH

§lted by (B)-

30

W

Table 1. Malysis of variaIICe for the RILS at the F
ulatiot‘léior yield and 100 seed weight at Saginaw,

3:5 generation in tbe L88 and L9!

M1 2000.
L91

 

PBL/f

 

fl—

 

 

 

 

 

L88
Source A DF MS F 1- OF MS
____————-'"" est
Grand Me an 3512 3042
LSD (0.05) 1 056 977
cv 19 20
Replication 2 837809 1.94 2 1 191759 3'26:..
Genotype 80 898733 2.09"" 58 759650 2'08
Stand 1 9923410 23,03We 1 17782870 48.58“**
Error 1 59 430847 135 366035
100 Seed weight (g) ‘00 Seed weight i9)
Grand Mean 23 0 24 9
LSD (0.05) 1'6 1'9
CV 48 '
Replication 2 ' m- 4.8
G 19.9 20.8 2 3.3 2,7
enotype 80 9 4 9 aura ”in
St :1 ' ' 68 14.5 10.2
a“ 1 0 6 0 6
E ' ‘ 1 0.1 0'1
x ' 135 1930/

 

 

 

 

*P<.05; **P<.01; '*'P< - 001- **“P< 0001

and low soil fertility , considerably decreased Yp com ared '
comparisons among most genetic traits COUId b3 conducted F
o requenc d . ' ' f
. tn ons 0
mean yields Showed the trends toward moisture Stress and n Y 15 but]
on‘SlTess -
Condl tions (F igur e

2). Distribution of yield was skewed With only 15 RILS yield'
mg above 4
S

y1€1dS for L88 RILS ranged from 77 to 842 kg/Ila in the stress treatm Q kg/ha. M6211
ent (

‘

overall mean of 317 kgfha. In the non-stress treatment, mean yields in ‘h lgure 2) with an
p0pulation ranged from 1441 to 2922 kg/ha With an overall mean of 206: Same
frequency distribution of RILs for Yp appears to resemble a normal Ga 1.(g/ha. The
Significant differences were recorded among genotypes for Yield in USSIan curve.
but not in the non-stress treatment (Table 2). the stress treatment,
The frecmency distribution in L91 followed a similar Pattern t .
The hiStogr am Showing Yd in L91 was also skewed, yet only three R: L88. (Flgure 3).
31 Le yielded above

 

 

 

Frequency Distribution for Yd
in L88, ZamoranO

 

 

N
O

 

Frequency
'4 -—l
C><n

 

‘,A I.

150 250 350 450 550 650 750 850
Yield (kglh a)

 

001

 

 

‘

F requency Distribution for Yp
in L88, Zamorano

 

 

 

N N
ES 0 0‘
l

 

10‘

Frequency

 

 

1550 1750 1950 2150 2350 2550 2750 2950

Yield (kglha)

 

J

 

Figure 2. Frequency Distributions for yield under stress (Yd) and no“ .
“3‘18th means from eachRU-a from population L88 in Honduras. 1V1 ‘stress (YD) usmg the
parents, B98311 (B) and TLP 19 (T) are noted. ean yleld 0f the

32

A

Frequency Distribution for Yd
in L91, Zamorano

 

 

Frequency

0

 

 

 
 

     

Vl

50 150 250 350 450 550 650

Yield (kglh a)

 

N
A
010

Frequency
8

O U7
l

 

 

 

F requency Distribu tion for Yp

in L91, Zal'norano

 

 
    

    

 

rm '

.'_’ i'
i B

I" . ,1

: :1

. “A I

i

3 7'?!
.."‘~

12501450 1650 1850 2050 2250 2450 2550
Yield (kglha)

 

Figure 3 . Frequency Distributions for yield “Pd“ Sm?“ (Yd) 311d 11
adjusted means from each RIL, from population L91 m Honduras,

J

or1~stress (Yp) using the

Parents, B98311 (B) and V AX 5 (V) are noted. I\’Ieaii yield of the

33

Table 2. Analysis of variance for 83 genotypes of population L88 in stress and non-stress
treatments from Honduras 2001.

 

 

 

 

 

 

Stress Non-stress

Source DF MS F Test MS F Test
Yield (kg/ha)
Grand Mean 317 2060
LSD (0.05) 357 916
CV 70 28
Replication 2 125133 2.55 1053485 3.27"
Genotype 82 72286 1 .47' 308196 0.96
Stand 1 864683 1 7.60”" 1 088310 3.37
Error 163 49127 322544
° Biomass (kg/ha)
Grand Mean 1676 3974
LSD (0.05) 1 165 1582
CV 37 25
Replication 2 2075102 5.36" 4953171 5.17“
Genotype 82 539428 1 .39’ 826276 0.86
Stand 1 22548148 58.28”“ 1576618 1 .64
Error 163 386921 958522
100 Seed Weight (9)

Grand Mean 17.9 18.9
LSD (0.05) NA 2.7
CV 6.9 8.9
Replication 2 2.8 1 .80 341 .6 120.39**“
Genotype 81 4.9 3.14“" 6.9 2.43““
Stand 1 2.0 1.26 0.1 0.04
Error 125 1.6 2-3

 

‘P<.05; “P<.01 ; *"P<.001; “”P<.0001

450 kg/ha. Mean yields for L9] under stress ranged from 2 to 599 kg/ha with an overall
mean of 211 kg/ha (Figure 3). RILs in the non-stress treatment ranged from 1130 to 2587
kg/ha and averaged 1863 kg/ha. Significant differences existed for yield among genotypes
in the stress treatment but not among genotypes in the non-stress treatment (Table 3). As
in Saginaw, mean yield in L88 was greater than in L91 for both stress (p<.05) and non-
stress conditions (not significant) in Honduras.

Data for 100 sw followed trends similar to the results in Saginaw (Table 2). In
L88, mean values for 100 sw ranged fi'om 13.8 to 21.6 g in the stress plots and 14.2 to

27.9 g in the non-stress plots with overall means of 17.9 and 18.9 g respectively. In L91,

34

Table 3. Analysis of varianoe for 71 genotypes of
Wm Honduras 2001,

population L91 in Stress and non-stress

#/__\__/__/

 

 

 

 

 

 

 

 

Stress W
W DF MS F Test
Yield (kg/ha) 1863
Grand Me?" 211 931
LSD (0.05) 303 31
cv 89 1.78*
Replication 2 104563 2.97 3:33: 1 .08
Genotype 70 49137 1 .40" 661 11 6 1 99
Stand 1 181353 515* 332432
Error 1 39 35199 /
Biomass Wm”
Grand Mean 1661 3709
LSD (0.05) 1267 1379
CV 43 23
Replication 2 2600809 5.17" 6465282 9.12“”
Genotype 70 527540 1.05 894284 1.26
Stand 1 1 6358738 32.51 **** 3680182 549*
Error 139 5031 17 708965 /
100 S -
Grand Mean 19.9 eed Weight (9) 21.5
LSD (0-05) NA 3 1
CV 2 6.9 9 0
Replication 21 .2 1 1 .33**~- ' 5""
Genotype 62 8.9 4.76**** 278-5 7232""
Stand 1 3.6 1.94 12-7 '
Error 73 1 .9 7.2 1 '95

 

 

‘

 

*iiv
*P<.05; “FR-0"- P<.001; ****P<.0001 W

seed weight values ranged from 16.3 to 27.8 g in the stress treatm
ent and
130m 17.4 to

26.7 g in the n(HI-stress treatment With overall means of l 9 9 and 21 5
. . g P

. g .
Populatlon L91 had a greater CV value than L8 8 in every categ Pecttvely.
or)»

. e"Cont b‘
under non-stress (Tables 2 and 3)- RILS 1n L91 must have greater standard d ‘Omass
eviat‘
the mean or a lower mean value when compared to L88 RILs The fa t th th Ion from
‘ C at e CV value
8
were so high can be attributed to a large variance due to envir
onmental ' ‘
conditions of stress
and a lack of control of experimental error in the CRD.

Despite high CV values, significant genotypic differences We (1
re observe for yield,

biomass and 100 sw (Tables 2 and 3). For both Populations, ‘00 8w h
ad significant

35

 

 

FreCIuency Distribution Of 63"
Pop L88 and L91, Zamora“
60 ’

5O

 

 

A
Q

 
 

Frequency
09
O

2

0 5 74‘
10 I . (.1 . ~
0 I .3 7-1' ._ '~
ZOO 400 600 800 1000 1200 1400 1 600
Geometric Mean (kg/ha)

Figure 4, Frequency Distribution of 150 RILS ShOWing selection Ofresi

susceptible (S) genotypes based on geometric mean, Parents VAX 5 (vitaht (R) and
B98311 (B) are included. ’ LP 19 (T), an

genotypic differences in each treatment. In population L88, significant génotyp'
1C

 

 

 

 

differences in yield and biomass were observed in the stress treatment, but not in the non‘
stress treatment. In population L91, only Yd showed significant genotypic differences
whereas no significant genotypic differences were observed for Yp and b i amass. Overal 1,
the yield data fi-om the stress treatment showed a separation of resiStance and susceptible

genotypes and yield potential of 6219“ genotype was (“pressed in the hon-stress treatment.

36

 

To select genotypes with drought resistance, geometric mean (GM) between
treatments was calculated. The frequency distribution of GM (Figure 4) appeared to
follow a Gaussian curve. Resistant (R) and susceptible (S) RILS were selected based on
the top and bottom 10 % of the curve. Other characteristics such as drought susceptibility
(D81) and harvest index (HI) for moisture stress and non-stress treatments were calculated
(Table 6) but not used directly in the selection of resistant and susceptible individuals.
Mean values for GM, D81 and HI were contrastingly different. The average GM value of
the resistant RIL’s was more than threefold greater than the susceptible RILs. A lower
DSI value signifies a lower susceptibility to drought or a greater ability to tolerate
moisture stress. The resistant RILS showed a greater tolerance to drought than the
susceptible RILS. The resistant RILS had higher HI values and were more efficient than
susceptible RILS in the stress and non-stress conditions.

Additional comparisons of the selected resistance and susceptible RILS from each
population with checks is based and ranked on GM are shown in Tables 4 and 5. Parental
values are also included to illustrate the transgressive segregation for different traits
among the RILs. In every category, the parents are ranked in the middle while the
resistant and susceptible RILS trended towards the extremes. The parents were not
significantly different for yield or biomass in both treatments. The drought resistant
parent, B98311, ranked higher than the susceptible parent in every category except
biomass in L88. TLP 19 produced greater above-ground biomass than B98311 in both
stress and non-stress treatments. VAX 5 out-yielded TLP 19 under stress, yet when

comparing population mean, L88 out-yielded L91 in every category. In comparison to

TLP 19, VAX 5 yielded much less under non-stress
37

Table 4. Y' ield (Yd or Yp), Biomass (EM), and 100 $6
RILS selected as highest and lowest yielding based on

‘tgoptllatiOI'ZJAL88 gr Own under moisture stress and non-SW
iii—95$

—¥

 

 

 

ed weigh? (1 00 SW) offlle Sllrteen
Geometrlc Neal] (GM) in
n Honduras 2001 .

W
100 sw

 

 

 

 

 

 

 

 

 

 

 

 

G 4’» "'YBM
RIL k9/ha kg/ha kg/ha g 17 6
_§. 4410 (26) -
L88-63 1473 (1); 842 (1) 2385 (15) 16.7 225;): (8)2) 4725 (7) 17.9
L88-74 1362(2) 740 (3) 2610 (7) 16.9 2263 (30) 4050 (52) 18.4
L88-30 1328(3) 779 (2) 1800 (60) 18.4 2432118) 4500 (20) 18.2
L88-69 1286 (4) 830 (4) 1890 (46) 17.6 2922 (1) 5445 (1) 195
L88-13 1285 (5) 565 (9) 2250 (20) 19.0
L88-66 1205 (6) 561 (10 1575 (90) 13.3 2539 )4) 4590 “6) ”'4
L88-19 1 126 (7 ) 579 (8) ) 1800 (59) (9.3 2188 (36) 4095 (46) 20.4
L88-3 1082 (8 ) 583 (7) 3015(2) 19.4 2007 (70) 4050 (49) 18-7
LBS-61 1050 (1 1) 507 (13) 1575 (91) 17.1 2173 (39) 3960 (52) 18-2
L88-31 1048 (1 2) 635 (5) 1935 (39) 18.9 1730 (117) 2970 (150) 18-8
L88-59 1033 (1 3) 455 (1 a) 2115 (28) 19.8 2344 (25) 4535 (12) 18.1
“8'37 577 (1 ‘12) 224 (93) 990 (152) 21 3 1486 (151) 3600 (107) 21-8
L884 55‘ (1 22) 168 (1 19) 1440 (107) 19.7 1802 (105) 3800 (104) 18-2
“3'64 457 (1 39) 113 (141) 1215 (134) 18.3 1849 (97) 3825 (78) ‘87
L88-18 368 (1 46) 90 (144) 1755 (84) _ 1501 48 3420 (125) 21.6
fizz—— 364%) 77 (150) 945 (155) 16.7 1729 (118) 4140 49 19.4
Parents 951
898311 (22) 375 (33) 1755 (71 )
TLP 19 6m) 169 (118) 189% :3“? 3:97; (20) 4275 (36) 112%
. 2 ‘
Checks 2 4545 19
Tacana 667 (37) 213 (97) 1800 (62)
15.
\T/80C2:5 612 (102) 210 (99) 1395 (1 1 6) 16.3 123:7 (54) 400 9
1o anela 502 (105) 218 (95) 1845 (57) 17 8 16 3 (103) 32 S (61) 18.
EAP9510—77 699 (73) 232 (89) 1710 (75) ,9'5 2 57 (129) 38:0 (135) 17.3
2.47 477 _ 734 (54) 400 (27) 2555 (e) 19', 1; g 2 (50 S o (75) 20, 2
S a Tébagl 886 (31) 372 (34) 3330 (1) 16:2 2102 g?) Zq S (124) 22.7
W 20.9 1521 (145) 32;:(1138 12g}?
lfillsean 699 269 1683 18.7 1957 5 155 25.4
CV0 (0-05) 3373 1098 NA 933 3827 20 1
V 7 41 6.9 1484 -
T Values 0f parents, checks and LSD values are included. 30 A

I Rankings (in parentheses), mean, LSD and CV values are derived fro 160
m genotypes.

38

Table 5 . Yield (Yd 0r Yp), Biomass (PM), an
RILS selected as highest and lowest yielding b

d 100 seed weight ( l ()0 SW) oft)“, fifleen
ased on Geometric New, (GM) 11:
in Hondwas 2001 7‘.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

- moisture stress and non—$31”:SS 2.77355
population L91 grown under Stress 0 sw Yp Nan SA; 100 sw_
Genotype GM Yd Mr kg/ha kg/ha Q
_7 kglha kglha kglha 7 1 922 (83) 3690 (98) 23.8

1 4. 25 80 2
Eelifso 1073 (9)1 599 ‘5) 1845(5)?) $9.6 2202 (2:) 34310 ((27i 13':
L91-25 1064 (10) 514 “2) 2520(105) 20.3 2406 ((46) 4455(23) 20.9
L91'3 1°23 (14) 435 (20) 1485(50) 19.9 2126 31) 4050 (54) 2 ‘
L91-59 1016 (15) 486 (15) 139g ‘47) 21.3 2250 t 2.8
L91-10 1004 (16) 448 (19) 139 (

0 _ 2092 (56) 3960 (67) 22 .5

(91.37 402 (144) 77 (149) 1755 (1,0)6) 18.4 1938 (82) 3555 (111) 21 .3

L91-22 3921145) 79 (148) 1:32:53) 17.8 1690 (120) 3690 (99) 21.2

L91-41 282 (155) 47 (155) 8845 (56) - 1391 (157) 3060 (145) 21,0
L91-53 250 (156) 45 (156) 1755(69) _ 1857 (96) 3780 (85) 20 9
L91-13 1 15(158) 7 (158) 1 .
5(149) ’ 12$1058) 2655 158 22-7

149 316 151 80 (147) 103 ( )
$152 298 (154; 56 (154) 1125 (1:2) 18:3 1574 (139) 3500(103) 19.9
L91-68 1 76 (157) 17 (157) 171000) _ 1 92(105) 3735(94) 20.5

-19 64 159 3 (159) 1935(4) 528(143 645 102) 255
L91 ( ) 45(4) - 15 ) 3 ( 67

L91-89 60 (160) W 34 142 5175 3 2 -

Parents

398311 951 (8) 375 (9) 1755 (3;) 3'8 241 1 (6) 42-15(12) 18-9

VAX5 _663(24) 249 26 1665 - 1765 39 3645 40 22.4

Checks

_ 2) 15.6 2097

Tacana 667 87 213 (97) 1800 (6 (54) 4005

V8025 612i10)2) 210(99) 1395 (13,5) 1‘75; 1783 (108) 3240 (fig) 113;)

“0 Canela 602 (105) 218(95) 1845(35) 195 3:2 (539) 387° (75) 2 '

EAP9510—77 699 (78) 232 (89) 1710( ) 19-7 1536 (14: 34 5 (124 0.2

BAT 477 784 (54) 400 (3:; $3533 (‘15; "3'2 2108251)) §§>5(160)) 22557

Rio leagi 335(31) 372 ° 1 145 70(120 '

SEAS 393 (30) W

683 18.7 1957 38
Mean 699 269 1 933 27 2
3 1098 NA 1434 0.7
BSD (005) 37:3, 41 6.9 30 24 NA

 

‘ included.
1 Values 0 hecks and LSD values are .
I Rankingsfgzrggsenfheses), mean, LSD and CV values are denved from 160 genotypes.

39

 

Table 6. Resistant (16) and susceptible (15) RILS with geometric the“ (GM), drought
susceptibility index (D81) and harvest index (HI) under stress and 110n_Sa—ess 6011;101:1115,
days to maturity (DTM) under the stress treatment and height and er the stress an no
stress trejgments in Honduras, 2001.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

40

s Non-stress
Resistant Stress Non-stress 0:5938 5; a: t H 8'. ht
M GM 9% H' H' days cm cm ‘
kg/ha 82 46 43
L88-63 1473 0.80 0.35 0.58 82 45 45
L88-74 1362 0.34 0.28 0.53 83 42 47
L88-30 1328 0.78 0.43 0.56 83 39 45
L88-69 1286 0.85 0.36 0.54 81 40 44
L88-13 1285 0.96 0.25 0.54
81 38
L88-66 1205 0.93 0.36 0.56 82 43 it;
1.88-19 1126 0.87 0.32 0.53 82 45
0.19 0.50 45
L88-3 1 082 0.84 82 43
L91-3o 1073 0.82 0.32 0.52 82 45
(31-25 1064 0.91 020 0°58 48 45
L88-61 1050 0.91 0-32 0-55 81 38 43
L88 31 1048 0 75 0-33 0'58 82 43
‘ ' . 83 40
L88-59 1033 0.95 0.22 0 51 46 52
L91-3 1023 0.97 0.29 0. 55 81 40 53
L91-59 1016 0.91 0.26 0.48 82 52
491-10 1004 0.95 0.24 0- 56 82 49 49
Mean 1153 0.88 0.30 0- 54 M03 Zr
Susce tible ‘ Stress Non-stress M _
M -
k /ha a s
L88-37 377 1 .01 0.23 0.41 a; 32‘ cm
L88-4 551 1.07 0.12 0-50 82 37 47
L88-64 457 1,1 1 0.09 0.48 83 30 44
L91-37 402 1,14 0.04 0.53 33 36 44
L91-22 392 1_14 0.05 0.55 82 37 51
46
L88-1 8 368 1.1 1 0.05 0.44 83 42
L88-2 364 1 _ 1 3 0 .08 0.42 82 34 5 1
L91-49 315 1,11 0.08 0.47 83 38 45
L91-52 293 1,14 0.05 0.44 83 39 :97
L91~41 232 1.15 0.06 0.46 81 37 45
L91-53 1.15 0.02 0.45 82 36
”1‘68 1236) 1.17 0.01 0.48 82 37 fig
L91-13 115 1.1 8 0.00 0.49 83 39 47
L91-19 64 1_13 0.00 0.42 84 42 53
L91-69 60 1_13 A 0.00 0.30 g1 41 51 fl
Mean 31 1 1f}? 0.06 0.46 82.6 37 .4 47 .2

 

Table 7 - Analysis of variance for 36 genotypes grown under stress and non—stress

 

 

 

 

 

 

 

 

 

 

 

 

treatments in Montcalm, MI 2001. J
Stress 4 Wress *
Source OF M5 F Test MS F has;
Yield (kg/ha) 3006
Grand Mean 2950 965
LSD (0.05) 808 19.5
CV 16.7 ”.1 249.8 9.11 *v-u
Replication 2 2334.3 121-19 1 05.3 3.34“,
Genotype 35 68.8 3-57". 131.4 4 79...,
Block 1 5 62 3.22m 27.4 ' *
Error 55 19.3 [M l
100 Seed We‘g (9
Grand Mean 29.3 28-7
LSD (005) 2.1 22
CV 4.3 4.7
Replication 2 21.8 1352"" 8'4 4 54*
Genotype 35 21.1 1343"" 20's 1138*"
Block 1 5 2.2 1-36 3-4 1.86
Error 55 1.6 fl '

 

 

 

*P<.05', **P<.01 ; ***P<.OO1

conditions. VA} 5 might be better adapted to stress conditions yet I k th ' 1d I’Ot‘mfial
ac e yie
to remain competitive under non-stress conditionS-
The Bl Selected RILS, parents and local CheCkS were evaluat d f drought
e or
resistance in Montcalm, M1 in 200 1. Late rainfall during the
Season all(“Wed for genotypes

to negate the effects of the early drought stress in Montcalm, The DH for th
e Montcalm

experiment was extremely low at 0.02. Treatment means were not signific
antIy diff
erent

and only v ari ed by 56 kg/ha (Table 7). Mean yield under stress ranged fi‘om 19
26 to 40
1 s

kg/ha 3111011 g the 36 genotypes. In the non-stress treatment, mean yield ranged fr
0m 1682
‘0 4340 kg/ha. Significant genotypic differences were Present among stress and no
n-Stl‘ess
conditions for yield and 100 sw. Coefficients of variation for yield were moderately low

16.7 and 19.5 0/0, for stress and non-stress treatments respectively and low LSD Values

allowed the separation of high and low yielding genotypes within both populations.

41

Table 8. Yield under stress (Yd) and non-stress (Yp) and 1 00 seed

- . Weight (100 SW) for
sixteen genotypes ranked by Geometric Mean (GM) 1n P0pulatio 17

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Lgsgrovm 1'”
- trnents .
Montcalm, MI 2001 under stress and non 3:28;: 11634 T Non-Str‘leos;
Geno e 4M T 100 SW Y" 5'”
A kg/ha kg/ha g kg/ha Q
RlLs 3590 (5) 29.5
L88-69 3849 (2): 4015 (1) 28-3 4318 (2) 25.8
L88-63 3596 (3) 2995 (16) 25-5 3365 (12) 26.1
L88-30 3398 (9) 3432 (8) 25-9 3342 (13) 26.6
L88-6 1 3387 (10) 3432 (9) 27-2 3443 (10) 29 0
L88-59 3341 (11) 3241 (13) 28-9 '
L88-74 3241 (15) 2838 (21) 27 .2 31%; ll: 27.0
L88-66 3333 (12) 3488 (5) 26 -3 3398 ) 24-9
L88-19 3284 (13) 3174 (14) 23 -9 2 (11) 28.3
L8831 2910 (19) 2894 (20) 27 .3 927 (22) 26.3
L88-37 3191 (17) 3376(10) 3“ .3 3017(19) 30.4
L88-64 2901 (20) 2759 (23) 28 .6 3051 (18) 27.4
L88-2 2845 23 2703 (26) 29 -3 2
L88-13 2748 ((26; 2748 (24) 28 .9 2323 (20) 26.8
L88-3 2393 (30) 2322 (31) 30 .4 24 (24) 27.3
L88-4 2389 (31 ) 2355 (30) 3% '3 243; fig; 33;
L88-18 2075 35 . -
f 2001 (34) 1929 33 2&4
Parents 8
B983“ 3495 (6) 3903(2) 2 -6
TLP 19 A 3466 (7) 28.0 3129 15) 27-5
was) 3589 ((6) 26_4
Checks
T 39 3533 4 3544(4) 23-8
Phantom 3494 :7; 3466 (6) 26.5 33:: (g) 24.8
26.3
Mean 903 2950 29.3
LSD (0.05) 2 808 2.1 3226 28.7
cv 16.7 4.3 19.5 2.2
3; Values of parents are included. 4.7

Rankings (in parenthesis), mean, LSD and CV values are derived from 6
genotypes

42

 

 

Table 9 - Yield under stress (Yd) and non-stress (Yp) and 1 OO seed

 

 

 

 

 

 

 

 

 

 

 

 

 

. . Wag/1f (100 SW) for
fifteen genotypes ranked by Geometric Mean (GM) 1n p0pulat1017 L91 grown m
MontcaM, M12001 under stress and non-stress treatments To

Stress YNon-str 19033 SW
Genotype GM Yd 100 sw kg”; 9 fi
kg/ha kg/ha 9

RlLs 4340 (1) 29.9

L91-1 o 4056 (1 )x 3791 (3) 31.7 3970(3) 33,0
L91-3O 3480 (8) 3051 (15) 33.8 3118 (16) 316
L91—49 3244 (14) 3376 (11) 32-9 3073 (17) 27 1
L91-3 3210 (16) 3353 (12) 29-9 3533 (7) 31'9
L91-25 3038 (1a) 2613 (28) 32-8 -

L91-22 2875 (22) 2781 (22) 29 .1 23;: (21) 28.8

L91-52 2814 (24) 2905 (19) 3 1 .1 2 (26) 30.9

L91-41 2731 (27) 2714 (25) 29.3 748 (25) 27,1

L91-53 2765 (25) 2938 (17) 265 2602 (28) 26.1

L91-68 2888 (21) 2927 (18) 27-5 2349 (23) 27.6

L91-13 2611 28 2501 (29) 33.2

L91-59 2401 ((29; 2636 (27) 30.6 $173 (27) 27.0

L91-37 2026 (33) 2288 (32) 30.7 17 (32) 30.3
L91-19 1900 (35) 2131 (33) 35.6 94 (34) 30.4
L91-69 1801 (36) 1829436) 344 1694 (35) 36-3
1682 36 33.8
Parents 28
898311 3495 5 3903 (2) -6
VAXS A 2244(3)2 2097 34 27.4 3129 (15) 27-5
W
Checks
T 39 3533 4 3544(4) 23-8
Phantom 34947; 3466(6) 26.6 33:: (g) 24.3
26.
Mean 2908 2950 m
LSD (0,05) 808 2.1 965 28. 7
CV 16.7 4.3 1 9.5 2.2
l Values of parents are included- 4.7
I Ranki

 

43

n gs (in parentheSiS), mean, LSD and CV values are derived from 36g
e”O‘B’pes.

 

 

 

 

 

 

mansion of Geometric Man In SCI-curd all-s

 

 

 

 

 

 

 

 

 

  
 
  

 

 

 

botluun Honduras and mama-n
4500
"°°° E3151 317:2) ‘ , .
3500 M
-— r = 0:6.3, . - ' ..
g 30001 E” ........ §
"" 2500 9% ‘ ‘ El
: . o
is r: 0 20*
$2000 I T
‘5‘.
1500 r = 0.55*
1000 ---- unmoununed)
500 "Uneat (Resistant)
, ‘ " -Unear (Susceptible)
0 fl 4 4.1— r .
0 200 400 600 800 1000 12m
Geometric Im- a WM) 1400 1600
Hondums

 

 

F191“? 5- chl'ession analysis of Resistant and Susceptible . Mean
across the Honduran and Montcalm, MI locations. RILS for Geometric

Comparisons between P0P“13ti°ns’ individual genotypes and checks were
performed (Tables 8 and 9). RILS from L88 yielded ‘0 % more than L9 1 RILS b d
386 01]
GM. Means for 100 sw in L91 were 10 % larger than in L88. These Teen]
‘3 Were
COHSiStent with previous results, but must be considered with the informal.” tha
t tWo-
thirds of the resistant RILS were from L88 while two-thirds of the susceptibze
RILS canl
from L9 1 ,
Comparisons of the selected RILS between locations was performed by regression
analysis. Yield data obtained in Montcalm was used to validate the results obtained in
Honduras. Geometric mean was moderately correlated between locations, r = 0'6?

(Figure 5)- This Value is supported by the higher 60116130011 of suscepfible genotypes in

44

 

 

 

 

Regression of Yield under Stress In Selected RIL:
batman Honduras and Michigan

 

 

 

 

 

 

 

 

 

 

 

 

 

4500
4000
3500 a. B """""""""
a / r=o45* "Lflrflfi
aooo - .................. E]
5 m ' EBB .
3 w a
1 r=0.56*
1000
500
o .
0 100 200 30° ‘°° 500 50°
“clam-u)
m":

 

 

 

 

 

Figure 6. Regression of Selected RILS for yield under Stress d
Montcalm, NH locations. among the Honduran an

GM (r = 055*) and Yd (r = O 56*) regression analyses (Fi
' e 5)- Resistan
gm tgcnotypes

autilises. All

genotypes were moderately correlated in the non-stress treatments (r = O 5
‘ 2 '9 (Fi
gure

Resistant and susceptible genotypes were weakly correlated in the non-shes 7) ‘
. s "cannen
ts

h

= 015*; r = 007*) (Figure 7).
Agronomic traits may have contributed to yield in a positive or negativ

(Tables 1 O and 1 1). Biomass had a significant impact on yield in stress and DOD-stress

treatments. One hundred seed weight had a larger impact on yield and biomass in the
stress treatment than in the non—stress treatment. In the stress treaunem, plant stand was
positively associated with yield and biomass, but was negatively associated Wi‘h inc'ldence

45

 

 

 

vi

 

 

 

 

 

 

 

 

 

 

 

 

 

“09'0““! at Yield under Nan-stress In 80th an...
between Honduras and Michigan
5000 A
4500 ~
4000 -
3500 «
as 3000
5 e v :1. . 0
2500 . ' g.
3 2000 m E
Q E
a 9*» Resistant
1500 t s
r = 0.07" _ . . . W
100° 'U'W (Combined)
500 ~ ‘ 'UI'IUHSmcoptlbte)
o .f
° 500 1000 1500 2000
“old Who)
Honduras

 

 

 

 

Figure 7. Regression of selected RILS for yield under non-Stress conditions within the

Honduran and Montcalm, Ml locations.

OfMacrOphomina phaseolina at 45 311d 75 d3? Disease inCidence was 11 .
egatIVer
associated with yield, biomass, 100 sw and stand in both populations Within
the Stress
treatment- The negative association of D1 to stand also affected yield and bio
mass.
Yield and biomass in L88 were closely associated and were aSSOCiated .
with Plant
stand than in L91 (Table 10). Yet, 1 00 sw has a stronger association to yield and biom
ass
in L91 rather than L88. Seed size in L91 is greater than L88, but did not show Significant
differences (Tables 4 and 5). plant stand was affected more by D1 in L88 (1' = '0-41****)
than in L91 (r = 0.33“"). At 7 5 (hp, Yd was more negatively affected by D‘ in L88 0 =

-0.36****) than in L9 1 (r = -0 30““).
46

Table 10. Correlation between yield, biomass, 100 seed weight (100 sw), plant stand at
harvest and disease incidence (DI) at 45 and 75 days after planting in the moisture stress
and non-stress treatments for L88 and L91 RILs in Honduras, 2001.

 

 

 

 

 

 

 

 

 

 

 

Pop L88
Stress Yield Biomfis 100 sw Sfind DI 45
Yield - - - - -
Biomass 0.61 "“ - - - -
100 sw 0.31 ““ 0.35"" - - -
Stand 0.33”“ 0.46"” 0.12 — -
DI 45 -0.23” -0.28"* -0.01 -0.27“" -
DI 75 036““ -0.48*”* -0.16* -0.41“** 0.58“”
Non-sires Yield Biomass 100 sw Stand DI 45
Yield - - - - -
Biomass 0.91““ - - - -
100 SW 014* 0.20“ - - -
Stand 0.05 0.03 -0.09 - -
DI 45 016* -0.18* 0.07 -0.23" -
DI 75 -0.18“ -0.24***" -0.02 -% 0.81"”
Pop L91
Stress Yield Biomass 100 sw Stand DI 45
Yield - - - - -
Biomass 0.53"” - - - -
100 sw 0.37"” 0.45"“ - - -
Stand 0.21 *"' 0.40“” 0.1 1 - -
DI 45 -0.25"* -0.32"m -0.07 -0.26“ -
DI 75 -0.30“** -0.48“*" -0.05 -0.33***' 0.53””
Non-stres Yield Biomass 100 sw M DI 45
Yield - - - - -
Biomass 0.86"“ - - - -
100 sw -0.07 0.11 - - -
Stand 0.05 0.10 -0.08 - -
DI 45 -0.14 -0.15 0.05 -

' DI 75 -0.20** -0.24“" -0.09 -0.07 0.81““

 

'P<.05; *‘P<.01?WP<.001; “**P<-0001

Yield under stress was potentially compromised in Honduras due to a severe
infestation of; ASB caused by Macrophomina phaseolina. The D1 of ASB was
characterized by 99 % of the stress plots having at least one dead plant compared to 50%
in the non-stress treatment. DI values ranged from 0.05 to 0.54 across the 160 genotypes
grown under stress (Figure 8). In population L88 and L91, a negative correlation was
observed between plant stand and DI at 75 dap (r = -O.41**** and r = -O.33****). In the

non-stress treatment, no correlations existed between plant stand and DI at 75 dap.

47

Table 11 Correlations between yield-based traits including 1 00 Seed weight (100 sw) and
harvest Index (H1) and agronomic traits including desirability score (DS) in L88 (below
diagonal! and L91 (above dia agonal) m Honduras 2001.

 

 

 

 

 

 

 

 

gLe§S__ Yield Biomass 100 sw HI Flowerin Hei ht
Yield — 0.53"“ 0.37“" 0.78"“ -0.32“" 0.53""
Biomass 0.61“" — 0.45““ ns ns 0.53""
100 SW 0.31 ““ 0.35"" - ns ns 0.37“"
HI 0.77“" ns 0.14‘ - -0.39*“" 0.29""
Flowering -0.16* ns ns -0.29“' - ns
Height 0.49"" 0.62"" 0.27"“ 0 18“ ns -
Lodging 025*" 0.26““ 0.31“" 0.17' ns 0.20"
Maturity -o_14' ns ns -0. 25“" 0.29“" ns
03 0.78"“ 0 46““ 0.21" 0.72"“ -0.22* 0.42“"
Biomass

  

 
    
 
  
    
 

.. 0.86"" ns 0.64“" -0.19* 0.29"“

Biomass 0-9 1 "“ - ns 0.19" ns 0.39"“
100 SW 0. 1 4' 0.20“ - -0.32“" 0.31““ 0.28““
HI 0-42"“ ns ns - -0.38*“* ns
Flowering n 5 ns 0.29“" -022" - 0.33"“
Height 0 . 33“... 0'46gnwt 0.22.01 _o_19fi* 0.32.1111. _
Lodging 0 .30""" 0.37"" ns ns ns 0.19"
Maturity ns 0 18“ 0.65““ -0.18“ 0.49"“ 0. 29“"

DS ns 0.29"" ns -0. 26““

    

“S
.P< 05 tiP< ‘01 ".P<. 001; t'iiP< ooons1

  

 

 

0 .36‘.i‘ . 0‘69ttifi
B\OmaS$ O .38...‘ n5 0_51cn\u
1 00 sw 0 .26“ ns ns
HI 0.18‘ -0 22“ 0.56““
Flowering ns 0 .42.... 0.27..
Height 0.38"" ns 0.48""
Lodging - ns 0.29.”...
Maturity -0.14" - _0.15.
___05 0.16' -0.14* -
Non-stres Lod in Maturi 08
Yield 0.50"" -0.14* 0.20“
Biomass 047“" ns n s
1 00 SW -0.18" 0.61“" -0.26""*
Hi 0.29.1”: '0.30"“ 0.25,",
Flowering n5 047.... 0.32...
Height ns 0.28"" -0.16’
Lodging _ -0'30'lmn: ns
Maturity -0.1 71v- _ 4137“"
DS -0 49h"

 

*P<. 05 ‘*F’< .;o1 "*P<. 001; n"""P< 000‘

48

 

Disease incidence data revealed a greater resistance to ASB in L91 than in L88,
derived from the ASB resistant parent, TLP 19. Population L88 had a 6 % higher DI than
L91 yet averaged 113 kg/ha more in yield (Table 12). Population and parental means were
not significantly different (p<0.05). In the moisture stress and non-stress treatments,

B98311 yielded more, yet had a two-fold higher D1 in comparison to the two other
parents. TLP 19 and VAX 5 had moderately low DI values at 0.15 and 0.20. Two RILs,
L9] -45 and L88-76, had the lowest D1 in each population, Whereas two other RILS, L91-
30 and L88-69 that were selected as drought resistant based on GM had moderately 10W
DI values.
Correlations between agronomic data and yield can assist breeders in designing
plant phenotypes that thrive under moisture stress. Harvest index (HI) was highly
significant and strongly associated with yield in both populations in each treatment. The
moderate associations of height and lodging to yield and biomass suggest that tall plant
that lodge positively influence yield and biomass. Desirability score was highly
associated with yield and moderately associated with biomass in the stress treatment, but
not in the non—stress treatment. In the non-stress treatment, 100 sw was highly associated
with days to maturity. Days to maturity was negatively associated to yield in both
populations and treatments except for L88 in the non-stress treatment. Height was
positively associated to DS in the stress treatment and negatively associated to DS in the
non-stress treatment. Individual agronomic traits did not associate strongly enough with

yield under stress to support indirect selection, so direct selection based on yield is

required in breeding for drought resistance in common bean.

49

 

 

 

 

 

 

I L88 RlLs

 

 

 

 

 

 

 

 

0 L91 RlLs
x Checks and Parents
_.___# L88 Mean
- - - - L91 Mean
0.00 0.20 0.40 0.60
Disease Incidence
/

 

 

Figure 8. Field incidence of Macrophomina phaseolina compared to yield under stress in
160 genotypes grown in Honduras in 2001.

Table 12. Selected genotypes and means compared for their disease incidence (DI), plant
stand at harvest, yield under stress (Yd), yield under non-stress (Yp), and geometric mean
(GM) of moisture treatmentigrown in Honduras, 2001.

 

 

 

 

 

 

 

 

 

Genoype DI Stand Yd Yp GM
°/o —— kg/ha ——
L91-45 0.05 (1 )1“ 97 266 2468 810
L88-76 0.09 (4) 90 363 1862 822
V8025 0.13 (13) 93 210 1783 612
L91-30 0.13 (15) 90 599 1922 1073
SEA 5 0.14 (20) 77 524 1521 893
Tlo Canela 75 0.1 4 (22) 90 218 1657 602
TLP 19 0.1 5 (24) 100 169 2399 637
L88-69 0 . 1 6 (31 ) 90 680 2432 1 286
VAX 5 0.20 (61 ) 90 249 1765 663
Rio Tibagi 0.23 (75) 90 372 2108 886
BAT 477 0.24 (88) 90 400 1536 784
EAP 9510-77 0.29 (120) 90 232 21 12 699
Tacana 0.30 (125) 93 213 2097 667
__B_98311 0.42 (154) 90 375 241 1 951 .
Mean. L88 0.27 320 2057 791
Mean, L9 1 0.21 207 1858 591

 

:17 Ranking based on DI for 160 gen0types.

50

 

Root Study

Root traits were expected to differ due to associated differences in growth habit
between the parents. TLP 19 has a type HI grth habit while B98311 and VAX 5 exhibit
a type H habit. At nine days after transplanting (dat), the root system of TLP 19 was
significantly smaller than either B98311 or VAX 5. Due to root length differences
between B98311 and TLP 19 and the lack of differences between B98311 and VAX 5,
only the root characteristics of population L88 were studied.

The root characteristics measured in were total root length, length according to
diameter class and fractal dimension. Significant genotypic differences were found for
total root length and fractal dimension in L88 (Table 13). Within the ten root length
diameter classes, the two classes for each extreme, A and B; I and I, showed significant
genotypic differences while classes C through G did not. The ten different root diameter
classes previously reported in common bean (Y abba, 2001) were grouped into fine (A-C),
intermediate (D-G), and taproots (H-J). Fine roots described length for roots with 0-1.50
mm in diameter. Intermediate roots were classified as having diameters 1.51-3.50 m.
Taproot length is characterized as having a diameter greater than 3.51 mm. The extreme
classes had low CV values whereas the intermediate classes had CV values that exceeded
100 %.

The correlations of root characteristics to yield data showed unexpected results.
Beans having a high Yd were expected to have a deep taproot. The fine roots, which
accounted for 99 % of the total root length, correlated to Yd and GM in Honduras whereas
taproot length correlated to Yp (Table 14). The negative associations of the fine roots in

class B suggest that as root length with a diameter of 0.50 mm to 1.00 mm decrease, yield

51

Table 13. Analysis of Variance for the 81 RILS in population L88 for Total root length,
Fractal Dimension, and root length according to 10 different diameter widths (A-J).

Source l

genotype 77 295173

Total Root Length

F Test
1 .62“

MS

MS
0.0029

Fractal Dimension

F Test

1 .44‘

 

 

 

 

 

 

block 3 3135560 17.20"“ 0.0746 36.91“"

error 162 182248 0.002

Fine Roots At 8 C

Source i MS F Test MS F Test MS F Test

genotype 77 149161 1.79“ 21685 1.43’ 701 1.18

block 3 1980019 23.73““ 270907 17.88““ 28355 47.87"“

error 162 83422 1 51 53 592

Intermediate

Roots D E F G
Source i MS F Test MS F Test MS F Test MS F Test
genotype 77 68 1.11 14 1.00 2 1.08 0.36 1.21
block 3 4977 81.29““ 851 58.52“" 135 66.45““ 17.00 57.29“"
error 162 61 15 2 0.30

Tag Roots H I J

Source _D_F_ MS F Test MS F Test MS F Test

genotype 77 0.12 0.99 0.05 1.48‘ 1.59 150*

block 3 3.75 30.50“” 0.44 11.66"" 5.98 5.64“

error 162 0.12 0.04 1.06

 

'P<.05; **P<.01; ***P<.001; ““P<.0001
1 Root diameter classes A, B, C, D, E, F, G, H, I, J are 0-0.5, 0.5-1.0, 1.0-1.5, 1.5-2.0, 2.0-
2.5, 2.5-3.0, 3.0-3.5, 3.5-4.0, 4.0-4.5, and greater than 4.5 mm, respectively.

(A)

N
\

/
/
/\
/

 

/

\

 

Figure 9. Representations of herringbone (A) and dichotomous (B) topologies.

52

Table 14. Correlation values between root characteristics and yields in the Saginaw 2000

and Honduras (Hon) 2001 egeriments for the 81 R_ILs of opulation L88.
Total Root Length Fractal Dimensign_

 

 

 

 

 

 

 

 

Hon Yd ns -0.13'
Hon Yp ns ns
Hon GM ns ns
Saginaw ns ns
Fine roots All: __3— __§___
Hon Yd ns -0.12t ns
Hon Yp ns ns ns
Hon GM ns 0191 "5
Saginaw ns 0.1 11 "5
Intermediate Roots D __E__ _F—. _G_
Hon Yd ns ns ns ns
Hon Yp ns ns ns ns
Hon GM ns ns ns ns
Saginaw ns ns ns ns
Tap roots H | J
Hon Yd ns ns ns
Hon Yp ns 0.19“ ns
Hon GM ns ns ns
figinaw ns ns ns

 

1' P<.10, *P<.05; **P<.01; ns - non-significant
1 Root diameter classes A, B, C, D, E, F, G, H, I, J are 0-0.5, 0.5-1.0, 1.0-1.5, 1.5-2.0, 2.0-
2.5, 2.5-3.0, 3.0-3.5, 3.5-4.0, 4.0-4.5, and greater than 4.5 mm, respectively.

will increase. The I class, which is representative of taproots, was positively associated to
Yp in Honduras (r = 0.19“). Fractal dimension which measures root architecture was
significantly correlated to Yd in Honduras (r = -0. 13*). Significant differences were
observed for the top and bottom five genotypes in total root length, fi'actal dimension and
each root diameter class. The mean values of drought resistant and susceptible genotypes
corresponded to the correlation values for each measurement. Drought resistant genotypes
had less root length than drought susceptible genotypes in every root trait category except
the I class (4.0-4.5 m) (Table 15). Fractal dimension was also a lower value in drought
resistant genotypes. The opposite effect was observed in the parents, in that, B98311 was

greater than TLP 19 in every root length measurement and fractal dimension. Only in the

53

 

Table 15. Mean values of total root length, fractal dimension, fine roots (A-C) and
taproots (H-J) of drought resistant and drought susceptible RILS and parents of population
L88 obtained by the root pouch method.

 

 

 

 

 

 

 

 

 

genogge Total Fractal A1 5 C H I J
mm ------ mm ------ ------- mm ------
Drought Resistant
L88-30 2106 1 .52 1506 529 55 0.02 0.26 3.54
L88-63 2054 1.47 1645 366 32 0.28 0.10 2.95
L88-74 2019 1 .47 1 560 409 36 0.12 0.00 3.45
L88-13 1928 1.47 1464 413 38 0.15 0.13 3.51
_L88-69 153g 1 .46 1 163 330 29 0.31 0.08 LE
Mean 1928 1.48 1468 410 38 0.18 0.11 3.12
Drought Susceptible
L88-64 2368 1.56 1648 614 79 0.25 0.04 4.17
L88-37 21 19 1 .53 1566 483 57 0.21 0.07 3.35
L88-18 2056 1 .50 1562 441 42 0.26 0.08 2.41
L88-02 2007 1 .49 1564 392 37 0.20 0.09 3.29
_LB8-04 1948 1.50 1391 504 41 0.33 0.16 3.09
Mean 2100 1 .52 1546 487 51 0.25 0.09 3.26
Parents and Ranges
89831 1 2295 1 .53 1660 547 62 0.32 0.19 3.56
TLP 19 1618 1.48 1174 391 40 0.16 0.06 2.52
Maximum 2660 1 .61 1965 732 123 1 .13 0.55 4.72
Minimum 1285 1.44 908 329 24 0.04 0.01 1.77
Mean: 1959 1.50 1415 475 51 0.26 0.16 3.07
LSD (0.05) 688 0.07 466 198 39 0.57 0.31 1.66
CV 22 2.99 20 26 47 133.54 119.64 33.58

 

 

 

1' Root diameter classes A, B, C, H, I, J are 005, 0.5-1.0, l.0-1.5, 3.5-4.0, 4.0-4.5, and

greater than 4.5 mm, respectively.
I 83 genotypes are included in the calculations of mean, LSD and CV values.

I class (4.0-4.5 mm) did B98311 and TLP 19 correspond to the correlation values
observed among the RILS. A lower fractal dimension correlated to yield under drought
stress, yet the drought resistant parent, B98311, had a greater fractal dimension and Yd
than the susceptible parent, TLP 19. Significant differences between the parents were
observed only in the A class (0-0.5 mm) category. Although the resistant and susceptible
genotypes were not significantly different in the B and I root classes and fractal dimension
, the correlation values indicate that these root characteristics are important in the yield

performance of beans grown under drought stress in the lowland tropics.

54

Marker Study

Molecular markers previously associated with drought resistance in pinto bean
(Schneider et al., 1997a) were tested across both black bean populations. Only 40 of the
70 reported RAPD markers (Table A8) were tested for polymorphisms between T-3016
and Raven, the parents of B983 1 1. The 40 RAPD markers selected for testing represented
those that were significantly associated with GM in a previous study (Schneider et al.,
1997a). Nine of the 40 RAPD markers tested were polymorphic between T-3016 and
Raven suggesting that B98311 had the same marker phenotype as T-3016. Three of these
markers were present on linkage group (LG) 4, while the other six were unlinked or on a
LG with low association to performance under stress (Table A8). Of the nine reported
LGs, only two (4 and 9) significantly associated with GM in a combined analysis across
five locations (Schneider et al., 1997a). Markers on LG 9 were monomorphic between the
parents, whereas markers on LG 4 were polymorphic (Table 16). RAPD primers 1706970,
103”30 and A16850 from LG 4 were screened across both populations. The lack of clear
amplification of A16850 made the marker unscoreable. F 06970 and I03”30 were 100 %
linked in population L88 while 3 recombinants existed in population L91. In L88, both
primers were significantly associated with yield potential (R2 = 0.05*) (Table 17). In
multiple regression analysis, 103ll 30 accounted for 5 % of the variation (R2 =0.05; p<0.10)

in population L91.

55

Table 16. Preseice/abgence of RAPD markers in six bean genotypes
LGt RAflDt T-3016 Sierrg 898311 _R_ayen VAX 5 TL
9 A318650 + + + + +
H1959,
4 v01”,
F0697o
A165,,o
I03“...

1 LG - Linkage Group
IRAPD markers were previously associated to drought resistance in one pinto bean
population (Schneider et al., 19973).

i

 

w
—L
‘D

 

 

+
+

-...+i

++++.

+++II

Table 17. Coefficients of determination (R2) accounting for the variation in yield under
drought (Yd), yield under non-stress (Yp) and geometric mean (GM) for two RAPD

markers.

 

L88 L91
F06m '031139 F0691n '031130
Yd 0 0 0.03 0.03
Yp 0.05" 0.05’ 0.05 0.051
GM 0 0 0.03 0.03

 

T P<.10, *P<.05

Multiple Regression Analysis

For population L88, fractal dimension, root classes B and I and primer F06970 were
combined in a multiple regression analysis to identify which combination would explain
the greatest variation for performance under drought. Root class I (4.0-4.5 mm) explained
12.7 % of the variation alone for Yp in Honduras (Table 19). A larger percentage, 16 %,
of the variation was explained when additional measurements of fiactal dimension, root
class B (0.5-10. mm) and primer F06970 were included. Statistically, the best model has
the highest adjusted R2 value or the lowest C(p) value. Including primer F06970 with the I
class increased the R2 value by 2 %. This same effect occurred to a greater extent in the
multiple regression analysis for Yd in Honduras. Alone, fractal dimension and F06970
explained 3.5 and 3.1 % of the variation respectively, but together they explained 7.6 % of

the variation (Table 18). This two-variable model also had the highest adjusted R2 value

56

and the lowest C(p) value. Using the four variables together in one model explained 8.8 %
of the variation in Yd. The combination of root characteristics and marker values

explained a larger amount of the variation than any single variable.

'57

Table 18. Coefficient of determination (R2) selection method for yield under stress in
Honduras evaluating fractal dimension, root classes B (0.5-1.0 mm) and I (4.0-4.5 mm),

 

 

 

 

 

and RAPD marker F06970 (F 06).
Number in Adjusted
Model R-Sguare R-Sguare C(g) MSE Varia_bles in Model
1 0.035 0.0228 3.41 14 3099.921 fractal
1 0.0314 0.0191 3.71 3111.433 F06
1 0.0297 0.0174 3.8552 3117.032 B
1 0.0001 -0.0125 6.316; 3211.906 I
2 0.0762 0.0525 1.9752 3005.5 fractal F06
2 0.0727 0.0489 2.2693 3016.98 B F06
2 0.0427 0.0182 4.7647 3114.414 fractal l
2 0.0386 0.0139 5.1141 3128.058 B I
2 0.035 0.0103 5.4067 3139.48 fractal B
2 0.0316 0.0068 5.6926 3150.645 I F06
3 0.0859 0.0503 3.1685 3012.624 fractal I F06
3 0.0851 0.0494 3.2365 3015.314 B l F06
3 0.0765 0.0405 3.9558 3043.764 fractal B F06
3 0.0431 0.0058 6.7394 4153.861 fractal B I
4 0.0879 0.0399 5 3045.512 fractal B I F06

 

Table 19. Coefficient of determination (R2) selection method for yield under non-stress in
Honduras evaluating fractal dimension, root classes B (0.5-1 .0 mm) and I (4045 mm),

 

 

 

 

 

 

 

 

58

and RAPD marker F069m (FO6).
Number in Adjusted .
Model R-Sguare R-Sguare C(g) MSE Variables in Model
1 0.1273 0.1163 2.0153 9974.918 l
1 0.0166 0.0042 12.0397 1 1240 F06
1 0.013 0.0005 12.3658 11282 fractal
1 0.012 -0.0005 12.4587 1 1293 B
2 0.1424 0.1204 2.6537 9928.709 I F06
2 0.1362 0.114 3.2161 10001 B l
2 0.1313 0.109 3.6562 10057 fractal I
2 0.0263 0.0013 13.1658 1 1273 fractal F06
2 0.0247 -0.0003 13.3061 11291 B F06
2 0.0131 -0.0122 14.3601 11425 frac_tal B
3 0.1562 0.1233 3.4011 9895.425 B I F06
3 0.1489 0.1157 4.0659 9981.525 fractal I F06
3 0.1394 0.1059 4.9188 10092 fractal B I
3 0.0263 -0.0116 15.1636 11419 fractal B F06
4 0.1606 0.1165 5 9972.991 fractal B I F06

 

DISCUSSION

Yield

The main objective of this study was to evaluate the performance of two RIL
populations under terminal drought stress and non-stress conditions in the lowland tropics
of Central America. The lowland tropics of Honduras provide appropriate conditions to
evaluate terminal drought. By identifying genotypes resistant to terminal drought,
breeding programs located in Honduras and other Central American countries will be able
to use these genotypes to improve local varieties for performance under drought. RILs
selected as drought resistant out-performed all local checks, drought resistant checks and
the drought resistant parent based on GM (Tables 4, 5 and 6). Likewise, drought
susceptible RILS performed well below all checks and parents. The stability of the
drought resistant RILS was evaluated across locations with different intensities of drought
stress. An experiment that included the resistant and susceptible genotypes selected in
Honduras was conducted in Montcalm county, Michigan under moisture stress and non-
stress conditions. Despite the absence of a significant drought stress in Montcalm in 2001,
GM values across treatments were moderately correlated between experiments in
Honduras and Montcalm (Figure 5). Drought resistant genotypes selected in Honduras
were among the highest yielding genotypes while drought susceptible genotypes were
among the lowest yielding in Montcalm (Tables 8 and 9).

The Honduran field experiment in the lowland tropics experienced a severe
terminal drought stress, DII = 0.82. This stress was more severe than previous

experiments conducted with beans under rain-fed conditions in the Mexican highlands

59

(D1

in‘.‘

erg

cm

W

be

 

(DII = 0.49; Schneider et al., 1997b) and under rain shelter controlled, drought treatments
in Michigan (DII = 0.63; Ramirez-Vallejo and Kelly, 1998). Lowland tropical areas can
experience decreasing soil moisture and increasing temperatures, both of which
contributed to the substantial reduction in Yd in Honduras.

Resistant and susceptible RILS were identified among the populations based on
yield in relation to the parents. B98311 was selected as the drought resistant parent
because it was the highest yielding genotype in drought experiments conducted in
Michigan in 1998 (Kolkrnan and Kelly, 1999). TLP 19 and VAX 5 were selected for
adaptation to lowland tropics without prior knowledge of their response to drought stress.
TLP 19 and VAX 5 yielded 50 and 30 % less than B98311 under stress. The
corresponding population means of each parent did not follow the same relationship.
Even though TLP 19 had a lower yield mean under stress than VAX 5, the corresponding
population L88 mean was greater than the population mean of L91 where VAX 5 was the
parent (Tables 4 and 5). Population L91 had a lower mean yield than its drought
susceptible parent, VAX 5. Many of the RILs in L91 in the drought stress treatment had
high biomass but low seed yield. This relationship was displayed by RIL L91-69, which
under stress, ranked last in yield, but first in biomass (Table 15). Under stress, this RIL
remained longer in a vegetative stage and began to flower late into the period of stress
producing a low yield as a result of low HI (Table 6). Many RILS flowered late in
population L91 in Honduras and exhibited low H1. The biomass means between
populations differed by 15 kg/ha whereas the yield means differed by 106 kg/ha. Harvest

index under stress for populations L88 and L91 was low, 0.19 and 0.13, respectively, yet

60

the RILs in population L88 were more efficient in partitioning nutrients to the seed than
those in population L91.

Biomass has been suggested as an indirect selection criterion for Yd since it is
highly correlated with biomass (Acosta-Gallegos, 1988). This correlation is logical in that
high yielding genotypes need to fix greater biomass to partition to the seed. Yet, selection
for biomass can indirectly increase days to maturity. In populations L88 and L91, biomass
was moderately correlated with Yd (r = 0.61"" and r = 0.53”“) and highly correlated
with Yp (r = 0.91"“ and r = 0.86""), respectively (Table 10). These values are greater
than the correlation between Yd and Yp in each population. Therefore, under severe
stress, biomass would be a better indirect measurement of Yd than Yp. For temperate
climates less affected by drought like Michigan, selection for biomass would be counter-
productive since days to maturity would increase. In Michigan, HI would be a better
selection criteria for Yd. HI and biomass should always be used together when selecting
for performance under stress, so that plants with high biomass and low reproductive
efficiency are not selected.

Growth habit in common bean has also been associated with Yd. Indeterminate
genotypes are more suitable than determinate genotypes for production in semi-arid areas
(Samper, 1984; Acosta-Gallegos and Adams, 1991). Indeterminate genotypes exhibit
early vigorous establishment and accumulate a greater biomass with the ability to transfer
assimilates to the seed (Samper, 1984; Acosta-Gallegos and Adams, 1991). All three
parents in this study were indeterminate, yet they exhibited different types of
indeterminancy. B98311 and VAX 5 have a type II, short vine, erect grth habit while
TLP 19 has a type IH, prostrate vine. Comparisons between type II and type III

61

indeterminacy for drought resistance were not made in this study. Population L88
segregated for growth habit whereas population L91 exhibited only type 11 growth habit.
The range of growth habits present in L88 might have given individual genotypes an
opportunistic edge over the type II individuals in population L91. Therefore, a strict
adherence to type II growth habit may not be beneficial in selecting for drought resistance
in Central America.

The field experiment in Montcalm, Michigan was conducted to validate the
Honduran results. The resistant and susceptible RILS selected in Honduras experienced
minimum drought stress in Michigan and consequently the effect of stress could not be
evaluated at a second location. Despite the lack of drought in Michigan, relationships of
yield between Honduras and Montcalm were compared using regression analysis. The
strongest correlation for Yd was shown within the susceptible genotypes. The negative
affects of late flowering and inability to partition nutrients to the seed were also
detrimental to yield in Michigan. Resistant genotypes showed a higher correlation
between locations than susceptible genotypes for Yp (Figure 7). The moderate
correlations of Yd (r = 0.45’), Yp (r = 0.52*), and GM (r = 0.63") supported the
adaptability of selected RILs to temperate conditions and the consistent performance
between locations. Selection under drought conditions in Honduras was successful in
identifying high and low yielding RILs that expressed similar potential in the Michigan
environment.

In Honduras, a high CV was recorded due to high experimental error resulting
from genetic susceptibility of RILs to diseases such as ASB in the stress treatment and the
higher environmental variation. The high CV in our experiment was considerably higher

62

than previously published studies of bean grown under drought stress (Schneider et al.,
1997b). The experimental design used in Honduras was a CRD which accounts for less of
the total error than a RCBD or lattice designs that better control environmental variation.
Plant stand was reduced and consequently, yield due to ASB infestation under stress
conditions only. The mean yield reduction due to stress was 85 and 89 % for populations
L88 and L91, respectively. In breeding for resistance to terminal drought stress, ASB
resistance among genotypes must be considered.

Although variation was high within the drought stress treatment, the 150 RILS in
the non-stress treatment can be directly compared between locations. The top yielding
genotypes in Honduras were poor performers in non-replicated trials in Saginaw,
Michigan in 2000. The lack of consistent performance between these locations can relate
to adaptation of the genotypes, environmental differences and different planting
populations. All plots were adjusted by plant stand using covariate analysis, so that valid
comparisons could be made. One genotype, L88-69, had consistent high yield in all
locations and treatments.

Bean genotypes that yield well under stress do not always yield well under non-
stress. A strong relationship between Yd and Yp was not reported in previous experiments
(Ramirez-Vallejo, 1992). Different mechanisms within the plant contribute to high Yd
and Yp. A breeding strategy for drought resistance should combine in a genotype both
high Yd and high Yp (Schneider et al., 1997b). This strategy would be more effective in
population L88. B98311 yielded above average under drought (375 kg/ha) in Honduras
while TLP 19 (169 kg/ha) yielded below the mean. TLP 19 was a poor yielding genotype
under stress in Honduras, but was equivalent to B98311 in Yp in Honduras. TLP 19 out-

63

 

yield
seve

Mir

yielded B98311 in Montcalm, Michigan by 13 %. This parental combination resulted in
several progeny L88-63 and L88-69 with superior yield in each treatment in Honduras and
Michigan (Tables 4 and 5).

The parent of population L91, VAX 5, was derived from tepary bean which is
known to be drought tolerant (Thomas et al., 1983). Genes from tepary bean present in
VAX 5 could have aided in tolerance to stress. VAX 5 produced a moderate Yd, yet poor
combining ability for performance with B98311 was observed that could have resulted
from the tepary ancestry.

Both populations belong to common bean race, Mesoarnerica in the Middle
American gene pool. Gerrnplasm from race Mesoarnerica is recognized as a source of
yield genes for stressed or non-stressed environments in Central America (White et al.,
1994a). The two other races of common bean in the Middle American gene pool are
Durango and Jalisco. These races have been exclusively screened for additional drought
resistant genes (Singh, 1995; Teran and Singh, 2002). Durango race cultivars have shown
a higher yield under drought stress than J alisco race cultivars (Teran and Singh, 2002).
Moderate success in breeding for drought resistance has been achieved in the Durango
race (Acosta-Gallegos and Adams, 1991; Schneider et al., 1997b), which could result from
its limited adaptation. Mesoamerican genotypes have a broader adaptation than Durango
genotypes. Both races could endure drought periods by different mechanisms. Interracial
crosses of Durango and Mesoamerican genotypes may provide a strategy to improve
drought resistance by combining different adaptation and yield performance traits from

both races.

64

Mesoamerican genotypes deriving their drought resistance from Durango varieties
could have an advantage in complementation of drought resistance from both races.
B98311 from race Mesoarnerica was derived from a drought resistant Durango genotype,
T-3016 which is a non-commercial, Durango race genotype previously identified as the
most drought resistant genotype based on GM from a cross of Sierra/AC1028 (Schneider
et al., 1997b). Drought resistance was transferred from the Durango genotype to the
Mesoamerican genotype as the other Mesoamerican parent, Raven, demonstrated no
drought resistance. The most drought resistant RILS in both populations, L88-63 and L91-
30, exceeded the yield of previously recognized drought resistant genotypes BAT 477,
V8025 and Rio Tibagi by 40 and 17 %, respectively. Mesoamerican drought resistant
genotypes BAT477, V8025 and Rio Tibagi have shown moderate Yd and high Yp (White
et al., 1994a). The stability of performance under stress for the drought resistant RILs will
be determined across additional locations and years. A Mesoamerican genotype with
complimentary drought resistance genes from a Durango genotype might have a greater
impact on drought-prone areas than either a Mesoamerican or Durango genotype.

The goal in breeding for drought resistance in common bean was to combine
mechanisms of drought tolerance and avoidance into a broadly-adapted genotype that
produced high yields under stress and non-stress conditions. The resistant mechanisms of
avoidance and tolerance are so integrated that separation is not always possible.
Improving drought resistance will require combining plant traits known to be beneficial in
performance under stress. In traditional breeding, selection based on GM across
treatments accounts for all traits contributing to yield without distinguishing between

drought resistance mechanisms. If avoidance traits could be combined with tolerance

65

traits, drought resistance could be improved in common bean. Traits such as growth habit,
root architecture, osmotic adjustment, or indirect selection using linked molecular markers

could be useful in population development for drought resistance in common bean.

Root Study

The hypothesis that deep-penetrating roots contribute to drought resistance in
common bean was tested. Separate field experiments have supported this hypothesis
(Sponchiado et al., 1989; White and Castillo, 1989; White and Castillo, 1992). Since only
a small numbers of genotypes have been compared and BAT 477 was mainly used in
associating drought resistance to roOting depth, new studies were deemed necessary.

The pouch method was used to ascertain whether seedling root growth correlated
to field performance. Genetic differences in root systems can be measured in early stages
of development of common bean (less than 20 dap) . The pouch method was designed to
study root vascular systems (McMichael et al., 1985) and was modified to study drought
resistance in bean (Yabba, 2001). It was also used to study plant response to phosphorus
availability (Liao et al., 2001). Genotypes responding in a phosphorus-efficient manner
allocated roots to shallow soil horizons during phosphorus stress. Inefficient genotypes
would continue to grow deeper. These differences in root length for phosphorus
accumulation substantiate an investigation of root length correlations to drought stress.

One disadvantage of the pouch study is that it yields a two-dimensional root.
Although not representative of natural conditions where roots can grow in three-
dimensional directions, it makes digital scanning easier. Digital root images from the

pouches were used to measure root length and root architecture. Computer analysis using

66

 

 

WinRhionM increases root measurement efficiency so that more tedious measurements
can be performed in a shorter time and with less error.

Root length and root architecture were measured in population L88 in which the
parents contrasted in root length (Table 15). Root architecture among RILS did correlate
with yield performance under stress, whereas total root length did not (Table 14). Using
fractal dimension to describe root architecture, drought resistant genotypes exhibited a
root structure equipped for deep soil-water extraction while drought susceptible genotypes
did not possess the same root architecture. This relationship was not observed among the
parents. The drought resistant parent, B98311, exhibited a highly branched root system
unlike the deep taproot structure considered to be important in terminal drought conditions
(Sponchiado et al., 1989). Additional drought resistant genes unrelated to root structure

must be present in B98311 to account for its yield performance under stress.

Root Length

The measurement of root length for the whole root system and for specific
diameter classes provides usefirl information into root components. Ten different root
classes based on root diameter and used in previous root studies (Yabba, 2001) were
measured to provide a better understanding of root architecture in different bean
genotypes. Root components were primarily classified into four classes: adventitious,
basal, tap and lateral roots (Stoffella et al., 1979). The acquisition of root data through
WinRhionM in the present study did not measure root length based on these root

components, but on different components that have the same diameter. Even though

 

67

correlations can not be made to individual root components, the diameter of root segments
will govern its importance and use in water accumulation.

Results were ascertained based on the assumption that root diameters will continue
to expand such that the large diameters will always be larger than the small diameters.
Therefore, root classes of small diameters (0-1 .5 mm) represent fine roots and large
diameters (>3.5mm) represent taproots. Taproots positively correlated to Yp and fine
roots negatively correlated to Yd (Table 14). The correlation with taproots were opposite
than expected as taproots were expected to associate with Yd. This data showed that short
fine root length supports yield performance under stress conditions and long taproots
contribute to yield potential. Roots that are longer at a greater diameter have more
potential to transport increased volumes of water from the soil. Potentially these roots
would continue to increase in grth so that the root genotype is representative of a deep
and large-width taproot. Large taproots are able to keep the shoot well supplied with
water. The I class (4045 mm), representing the taproots, was significantly correlated
.with Yp, r = 0.19"""r (Table 14). Since the parents showed the same relationship as the
drought resistant and susceptible RILs in the I class, this taproot measurement is suggested
as a selection criterion for Yp.

Fine roots were negatively correlated to yield performance under stress. Greater
fine root length would be detrimental to performance under stress. Although fine roots,
measured at the 0.5-1 .0 mm diameter, negatively correlated to Yd in Honduras, they
positively correlated to yield in Saginaw, Michigan. Since the plots in Saginaw were

space-planted, individual plants grew without competition nor water stress. More fine

68

 

roots allow a more extensive exploration of the soil which would logically account for
more yield at this location.

This conclusion was supported by a previous study in which root characteristics
were studied in drought resistant and susceptible bean genotypes (Yabba, 2001). Eight
genotypes were compared in the pouch method under non-stress and water stressed
treatments induced by abcisic acid. Drought resistant genotypes had less fine root length
than drought susceptible genotypes suggesting that fine roots did not impart resistance to
drought (Yabba, 2001). Total root length of these eight genotypes averaged at 1899 mm
(Yabba, 2001). The RILs of population L88 had a slightly larger total root length of 1959
mm. The larger root length can be attributed to the root length of B9831 1. This breeding
line was derived from Michigan germplasm whereas the eight other genotypes were
mainly derived from Latin America (Yabba, 2001). The range of total root length was
also greater in pOpulation L88 than among the eight genotypes. In population L88 mean
values ranged from 1230 to 2694 compared to 1530 to 2350 mm from the eight tested
genotypes (Yabba, 2001). Greater variation was recorded in population of L88, which
encourages the use of populations segregating for root characteristics. However, the
differences among individual genotypes for total root length were not significant enough

to be used as an indirect selection criterion.

Root Architecture
Fractal dimension was first described to mathematically explain complex objects
of nature (Mandelbrot, 1977). Recently in common bean, fractal dimension has been used
to describe root architecture which aids in drought avoidance (Lynch and van Beem, 1993;

69

 

 

 

Niel

des

€85

l9

Nielsen et al., 1997). Fractal dimension is promising as a selection criterion because it
describes root branching patterns independently of root size (Fitter et al., 1988). It is also
easy to measure and is a single number amenable to mass screening (Lynch and van Beem,
1993)

In a breeding program, screening methods cannot be time consuming nor laborious
as large numbers of individuals need to be evaluated. The pouch method and WinRhionM
computer analysis make data collection for fractal dimension relatively easy while the
digital root image provides a permanent record. The WinRhionM program has simplified
a difficult mathematical calculation. Complete root systems must be measured and need
to be in a two-dimensional format. This orientation is not representative of the root
system growth in situ. In order to fully exploit the genetic variation of root architecture,
root measurement technology must consider three-dimensional models (Lynch and van
Beem, 1993).

Measurement of fractal dimension for roots grown in a narrow space or excavated
and flattened prior to analysis may be problematic (Nielsen et al., 1997). Three-
dimensional root structures can be measured in trench excavations by marking the root
intersections on the exposed planes of soil. This method is not suitable for screening large
numbers of individuals. Currently, screening for root characteristics in a breeding
program is only possible with two-dimensional representations of roots.

Fractal dimension is always reported in values between one and two. Roots
exhibiting a deep penetrating taproot or herringbone structure (Figure 9) should have a low
fractal dimension, less than 1.50 whereas, highly branching roots exhibiting a dicotymous
structure should have a high fractal dimension greater than 1.50. This derivation was

70

supported in a study (Lynch and van Beem, 1993), where root systems from short, bush
type I bean plants had a larger fractal dimension than climbing type IV plants with a long
taproot and fewer root meristcms. The range of fractal dimension in the 81 RILS of
population L88 ranged from 1.41 to 1.65 with a mean of 1.50. This range is similar to the
reported fractal dimension of other leguminous species, peanut (1.56), garden pea (Pisum
sativum L.) (1.57) (Tatsumi et al., 1989), and common bean (1.33 to 1.59) (Lynch and van
Beem, 1993). The top five drought resistant RILs had a mean fractal dimension of 1.48
whereas mean fractal dimension of the corresponding five susceptible RILs was 1.52. The
resistant RILs were not significantly different than the susceptible RILS for fractal
dimension.

The negative correlations (r = 0.13*) of fiactal dimension to Yd indicates that as
fractal dimension gets smaller, Yd increases. Fractal dimension did not adequately
separate drought resistant genotypes from susceptible genotypes. Since only field data
under stress was available from one location, results are tentative.

Root characteristics associated with drought resistance allow breeders to combine
root traits with tolerance traits and to select genotypes with specific root characteristics for
drought resistance. Detailed root measurements can help explain the plant’s response to
drought stress. Avoidance and tolerance traits could be separately identified and
combined to improve drought resistance in common bean.

Since the fractal dimensions of B98311 and TLP 19 do not correlate with drought
resistant and drought susceptible RILs, fiactal dimension as a technique for selecting
parents is not suggested. Out of the three root characteristics that correlated to yield
performance under stress only the I class (4.0-4.5 mm) reflected the parental values.

71

 

 

The

tecl

19

CE

tl

Therefore, selection for the larger taproot (1 class) is suggested as an indirect screening

technique for selecting parents of fiiture drought resistant populations.

Molecular markers

RAPD markers previously associated with drought resistance (Schneider et al.,
1997a) were tested for their usefulness in this study. Two linkage groups (4 and 9) which
explained more than 10 % in yield variation under drought stress in pinto bean populations
were analyzed in the two black bean populations L88 and L91 (Table A8). Only one of
these LGs was polymorphic between three parents.

RAPD markers associated with drought resistance in common bean were
previously used in MAS (Schneider et al., 1997a). Nine LGs were present in the one
population. Although two LGs (4 and 9) explained over 10 % in the combined analysis
across seven locations, only LG 9 with two flanking markers was used in MAS. A 3 %
gain in Yd was obtained in one pinto population using the four markers as selection
criteria. The markers were inefficient in previous studies because Yd was a moderately
heritable trait in one pinto population being tested (Schneider et al., 1997a).

The most resistant genotype, T-3016 from one pinto population, was a parent of
B98311. Most markers linked to drought resistance in T-3016 were not passed on to
B98311, except for three of four markers on LG 4 (Table A8). Linkage group 4 explained
12 % of the variation in Yd and GM across seven locations (Schneider et al., 1997a).
Only two markers from LG 4 were tested in populations L88 and L91 since the other two
markers were either not scoreable (A16850) or monomorphic between parents (V 01330)-
The two markers, F06970 and 103,130, explained 5 % of the variation in Yp in populations

72

 

L88

ior 1

CM

V3

it

L88 and were tightly linked with no cross-overs. Every RIL that showed a band presence
for F06970 also showed a presence in 1031 130 and vice versa, whereas they were mapped 8
cM apart in a previous study (Schneider et al., 1997a).

The linkage relationship of F 06970 and 1031130 in population L88 and the extent of
variation explained suggest that a QTL was transferred to the L88 population from T-3016
through the resistant parent, B98311. Further research is needed to validate the continued
usefulness of these markers in MAS for drought resistance and locate the markers on the
bean core map (Freyre et al., 1998).

Regression analysis was conducted to determine the best variables to use for
indirect selection for Yd. Root characteristics and molecular markers associated with
yield potential, explained 9 % of the variation of Yd and 16 % of the variation in Yp. The
reliability of screening techniques is necessary for plant breeders to consistently select for
the desired trait. Since drought resistance is a quantitative trait, indirect selection using
plant attributes or molecular markers will vary according to the environmental conditions.

Selection criterion must be highly associated to drought resistance across multiple

locations to substantiate its potential in breeding for drought resistance in common bean.

73

110

p6

UT.

iii

 

CONCLUSIONS

 

Terminal drought stress negatively affected bean productivity in the lowland
tropics. Transgressive segregation for yield was shown by 15 % of the RILS that out-
performed the three parents and seven checks. Drought resistant RILs yielded 700 kg/ha
under stress and 2500 kg/ha under non-stress conditions in Honduras. In Montcalm,
Michigan, drought resistant genotypes yielded up to 4000 kg/ha in minimal drought stress
conditions preventing confirmation of drought resistance. Broad adaptation and
performance of genotypes under stress across different environments could not be
evaluated. Drought resistance results are based on one location and will be confirmed.
However, the drought resistant genotypes show great potential for increasing yields of
common bean in Honduras and Michigan. Breeding programs can incorporate yield and
drought resistance traits from the black bean RILs into locally adapted material with
commercially acceptable seed types. Areas of the Latin American/Carribean region where
common bean is planted into dry season will benefit the most. Further testing in different
locations will identify the zone of adaptation for the drought resistant RILs.

Root length, root architecture and molecular markers were correlated to yield in
common bean. Under drought stress conditions, root length at the 0.5-1 .0 mm diameter
and fractal dimension correlated to yield. Fine roots correlating to yield under stress
suggesting that highly branched root systems are not favored in drought conditions in
Honduras. Low fractal dimension values suggest that root architecture designed for deep
soil exploration are present in drought resistant RILs. Taproot length had the highest
correlation to yield potential. In conjunction with molecular markers associated to yield

potential, the root measurements explained 9 % of the variation of yield under stress and

 

74

 

mar

COI

16 % of the variation of yield potential. Since the combination of root traits and molecular
markers explained greater variation than either trait alone, bean breeders should consider

combining different approaches to improve drought resistance in common bean.

75

 

 

 

 

Abebe. l
i

Acosta

Acost-

Acos

Acc

Af

REFERENCES

Abebe, A., M. A. Brick and R. A. Kirkby. 1998. Comparison of selection indices to
identify productive dry bean lines under diverse environmental conditions. Field
Crops Research 58: 15-23.

Acosta-Diaz, E. 1998. Early vigor as a selection criteria for adaptation to drought stress in
common bean, Annual Report of the Bean Improvement Cooperative: 153-154.

Acosta-Gallegos, J. A. 1988. Selection of common bean (Phaseolus vulgaris L.)
genotypes with enhanced drought tolerance and biological nitrogen fixation. Ph. D.
Michigan State University. East Lansing

Acosta-Gallegos, J. A. and J. K. Shibata. 1989. Effect of water-stress on growth and yield
of indeterminate dry bean (Phaseolus vulgaris L.) cultivars. Field Crops Research
20(2): 81-93.

Acosta-Gallegos, J. A. and M. W. Adams. 1991. Plant traits and yield stability of dry bean
(Phaseolus vulgaris L.) cultivars under drought stress. Journal of Agricultural
Science 117: 213-219.

Afanador, L., S. D. Haley and J. D. Kelly. 1993. Adoption of a "mini-prep" DNA
extraction method for RAPD marker analysis in common bean (Phaseolus vulgaris
L.). Annual Report of the Bean Improvement Cooperative 36: 10-11.

Aguirre-Medina, J. F., C. Trejo-Lopez, J. Kohasi-Shibata and J. A. Acosta Gallegos 1998.
Physiological response of bean Phaseolus vulgaris L. to water stress in a split root
system. Memorias del XVII Congreso de Fitogene'tica, Chapingo, México,
SOMEFI.

Box, J. E., A. J. M. Smucker and J. T. Ritchie. 1989. Minirhizotron installation techniques
for investigating root responses to drought and oxygen stresses. Soil Science
Society of America Journal 53: 115-118.

Box, J. E., Jr. 1996. Modern methods for root investigations. 193-237. Y. Waisel, A. Eshel
and U. Kafltafi Plant roots the hidden half. Marcel Dekker Inc. New York.

Brothers, M. E. and J. D. Kelly. 1993. Interrelationship of plant architecture and yield
components in the pinto bean ideotype. Crop Science 33(6): 1234-1238.

Champoux, M. C., G. Wang, S. Sarkarung, D. J. Mackill, J. C. O'Toole, N. Huang and S.
R. McCouch. 1995. Locating genes associated with root morphology and drought
avoidance in rice via linkage to molecular markers. Theoretical and Applied
Genetics 90: 969-981.

76

 

 

Chapman,
ma
110

Checkai.
re

Cotty, E

Ehlerir

Ehler

Farq

F151
Fit

Fi

Chapman, A. L. and R. C. Muchow. 1985. Nitrogen accumulated and partitioned at
maturity by grain legumes grown under different water regimes in a semi-arid
tropical environment. Field Crops Research 11(1): 69-79.

Checkai, R. T., W. A. Norvell, R. M. Welch and P. H. Brown. 1987. Using exchange
resins to impose nutrient treatments on hydroponically grown seedlings from
germination to transplanting. Journal of Plant Nutrition 10: 1447-1455.

Cotty, D., M. Garcia, 1. Estrada and E. Anchundia 2001. Proyecto de informacion agricola
y analisis de politicas: Indicadores basicos sobres el desempefio agropecuario
1971-2000. Zamorano,.1nstituto Nacional de Estadistica: 34, 38.

Ehleringer, J. R., J. W. White, D. A. Johnson and M. Brick. 1990. Carbon isotope
discrimination, photosynthetic gas-exchange, and transpiration efficiency in beans
and range grasses. Acta Oecologica-Intemational Journal of Ecology 11(4): 611-
625.

Ehleringer, J. R., S. Klassen, C. Clayton, D. Sherrill, M. Fuller-Holbrook, Q.-n. Fu and T.
A. Cooper. 1991. Carbon isotope discrimination and transpiration efficiency in
common bean. Crop Science 31: 1611-1615.

Farquhar, G. D., J. R. Ehleringer and K. T. Hubick. 1989. Carbon isotope discrimination
and photosynthesis. Annual Review of Plant Physiology and Plant Molecular
Biology 40: 503-537.

Fischer, R. A. and R. Maurer. 1978. Drought resistance in spring wheat cultivars. I grain
yield responses. Australian Journal of Agricultural Research 29: 897-912.

Fitter, A. 1996. Characteristics and functions of root systems. 1-20. Y. Waisel, A. Eshel
and U. Kaflrafi Plant roots the hidden half. Marcel Dekker, Inc. New York.

Fitter, A. H., R. Nichols and M. L. Harvey. 1988. Root system architecture in relation to
life history and nutrient supply. Functional Ecology 2: 345-351.

Foster, E. F., A. Pajarito and J. Acosta-Gallegos. 1995. Moisture stress impact on N
partitioning, N-remobilization and N-use efficiency in beans (Phaseolus vulgaris
L.). Journal of Agricultural Science 124: 27-37.

Freyre, R., P. W. Skroch, V. Geffroy, A. F. Adam-Blondon, A. Shirmohamadali, W. C.
Johnson, V. Llaca, R. O. Nodari, P. A. Pereira, S. M. Tsai, J. Tohme, M. Dron, J.
Nienhuis, C. E. Vallejos and P. Gepts. 1998. Towards an integrated linkage map of
common bean. 4. Development of a core linkage map and alignment of RFLP
maps. Theoretical and Applied Genetics 97(5-6): 847-856.

77

Gabelrnan, W. H., G. C. Gerloff, T. Schettini and R. Coltrnan. 1986. Genetic variability in
root systems associated with nutrient acquisition and use. HortScience 21(4): 971-
973.

Gollan, T., J. B. Passioura and R. Munns. 1986. Soil water status affects the stomatal
conductance of fully turgid wheat and sunflower leaves. Australian Journal of
Plant Physiology 13: 459-464.

Graham, P. H. and P. Ranalli. 1997. Common bean (Phaseolus vulgaris L.). Field CrOps
Research 53: 131-146.

Haley, S. D., L. Afanador and J. D. Kelly. 1994. Identification and application of a
random amplified polymorphic DNA marker for the I gene (potyvirus resistance)
in common bean. Phytopathology 84(2): 157-160.

Hall, A. E. 1993. Is dehydration tolerance relevant to genotypic differences in leaf
senescence and crop adaptation to dry environments? 1-10. T. J. a. E. A. B. Close
Plant response to cellular dehydration during environmental stress. The American
Society of Plant Physiologists.

Harris, G. A. and G. S. Campbell. 1989. Automated quantification of roots using a simple
image analyzer. Agronomy Journal 81(6): 935-938.

Jung, G., D. P. Coyne, P. Skroch, J. Nienhuis, E. Amaud-Santana, J. Bokosi, H. M.
Ariyarathne, J. R. Steadman, J. S. Beaver and S. M. Kaeppler. 1996. Molecular
markers associated with plant architecture and resistance to common blight, web
blight and rust in common beans. Journal of American Society of Horticultural
Science 121(5): 794-803.

Kao, W. Y., J. P. Comstock and J. R. Ehleringer. 1994. Variation in leaf movements
among common bean cultivars. Crop Science 34(5): 1273-1278.

Kebede, H., P. K. Subudhi, D. T. Rosenow and H. T. Nguyen. 2001. Quantitative trait loci
influencing drought tolerance in grain sorghum (Sorghum bicolor L. Moench).
Theoretical and Applied Genetics 103: 266-276.

Kelly, J. D., P. Gepts, P. N. Miklas and D. P. Coyne. 2002. Tagging and mapping of genes
and molecular marker-assisted selection for traits of economic importance in bean
and cowpea. Field Crops Research (in review).

Kelly, J. D., L. Afanador and S. D. Haley. 1995. Pyrarniding genes for resistance to bean
common mosaic-virus. Euphytica 82(3): 207-212.

Kelly, J. D., G. L. Hosfield, G. V. Vamer, M. A. Uebersax, S. D. Haley and J. Taylor.
1994. Registration of 'Raven' black bean. Crop Science 34(5): 1406-1407.

78

 

 

Kolkman, J. M. and J. D. Kelly. 1999. Drought stress and bean productivity during the
1998 growing season. Michigan Dry Bean Digest. 23: 2-3.

Lande, R. and R. Thompson. 1990. Efficiency of marker-assisted selection in the
improvement of quantitative traits. Genetics 124: 743-756.

Liao, H., G. Rubio, X. Yan, A. Cao, K. M. Brown and J. P. Lynch. 2001. Effect of
phosphorus availability on basal root shallowness in common bean. Plant and Soil
232(1-2): 69-79.

Lilley, J. M., M. M. Ludlow, S. R. McCouch and J. C. O’toole. 1996. Locating QTL for
osmotic adjustment and dehydration tolerance in rice. Journal of Experimental
Botany 47(302): 1427-1436.

Ludlow, M. M. 1989. Strategies of response to water stress. 269-281. K. H. Kreeb, H.
Richter and T. M. Hinckley Structural and functional response to environmental
stresses. SPB Academic Publishing. The Hague, The Netherlands.

Ludlow, M. M. and R. C. Muchow. 1990. A critical-evaluation of traits for improving
crop yields in water-limited environments. Advances in Agronomy 43: 107-153.

Lynch, J. and J. J. van Beem. 1993. Growth and architecture of seedling roots of common
bean genotypes. Crop Science 33: 1253-1257.

Mandelbrot, B. B. 1977. The fractal geomefl of nature. New York, W. H. Freeman and
Company.

Mayaki, W. C., I. D. Teare and L. R. Stone. 1976. Top and root growth of irrigated and
non-irrigated soybeans. Crop Science 16: 92-94.

Mayek-Pérez, N., C. Lopez-Castaneda, E. Lopez-Salinas and J. A. Acosta Gallegos.
2001a. Inheritance of genetic resistance to Macrophomina phaseolina (Tassi)
Goid. in common bean. Agrociencia 35: 637-648.

Mayek-Pérez, N., C. Lopez-Castaneda, E. Lopez-Salinas, J. Cumpian-Gutiérrez and J. A.
Acosta Gallegos. 2001b. Resistance to Macrophomina phaseolina (Tassi) Goid. in
common bean under field conditions in Mexico. Agrociencia 35: 649-661.

McMichael, B. L., J. J. Burke, J. D. Berlin, J. L. Hatfield and J. E. Quisenberry. 1985.
Root vascular bundle arrangements among cotton strains and cultivars.
Environmental and Experimental Botany 25(1): 23-30.

Mejia-Jimenez, A., C. Munoz, H. J. Jacobsen, W. M. Roca and S. P. Singh. 1994.
Interspecific hybridization between common and tepary beans: Increased hybrid

79

 

 

 

 

Miklas

Mikla

Mir

Mil

embryo growth, fertility, and efficiency of hybridization through recurrent and
congruity backcrossing. Theoretical and Applied Genetics 88: 324-331.

Miklas, P., V. Stone, C. Urrea, E. Johnson and J. Beaver. 1998. Inheritance and QTL
analysis of field resistance to ashy stem blight in common bean. Crop Science
38(4): 916-921.

Miklas, P. N., J. R. Stavely and J. D. Kelly. 1993. Identification and potential use of a
molecular marker for rust resistance in common bean. Theoretical and Applied
Genetics 85(6-7): 745-749.

Miklas, P. N., W. C. Johnson, R. Delorrne and P. Gepts. 2001. QTL conditioning
physiological resistance and avoidance to white mold in dry bean. Crop Science
41: 309-315.

Miklas, P. N., E. Johnson, V. Stone, J. S. Beaver, C. Montoya and M. Zapata. 1996.
Selective mapping of QTL conditioning disease resistance in common bean. Crop
Science 36: 1344-1351.

Newman, E. J. 1966. A method for estimating total length of root in a sample. Journal of
Applied Ecology 3: 139-145.

Nielsen, K. L., J. Lynch and H. N. Weiss. 1997. Fractal geometry of bean root systems:
Correlations between spatial and fractal dimension. American Journal of Botany
84(1): 26-33.

Nodari, R. O., S. M. Tsai, P. Guzman, R. L. Gilbertson and P. Gepts. 1993. Toward an
integrated linkage map of common bean .3. Mapping genetic-factors controlling
host-bacteria interactions. Genetics 134(1): 341-350.

Pandey, R. K., W. A. T. Herrera and J. W. Pendleton. 1984a. Drought response of grain
legumes under irrigation gradient: 1. Yield and yield components 76: 549-553.

Pandey, R. K., W. A. T. Herrera and J. W. Pendleton. 1984b. Drought response of grain
legumes under irrigation gradient: H. Plant water status and canopy temperature.
Agronomy Journal 76: 553-557.

Pandey, R. K., W. A. T. Herrera, A. N. Villegas and J. W. Pendleton. 1984c. Drought
response of grain legumes under irrigation gradient: 111. Plant growth. Agronomy
Journal 76: 557560.

Paterson, A. H., J. W. DeVerna, B. Lanini and S. D. Tanksley. 1990. Fine mapping of

quantitative trait loci using selected overlapping recombinant chromosomes, in an
interspecies cross of tomato. Genetics 124: 735-742.

80

Ramirez-Vallejo, P. 1992. Identification and estimation of heritabilities of drought related
resistance traits in dry bean (Phaseolus vulgaris L.). Ph. D. Michigan State
University. East Lansing

Ramirez-Vallejo, P. and J. D. Kelly. 1998. Traits related to drought resistance in common
bean. Euphytica 99: 127-136.

Régent Instruments Inc. 2000. WinRhizo. Quebec, Qc., Canada, Régent Instruments Inc.

Ribaut, J .-M., C. Jiang, D. Gonzalez-de-Leon, G. O. Edmeades and D. A. Hoisington.
1997. Identification of quantitative trait loci under drought conditions in tropical
maize. 2. Yield components and marker-assisted selection strategies. Theoretical
and Applied Genetics 94: 887-896.

Rosales-Sema, R., P. Ramirez-Vallejo, J. A. Acosta-Gallegos, F. Castillo-Gonzalez and J.
D. Kelly. 2000. Grain yield and drought tolerance of common bean under field
conditions. Agrociencia 34(2): 153-165.

Rosas, J. C., J. D. Erazo and J. R. Moncada. 1991. Tolerancia a la sequia en germoplasma
de fiijol comr'rn y fiijol tepari. CEIBA 32(2): 91-105.

Rosielle, A. A. and J. Hamblin. 1981. Theoretical aspects of selection for yield in stress
and non-stress environments. Crop Science 21: 943-945.

Samper, C. 1984. Effects of water stress imposed at mid-pod filling upon yield and dry
matter partitioning in dry beans (Phaseolus vulgaris L.). M.S. Michigan State
University. East Lansing

Samper, C. and W. M. Adams. 1985. Geometric mean of stress and control yield as a
selection criterion for drought tolerance. Annual Report of the Bean Improvement
Cooperative 28: 53-54.

SAS Institute Inc. 2000. SAS. Cary, NC, SAS Institute Inc.

Sax, K. 1923. The association of size differences with seed-coat pattern and pigmentation
in Phaseolus vulgaris. Genetics 8: 552-560.

Schneider, K. A., M. E. Brothers and J. D. Kelly. 1997a. Marker-assisted selection to
improve drought resistance in common bean. Crop Science 37(1): 51-60.

Schneider, K. A., K. F. Grafion and J. D. Kelly. 2001. QTL analysis of resistance to
fusarium root rot in bean. Crop Science 41(2): 535-542.

81

Schneider, K. A., R. Rosales-Sema, F. Ibarra-Perez, B. Cazares-Enriquez, J. A. Acosta-
Gallegos, P. Ramirez-Vallejo, N. Wassimi and J. D. Kelly. 1997b. Improving
common bean performance under drought stress. Crop Science 37: 43-50.

Schulze, E. D. 1986. Carbon dioxide and water vapor exchange in response to drought in
the atmosphere and in the soil. Annual Review of Plant Physiology 37: 247-74.

Singh, S. P. 1995. Selection for water-stress tolerance in interracial populations of
common bean. Crop Science 35(1): 118-124.

Singh, S. P. and C. G. Munoz. 1999. Resistance to common bacterial blight among
Phaseolus species and common bean improvement. Crop Science 39: 80-89.

Singh, S. P., M. A. Pastor-Corrales, M. A., C. Urrea and C. C. 1991. Independent,
alternate, and simultaneous selection for resistance to anthracnose and angular leaf
spot and effects on seed yield in common bean (Phaseolus vulgaris L.). Plant
Breeding 106: 312-318.

Snapp, S., R. Koide and J. Lynch. 1995. Exploitation of localized phosphorus-patches by
common bean roots. Plant and Soil 177: 211-218.

Specht, J. E., K. Chase, M. Macrander, G. L. Graef, J. Chung, J. P. Markwell, M.
Germann, J. G. Orf and K. G. Lark. 2001. Soybean response to water: A QTL
analysis of drought tolerance. Crop Science 41: 493-509.

Sponchiado, B. N., J. W. White, J. A. Castillo and P. G. Jones. 1989. Root growth of four
common bean cultivars in relation to drought tolerance in environments with
contrasting soil types. Experimental Agriculture 25: 249-257.

Staub, J. E., F. C. Serquen and M. Gupta. 1996. Genetic markers, map construction, and
their application in plant breeding. HortScience 31(5): 729-741.

Stoffella, P. J ., R. F. Sandsted, R. W. lobe] and W. L. Hymes. 1979. Root characteristics
of black beans. 11. Morphological differences among genotypes. Crop Science 19:
826-830.

Tatsumi, J ., A. Yarnauchi and Y. Kono. 1989. Fractal analysis of plant root systems.
Annals of Botany 64: 499-503.

Taylor, H. M., M. G. Huck, B. Klepper and L. Z. F. 1970. Measurement of soil-grown
roots in a rhizotron. Agronomy Journal 62: 807-809.

Tennant, D. 1975. A test of a modified line intersect method of estimating root length.
Journal of Ecology 63: 995-1001.

82

Teran, H. and S. P. Singh. 2002. Comparison of sources and lines selected for drought
resistance in common bean. Crop Science 42: 64-70.

Teulat, B., D. This, M. Khairallah, C. Borries, C. Ragot, P. Sourdille, P. Leroy, P.
Monneveux and A. Charrier. 1998. Several QTLs involved in osmotic-adjustrnent
trait variation in barley (Hordeum vulgare L.). Theoretical and Applied Genetics
96: 688-698.

Thomas, C. V., R. M. Manshardt and J. G. Waines. 1983. Teparies as a source of useful
traits for improving common beans. Desert Plants 5(1): 43-48.

Urrea, C. A., P. N. Miklas, J. S. Beaver and R. H. Riley. 1996. A codominant randomly
amplified polymorphic DNA (RAPD) marker useful for indirect selection of bean
golden mosaic virus resistance in common bean. Journal of the American Society
for Horticultural Science 121(6): 1035-1039.

White, J. W. and J. A. Castillo. 1989. Relative effect of root and shoot genotypes on yield
of common bean under drought stress. Crop Science 29: 360-362.

White, J. W. and J. A. Castillo. 1992. Evaluation of diverse shoot genotypes on selected
root genotypes of common bean under soil water deficits. Crop Science 32: 762-
765.

White, J. W., J. A. Castillo and J. R. Ehleringer. 1990. Associations between productivity,
root growth, and carbon isotope discrimination in Phaseolus vulgaris under water
deficit. Australian Journal of Plant Physiology 17: 189-198.

White, J. W., R. Ochoa M., F. Ibarra P. and S. P. Singh. 1994a. Inheritance of seed yield,
maturity and seed weight of common bean (Phaseolus vulgaris) under semi-arid
rainfed conditions. Journal of Agricultural Science 122: 265-273.

White, J. W., J. A. Castillo, J. R. Ehleringer, J. A. Garcia and S. P. Singh. 1994b.
Relations of carbon-isotope discrimination and other physiological traits to yield in
common bean (Phaseolus vulgaris L.) under rain-fed conditions. Journal of
Agricultural Science 122: 275-284.

Yabba, M. D. 2001. Common bean (Phaseolus vulgaris L.) yield, root growth, and N
fixation response to moisture deficits. Ph.D. Michigan State University. East
Lansing

Yabba, M. D. and E. F. Foster. 1997. Bean seedling root growth as an indicator of field
performance under moisture stress. Annual Report of the Bean Improvement
Cooperative 40: 63-64.

83

 

Yan, X., J. P. Lynch and S. E. Beebe. 1995. Genetic variation for phosphorus efficiency of
common bean in contrasting soil types: 11. Yield response. Crop Science 35: 1094-
1099.

Young, R. A. and J. D. Kelly. 1996. RAPD markers flanking the Are gene for anthracnose
resistance in common bean. Journal of the American Society of Horticultural
Science 121(1): 37-41.

Zheng, H.-g., R. C. Babu, M. S. Pathan, L. Ali, N. Huang, B. Courtois and H. T. Nguyen.

2000. Quantitative trait loci for root-penetration ability and root thickness in rice:
Comparison of genetic backgrounds. Genome 43: 53-61.

84

 

APPENDIX A

DATA TABLES FROM DROUGHT EXPERIMENTS IN MICHIGAN AND
HONDURAS

85

Table A1. Yield, rank and 100 seed weight for the 150 RILs in Saginaw, MI 2000.

 

 

 

 

Line entry Yield Rank 100 sw Line entry Yield Rank 100 sw
kg/ha kg/ha
L88-1 1 3847 21 26.5 L91-1 82 3106 32 24.5
L88-2 2 3048 62 24.3 L91-2 83 3272 24 26.3
L88-3 3 3295 52 23.5 L91-3 84 2006 67 23.4
L88-4 4 3923 16 22.8 L91-4 85 2867 46 28.3
L88-5 5 3171 59 22.4 L91-5 86 3541 10 26.4
L88-6 6 2981 66 23.0 L91-6 87 3186 27 24.5
L88-7 7 3662 35 25.6 L91-7 88 3285 23 28.0
L88-8 8 3529 41 22.2 L91-8 89 3444 14 25.2
L88-9 9 3249 56 21.5 L91-9 90 2926 45 24.5
L88-10 10 4583 3 21.7 L91-10 91 3303 21 24.1
L88-11 11 2491 80 20.0 L91-11 92 2046 66 20.7
L88-12 12 3281 54 23.6 L91-12 93 3142 29 26.7
L88-13 13 3418 49 24.0 L91-13 94 2930 44 25.6
L88-14 14 4540 4 24.8 L91-14 95 2208 65 25.4
L88-1 5 1 5 3588 39 23.4 L91-15 96 3337 19 26.8
L88-16 16 3840 22 23.3 L91-16 97 3594 9 29.6
L88-17 17 3483 44 23.7 L91-17 98 3699 7 24.8
L88-18 18 2644 78 22.5 L91-18 99 3139 30 24.7
L88-19 19 3644 37 22.6 L91-19 100 3865 5 26.2
L88-20 20 3951 15 21.5 L91-20 101 2813 49 26.4
L88-21 21 4926 1 25.4 L91-21 102 2393 61 23.2
L88-22 22 3459 45 23.3 L91-22 103 3074 33 23.7
L88-23 23 4071 13 24.1 L91-23 104 3163 28 22.5
L88-24 24 3649 36 23.4 L91-24 105 2422 60 24.3
L88-25 25 3012 63 20.7 L91-25 106 2538 58 28.8
L88-26 26 2863 71 21.3 L91-26 107 4042 3 26.6
L88-27 27 2257 81 19.0 L91-27 108 2628 56 22.5
L88-28 28 3663 34 21.9 L91-28 109 2219 64 21.3
L88-29 29 4159 9 25.4 L91-29 1 10 2951 43 22.5
L88-30 30 3485 43 22.1 L91-30 1 1 1 3488 13 28.8
L88-31 31 2694 76 21.5 L91-31 112 3505 11 23.7
L88-32 32 2671 77 23.8 L91 -32 1 1 3 3022 39 28.7
L88-33 33 2928 70 24.0 L91-33 114 3977 4 27.1
L88-34 34 3139 60 25.2 L91-34 1 1 5 2757 52 21 .3
L88-35 35 4621 2 22.6 L91-35 116 2670 54 25.8
L88-36 36 3788 25 24.1 L91-36 117 2991 42 26.8
L88-37 37 41 15 12 25.1 L91-37 1 18 2866 47 26.2
L88-38 38 3887 20 20.6 L91-38 119 2630 55 23.4
L88-39 39 3327 51 20.3 L91-39 120 3839 6 25.5
L88-40 40 2566 79 19.7 L91-40 121 3233 25 25.4
L88-41 41 3721 30 24.9 L91-41 122 3405 17 24.2
L88-42 42 3458 46 23.3 L91-42 123 1922 68 21.2
L88-43 43 4140 10 24.3 L91-43 124 3133 31 25.8
L88-44 44 3002 64 1 8.9 L91 -44 125 2242 63 23.8
L88-45 45 2713 74 22.0 L91-45 126 3050 35 22.1

 

 

86

 

 

Table A1. Continued.

 

 

 

 

 

Line entry Yield fink 100 sw Line entry Yield Rank 100 sw
kg/ha g kg/ha g
L88-46 46 4161 8 24.8 L91-46 127 3012 40 25.1
L88-47 47 2735 73 23.8 L91-47 128 3001 41 24 .4
L88-48 48 4367 6 24 .0 L91 -48 1 29 3291 22 24 .4
L88-49 49 3785 26 21.5 L91-49 130 4109 2 26.4
L88-50 50 3546 40 22.0 L91-50 131 3492 12 23.9
L88-51 51 2949 68 24.3 L91-51 132 3643 8 25.6
L88-52 52 3223 58 23.6 L91 -52 1 33 3429 16 22.9
L88-53 53 3907 1 7 26.0 L91-53 134 2564 57 22 .0
L88-54 54‘ 3779 27 25.8 L91-54 135 2261 62 19.3
L88-55 55 3450 47 22 .4 L91 -55 136 2757 51 24.1
L88-56 56 3517 42 20 .4 L91 -56 137 2814 48 25.4
L88-57 57 3241 57 22 .9 L91 -57 1 38 3022 38 28 .9
L88-58 58 2856 72 21 .8 L91-58 139 3217 26 25.0
L88-59 59 3968 14 23 .1 L91 ~59 140 3054 34 25.7
L88-60 60 3896 1 9 22 .9 L91 -60 141 3322 20 24.7
L88-61 61 2995 65 21 .8 L91-61 142 3026 37 25.0
L88-62 62 3441 48 23 .0 L91 -62 143 3040 36 21 .3
L88-63 63 3813 23 21 .8 L91 -63 144 4323 1 28.9
L88-64 64 3066 61 22.3 L91 -64 145 2459 59 24.4
L88-65 65 3293 53 25.8 L91-65 146 2779 50 21 .6
L88-66 66 3793 24 21 .7 L91 -66 147 2728 53 21 .9
L88-67 67 2963 67 19.6 L91-67 148 3440 15 28.9
L88-68 68 4277 7 23.2 L91-68 149 1868 69 22.4
L88-69 69 3697 33 24.8 L91 -69 1 50 3396 1 8 26.0
L88-70 70 2949 69 21 .7
L88-71 71 3603 38 23.4
L88-72 72 3738 28 20.7
L88-73 73 2706 75 21 .3
L88-74 74 3250 55 21 .8
L88-75 75 4453 5 25.2
L88-76 76 3389 50 25.3
L88-77 77 3720 31 23.3
L88-78 78 3724 29 25.5
L88-79 79 3902 1 8 23.8
L88-80 80 371 3 32 22 .3
L88-81 81 41 gs 1 1 fi26.8

 

 

 

87

 

Table A2. F reld data for 160 genotypes from drought treatment in_ Honduras 2001.

 

 

 

Line Flori tht Lodg Matr DS PM Mac Mac Stnd Yield Pct 100 Bio HI
45d 75d moist sw mass
d cm d d g/mz m2

L88-1 42 43 2 84 3 73 6 13 28 39.3 - 17.6 510.3 0.08
L88-2 42 34 2 82 2 71 5 21 30 21.2 - 16.9 198.5 0.12
L88-3 43 45 2 82 4 73 2 9 30 128.5 12.3 19.5 633.1 0.17
L88-4 42 37 2 82 2 72 2 19 25 23.8 - 19.6 302.4 0.07
L88-5 41 40 3 81 4 70 2 11 27 79.8 - 15.8 396.9 0.20
L88-6 41 41 1 82 5 71 6 12 28 79.1 10.2 18.5 396.9 0.23
L88-7 42 37 3 82 3 74 4 12 30 60.2 - 19.3 472.5 0.14
L88-8 42 38 2 82 3 73 1 19 26 46.2 - 16.5 321.3 0.13
L88-9 45 35 2 82 2 75 7 27 24 31.1 - 17.4 179.6 0.15
L88-10 42 34 2 82 4 72 10 14 29 71.2 - 17.4 302.4 0.24
L88-11 40 38 1 80 4 70 7 16 26 63.7 11.5 17.8 264.6 0.22
L88-12 43 45 2 84 2 76 3 11 30 22.3 - 16.8 368.5 0.07
L88-13 42 40 2 81 3 72 0 12 29 121.6 11.3 19.0 472.5 0.25
L88-14 42 43 2 82 3 74 4 9 31 43.9 - 18.4 406.3 0.14
L88-15 42 38 2 82 3 71 4 15 26 53.0 - 18.7 378.0 0.15
L88-16 42 36 3 82 5 71 7 8 30 89.4 - 18.3 387.4 0.22
L88-17 41 39 2 81 4 70 1 13 29 99.4 - 19.1 302.4 0.33
L88-18 43 42 1 83 2 75 4 19 25 7.4 - - 368.5 0.04
L88-19 41 43 2 82 4 74 5 11 27 117.3 - 19.2 378.0 0.31
L88-20 41 39 2 80 3 70 1 16 27 67.2 - 14.8 387.4 0.21
L88-21 44 37 2 82 3 73 1 8 28 37.5 - 18.9 585.9 0.07
L88-22 41 46 2 81 3 71 1 15 27 52.9 - 16.0 321.3 0.17
L88-23 42 32 1 81 3 72 3 11 28 36.1 - 15.2 207.9 0.17
L88-24 42 43 2 82 2 74 1 19 27 48.9 - 19.0 368.5 0.13
L88-25 42 44 2 81 4 72 6 14 30 85.2 11.5 18.6 434.7 0.16
L88-26 43 36 2 83 2 75 1 13 29 55.5 12.1 15.4 321.3 0.14
L88-27 42 41 2 82 4 73 1 12 30 80.4 - 15.7 340.2 0.27
L88-28 42 41 2 83 3 75 5 15 27 46.4 - 17.6 264.6 0.18
L88-29 42 35 2 83 3 73 6 18 26 48.9 - 17.6 226.8 0.19
L88-30 42 42 1 83 5 72 0 11 26 156.2 11.6 18.3 378.0 0.36
L88-31 41 43 2 82 5 72 2 10 29 135.4 11.6 18.9 406.3 0.33
L88-32 42 41 1 82 3 72 5 16 30 60.2 - 20.1 340.2 0.17
L88-33 42 40 2 82 3 71 11 16 22 58.1 - 17.2 311.8 0.18
L88-34 41 37 2 83 2 75 3 14 22 44.4 11.7 19.2 359.1 0.10
L88-35 41 38 1 81 3 72 1 10 30 61.6 - 18.1 283.5 0.22
L88-36 42 35 2 82 4 73 4 12 30 43.6 - 18.2 274.0 0.15
L88-37 43 36 2 82 2 72 4 16 24 34.4 - 21.0 207.9 0.14
L88-38 42 40 2 81 3 72 7 20 29 54.3 - 17.9 236.3 0.23
L88-39 42 37 2 82 3 72 3 13 29 47.3 - 15.8 255.2 0.17
L88-40 42 36 1 83 3 71 6 15 24 53.3 - 17.2 311.8 0.14
L88-41 43 44 2 84 2 74 3 12 28 46.0 - 19.9 368.5 0.12
L88-42 42 44 2 82 3 75 1 10 29 68.4 - 17.8 500.8 0.13
L88-43 43 31 2 82 2 73 5 11 26 33.5 - 19.6 302.4 0.08
L88-44 42 43 1 83 3 72 1 10 30 74.2 11.6 17.5 567.0 0.15
L88-45 41 38 2 82 3 71 6 17 30 39.5 - 16.3 283.5 0.15

 

 

00
00

 

Table A2. Continued.

 

 

 

Line Flor tht Lodg Mair 08 PM Mac Mac Stnd Yield Pct 100 Bio HI
45d 75d moist sw mass
d cm d d glm’ g glm2

L88-46 41 36 1 82 3 72 3 22 27 33.2 - 17.2 217.4 0.14
L88-47 42 39 2 82 2 73 3 9 30 56.4 - 17.9 472.5 0.11
L88-48 42 39 2 82 3 72 3 13 28 32.4 - 17.8 292.9 0.12
L88-49 42 41 2 83 3 74 6 12 30 76.7 - 18.5 491.4 0.16
L88-50 40 42 2 80 4 70 1 12 30 104.0 - 17.4 368.5 0.28
L88-51 41 37 2 80 5 70 2 14 30 92.2 - 19.0 321.3 0.28
L88-52 43 39 2 83 2 74 2 18 29 32.2 - 18.2 349.6 0.09
L88-53 41 36 2 82 3 70 2 8 30 79.2 11.8 19.9 340.2 0.23
L88-54 44 43 2 83 3 73 3 14 30 56.5 - 19.1 415.8 0.16
L88-55 41 35 2 82 3 71 4 13 30 42.8 - 17.7 283.5 0.14
L88-56 40 36 2 80 4 71 1 21 26 76.9 - 18.6 236.3 0.32
L88-57 42 34 1 82 3 72 5 14 30 35.4 - 16.3 255.2 0.14
L88-58 40 36 2 82 5 73 3 14 30 107.4 - 18.2 311.8 0.33
L88-59 42 46 2 83 4 74 1 11 30 100.6 11.6 19.9 444.1 0.23
L88-60 41 36 2 81 3 72 0 11 28 55.0 11.3 17.2 396.9 0.12
L88-61 41 38 2 81 6 72 4 12 30 111.6 - 17.1 330.7 0.33
L88-62 41 40 2 82 2 71 2 17 30 44.2 - 17.2 311.8 0.15
L88-63 41 46 2 82 5 72 3 12 30 181.8 19.9 16.7 500.8 0.34
L88-64 40 30 1 83 2 74 6 11 27 18.4 - - 255.1 0.07
L88-65 42 40 2 82 3 73 5 16 26 46.7 - 19.7 255.1 0.20
L88-66 39 38 2 81 5 71 2 10 30 122.8 - 18.3 330.7 0.38
L88-67. 42 39 2 82 2 73 2 15 26 41.9 - 14.5 207.9 0.21
L88-68 41 39 2 82 3 71 8 15 30 90.9 - 20.9 302.4 0.29
L88-69 42 39 2 83 4 73 2 8 27 138.5 12.7 17.5 396.9 0.28
L88-70 40 40 2 82 4 72 3 10 30 73.5 - 18.0 453.6 0.17
L88-71 41 41 2 83 3 74 2 16 30 53.3 - 15.9 255.2 0.20
L88-72 42 41 2 81 4 71 5 14 30 70.9 - 17.9 425.2 0.16
L88-73 42 33 2 80 3 72 0 11 25 44.9 11.2 18.3 311.8 0.15
L88-74 41 45 2 82 4 69 0 10 30 160.3 12.7 16.9 548.1 0.28
L88-75 42 45 2 82 4 73 1 13 30 84.8 - 19.7 472.5 0.19
L88-76 42 40 3 82 4 73 1 4 30 81.3 - 20.0 415.8 0.20
L88-77 42 34 2 84 3 74 14 17 24 50.0 - 16.6 217.4 0.20
L88-78 41 41 2 83 3 72 0 12 30 64.7 - 19.0 463.0 0.13
L88-79 43 40 2 82 4 75 6 13 30 62.9 - 19.0 349.6 0.17
L88-80 41 35 2 82 3 69 6 18 25 63.6 - 20.0 217.4 0.25
L88-81 42 40 2 81 3 70 1 10 29 56.9 - 16.2 349.6 0.17
L91-1 41 40 2 80 3 69 3 10 29 36.0 - 18.0 207.9 0.17
L91-2 44 43 2 82 3 74 2 7 30 21.1 - 19.5 368.5 0.06
L91-3 41 40 2 81 3 71 1 12 28 91.1 11.2 20.3 311.8 0.23
L91-4 41 42 3 82 2 72 4 18 29 38.7 - 22.7 330.7 0.09
L91-5 41 36 1 82 1 74 9 26 19 25.2 - 20.3 141.8 0.15
L91-6 42 46 2 82 3 73 2 7 30 43.4 - 19.1 321.3 0.15
\91-7 40 43 3 82 3 72 0 6 30 81.5 - 22.9 453.6 0.18
L91-8 43 43 2 83 2 73 4 10 27 17.3 - 19.9 283.5 0.07
L91-9 42 41 2 83 2 74 2 13 30 39.8 - 24.0 255.1 0.14

 

 

00
\O

Table A2. Continued.

 

 

 

Line Flor tht Lodg Matr o"s""'PM Mac Mac Stnd Yield Sci 100 Ea HI
45d 75d moist sw mass
d cm d d g/m2 9 g/m’

L91-10 42 48 2 82 3 73 2 1o 30 98.1 12.8 21.3 396.9 0.23
L91-11 43 43 2 82 2 72 1 17 28 36.2 - 17.2 217.4 0.15
L91-12 42 37 2 83 2 73 2 9 24 22.7 - 17.9 160.7 0.13
L91-13 43 39 2 83 2 74 1 7 33 13.9 - - 368.5 0.04
L91-14 48 43 2 83 3 72 o 7 28 21.9 - 20.2 415.8 0.06
L91-15 45 44 3 82 3 73 4 9 29 82.5 - 20.9 538.6 0.12
L91-16 44 42 2 83 3 75 1 7 30 57.3 - 21.4 425.2 0.14
L91-17 42 44 2 83 4 73 1 8 29 32.9 - 19.0 396.9 0.08
L91-18 41 43 1 79 3 70 4 8 30 80.8 - 17.9 283.5 0.21
L91-19 47 42 2 84 2 75 3 8 30 5.8 - - 408.3 0.01
L91-20 42 47 2 83 3 75 1 8 33 50.4 - 23.0 510.3 0.10
L91-21 42 43 2 82 3 73 2 8 27 50.0 11.6 19.1 349.8 0.11
L91-22 41 37 1 82 2 73 1 14 27 13.4 - 18.1 311.8 0.04
L91-23 45 51 2 85 2 74 1 7 30 41.9 - 21.4 510.3 0.08
L91-24 43 39 2 83 3 72 3 12 27 35.4 - 20.8 408.3 0.09
L91-25 41 48 2 82 4 70 o 5 30 113.0 10.8 19.7 529.2 0.23
L91-26 41 41 3 82 3 71 3 8 28 78.0 - 21.0 415.8 0.17
L91-27 41 39 2 81 3 71 0 12 22 38.8 - 18.8 189.0 0.19
L91-28 44 48 2 83 3 75 1 2 30 35.9 11.3 20.3 340.2 0.20
L91-29 4o 48 3 82 4 71 0 7 27 79.1 - 19.6 349.6 0.23
L91-30 42 43 2 82 4 73 1 7 30 130.7 - 24.7 387.4 0.33
L91-31 43 42 2 82 3 73 3 15 28 42.1 - 18.9 302.4 0.12
L91-32 42 39 2 82 2 72 1 21 29 14.2 - - 245.7 0.06
L91-33 41 50 2 81 3 71 2 8 30 82.3 10.5 20.0 453.8 0.18
L91-34 42 44 1 81 3 72 3 14 30 40.3 - 17.8 284.8 0.18
1.9135 41 37 2 82 2 72 1 8 30 48.5 10.5 17.2 292.9 0.14
L91-36 43 34 2 83 3 73 5 9 28 27.5 - 20.8 245.7 0.10
L91-37 44 38 2 83 2 75 1 7 27 10.9 - - 388.5 0.05
L91-38 42 38 2 82 4 73 4 18 30 57.8 - 18.6 284.8 0.22
L91-39 43 41 2 82 2 74 0 9 28 27.9 - 19.8 264.6 0.09
L91-40 42 42 2 83 3 72 3 12 30 80.0 12.1 19.2 359.1 0.15
1.9141 42 37 2 81 3 71 5 20 29 12.8 - 17.7 170.1 0.08
1.9142 42 44 2 82 3 72 1 4 30 37.9 - 17.5 491.4 0.08
L91-43 42 4o 2 83 3 72 1 1o 30 58.3 - 19.7 406.3 0.14
1.9144 48 48 2 83 3 75 1 8 30 34.9 - 17.3 434.7 0.09
L91-45 41 47 2 81 3 71 o 2 30 80.9 10.8 17.3 453.6 0.14
L91-46 44 45 2 81 3 72 1 8 30 88.9 - 24.5 538.8 0.14
L91-47 4o 38 1 80 2 70 2 7 30 43.4 - 16.3 274.0 0.18
L91-48 43 48 2 82 3 71 0 11 28 88.1 12.9 21.2 500.8 0.13
L91-49 45 38 2 83 3 73 2 14 27 12.5 - - 217.4 0.06
L91-50 40 39 2 83 3 72 2 15 29 44.1 - 18.5 228.8 0.20
L91-51 41 48 3 81 3 72 1 7 30 58.8 - 24.4 302.4 0.19
L91-52 47 45 2 83 2 75 1 1o 27 8.6 - 18.2 238.3 0.04
L91-53 43 41 2 82 3 74 1 4 30 14.5 - - 387.4 0.03
L91-54 4o 33 2 82 4 72 1 9 30 19.3 - 27.8 217.4 0.09
1.355 43 45 2 83 4; 73 1 10 28 52.9 10.2 23.5 349.8 0.1;

 

 

90

 

Table A2. Continued.

 

Line

L91-56
L91-57
L91-58
L91-59
L91-60
L91-61
L91-62
L91-63
L91-64
L91-65
L91-66
L91-67
L91-68
L91-69
Tacana
V8025
Tio Canela
898311
VAX 5
TLP 19
BAT 477
Rio Tibagi
EAP
9510-77
LEE 5

 

 

 

 

 

 

Flor tht Lodg Matr o' s'"P"M Mac Mac Stnd Yield ict 100 818 Hi
450 75d moist sw mass

d cm d dL g/mz m2

42 48 2 82 3 71 1 19 30 68.0 - 20.8 396.9 0.17
42 39 2 84 2 75 0 14 25 18.0 - 20.5 349.6 0.06
42 41 3 80 3 71 2 11 30 30.8 - 20.6 340.2 0.10
41 49 2 82 3 72 2 11 30 107.0 12.2 19.7 396.9 0.21
42 43 2 83 3 72 2 14 29 35.6 - 19.6 283.5 0.13
42 38 2 82 2 74 0 14 30 37.9 - 20.2 321.3 0.13
42 45 2 82 3 74 2 9 24 69.7 - 21.8 425.2 0.12
42 37 1 83 2 72 7 22 30 29.3 - 20.4 283.5 0.09
42 44 2 81 3 71 5 8 29 49.2 - 17.6 378.0 0.12
45 39 2 83 3 75 2 5 30 23.2 - 22.8 349.6 0.07
43 43 2 82 3 74 0 7 30 31.9 - 19.0 330.7 0.10
43 41 2 83 3 73 2 10 28 29.6 - 18.7 396.9 0.07
44 37 2 82 2 74 0 5 30 8.7 - - 359.1 0.02
47 41 2 84 2 75 2 11 30 4.5 - - 576.4 0.01
42 36 2 84 2 73 8 15 25 34.1 - 15.4 378.0 0.10
42 3O 3 81 3 72 2 6 30 49.1 - 16.8 292.9 0.18
40 33 2 80 3 71 1 7 30 49.9 - 17.6 387.4 0.14
41 47 2 80 3 69 3 21 26 71.4 11.8 17.8 368.5 0.19
41 39 2 81 3 72 2 10 27 48.0 - 21.8 349.6 0.15
45 35 2 84 3 76 1 7 30 40.6 - 18.5 396.9 0.10
41 33 3 80 2 71 0 12 26 76.7 - 19.7 557.5 0.10
41 43 2 82 3 72 2 11 26 69.8 - 16.1 699.3 0.11
41 29 2 82 3 72 4 15 26 41.2 - 19.4 359.1 0.11
35 37 2 72 4 65 1 7 27 104.8 - 20.8 264.6 0.40

1' Flor - days to flowering, tht - Height, Lodg - Lodg‘ing (1-5), Matr - days to maturity,
DS - Desirability Score (1-9), PM - Physiological maturity, Mac 45d - Macrophorrrina
incidence at 45 days, Mac 75d - Macrophomina incidence at 75 days, Stnd - Stand, Pct

moist - percent moisture, 100 sw - 100 seed weight, HI - Harvest Index

91

Table A3. Field data for 160 genotype es fiom non- -stre_ss treatment in Honduras 2001.

 

 

 

Line Flori tht Lodg Matr 08 PM Mac Mac Stnd Yield Pct 100 Bio HI
45d 75d moist sw mass
d cm d d g/m2 lm2

L88-1 43 45 3 84 3 75 1 1 30 304.1 7.5 20.1 604.8 0.50
L88-2 41 45 3 82 3 73 2 2 29 360.5 8.2 19.4 869.4 0.44
L88-3 42 45 4 82 3 75 1 0 30 423.0 7.8 18.7 850.5 0.49
L88-4 42 44 4 81 3 72 1 2 30 379.9 8.5 18.2 756.0 0.50
L88-5 41 44 4 83 3 72 2 2 30 453.4 8.0 18.7 822.1 0.55
L88-6 42 46 4 86 4 78 4 4 25 487.4 8.7 20.3 907.2 0.53
L88-7 42 41 3 82 4 76 0 3 30 342.4 7.6 20.0 699.3 0.49
L88-8 42 44 4 83 4 74 1 2 30 461.3 7.2 20.2 859.9 0.53
L88-9 44 44 3 85 4 77 0 2 30 382.4 8.1 18.7 812.7 0.47
L88-10 42 41 3 80 4 73 0 1 30 385.2 7.2 17.8 756.0 0.51
L88-11 41 42 3 79 5 70 1 1 30 437.2 8.3 16.8 746.5 0.59
L88-12 44 48 4 85 3 79 1 1 30 528.7 7.7 19.8 973.3 0.54
L88-13 42 44 4 82 5 75 0 1 30 615.2 7.4 19.6 1143.4 0.54
L88-14 43 51 3 85 3 78 0 2 30 373.2 10.1 20.3 812.7 0.45
L88-15 41 42 5 79 2 71 0 1 30 395.5 8.0 18.4 680.4 0.62
L88-16 41 44 4 81 4 71 3 2 30 351.8 8.8 20.0 670.9 0.52
L88-17 43 44 3 83 4 75 0 1 30 414.3 8.3 17.1 841.0 0.49
L88-18 44 51 3 85 2 78 0 3 30 316.8 7.1 21.6 718.2 0.43
L88-19 43 46 3 84 4 77 0 0 30 461.0 8.1 20.4 859.9 0.53
L88-20 42 45 3 83 4 72 1 2 30 431.0 8.2 17.4 812.7 0.53
L88-21 44 46 4 85 3 78 0 1 30 348.0 8.4 21.7 954.4 0.37
L88-22 42 48 3 81 4 72 0 0 30 413.8 8.0 17.1 859.9 0.48
L88-23 41 48 3 79 4 74 1 1 30 369.9 7.5 18.1 718.2 0.51
L88-24 41 48 3 83 3 77 0 2 30 447.2 7.2 19.4 850.5 0.53
L88-25 42 45 3 81 4 72 2 2 30 428.7 7.9 17.8 793.8 0.55
L88-26 42 46 3 82 5 76 0 0 30 454.5 8.3 17.6 888.3 0.51
L88-27 41 40 3 82 5 74 0 1 30 411.4 8.3 16.5 784.3 0.52
L88-28 41 50 4 83 3 76 1 0 30 496.0 7.8 16.9 963.9 0.51
L88-29 42 48 3 81 5 73 1 1 30 446.7 8.0 17.4 850.5 0.52
L88-30 42 47 3 83 6 75 0 1 30 476.7 7.5 18.4 850.5 0.56
L88-31 42 40 3 81 5 75 0 0 30 364.7 7.6 18.8 623.7 0.58
L88-32 42 41 3 82 4 74 1 2 30 452.4 8.6 19.9 841.0 0.54
L88-33 41 44 3 82 5 74 4 6 30 389.0 7.9 19.1 803.2 0.50
L88-34 42 45 3 82 4 77 1 0 30 466.4 8.4 21.7 916.6 0.51
L88-35 41 43 3 80 4 74 1 1 30 417.2 7.3 18.1 765.4 0.55
L88-36 42 44 3 85 5 74 4 3 30 351.2 7.9 19.0 756.0 0.47
L88-37 44 47 3 84 3 76 1 2 30 313.6 7.1 21.8 756.0 0.42
L88-38 41 45 3 81 3 73 2 3 30 317.4 8.8 18.6 576.4 0.56
L88-39 42 44 3 81 4 75 1 1 30 499.4 8.7 16.1 926.1 0.53
L88-40 42 46 3 82 4 74 1 1 30 395.2 7.6 16.4 812.7 0.48
L88-41 43 50 3 82 4 75 0 2 30 413.8 7.6 21.0 774.9 0.53
L88-42 42 45 4 83 4 76 3 3 28 415.1 8.1 19.7 803.2 0.50
L88-43 44 48 3 82 4 74 0 0 30 525.5 7.8 19.9 982.8 0.53
L88-44 42 47 2 83 6 76 0 1 30 510.4 9.0 16.2 935.5 0.54
L88-45 40 44 3 81 5 71 3 5 30 353.1 7.7 19.5 708.7 0.50

 

 

92

Table A3. Continued.

 

 

 

 

Line W tht Lodg Matr 05' PM Mac Mac Stnd Yieldect 100 1310 HI
45d 75d moist sw mass
<1 cm (I d gfin2 m2
L88-46 42 48 3 82 4 73 1 2 30 429.3 8.5 18.7 916.6 0.47
L88-47 42 44 4 82 3 75 o o 30 492.3 8.3 19.9 907.2 0.54
L88-48 42 48 3 81 4 72 4 3 30 424.2 7.3 20.3 841.0 0.51
L88-49 43 45 3 83 4 78 o o 30 385.8 7.1 20.5 727.8 0.53
L88-50 41 43 3 81 8 72 1 4 30 442.8 7.6 19.8 737.10.80
L88-51 4o 39 3 81 5 72 2 5 30 441.5 7.4 20.1 784.3 0.57
L88-52 43 49 4 85 3 78 o 0 30 402.3 7.8 19.7 878.8 0.46
L88-53 43 40 4 84 2 74 5 8 30 370.9 7.1 22.8 737.10.49
L88-54 43 50 4 87 3 77 0 0 30 803.8 8.9 21211245054
L88-55 41 51 3 80 4 72 1 1 30 522.3 8.0 20410017052
L88-56 41 43 4 81 4 78 1 1 30 481.9 8.1 17.3 889.4 0.56
L88-57 42 48 4 83 4 75 3 4 30 541.4 8.4 18.3 982.8 0.55
L88-58 41 42 2 81 5 73 1 4 27 338.4 9.4 17.8 823.7 0.55
L88-59 42 52 4 84 3 78 1 0 30 493.8 7.4 18.1 973.3 0.51
L88-60 42 48 4 83 3 77 0 3 30 545.3 8.0 21210017054
L88-61 42 43 3 82 5 74 3 3 30 457.9 7.9 18.2 831.8 0.54
L88-62 42 48 4 81 3 72 0 o 30 401.1 6.4 20.2 869.4 0.46
L88-63 41 43 3 81 8 74 0 2 30 542.5 8.0 17.8 926.10.59
L88-64 41 44 3 82 4 74 1 1 30 389.9 8.2 18.7 803.2 0.48
L88-65 42 43 4 82 4 77 5 5 30 383.8 8.5 18.3 784.3 0.49
L88-66 40 43 3 82 5 72 1 5 30 545.2 8.4 17.4 983.9 0.57
L88-67 42 44 2 80 4 74 2 3 30 399.3 10.1 15.7 785.4 0.52
L88-68 42 45 4 80 4 72 o o 30 440.2 7.9 17.9 897.7 0.49
L88-69 41 45 3 80 5 75 1 1 30 512.2 7.9 18.2 945.0 0.55
L88-70 41 44 3 83 4 75 o 1 30 448.0 8.3 19.1 850.5 0.52
L88-71 42 51 3 84 4 78 2 1 30 438.3 8.3 17.1 869.4 0.51
L88-72 42 39 4 82 4 71 1 2 30 402.5 7.9 18.3 784.3 0.51
L88-73 41 42 4 83 3 74 o 2 30 411.7 8.2 18.9 841.0 0.49
L88-74 42 45 4 82 3 73 0 0 30 528.3 8.0 17.9 992.2 0.54
L88-75 42 48 3 82 4 75 2 4 30 394.7 8.2 18.5 822.10.48
L88-76 42 43 3 80 4 74 2 2 30 392.6 7.7 20.3 718.2 0.55
L88-77 42 42 3 80 5 73 o o 30 383.7 7.7 18.4 881.5 0.57
L88-78 42 43 4 83 4 74 0 1 30 532.9 8.7 21.7 973.3 0.54
L88-79 42 47 4 82 3 74 3 3 30 425.8 8.1 18.8 831.6 0.51
L88-80 41 44 3 81 3 71 o 0 30 412.8 7.5 19.1 803.2 0.51
L88-81 42 45 3 81 5 73 1 1 30 384.8 7.3 17.0 880.4 0.53
L91-1 41 44 4 79 4 71 o o 30 480.0 7.8 18.5 841.0 0.56
L91-2 48 53 3 85 3 78 2 2 30 321.3 8.7 23.1 870.9 0.48
L91-3 42 53 3 80 5 72 o o 30 508.8 8.5 18.9 926.10.55
1.914 42 45 3 83 3 74 1 1 30 371.8 9.0 22.5 774.9 0.47
L91-5 42 42 2 85 3 77 1 2 30 349.7 8.3 21.3 852.0 0.53
L91-6 41 48 3 81 4 73 2 1 30 459.2 8.5 20.9 831.8 0.55
L91-7 41 47 3 82 4 72 0 o 30 535.3 7.1 24.5 973.3 0.55
L91-8 43 45 3 83 3 77 1 4 30 339.3 9.0 24.0 623.7 0.54
L91-9 43 53 3 83 3 78 0 1 30 442.9 8.0 23.2 831.6 0.53

 

\O
U)

Table A3. Continued.

 

 

Line Flor tht Lodg Matr o's' PM Mac Mac 51nd Yield Pct 100 8‘10 HI
45d 75d moist sw mass
d c_n_1 d d g/m2 /m2
L91-10 42 - 49 3 88 3 75 1 1 30 474.1 9.3 22.8 850.5 0.55
L91-11 41 47 3 79 5 72 0 o 30 402.8 8.2 18.2 737.10.54
L91-12 42 48 3 82 3 78 1 1 30 419.0 7.3 22.5 793.8 0.54
L91-13 43 47 3 81 4 74 2 1 30 391.8 7.4 20.9 793.8 0.49
L91-14 43 45 3 84 4 74 o 2 30 258.3 7.9 21.0 889.8 0.38
L91—15 44 48 3 85 3 74 1 1 30 343.8 7.6 22.7 774.9 0.44
L91-16 44 43 3 84 3 78 3 3 30 338.9 7.2 23.9 708.7 0.47
L91-17 43 48 4 83 3 75 0 1 30 414.8 8.0 20.2 841.0 0.49
L91-18 41 44 3 82 5 73 2 3 25 421.5 8.6 19.8 784.3 0.54
L91-19 47 53 2 86 3 78 4 3 30 322.4 11.1 28.8 765.4 0.41
L91-20 42 44 3 82 4 78 1 1 30 352.4 19.6 24.2 737.1048
L91-21 43 48 3 84 3 78 0 1 30 438.6 7.8 20.5 888.3 0.49
L91-22 42 48 3 85 3 75 1 2 30 408.5 10.0 21.3 746.5 0.52
L91-23 45 51 3 85 4 77 2 1 30 490.0 8.1 22.41020.60.48
L91-24 43 45 2 85 4 75 0 1 30 305.0 8.8 19.9 842.8 0.48
L91-25 42 48 4 84 3 72 3 2 30 484.0 7.9 24.2 803.2 0.58
L91-26 41 45 4 80 5 73 o 2 30 472.7 7.9 21.2 831.6 0.56
L91-27 43 45 3 80 4 72 0 o 30 413.9 8.7 19.9 748.5 0.55
L91-28 43 47 3 88 3 79 0 1 30 307.5 7.9 19.7 633.10.48
L91-29 41 43 3 84 4 71 o o 30 371.6 8.0 19.1 670.9 0.55
L91-30 42 48 2 83 3 74 1 1 29 399.1 8.8 23.6 774.9 0.51
L91-31 43 48 3 81 3 74 1 2 30 328.8 10.1 20.8 623.7 0.51
L91-32 42 5o 3 85 3 76 1 o 30 429.1 7.8 22.3 897.7 0.49
L91-33 42 47 3 83 4 72 2 1 30 544.9 8.3 23.1 992.2 0.55
L91-34 41 48 2 81 5 73 2 2 30 419.8 8.5 18.1 737.10.57
L91-35 41 39 3 85 3 73 o o 30 377.5 7.8 20.7 727.8 0.52
L91-38 42 44 3 83 4 74 1 1 30 315.8 8.8 24.2 727.6 0.43
L91-37 44 51 4 88 3 78 0 o 30 440.9 9.3 22.8 831.6 0.53
L91-38 43 45 2 82 5 75 1 3 30 316.3 7.2 19.7 633.10.50
L91-39 41 49 3 83 2 75 0 0 30 369.8 8.8 21.5 746.5 0.50
L91-40 43 48 3 83 3 74 1 1 30 336.1 9.2 20.0 670.9 0.50
L91-41 42 45 3 84 4 72 1 1 30 356.5 8.8 21.2 774.9 0.48
L91-42 42 42 3 83 4 74 1 2 30 350.9 9.6 18.8 870.9 0.52
L91-43 43 44 3 83 4 73 1 1 30 522.7 7.7 21.4 983.9 0.54
L91-44 43 45 3 84 3 78 0 0 30 402.8 7.7 21.2 803.2 0.50
L91-45 42 48 3 83 5 73 1 1 30 519.7 9.3 19.3 926.10.56
1.9146 42 50 3 86 3 74 0 1 30 427.4 10.0 24.9 859.9 0.49
L91-47 41 47 2 82 4 72 o 1 30 353.8 8.2 18.9 670.9 0.51
L9148 42 48 3 85 3 74 o 1 30 297.9 8.6 21.9 670.9 0.42
L91-49 43 47 3 84 3 78 0 0 30 284.2 8.4 22.7 557.5 0.47
L91-50 41 42 3 81 4 74 1 2 30 338.7 7.6 20.4 633.10.54
L91-51 42 45 3 84 3 74 2 2 30 429.8 8.5 23.0 793.8 0.53
L91-52 45 51 4 82 2 78 1 1 30 332.1 7.7 19.9 756.0 0.43
L91-53 44 49 2 88 3 78 4 3 30 293.5 7.8 21.0 642.6 0.48
L91-54 42 37 3 82 3 73 2 3 30 296.8 9.5 18.2 557.5 0.52
L91-55 43 47 3 84 3 75 2 1 30 384.4 7.5 22.6 784.3 0.49

 

 

94

Table A3. Continued.

 

 

 

Line Flor tht Lodg Matr DS_PM Mac MacJStnd Yield Pct 100 Bio HI
45d 75d moist sw mass
(1 cm d 0 gm? /m2

L91-56 42 48 3 81 3 73 o 0 30 537.8 8.2 23.1 935.5 0.57
L91-57 42 47 3 82 4 74 1 0 30 449.4 8.0 24.5 907.2 0.49
L91-58 44 48 3 87 3 77 0 0 30 341.2 11.3 23.8 822.10.40
L91-59 43 52 3 84 3 78 o o 30 448.1 11.2 20.9 935.5 0.46
L91-60 41 45 3 82 3 74 0 0 30 340.0 8.2 18.8 803.2 0.42
L91-61 42 45 3 80 4 73 0 1 30 397.3 8.3 18.9 758.0 0.53
L91-62 44 50 3 83 4 78 o 0 30 358.2 8.7 21.4 737.10.48
L91-63 40 44 3 84 5 74 1 0 30 318.6 8.6 22.5 680.4 0.48
L91-64 42 43 4 83 3 73 1 0 30 448.8 7.2 17.4 869.4 0.52
L91-65 44 45 3 88 3 78 0 0 30 238.8 8.4 22.5 578.4 0.41
L91-66 44 43 3 84 3 76 0 1 30 402.8 8.1 21.1 793.8 0.50
L91-67 43 48 3 84 3 74 4 4 30 460.0 9.5 20.5 935.5 0.49
L91-68 44 48 3 83 4 78 0 0 30 377.9 8.1 20.5 784.3 0.48
L91-69 44 51 3 87 3 79 0 1 30 323.8 12.1 26.71086.70.31
Tacana 43 45 3 83 4 74 0 0 30 441.8 9.1 18.9 841.0 0.51
V8025 42 48 4 80 3 73 0 2 30 376.0 7.8 17.3 680.4 0.58
Tio Canela 41 42 4 82 3 75 0 1 30 349.4 8.6 20.2 812.7 0.43
898311 41 48 3 83 5 73 2 1 30 507.9 6.9 18.9 897.7 0.56
VAX5 41 48 3 85 4 75 o 3 30 372.3 9.2 22.4 765.4 0.47
TLP19 45 44 3 87 4 80 3 3 28 493.1 7.5 19.8 954.4 0.51
BAT477 41 39 3 85 3 74 1 2 22 278.1 9.8 19.7 519.7 0.40
Rio Tibagi 42 49 3 84 5 75 0 0 22 397.8 7.8 16.7 737.10.54
EAP 40 33 3 80 4 75 0 3 27 428.8 10.8 22.8 727.8 0.57
9510-77

SEAS 34 39 2 82 5 89 7 5 28 308.7 11.0 25.5 595.3 0.53

 

 

T Flor - days to flowering, tht - Height, Lodg - Lodging (1-5), Matr - days to maturity,
DS - Desirability Score (1-9), PM - Physiological maturity, Mac 45d - Macrophomina
incidence at 45 days, Mac 75d - Macrophomina incidence at 75 days, Stnd - Stand, Pct
moist - percent moisture, 100 sw - 100 seed weight, HI - Harvest Index

95

Table A4. Yield under stress (Yd), yield under non-stress (Yp), and geometric mean
(GM) of 160 genotypes adjusted for plant stand by covariate analysis for the Honduras
experiment 2001.

 

 

 

 

 

 

 

 

 

 

Line Efiry Yd Yp GM Line Evy Yd Yp G_M—
k a k a k lha k a k a k a

L88-1 1 193 1441 527 L91-1 82 167 2279 617
L88-2 2 77 1729 364 L91-2 83 77 1523 342
L88-3 3 583 2007 1082 L91-3 84 435 2406 1023
L88-4 4 168 1 802 551 L91 -4 85 1 75 1 763 556
L88-5 5 400 2152 928 L91-5 86 264 1658 661
L88-6 6 382 2468 971 L91-6 87 183 2179 631
L88-7 7 263 1623 653 L91-7 88 364 2542 962
L88-8 8 250 2189 740 L91-8 89 108 1609 416
L88-9 9 213 1814 621 L91-9 90 166 2102 590
L88-10 10 335 1827 782 L91-10 91 448 2250 1004
L88-11 11 343 2075 844 L91-11 92 178 1910 583
L88-12 12 82 2510 454 L91-12 93 168 1988 578
L88-13 13 565 2922 1285 L91-13 94 -7 1857 1 15
L88-14 14 170 1770 549 L91-14 95 105 1213 357
L88-1 5 1 5 287 1 876 734 L91-1 5 96 289 1629 686
L88-16 16 402 1668 818 L91-16 97 249 1597 631
L88-1 7 1 7 459 1 966 950 L91 -17 98 142 1968 529
L88-18 18 90 1501 368 L91-18 99 270 2144 760
L88-19 19 579 2188 1126 L91-19 100 3 1528 64
L88-20 20 345 2045 840 L91-20 101 167 1671 528
L88-21 21 184 1650 551 L91-21 102 263 2072 739
L88-22 22 272 1963 731 L91-22 103 79 1938 392
L88-23 23 177 1 754 558 L91-23 104 175 2326 639
L88-24 24 248 2122 726 L91-24 105 184 1445 516
L88-25 25 382 2034 881 L91-25 106 514 2202 1064
L88-26 26 255 2157 742 L91-26 107 368 2244 908
L88-27 27 359 1952 837 L91-27 108 269 1964 726
L88-28 28 241 2355 754 L91-28 109 147 1457 463
L88-29 29 268 2120 754 L91-29 110 402 1762 842
L88-30 30 779 2263 1328 L91-30 111 599 1922 1073
L88-31 31 636 1730 1048 L91-31 112 211 1549 572
L88-32 32 263 2147 751 L91-32 113 64 2036 360
L88-33 33 371 1845 827 L91-33 114 368 2587 976
L88-34 34 311 2214 829 L91-34 115 168 1992 579
L88-35 35 270 1979 730 L91-35 1 16 207 1 791 609
L88-36 36 183 1665 553 L91-36 117 171 1497 506
L88-37 37 224 1486 577 L91-37 1 18 77 2092 402
L88-38 38 250 1 504 613 L91-38 1 19 251 1499 614
L88-39 39 221 2371 724 L91-39 120 143 1754 501
L88-40 40 324 1875 779 L91-40 121 262 1593 646
L88-41 41 225 1963 664 L91-41 122 47 1690 282
L88-42 42 317 2037 803 L91-42 123 157 1664 510
L88-43 43 195 2495 697 L91-43 124 253 2482 793
L88-44 44 334 2423 900 L91-44 125 142 1911 522
L88-45 45 164 1674 524 L91-451 126 266 2468 810

 

 

96

Table A4. Cofintinued. _
Line Entry Yd Yp GM

   

 

 

 

 

L88-51 51 41 5 2095 933 L91 -51 1 32 246 2039 708
L88-52 52 144 1909 525 L91 -52 1 33 56 1 574 298
L88-53 53 353 1 759 788 L91 -53 1 34 45 1391 250
L88-54 54 245 2868 838 L91-54 135 68 1406 309
L88-55 55 1 80 2480 668 L91 -55 1 36 258 1823 685
L88-56 56 406 2287 964 L91 -56 1 37 300 2553 875
L88-57 57 145 2571 610 L91-57 138 131 2133 528
L88-58 58 487 1681 905 L91-58 139 123 1617 445
L88-59 59 455 2344 1 033 L91 -59 140 486 21 26 1 01 6
L88-60 60 263 2590 825 L91—60 141 156 1612 501
L88-61 61 507 2173 1050 L91-61 142 157 1885 543
L88-62 62 1 87 1 903 596 L91 -62 143 397 1698 821
L88-63 63 842 2576 1473 L91-63 144 1 1 5 1 510 417
L88-64 64 1 1 3 1 849 457 L91 -64 145 225 2129 692
L88-65 65 258 1819 685 L91-65 146 86 1 130 312
L88-66 66 561 2589 1205 L91-66 147 1 28 1 91 1 494
L88-67 67 235 1894 667 L91-67 148 142 2183 556
L88-68 68 409 2089 924 L91-68 149 1 7 1 792 1 76
L88-69 69 680 2432 1286 L91 -69L 1 50 -2 1 534 60
L88-70 70 326 21 1 7 831 Tacana 1 51 21 3 2097 667
L88-71 71 230 2080 692 V8025 1 52 21 0 1 783 612
L88-72 72 314 1 91 0 774 Tio Canela 1 53 21 8 1 657 602
L88-73 73 264 1953 718 89831 1 154 375 241 1 951
L88-74 74 740 2508 1362 VAX 5 1 55 249 1765 663
L88—751 75 380 1872 843 TLP 19 1 56 1 69 2399 637
L88-76 76 363 1862 822 BAT 477 1 57 400 1 536 784
L88-77 77 298 1 820 737 Rio Tibagi 1 58 372 2108 886
L88-78 78 284 2530 848 EAP 1 59 232 21 12 699
951 0-77
L88-79 79 280 2021 753 SEA 5 1 60 524 1 521 893

 

L88-80 80 348 1959 825
_LB8-81 81 gm 1730 879

97

Table A5. Field data of 36 genotypes under drought stress at Montcalm, MI 2001.
Line Tantry Yield 100 swt Er matr lodg hght F
J d d gm
25.5 45.2 97.7 3.0 39.7 3.6
27.2 44.0 102.7 3.7 36.4 3.7
28.9 44.8 98.0 3.0 34.7 4.7
26.9 46.2 106.7 3.0 41.4 4.0
28.3 42.9 104.7 3.0 38.3 4.1
26.3 41.6 97.0 3.0 35.1 3.7
30.4 46.8 107.0 3.0 38.9 3.6
28.9 45.1 104.7 3.7 38.8 3.9
32.8 44.6 107.7 2.7 41.3 4.5
33.8 45.7 100.0 2.7 44.2 4.0
27.2 44.0 102.0 2.7 39.3 4.0
28.9 45.5 105.7 3.7 40.3 3.7
27.3 43.8 94.7 2.7 42.2 4.6
30.6 46.0 104.7 3.0 40.2 4.6
31.7 44.6 104.7 3.0 42.6 3.9
29.9 45.1 98.3 2.7 42.8 4.8
23.8 45.5 101.7 4.0 30.7 2.6
28.6 44.1 107.0 2.7 43.3 5.0
28.0 44.7 107.0 3.3 35.4 3.0
27.4 45.3 95.7 2.3 42.5 4.8
26.6 46.9 108.0 2.7 46.5 4.7
31.3 45.8 106.7 3.3 35.6 2.7
29.6 44.8 108.0 4.0 34.6 2.2
29.3 44.9 105.0 4.0 32.1 3.7
28.6 45.1 104.0 3.7 37.9 2.9
29.1 44.5 99.7 4.0 35.3 2.5
33.2 45.2 108.7 3.7 36.8 2.7
30.7 46.9 106.0 3.3 36.6 2.9
29.3 44.9 105.7 3.0 42.8 3.1
26.5 44.7 103.7 3.0 38.7 3.6
27.5 45.8 97.7 2.7 46.6 4.7
32.9 46.1 103.0 3.7 38.6 3.0
31.1 45.1 95.7 3.0 38.7 4.8
27.8 46.7 108.7 3.7 46.0 1.7
. 35.6 48.7 112.7 3.7 41.3 1.7
L91-69 36 17.2 34.4 47.0 107.7 3.0 41.6 2.0

T 100 sw - 100 seed weight, flor - days to flowering, matr - days to maturity, lodg -
lodging (1-5), hght - height and DS - desirability score (1-9)

 

 

98

Table A6. F_ield data of 36 genotypes under non-stress at Montcalm, MI 2001.
Line Entry Yield 100 sw’r flor matr lodg hght DS

cwt/acre g d d cm

 

 

L88-63 1 38.5 25.8 45.0 103.5 3.3 40.1 3.4
L88-74 2 33.0 27.0 44.3 108.8 3.4 39.7 2.9
L88-13 3 24.5 27.3 44.4 97.9 3.5 39.8 3.0
L88-30 4 30.0 26.1 44.6 107.6 3.8 44.1 3.6
L88-69 5 32.9 29.5 43.7 107.0 3.2 43.6 2.9
L88-66 6 28.4 24.9 43.0 101.3 3.0 37.3 3.3

L88-3 7 22.0 30.1 47.0 110.9 3.2 43.1 2.1
L88-19 8 30.3 28.3 44.0 109.6 3.3 37.9 3.2
L91-25 9 31.5 31.9 43.7 107.5 3.2 41.2 3.4

L91-30 10 35.4 33.0 44.6 105.0 3.3 44.2 3.0
L88-61 11 29.8 26.6 43.1 104.6 3.1 42.4 3.6
L88-59 12 30.7 29.0 45.0 107.7 3.8 37.3 2.9
L88-31 13 26.1 26.3 44.3 96.2 3.1 42.9 2.5
L91-59 14 19.5 30.3 47.0 109.4 3.2 45.4 3.4
L91-10 15 38.7 29.9 45.7 109.3 3.0 42.4 4.2
L91-3 16 27.4 27.1 45.7 101.3 2.4 46.0 4.6
|81066 17 31.4 24.8 45.4 109.8 4.0 36.6 2.1
898311 18 27.9 27.5 44.0 107.8 2.9 44.6 3.1
TLP 19 19 32.0 26.4 45.3 107.5 3.2 42.4 3.2
VAX 5 20 21.4 26.9 45.1 97.0 2.7 45.5 3.3
B95204 21 31.4 26.3 46.3 107.4 2.3 44.2 5.0
L88-37 22 26.9 30.4 45.6 105.9 3.4 38.8 2.9
L88-4 23 21.6 29.9 46.0 111.3 4.1 39.0 2.2
L88-2 24 26.7 26.8 43.4 108.6 3.7 36.4 2.7
L88-64 25 27.2 27.4 44.0 110.5 3.6 37.7 2.1
L91-22 26 26.5 28.8 44.4 109.0 3.6 39.6 2.0
L91-13 27 24.3 31.3 45.6 108.8 3.4 38.5 2.4
L91-37 28 16.0 30.4 46.7 112.8 3.8 41.9 2.4
L91-41 29 24.5 27.1 45.7 109.6 3.1 42.0 2.9
L91-53 30 23.2 26.1 45.7 107.0 3.4 39.1 2.8
L91-68 31 25.4 27.6 46.3 105.4 2.3 49.2 4.1
L91-49 32 27.8 31.6 46.3 107.0 3.7 39.3 3.5
L91-52 33 24.3 30.9 46.0 106.6 3.8 38.0 2.0
L88-18 34 17.2 28.4 47.0 117.5 4.5 46.4 1.6
L91-19 35 15.1 36.3 48.3 113.9 4.2 42.4 1.4
L91-69 36 15.0 33.8 46.3 111.5 3.9 43.4 2.0

‘r 100 sw - 100 seed weight, flor - days to flowering, matr - days to maturity, lodg -
lodging (1-5), hght - height and DS - desirability score (1-9)

 

99

Table A7. Mean values of Fractal Dimension, total root length, fine roots (A-C) and

 

 

 

ta roots - for the 81 RILS of population L88.
Line Fractal TotalTRoot AT 8" "—c H I J
Dimension Len th mm mm mm mm mm mm
L88-1 1 .53 2455.57 1854.07 521.04 58.39 0.19 0.13 3.61
L88-2 1 .49 2007.44 1563.84 392.08 37.31 0.20 0.09 3.29
L88-3 1 .52 2212.16 1633.82 496.22 60.73 0.43 0.1 1 2.76
L88-4 1 .50 1947.68 1391 .05 503.58 41 .34 0.33 0.16 3.09
L88-5 1 .48 1899.39 1471 .23 380.29 36.55 0.15 0.04 2.69
L88-6 1.51 2185.40 1653.08 472.95 46.02 0.12 0.09 2.30
L88-7 1 .42 1271 .89 914.79 331 .65 20.60 0.13 0.08 1 .95
L88-8 1.52 1876.98 1338.21 458.98 55.97 0.15 0.22 2.64
L88-9 1 .55 1970.53 1340.70 503.64 83.74 0.92 0.15 2.30
L88-10 1.51 2396.09 1887.57 444.67 48.55 0.07 0.17 2.87
L88-11 1.45 1711.56 1357.87 315.11 28.45 0.10 0.13 2.15
L88-12 1 .46 1972.86 1579.40 358.10 26.26 0.09 0.16 2.67
L88-13 1.47 1927.67 1463.76 413.17 38.34 0.15 0.13 3.51
L88~14 1.51 1998.46 1417.98 516.58 51.30 0.05 0.18 3.13
L88-15 1.49 1767.68 1223.30 494.38 39.55 0.33 0.12 3.18
L88-16 1.49 1919.91 1425.55 430.37 49.10 0.19 0.08 3.06
L88-17 1 .48 1703.16 131 1 .29 339.99 38.76 0.20 0.20 2.67
L88-18 1 .50 2056.45 1561 .53 441 .47 42.38 0.26 0.08 2.41
L88-19 1.51 1834.76 1284.52 485.45 52.37 0.14 0.17 2.50
L88-20 1 .52 2065.64 1480.86 510.47 59.37 0.52 0.13 2.86
L88-21 1.49 1940.02 1394.31 491.68 43.81 0.10 0.06 2.58
L88-22 1.50 2073.12 1517.26 500.21 42.85 0.13 0.05 4.51
L88-23 1.54 2398.38 1710.68 615.65 56.83 0.22 0.23 3.28
L88-24 1.46 1716.39 1241.52 431.61 33.62 0.04 0.11 3.21
L88-25 1.47 1316.23 922.90 353.02 32.44 0.16 0.18 2.19
L88-26 1.48 1719.07 1182.08 483.59 43.07 0.13 0.08 2.66
L88-27 1 .52 1638.26 1 160.32 397.33 57.34 0.59 0.00 2.56
L88-28 1 .53 2271.56 1621 .88 572.69 62.31 0.08 0.22 2.68
L88-29 1.51 2200.98 1558.71 578.17 50.80 0.17 0.24 3.24
L88-30 1.52 2105.69 1506.48 529.22 54.99 0.02 0.26 3.54
L88-31 1.50 2139.44 1599.37 473.33 48.35 0.14 0.31 3.66
L88-32 1 .57 2373.22 1659.38 592.89 83.36 0.54 0.32 2.87
L88-33 1 .50 2001 .49 1476.10 478.29 36.86 0.30 0.30 3.27
L88-34 1.47 1366.42 997.94 336.13 26.24 0.09 0.16 1.54
L88-35 1.52 2384.38 1730.78 575.34 58.57 0.17 0.14 3.82
L88-36 1 .46 1538.65 1041 .41 449.87 36.44 0.32 0.34 2.72
L88-37 1.53 2119.42 1565.76 482.71 57.1 1 0.21 0.07 3.35
L88-38 1 .50 1965.69 1420.34 500.34 35.23 0.27 0.16 2.45
L88-39 1.49 1786.83 1283.44 449.15 43.56 0.17 0.05 2.45
L88-40 1 .60 2423.41 1539.43 705.47 1 13.00 0.81 0.56 3.30
L88-41 1.55 2109.80 1500.17 518.63 70.38 0.41 0.08 3.13
L88-42 1.54 2213.90 1570.30 550.31 72.55 0.07 0.17 3.21
L88-43 1 .49 1554.04 1095.50 402.06 42.32 0.10 0.21 3.39
L88-44 1.46 1450.88 1034.19 380.83 28.64 0.15 0.18 1.91
L88-45 1 .51 1732.21 1252.49 420.01 46.44 0.32 0.22 2.18

 

 

100

Table A7. Continued.

 

 

Line Fractal Totfimt AT 8 C H l J
Dimension Len th mm mm mm mm mm mm
L88-46 1 .50 1995.22 1452.22 485.73 46.35 0.28 0.26 2.49
L88-47 1.51 2506.79 1949.99 496.44 44.63 0.16 0.12 4.89
L88-48 1.55 2245.70 1585.05 576.18 64.98 0.13 0.05 4.59
L88-49 1.45 1252.36 882.84 339.38 23.70 0.30 0.17 1.65
L88-50 1 .45 1372.76 974.05 368.06 24.75 0.07 0.03 1 .81
L88-51 1.50 2364.49 1849.15 456.56 42.81 0.26 0.13 2.95
L88-52 1.50 1803.74 1322.38 427.06 42.26 0.18 0.10 3.16
L88-53 1 .54 2171 .95 1507.55 578.22 65.40 0.23 0.03 3.59
L88-54 1 .59 2375.49 1421 .89 720.48 138.1 1 2.13 1.20 4.83
L88-55 1 .52 1939.71 1371.82 497.38 53.28 0.07 0.12 3.83
L88-56 1 .51 1835.58 1337.55 425.94 56.86 0.10 0.23 3.42
L88-57 1.50 1804.26 1282.39 453.78 53.46 0.16 0.09 3.15
L88-58 1.45 1770.94 131 1.95 421.33 29.73 0.24 0.07 2.88
L88-59 1 .60 2369.07 1495.41 693.40 1 19.46 0.86 0.34 3.85
L88-60 1.50 1987.14 1407.94 519.93 46.82 0.10 0.01 3.93
L8851 1.49 2128.10 1641.02 427.71 44.60 0.35 0.10 3.29
L88-62 1 .44 1421 .27 1030.18 361 .12 23.28 0.07 0.26 2.66
L88-63 1 .47 2053.61 1644.52 365.70 32.35 0.28 0.10 2.95
L88-64 1.56 2367.52 1648.38 614.17 78.64 0.25 0.04 4.17
L88-65 1 .45 1666.80 1206.45 421.87 29.64 0.34 0.04 2.61
L88-66 1 .45 1355.56 960.30 352.16 35.33 0.04 0.09 2.19
L88-67 1.60 2519.12 1715.64 660.12 96.69 0.62 0.28 3.93
L88-68 1 .54 2694.40 2080.43 533.31 60.04 0.16 0.00 4.22
L88-69 1.46 1531.77 1 163.28 330.08 29.23 0.31 0.08 2.16
L88-70 1 .43 1456.09 997.63 419.92 30.43 0.21 0.32 2.1 1
L88-71 1 .45 1566.58 1 152.80 377.83 29.55 0.07 0.24 1 .95
L88-72 1 .44 1229.69 859.99 332.48 28.53 0.30 0.09 2.26
L88-73 1 .53 2385.51 1790.29 509.82 61 .67 0.08 0.08 3.91
L88-74 1 .47 2019.02 1 560.40 409.47 36.38 0.12 0.00 3.45
L88-75 1 .65 3122.16 2050.44 862.90 137.70 0.88 0.20 4.59
L88-76 1 .49 1570.25 1096.46 423.28 39.30 0.05 0.13 3.58
L88-77 1.53 1984.24 1421.91 478.41 65.57 0.06 0.25 3.56
L88-78 1 .51 1887.86 1392.12 426.98 55.85 0.18 0.05 3.99
L88-79 1 .61 2403.66 1556.08 676.27 108.40 0.87 0.25 4.71
L88-80 1.51 1832.25 1319.99 453.87 45.41 0.19 0.05 2.54
L88-81 1.48 1971.53 1451.65 468.34 40.65 0.17 0.22 3.22

 

 

1' Root diameter classes A, B, C, H, I, J are O-O.5, O.5-1.0, 1.0-1.5, 3.5-4.0, 4.0-4.5, and
greater than 4.5 mm, respectively.

101

Table A8. Presence/absence of RAPDs associated with drought resistance in T- 3016 (T),
Sierra(S), B98311 (B), Raven (R), N98122 (19, Huron (H), VAX 5g!) and TLP 19 (PPL

 

 

 

 

 

 

 

 

 

 

 

 

LGRAPD P unlinkedTSBRNHV
9 H19 960 - + - - - - - - A09. 860

AB18.650 + + + + + + + A16.1220

v01.830 - + - - - - A18.800 - + + -

F06.970 + - + - - - - A18.1400

A16.850 + - + - - - - 1:01.520 + + - -

103.1130 + - + - - - - F10.1000

A08.510 - + - - + + 604.330 - + + -

1:05.440 - + - - - - 608.1240

l18.1400 - + - - - - 608.720

H03.1060 609.1070 + + + +

l03.870 - + - - - - - H01

R16.1180 + - + + + + H12.523 - - + +

H02.760 H18.520

114.770 + - + + + + H18.710 + + - -
"'_002.101o + + - - + - 1.07.900

203.1010 M05

606.400 - + - - - N03

201.780 - + - - - - 006.970

G11.500 + - + + ‘1' 4' T01

AB18.600 + - + + + + + T16

1.12.420 003

119.840 003

605.620 + - + - + + 010.1600 - + - -

P03.700 w20.3oo

A09.600 + + + + + + + wzo.1300 - - - -

610.550 - + + + + + + x01.850 + + + -

T18.550 + - + + x03.850 + + + -

L08.1090

N09.860 + - + + + SL-1

A09.500 A04.560 + + +

12101000 - + - + - + X11680 + + + +

006.900 X18980 + + + +

A603.570 A08.78o + + - +

103.830 - + - - - 208.750

110.500

110.950

A07.740 + + + + + +

AB14.450

R11.540

602.1010

204.580 + - + + - -

H08.490 + - + + + +

 

102

1 LG - Linkage groups identified in a previous study (Schneider et al., 1997a).

APPENDIX B

INT ROGRESSION OF ROOT ROT RESISTANCE FROM MIDDLE AMERICAN
LANDRACE BEAN TO CULTIVATED ANDEAN BEAN GENOTYPES

103

INTRODUCTION

Large-seeded dry beans (SS-65 grams per 100 seeds) are highly susceptible to root
rot, F usarium solam' pv. phaseoli (F sp). Irrigation increases production and provides
suitable conditions for F sp to proliferate. Relatively, no resistant sources can be found
among large-seeded kidney and cranberry beans in the Andean gene pool. Resistance to
root rot in small-seeded genotypes in the Middle American gene pool behaves as a
quantitative trait (Schneider et al., 2001). Due to intrinsic genetic differences, the
transference of quantitative traits across gene pools of common bean is a difficult task
(Kelly, 1988).

A major barrier for gene exchange across gene pools is the phenomenon of dwarf
lethality. Andean germplasm is characterized as possessing dl,d1,D12D12 genes, whereas
Middle American germplasm possess D1,D1,d12d12 genes for dwarf lethality. The D11 and
D12 genes are differentially expressed in bean roots and shoots, respectively (Shii etal.,
1981). When both loci are dominant, lethal and sub-lethal phenotypes occur (Shii et al.,
1980). The symptoms of dwarf lethality are stunted growth, chlorosis, crippled leaf
formation and plant death (Shii et al., 1980). The restricted root growth experienced by
dwarf lethal Fl hybrids between two gene pools has been overcome by hormonal
treatment (Beaver, 1992).

The inbred backcross method can be used to introgress the quantitative traits from
the wild source to elite cultivars of bean (Bliss, 1993). Afier the initial cross is made, Fl
plants are crossed back to the recurrent parent. The favorable genes of the recurrent
parent are recovered more quickly when the recurrent parent is used as the female. When

a suitable number of backcrosses have been made, the lines are advanced to near

104

homozygosity by single seed descent without selection. The desired quantitative trait can
be evaluated among the progeny lines in the F4 or later generations in common bean.

Advanced backcross QTL analysis is a method that combines QTL analysis with
the inbred backcross method (Tanksley and Nelson, 1996). Mapping of QTLs is delayed
until the BC2 or BC3 generation. RAPD markers conferring root rot resistance have
already been discovered in bean populations (Schneider et al., 2001). These markers
could be used to check for QTL presence in the inbred backcross populations developed in
the project. An additional backcross to the recurrent parent, inter-mating between sister
lines or crossing to other genotypes that acquired root rot resistance from a different
source could be facilitated through MAS to develop a superior variety with improved
levels of root rot resistance.

In this project, a source of root rot resistance was identified in the Mexican
landrace, Negro San Luis (NSL). It is a small-seeded black bean from the Middle
American gene pool. Its lack of adaptation in temperate latitudes make it difficult to
obtain viable offspring when crossed to Andean genotypes. Parental crosses were made
between the Middle American source of resistance, NSL, and the elite Michigan Andean
lines, Redhawk (dark red kidney) and C97407 (cranberry). This project was initiated in

order to introgress root rot resistance genes into the large-seeded bean class.

105

MATERIALS AND METHODS
Backcross #1

NSL was crossed to Redhawk and C97407. Since crosses between gene pools is
known to produce dwarf lethals a modified protocol (Beaver, 1992) was implemented on
October 4‘“, 1999. Three F, seeds from each cross were planted, one seed per pot. F,
seed was planted 2 cm deep in a shallow 4 cm soil. Significantly less root growth
develops in F, since it produces deleterious effects for the plant (Shii et al., 1981). The
stem grew unusually long. Soil and Hormex, 1000 ppm Indol 3-Butyric Acid was added
to encourage adventitious root growth in the F, hybrids.

A shorter day length was required to induce flowering. Two large trash cans were
placed on top of each other to create the dark period. Three pots were placed inside the
trash cans and the brims were sealed with duct tape to prevent any light from entering. A
photoperiod of 14 hours of darkness was kept in order to induce flowering. Recurrent
parents were planted at different time intervals to ensure that viable crosses would be
made. All crosses were made without emasculation. All F, progeny from a cross

between gene pools are expected to be dwarf lethals showing semi-lethality (Figure B1).

Backcross #2
In January 2000, 24 BC,F, individuals were planted at three seeds per pot.
Parental material and eleven genotypes consisting of commercial varieties and breeding
lines was planted along with the BC,F , seeds. BC,F , plants that resembled recurrent
parent phenotypes were preferentially crossed to the recurrent parent. Morphological

information was recorded to ensure a cross-fertilization was made (Table B1).

106

DNA Preparation
Leaf tissue from NSL, Redhawk and C97407 was collected and ground into a
powder using liquid Nitrogen. DNA samples were extracted according to the mini-prep
method (Afanador et al., 1993). The DNA concentration of each sample was quantified
using a fluorometer (Hoefer TKOlOO, Hoefer Scientific, San Francisco, CA). This stock

sample was diluted to a 10 ng/ml working solution for amplification by PCR.

PCR Protocol
The fragment size of the RAPD markers associated to root rot resistance
(Schneider et al., 2001) varied from 2000 base pairs to 800 base pairs. Gibco enzyme
was used for RAPD fragments greater than 1200 base pairs while the Stoffel enzyme was
used for fragments less than 1200 bp. Each RAPD primer was ran across NSL, Redhawk
and C97407. PCR reactions were performed only for primers that were present in the lab.

Samples were separated by electrophoresis on a 1.4% agarose gel.

107

RESULTS AND DISCUSSION
Backcross Study

Flowering occurred in seven days after the short day length treatment began.
Within 14 days, all F,s were flowering. A total of 22 successful cross-fertilizations were
made in the BCl generation. From the successful crosses, 50% of the BC,F,s were
expected to be dwarf lethals (Figure B1) and the actual percentage was 42%.

A concern of backcrossing is that seed might have resulted fi'orn self-fertilization.
Growth habit and flower color were used as morphological markers to identify cross-
fertilizations in the BC ,F,s (Table B1). NSL has two characteristics that are fixed at
homozygosity and dominant in nature; indeterminancy (GG) and purple flowers (FF)
(Figure BZ). The other parents, Redhawk and C97407, are homozygous recessive in both
loci (ggff) and show a determinate grth habit. Redhawk has white flowers while
C97407 has pink flowers due to epistatic interactions. All of the F,s (GgFt) expressed
indeterrninancy and purple flowers. Cotyledon color was also recorded, yet didn’t
provide full proof that a cross was made.

Only five plants from two different pods were determined to be self-pollinations
from a total of 62 BC,F, seeds from 24 pods that were planted. The other 22 pods were
confirmed to be crosses due to plant characteristics of indeterrninancy, purple flower
color or dwarf lethality. Variation in RNB 1-8 (Table B1) was recorded by a determinate
plant, an indeterminate plant and a dwarf lethal plant all coming from the same pod. One
BC,F , individual, RNB 1-9, yielded only one seed. This seed grew into a determinate
plant with white flowers. It was thought to be a self until its seed was examined. The
BC,F2 seed from RNB 1-9 was smaller than the seed of the recurrent parent, Redhawk,

108

having blunt ends and a darker seed color. Based on these seed characteristics, RNB 1-9
was determined to be a cross.

Two projects were developed in the second backcrossing scheme (Table B2). The
genetics project continued using the recurrent parents, Redhawk and C97407, to create
BCZF, individuals. A second project was devised to introgress the root rot genes from
BC,F, plants into other commercial kidney and cranberry seed types. BC,F, plants were
crossed into eleven diverse large seeded elite lines (Table B3). An average of ten

individual plants were planted for each of the 680 BCZF4 line.

RAPD Analysis

Molecular markers conferring root rot resistance were screened on the parents
(Schneider et al., 2001). Only 11 of the 16 RAPD markers (Schneider et al., 2001)
associated to root rot resistance were tested. Polymorphisms between NSL and the two
susceptible parents, Redhawk and C97407 were expected to show the same
absence/presence as FR266 and Montcalm, the parents used to discover the markers
associated to root rot resistance. Six RAPDs showed identical marker phenotypes as in
FR266 and Montcalm (Table B4). RAPD markers P7700 and G6, ,00 were previously
mapped to chromosome BZ on the bean core linkage map and were shown to encompass
the PvPRZ locus (Schneider et al., 2001). The PR proteins translated fiom this locus
were suggested to aid in root rot resistance. These markers can be used in the current
populations to identify potentially resistant lines. Advanced Backcross QTL analysis can
also be initiated to further characterize and identify durable QTL for root rot resistance in

Andean bean germplasm.

109

Parents: Redhawk dl,dl,Dlle2 xflNegro San Luis D1,D1,d12d12
F,: Redhawk x Dl,dl,Dlzdl2 (used as pollen parent)

BC,F,: (possible genotypes)
dl,dl,D12Dl2 - Healthy
dl,dl, Dlzdl2 - Healthy
Dl,dl,D12dl2 - Semi-lethal
Dl,dl,D12Dl2 - Semi-lethal

Figure B 1. Diagram of inheritance of dwarf lethal genes D1, and D12 from the initial cross
to the BC,F,.

Parents: Redhawk ggff x Negro San Luis GGFF
F,: Redhawk x GgFf
BC,F,: (possible genotypes)

ggff Determinate, White
ggFf Deterrninate, Purple
Ggff Indeterminate, White
GgF f Indeterminate, Purple

Figure 82. Diagram of the inheritance of morphological markers between kidney and
black bean parents.

110

Table B1. Morphological characteristics of BC lFl cranberry and kidney plants.

 

 

 

 

 

 

Cranberry Cotyledon flower determinate indeterminate dwarf total Results!

color lethals plants
CNB 2-‘l green purple 3 0 0 3 cross
CNB 2-2 green - 0 0 3 3 cross
CNB 2-3 green pink 2 1 0 3 cross
CNB 2-4 purple purple 0 2 0 2 cross
CNB 3-1 green pink 0 2 0 2 cross
CNB 3-2 purple pink 0 2 1 3 cross
CNB 3-3 purple purple 1 2 0 3 cross
CNB 3-4 green - 2 0 0 2 self
CNB 3-5 purple purple 0 2 0 2 cross
RNB 1-1 green - 0 0 3 3 cross
Kidney fiCotyledon flower determinate indeterminate dwarf total W

color lethals plants
RNB 1-2 purple white 1 2 0 3 cross
RNB 1-3 green white 0 1 2 3 cross
RN B 1-4 green - 0 0 3 3 cross
RNB 1-5 green - 0 1 2 3 cross
RNB 1-6 green - 1 1 1 3 cross
RNB 1-7 green - 0 0 1 1 cross
RNB 1-8 purple white 2 1 1 3 cross
RNB 1-9 green white 1 0 0 1 cross
RNB 1 -10 purple - 0 0 3 3 cross
RNB 2-1 purple white 2 1 2 3 cross
RNB 2-2 purple purple 2 0 1 3 cross

and

white
RNB 2-3 green white 3 0 0 3 self
RNB 3-1 green - 0 1 cross
RNB 3-2 green pugple g 0 3 cross

 

 

 

 

111

Table B2: Summary of seed increase methods.

Season
Summer 2000

F all 2000
Spring 2001

Summer 2001

Genetics and Commercial Projects

BC,F, seed planted for increase in Montcahn. Harvested by
plant.

BCZF2 seed increased in Greenhouse

BC2F3 seed increased in Greenhouse. Harvested one plant
per pot

Over 680 BC,F ,. 4 lines pplanted in Montcalm

Table B3: Commercial varieties and breeding lines crossed with BC,F, individuals.

Variety Name
Chardonnay
Chinook 2000
Red Kanner
Montcalm
K99968
K99973
K99974
K99983
Hooter

T. Hort
199134

Seed Type

Light Red Kidney
Light Red Kidney
Light Red Kidney
Dark Red Kidney
White Kidney
White Kidney
White Kidney
White Kidney
Cranberry
Cranberry
Cranberry

112

Table B4: Presence or absence of RAPD markers present in Negro San Luis (N SL),

Redhawk and C97407 compared to the check varieties, FR266 and Montcahn.
.l_.G’r RAPD FR266 Montcalm NSL Redhgwk C97407

 

   

 

 

2 P7.700 - +

P101600 - +

66.1100 - + -
+
4.

I
+++.
+++1

 

3 l18.1800
l18.1700
5 A62.800
@7900
6 G3.800|

+
I
l

 

 

63.2000
P9.1550
4 Y11.600
01g00
7 58.500
V3.1100

1 LG = Linkage Group
I RAPD markers previously identified (Schneider et al., 2001).

'+++++i
l
+I++
11
i+l+

+ +
I
4.
I
I

 

113

 

REFERENCES

Beaver, J. S. 1992. A simple method for producing seed from hybrid dwarfs derived from
crosses between Middle American and Andean gene pools. Annual Report of the
Bean Improvement Cooperative 35: 28-29.

Bliss, F. A. 1993. Breeding common bean for improved biological nitrogen fixation.
Plant and Soil 152271-79.

Kelly, J. D. 1988. The impact of the dwarf lethal (D1) genes on bean breeding programs at
MSU. Annual Report of the Bean Improvement Cooperative 31: 192-193.

Schneider, K. A., K. F. Grafton and J. D. Kelly. 2001. QTL analysis of resistance to
fusarium root rot in bean. Crop Science 41 (2): 535-542.

Shii, C. T., M. C. Mok and D. W. S. Mok. 1980. Expression of developmental
abnormalities in hybrids of Phaseolus vulgaris L. The Journal of Heredity 71:
218-222.

Shii, C. T., M. C. Mok and D. W. S. Mok. 1981. Developmental controls of
morphological mutants of Phaseolus vulgaris L.: Differential expression of
mutant loci in plant organs. Developmental Genetics 2: 279-290.

Tanksley, S. D. and J. C. Nelson. 1996. Advanced backcross QTL analysis: A method
for the simultaneous discovery and transfer of valuable QTLs from unadapted
germplasm into elite breeding lines. Theoretical and Applied Genetics 92: 191-
203.

114

APPENDIX C

PRESENCE OF TWO SCAR MARKERS LINKED TO RESISTANCE FOR COMMON
BACTERIAL BLIGHT IN POPULATION L91

115

INTRODUCTION

Common bacterial blight (CBB), caused by Xanthomonas campestris pv. phaseoli
(Smith) Dye., is a serious seed-borne disease endemic to common bean production in
Michigan and The only effective means of control is by planting disease-free seed.
Development of genetic resistance to CBB will combat the negative impact on bean
production in Michigan.

Quantitative trait loci regions conferring resistance to CBB have been identified
and localized on the bean core map (Miklas et al., 2000). SCAR markers tightly linked to
QTLs on bean linkage groups, BS and B10, have been developed. The SAP6 marker
from common bean is located on linkage group B10, whereas SU91 on B8 has tepary
bean ancestry. These markers can be used to screen for CBB resistance in breeding
programs where phenotypic screening for CBB is not routinely used. VAX 5, parent of
population L91, has been bred with pyramided CBB resistance and carries both SCAR
markers (Singh and Munoz, 1999). SCAR markers linked to QTL for resistance to CBB

were tested in population L91 to identify RILs possessing CBB resistance.

116

MATERIALS AND METHODS

DNA was extracted from leaf tissue of each F 3,4 RIL and parental genotype was
harvested, lyophilized and granulated. Lyophilized and granulated tissue was allocated
into 100 ml samples and DNA was extracted following the mini-prep procedure
(Afanador et al., 1993). The DNA concentration of each sample was quantified using a
fluorometer (Hoefer TKOIOO, Hoefer Scientific, San Francisco, CA). This stock sample
was diluted to a 10 ng/ml working solution for amplification by PCR.

SCAR protocol for one reaction in PCR, totaling 30.0111 was as follows: 17.85 111
H20, 3001 10X Buffer (Gibco), 2.25 111 MgC12 (Gibco), 0.601.11dNTP mix, 0.30111 Gibco
Taq Polymerase, 3.0111 10 ng/ul Primer mix, 3.0111 10 ng/ul DNA Template. The dNTP
mix consisted of 10111 of lOOmM of each dinucleotide (4) diluted in 60111 H20. The
primer mix included 10 ng/ul for the forward and reverse primers added in a 1:1 mixture.
This protocol was modified so that both markers could be amplified in the same PCR
reaction. This multiplexing step consists of halving the primer mix so that in each
reaction 1.5111 of SAP6 and 1.5111 of SU91 were added to make the standard 3.0111 of
primer mix. SAP6 and SU91 were multiplexed in the following therrnocycler regime: 34
cycles of 10s at 94°C, 40s at 57°C, and 2 min at 72°C followed by one cycle of 5 min at
72°C. Samples were ran by electrophoresis on a 1.4% agarose gel and viewed by ultra-

violet fluorescence. SAP6 and SU91 have fragment sizes of 820 and 700 bp respectively.

117

 

RESULTS AND DISCUSSION

VAX 5 was crossed to B98311, to generate population L91. Among the 69 RILS
from population L91, only 11 had both SCAR markers (Table C1). Some bands varied in
brightness of the flourescense. This could be explained by the collection of the DNA
samples. Tissue from three plants of each line was collected. Variation for the SCAR
marker might exist between plants in the F3,4 generation such that one two or three plants
may or may not have had the markers.

Field data from Saginaw 2000 and SCAR marker data were compiled for
evaluation and selection of the elite eleven lines (Table C2). Two lines, L91-47 and L91-
45, were selected based on the presence of both SCAR markers, the band intensity and
agronomic characteristics. Each was used into the crossing program with other elite
black lines. Segregation for seed brightness occurred in the L91 population. B98311 has
a dull seed coat while VAX 5 has a shiny seed coat. Further selection for dull seed coat
was made in L91-45 because it segregated for seed coat appearance. Resistance to CBB
has not been confirmed in selected lines through direct field or greenhouse screening.
Marker technology has allowed indirect selection for CBB resistance as the MSU Bean
Breeding Program is not routinely testing for resistance directly in the greenhouse or

field.

118

 

Table C1. Presence/absence of the SAP6 and SU91 SCAR markers for 69 RILs in

 

 

population L91.

RIL ‘ SAP6 SU91 RIL ‘ SAP6 SU91
L91 - 1 - - L91 - 40 + -
L91 - 2 - + L91 - 41 - -
L91 - 3 + - L91 - 42 + +
L91 - 4 - - L91 - 43 + -
L91 - 5 + - L91 - 44 - +
L91 - 6 + + L91 - 45 + +
L91 - 7 + - L91 - 46 + -
L91 - 8 - + L91 - 47 + +
L91 - 9 + 4' L91 - 48 + -
L91 - 10 + - L91 - 49 - -
L91 - 11 - - L91 - 50 + -
L91 - 12 - - L91 - 51 + -
L91 - 13 - - L91 - 52 - -
L91 - 14 + - L91 - 53 - -
L91 - 15 - + L91 - 54 - *-
L91 - 16 - - L91 - 55 + -
L91 - 17 - - L91 - 56 - -
L91 - 18 + - L91 - 57 - -
L91 - 19 - - L91 - 58 - -
L91 - 20 + - L91 - 59 + +
L91 - 21 - - L91 - 60 - +
L91 - 22 + + L91 - 61 - -
L91 - 23 - + L91 - 62 - -
L91 - 24 + - L91 - 63 + -
L91 - 25 + - L91 - 64 - -
L91 - 26 + - L91 - 65 + -
L91 - 27 + - L91 - 66 + +
L91 - 28 - + L91 - 67 - -
L91 - 29 + - L91 - 68 + +
L91 - 30 + - L91 - 69 - +
L91 - 31 - -

L91 - 32 + -
L91 - 33 - -
L91 - 34 + +
L91 - 35 + *-
L91 - 36 - -
L91 - 37 + -
L91 - 38 - -
L91 - 39 - -

 

 

119

 

Table C2. Marker and agronomic characteristics of eleven genotypes possessing both
SCAR markers along with the parents.

 

 

 

finotype Intensity Intensity flor’r hght lodg matr IDS Seed
of SAP6 of SU91 d cm 9 brilliance
L91-6 21 1 52 49 2 103 3 Dull
L91-9 3 2 51 50 2 102 4 Shiny
L91-22 2 2 50 33 3 99 3 Dull
L91-34 2 1 51 45 3 102 3 Shiny
L91-35 2 1 49 42 2 102 3 Dull
L91-42 4 4 50 52 2 102 5 Shiny
L91-45 4 4 ~ 52 59 3 104 3 Mixed
L91-47 4 4 51 44 2 99 5 Dull
L91-59 3 2 54 49 3 106 3 Shiny
L91-66 5 5 53 44 3 107 3 Shiny
L91-68 4 3 53 46 2 105 3 Shiny
VAX 5 5 5 - 53 2 1 10 5 Shiny
@9831 1 0 0 51 45 3 103 3 Dull

 

T flor - days to flowering, hght - height, lodg - lodging (1-5), matr - days to maturity, DS
- desirability score (1-9).

I F lourescense of band is characterized by 1 = very faint to 5 = very bright and 0 = no
amplification.

120

REFERENCES

Afanador, L., S. D. Haley and J. D. Kelly. 1993. Adoption of a "mini-prep" DNA
extraction method for RAPD marker analysis in common bean (Phaseolus
vulgaris L.). Annual Report of the Bean Improvement Cooperative 36: 10-11.

Miklas, P. N., J. B. Smith, R. Riley, K. F. Grafton, S. Singh, C. Jung and D. P. Coyne.
2000. Marker-assisted breeding for pyramided resistance to common bacterial
blight in common bean. Annual Report of the Bean Improvement Cooperative 43:
39-40.

Singh, S. P. and C. G. Munoz. 1999. Resistance to common bacterial blight among
Phaseolus species and common bean improvement. Crop Science 39: 80-89.

121

   

111111111111111.11111