ROW WIDTH AND PLANT POPULATION EFFECTS ON PLANT MORPHOLOGY,
WEED CONTROL, AND YIELD IN TYPE II BLACK BEAN AND SMALL RED BEAN
By
Ryan Charles Holmes

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
MASTER OF SCIENCE
Crop and Soil Sciences
2012

ABSTRACT
ROW WIDTH AND PLANT POPULATION EFFECTS ON PLANT MORPHOLOGY,
WEED CONTROL, AND YIELD IN TYPE II BLACK BEAN AND SMALL RED BEAN
By
Ryan Charles Holmes
The development of upright black and small red bean varieties gives dry bean growers
opportunities to plant in narrower rows, which has been associated with yield and weed control
benefits in many crops, and to use direct harvest methods with a standard combine. Field studies
were conducted in 2010 and 2011 at two locations in Michigan to examine the effect of row
width, population, and herbicide combination on canopy closure, weed suppression, pest
management, plant architecture, and yield in two upright varieties, ‘Zorro’ black and ‘Merlot’
small red beans. In one set of studies, each variety was planted in 38- and 76-cm rows, as well
as 51-cm rows at one location. Populations were 196500, 262000, and 327000 plants ha-1 in
black bean and 148000, 196500, and 262000 plants ha-1 in small red bean. In a second set of
studies, ‘Zorro’ only was planted at 262000 plants ha-1 in row widths of 38- and 76-cm, and six
weed control strategies were examined: S-metolachlor + halosulfuron (PRE), S-metolachlor
(PRE) fb. bentazon + fomesafen (POST), halosulfuron (PRE) + clethodim (+fomesafen in one
site-year) (POST), imazamox + bentazon (POST), weed-free, and no weed control. Narrow rows
generally increased canopy closure, enhanced weed suppression, did not increase pest pressure,
increased branching and pod formation, and increased yield. Population had little effect on
canopy closure, weed suppression, pest pressure or yield. Pod formation, seeds per pod, and
branching decreased as population increased. All herbicide combinations suppressed weeds and
increased yields compared to the untreated but varied in their control of specific weeds.
Nomenclature: Dry bean, Phaseolus vulgaris L.; preemergence; postemergence

DEDICATION
To John Casey Chumley, Derek Sova, Amanda Harden, Alex Lindsey, and Laura Bast.
All of these are daring in their pursuit of scientific knowledge and are therefore worthy.
Also to my parents, Bob and Debbie Holmes, without whom this thesis would not exist.

iii

ACKNOWLEDGEMENTS

I would like to thank Dr. Christy Sprague for taking me on as a graduate assistant, for
serving as my major professor, and for believing in me as I learned to be an agricultural
researcher. I would also like to thank my committee members, Dr. Christina DiFonzo, who
educated me on western bean cutworm, and Dr. James Kelly who supplied valuable information
on dry bean as a crop. Thanks are also in order to Dr. Karen Renner for allowing me to serve as
her teaching assistant and to Dr. Linda Hanson for help with bean disease identification.
It is difficult to imagine trying to undertake this work without the assistance of Gary
Powell in implementing the field research and Erin Taylor in keeping the lab group running
smoothly; both are truly devoted to doing excellent work.

Also, Paul Horny and Dennis

Fleischmann at the Saginaw Valley Research and Extension Center were consistently hospitable
and maintained a top-notch facility for my off-campus work. Thanks go to Meg Chludzinski for
running the cutworm trap network and David Muñoz for help with statistics. I must also thank
the Michigan Bean Commission and Project Greeen for granting the funds to do this research.
I would like to thank my fellow graduate students Kelly Barnett and Molly Buckham for
introducing me to the ways of weed science, and Alicia Spangler and David Powell for being
there to help me in my endeavors. In addition, this work would have been impossible without
the dedicated work of a number of undergraduates: Nicole Phillips, Dave Reif, Mark Dingee,
Megan Tomlin, Canton Brissette, Meghan Bonthuis, Amanda Anderson, Amanda Harden,
Heather Van Lieu, and Dave Houghtaling have all contributed to making this happen. Many of
those named have also become great friends, and I would be hard-pressed to find a better group
of co-workers. Lastly, major thanks go to my family and friends in Dixie for supporting me in
my northern sojourn and above all to Jesus Christ for saving my life and making me who I am.
iv

TABLE OF CONTENTS

LIST OF TABLES ..................................................................................................................... viii
LIST OF FIGURES ................................................................................................................... xiii
CHAPTER 1
LITERATURE REVIEW .............................................................................................................1
Introduction ........................................................................................................................1
Dry Bean Production and Weed Control.........................................................................2
Effects of Row Width and Population on Canopy Development ...................................5
Effects of Planting Practices and Canopy Development on Yield ...............................10
Effects of Planting Practices and Canopy Development on Weed Suppression ........22
Disease Considerations with Regard to Planting Practices and Canopy
Development .....................................................................................................................27
Effects of Planting Practices on Yield Components and Morphological
Characteristics..................................................................................................................31
Weed Control Systems in Dry Bean ...............................................................................36
Literature Cited ...............................................................................................................50
CHAPTER 2
EFFECTS OF ROW WIDTH AND PLANTING POPULATION ON CROP AND WEED
DEVELOPMENT IN TYPE II VARIETIES OF BLACK BEAN AND SMALL RED
BEAN ............................................................................................................................................58
Abstract .............................................................................................................................58
Introduction ......................................................................................................................59
Materials and Methods ....................................................................................................64
Dry bean populations...........................................................................................66
Crop canopy development ...................................................................................67
Weed emergence and biomass ............................................................................68
Insect monitoring and feeding ............................................................................68
Disease monitoring ...............................................................................................68
Maturity, lodging, and yield ................................................................................69
Economic analysis ................................................................................................69
Statistical analysis ................................................................................................70
Results and Discussion.....................................................................................................70
Growing conditions ..............................................................................................70
Crop canopy development ...................................................................................71
Row width effects on canopy development.................................................71
Population effects on canopy development ................................................73
Weed suppression.................................................................................................75
Weed biomass.............................................................................................75
Weed emergence ........................................................................................77
Summary ....................................................................................................78
v

Disease development and humidity ....................................................................78
White mold .................................................................................................78
Alternaria infection ....................................................................................78
Humidity .....................................................................................................79
Insect feeding ........................................................................................................80
Lodging .................................................................................................................83
Maturity ................................................................................................................84
Dry bean yield ......................................................................................................85
Economic returns .................................................................................................88
Conclusions ...........................................................................................................88
Appendix .........................................................................................................................110
Literature Cited .............................................................................................................113
CHAPTER 3
WEED CONTROL IN WIDE- AND NARROW-ROW BLACK BEANS ...........................118
Abstract ...........................................................................................................................118
Introduction ....................................................................................................................119
Materials and Methods ..................................................................................................122
Crop Injury and weed control, density, and biomass .....................................124
Yield ....................................................................................................................125
Economic returns ...............................................................................................125
Data analysis .......................................................................................................126
Results and Discussion...................................................................................................126
Growing conditions ............................................................................................126
Crop injury .........................................................................................................127
Weed control by species.....................................................................................128
Common lambsquarters ...........................................................................128
Pigweed ....................................................................................................130
Velvetleaf..................................................................................................131
Eastern black nightshade .........................................................................132
Common ragweed ....................................................................................133
Annual grass ............................................................................................134
Common purslane ....................................................................................135
Summary ..................................................................................................136
Mid-season weed populations ...........................................................................136
Final weed biomass ............................................................................................138
Dry bean yield ....................................................................................................140
Economic returns ...............................................................................................142
Conclusions .........................................................................................................143
Appendix .........................................................................................................................159
Literature Cited .............................................................................................................161
CHAPTER 4
EFFECTS OF ROW WIDTH AND PLANTING POPULATION ON TYPE II BLACK BEAN AND
SMALL RED BEAN ARCHITECTURE AND YIELD COMPONENTS ........................................ 165

Abstract ...........................................................................................................................165
Introduction ....................................................................................................................166
vi

Materials and Methods ..................................................................................................168
Results and Discussion...................................................................................................171
Growing conditions ............................................................................................171
Yield ....................................................................................................................172
Lodging ...............................................................................................................172
Plant height .........................................................................................................173
Pods per plant .....................................................................................................173
Seeds per pod ......................................................................................................175
Weight per hundred seeds .................................................................................176
Branching............................................................................................................177
Pod distribution between main stem and branch ...........................................179
Main-stem pods ........................................................................................180
Branch pods .............................................................................................181
Ratio of branch pods to main-stem pods .................................................182
Distribution of seeds per pod between main stem and branch ......................184
Seeds per main-stem pod .........................................................................184
Seeds per branch pod ...............................................................................184
Ratio of branch seeds per pod to main-stem seeds per pod .....................185
Conclusions .....................................................................................................................186
Literature Cited .............................................................................................................202

vii

LIST OF TABLES

Table 1.1 Summary of studies examining row width and population in dry bean ..........40
Table 1.2. Summary of studies examining row width and population in crops other than
dry bean ............................................................................................................................43
Table 2.1. Target within-row plant spacings for Type II black and small red beans
planted in three row widths and at three planting populations .........................................90
Table 2.2. Early season stand counts in 38- and 76-cm rows at East Lansing and 38-,
51-, and 76-cm rows at Richville compared with target planting populations .................91
Table 2.3. Monthly and 30-yr average precipitation at the Harry and Hazel Box Farm in
East Lansing, MI and the Saginaw Valley Research and Extension Center in Richville,
MI in 2010 and 2011 ........................................................................................................92
Table 2.4. Main effect of row width on weed biomass in black and small red bean plots
at East Lansing and Richville in 2010 and 2011 ..............................................................92
Table 2.5. Interaction of row width and planting population with regard to weed
biomass in small red bean at East Lansing in 2011 ..........................................................93
Table 2.6. Main effect of planting population on weed biomass in black and small red
bean plots at East Lansing and Richville in 2010 and 2011 .............................................93
Table 2.7. Effects of row width and planting population on average humidity under the
black bean canopy during the period in which the canopy was well-developed at East
Lansing and Richville in 2010 and 2011 ..........................................................................94
Table 2.8. Main effect of row width on western bean cutworm damage in black and
small red bean in at East Lansing and Richville in 2010 and 2011 ..................................94
Table 2.9. Interaction of row width and population with regard to western bean
cutworm damage in black bean at East Lansing in 2010 .................................................95
Table 2.10. Interaction of row width and planting population with regard to western
bean cutworm damage in black bean at Richville in 2011 ...............................................95
Table 2.11. Main effect of planting population on western bean cutworm damage by
planting population in black and small red bean at East Lansing and Richville in 2010
and 2011 ...........................................................................................................................95

viii

Table 2.12. Main effect of row width on lodging in black and small red bean at East
Lansing and Richville in 2010 and 2011 ..........................................................................96
Table 2.13. Main effect of planting population on lodging in black and small red bean at
East Lansing and Richville in 2010 and 2011 ..................................................................96
Table 2.14. Main effect of row width on maturity in black and small red bean at East
Lansing in 2010 and 2011 ................................................................................................97
Table 2.15. Main effect of planting population on maturity in black and small red bean
at East Lansing in 2010 and 2011 ....................................................................................97
Table 2.16. Interaction of row width and planting population with regard to late-season
maturity in small red bean at East Lansing in 2011 .........................................................98
Table 2.17. Main effect of row width on yield of black bean and small red bean at East
Lansing and Richville in 2010 and 2011 ..........................................................................98
Table 2.18. Main effect of planting population on yield of black and small red bean at
East Lansing and Richville in 2010 and 2011 ..................................................................99
Table 2.19. Main effect of row width on economic return of black and small red bean at
East Lansing and Richville in 2010 and 2011 ..................................................................99
Table 2.20. Interaction of row width and planting population with regard to economic
return in small red bean at Richville in 2011 .................................................................100
Table 2.21. Main effect of planting population on economic return of black and small
red bean at East Lansing and Richville in 2010 and 2011 .............................................100
Table A2.1. Schedule of activities in the field for population-row width studies .........111
Table A2.2. Dates of light meter readings and weed emergence counts in populationweed control studies .......................................................................................................112

ix

Table 3.1. Weed control systems, rates, and application timings in black beans planted
in 38- and 76-cm rows ....................................................................................................146
Table 3.2. Monthly and 30-yr average precipitation at the Harry and Hazel Box Farm in
East Lansing, MI and at the Saginaw Valley Research and Extension Center in
Richville, MI in 2010 and 2011......................................................................................147
Table 3.3. Summary of the effect of herbicide treatments and of narrow rows on crop
injury and control of various weed species ....................................................................148
Table 3.4. Main effect of herbicide treatment on control of common lambsquarters,
pigweed, and velvetleaf at 28 DAT at East Lansing and Richville in 2010 and 2011 ...149
Table 3.5. Main effect of row width on control of common lambsquarters, pigweed, and
velvetleaf at 28 DAT at East Lansing and Richville in 2010 and 2011 .........................150
Table 3.6. Interaction of row width and herbicide combination with regard to control of
common lambsquarters at 28 DAT at East Lansing in 2011 ..........................................150
Table 3.7. Main effect of herbicide treatment on control at 28 DAT of eastern black
nightshade and common ragweed at East Lansing in 2010 and of annual grass and
common purslane at East Lansing in 2011 .....................................................................151
Table 3.8. Interaction of row width and herbicide treatment at 28 DAT with regard to
control of eastern black nightshade at East Lansing in 2011 .........................................152
Table 3.9. Main effect of row width on control of eastern black nightshade and common
ragweed at 28 DAT at East Lansing in 2010 and of annual grass and common purslane
at East Lansing in 2011 ..................................................................................................152
Table 3.10. Main effect of row width on mid-season weed populations at East Lansing
and Richville in 2010 and 2011 ......................................................................................153
Table 3.11. Main effect of weed control system on mid-season weed populations at East
Lansing and Richville in 2010 and 2011 ........................................................................153
Table 3.12. Main effect of weed control system on end-of-season weed biomass at East
Lansing and Richville in 2010 and 2011 ........................................................................154
Table 3.13. Main effect of row width on end-of-season weed biomass at East Lansing
and Richville in 2010 and 2011 ......................................................................................154
Table 3.14. Main effect of weed control system on bean yield at East Lansing in 2011
and Richville in 2010 and 2011 ......................................................................................155

x

Table 3.15. Main effect of row width on bean yield at East Lansing in 2011 and
Richville in 2010 and 2011 ............................................................................................155
Table 3.16. Interaction of row width and weed control system with regard to bean yield
at East Lansing in 2010 ..................................................................................................156
Table 3.17. Main effect of row width on economic return at East Lansing in 2011 and
Richville in 2010 and 2011 ............................................................................................156
Table 3.18. Main effect of weed control system on economic return at East Lansing in
2011 and Richville in 2010 and 2011 .............................................................................157
Table 3.19. Interaction of row width and weed control system with regard to economic
return at East Lansing in 2010 ........................................................................................158
Table A3.1. Schedule of activities in the field for weed control-row width studies ......160
Table 4.1. Within-row plant spacing for Type II black and small red beans planted in
three row widths and at three planting populations ........................................................190
Table 4.2. Early-season stand counts of black and small red beans compared with target
planting populations at East Lansing and Richville in 2010 and 2011 ..........................191
Table 4.3. Monthly and 30-yr average precipitation at the Harry and Hazel Box Farm in
East Lansing, MI and at the Saginaw Valley Research and Extension Center in
Richville, MI in 2010 and 2011......................................................................................192
Table 4.4. Main effect of row width on yield of black and small red bean at East
Lansing and Richville in 2010 and 2011 ........................................................................192
Table 4.5. Main effect of planting population on yield of black and small red bean at
East Lansing and Richville in 2010 and 2011 ................................................................193
Table 4.6. Main effect of row width on lodging in black and small red bean at East
Lansing and Richville in 2010 and 2011 ........................................................................193
Table 4.7. Main effect of population on lodging in black and small red bean at East
Lansing and Richville in 2010 and 2011 ........................................................................194
Table 4.8. Main effect of row width on plant height in black and small red bean at East
Lansing and Richville in 2010 and 2011 ........................................................................194
Table 4.9. Main effect of planting population on yield components and plant
architecture in black and small red beans at East Lansing and Richville in 2010 and 2011
........................................................................................................................................195

xi

Table 4.10. Main effect of row width on number of pods per plant in black and small
red bean at East Lansing and Richville in 2010 and 2011 .............................................195
Table 4.11. Main effect of row width on number of seeds per pod in black and small red
bean at East Lansing and Richville in 2010 and 2011 ...................................................196
Table 4.12. Main effect of row width on weight per hundred seeds in black and small
red bean at East Lansing and Richville in 2010 and 2011 .............................................196
Table 4.13. Main effect of row width on number of branches per plant combined across
black and small red bean at East Lansing and Richville in 2010 and 2011 ...................196
Table 4.14. Interaction between row width and planting population with regard to
number of branches per plant in black and small red bean at Richville in 2010...........197
Table 4.15. Main effect of row width on number of pods on the main stem in black and
small red beans at East Lansing and Richville in 2010 and 2011 ..................................197
Table 4.16. Main effect of planting population on number of pods on the main stem in
black and small red beans at East Lansing and Richville in 2010 and 2011 ..................198
Table 4.17. Main effect of row width on number of pods on the branches in black and
small red beans at East Lansing in 2010 and 2011 and in black beans at Richville in
2011 ................................................................................................................................198
Table 4.18. Interaction of row width and planting population with regard to number of
pods on the branches in black and small red beans at Richville in 2010 .......................199
Table 4.19. Interaction of row width and planting population with regard to number of
pods on the branches in small red bean at Richville in 2011 .........................................199
Table 4.20 . Main effect of row width on the ratio of branch pods to main-stem pods in
black and small red beans at East Lansing and Richville...............................................200
Table 4.21. Main effect of row width on the ratio of number of seeds in branch pods to
number of seeds in main-stem pods in black and small red beans at East Lansing and
Richville in 2010 and 2011 ............................................................................................200
Table 4.22. Main effect of planting population on the ratio of number of seeds in branch
pods to number of seeds in main-stem pods in black and small red beans at East Lansing
and Richville in 2010 and 2011 ......................................................................................201

xii

LIST OF FIGURES

Figure 1. Main effect of row width on canopy closure for black beans planted at East
Lansing, MI in 2010. Vertical bars represent Fisher’s protected LSD at the α < 0.05 level
of significance. .................................................................................................................101
Figure 2. Main effects of (a) row width and (b) planting population on canopy closure
for black beans at East Lansing, MI in 2011. Vertical bars represent Fisher’s protected
LSD at the α < 0.05 level of significance.........................................................................102
Figure 3. Main effects of (a) row width and (b) planting population on canopy closure
for black beans at Richville, MI in 2011. Vertical bars represent Fisher’s protected LSD
at the α < 0.05 level of significance .................................................................................103
Figure 4. Main effects of (a) row width and (b) planting population on canopy closure
for small red beans at East Lansing, MI in 2010. Vertical bars represent Fisher’s
protected LSD at the α < 0.05 level of significance.........................................................104
Figure 5. Main effects of (a) row width and (b) planting population on canopy closure
for small red beans at East Lansing, MI in 2011. Vertical bars represent Fisher’s
protected LSD at the α < 0.05 level of significance.........................................................105
Figure 6. Main effects of (a) row width and (b) planting population on canopy closure
for small red beans at Richville, MI in 2010. Vertical bars represent Fisher’s protected
LSD at the α < 0.05 level of significance ........................................................................106
Figure 7. Main effects of (a) row width and (b) planting population on canopy closure
for black beans at Richville, MI in 2010. Vertical bars represent Fisher’s protected LSD
at the α < 0.05 level of significance .................................................................................107
Figure 8. Main effects of (a) row width and (b) planting population on canopy closure
for small red beans at Richville, MI in 2011. Vertical bars represent Fisher’s protected
LSD at the α < 0.05 level of significance. .......................................................................108
Figure 9. Effect of planting population on canopy closure for black beans planted in (a)
38- and (b) 76-cm rows at three populations at East Lansing, MI in 2010. Vertical bars
represent Fisher’s protected LSD at the α < 0.05 level of significance ...........................109

xiii

CHAPTER 1
LITERATURE REVIEW
Introduction
Dry bean (Phaseolus vulgaris L.) is a leguminous plant grown as a short-season summer
annual crop. It is divided into over a dozen market classes, each of which has a set of distinct
characteristics and is marketed under a different name. Among the most widely grown market
classes in the United States are black beans, small red beans, navy beans, and pinto beans. In
temperate production regions, dry beans are typically planted in late spring, and the seeds are
harvested after the plants have dried down in early fall. Dry beans are widely used as a human
foodstuff, serving as a relatively inexpensive source of protein in many parts of the world,
especially in developing countries where meat is difficult to obtain (Robertson and Frazier 1978).
They are the third most important food legume in the world, with only soybean and peanut
exceeding production levels of dry bean. Annual world dry bean production is estimated at 11.8
million tonnes on almost 13 million hectares (Schwartz et al. 2004). In the United States, dry
bean was planted on an average of 634,000 hectares per year between 2006 and 2010, and total
production was valued at 500 million to 900 million dollars annually (USDA-NASS, 2011).
Michigan is the second largest dry bean producer in the nation, with 80,000 to 100,000
hectares planted annually. Michigan is the nation’s number one producer of two market classes,
black beans and cranberry beans, and is the number two producer of two others, small red beans
and navy beans (Michigan Bean Commission, 2011). Four dry bean growth habits, referred to as
Types I, II, III, and IV, are recognized (Kelly 2001, 2010; Urwin et al. 1996). Type I are
determinate bush forms, Type II upright, indeterminate short vines, Type III long-prostrate or

1

semi-prostrate vines, and Type IV are climbing vines. Type IV varieties are rarely grown in the
United States.

Dry Bean Production and Weed Control
Dry bean is a low-growing crop that is highly susceptible to competition from weeds
(Burnside et al. 1998; Hekmat et al. 2008). Reported losses from weed competition have been as
high as 80-85% (Blackshaw and Esau 1991). Others have reported losses ranging from 16-52%
and from 40-71% when weeds were left uncontrolled (Chikoye et al. 1995; Wall 1995). The
variability in yield loss was partly a function of different predominant weed species; the 16-52%
loss was largely due to common ragweed (Ambrosia artimesiifolia L.), the 40-71% yield loss to
redroot pigweed (Amaranthus retroflexus L.) and common lambsquarters (Chenopodium album
L.), and the 80-85% yield loss to hairy nightshade (Solanum sarrachoides Sendtner). Two of the
most problematic weeds in North American dry bean production are common lambsquarters and
redroot pigweed (Wall 2005). Common ragweed and foxtails (Setaria spp.) are also considered
significant weed problems in dry bean production systems (Chikoye et al. 1995; Lamey et al.
1991). Other weeds, such as nightshades (Solanum spp.), can affect harvest efficiency, stain
beans, or cause spoilage in storage in addition to reducing yields (Blackshaw et al. 1999;
Burnside et al. 1998; Quackenbush and Andersen 1984; Robertson and Frazier 1978).
Historically, many dry bean growers have relied heavily on inter-row cultivation (Arnold
et al. 1993; Blackshaw et al. 2000; Burnside et al. 1998, Robertson and Frazier 1978) for weed
control. Chemical weed control has relied heavily on preplant incorporated herbicides (Arnold et
al. 1993; Blackshaw et al. 2000; Burnside et al. 1994; Goulden 1976; Robertson and Frazier
1978). Herbicide options available in the past were sometimes inadequate for good weed control

2

in the absence of cultivation (Arnold et al. 1993; Blackshaw and Esau 1991). For a variety of
reasons, dry beans have historically been grown in wide rows, often 71- or 76-cm. A key reason
for this was that inter-row cultivation was needed for weed control and, in some systems, for
hilling the beans in preparation for harvest (Blackshaw et al. 2000; Goulden 1976; Malik et. al.,
1993).

In addition, dry beans were grown in wide rows because growers desired to use

equipment shared with other crops planted in wide rows (Redden et al. 1987; Schwartz et al.
2004).

Moreover, there were concerns about increases in disease, especially white mold

(Sclerotinia sclerotiorum (Lib.) de Bary ), in narrow rows (Blackshaw et al. 2000; Saindon et al.
1995; Schwartz et al. 2004). Most traditional dry bean varieties grown in the United States prior
to the last decade of the twentieth century had a prostrate vine habit (Type III) that was not
amenable to direct harvest by combine (Horn et al. 2000; Kelly et al. 2009a; Schwartz et al.
2004). Thus, they were typically harvested with specialized equipment that cut the plant about 5
cm underground, pulled the cut plant, and placed the plants in a windrow, or that simply pulled
the whole plant and windrowed it; the plants were allowed to dry, often for several days, and
were then harvested with a combine equipped with special grates to prevent damage from rocks
and soil clods (Robertson and Frazier 1978; Schwartz et al. 2004). Type I dry beans, which were
widely grown in the mid-twentieth century, have also proven not to be amenable to direct harvest
methods (Adams 1995).
Type I and Type III dry beans are still grown in the United States. However, beginning
in the 1970s, breeders have been developing more upright Type II dry bean varieties with a
growth habit similar to the natural habit of soybean (Glycine max L.) that are suitable for direct
harvest (Hosfield et al. 2005; Kelly 1994, 2001, 2010; Mooers 1909; Welacky and Park 1987).
Direct harvest is an attractive option because a single pass across the field is more efficient,

3

because yield losses often occur in the windrow system, and because less specialized equipment
is needed (Kelly 2010; Robertson and Frazier 1978; Schwartz et al. 2004). Growers using a
direct harvest system are unlikely to use inter-row cultivation because it brings up rocks and
creates mounds, which interfere with combine operation (Schwartz et al. 2004). Inter-row
cultivation also requires more passes across the field, raising concerns about the possible erosion
and compaction caused by repeated cultivation (Blackshaw et al. 2000). In addition, cultivation
is no longer necessary to hill beans for pulling in a direct harvest system (Robertson and Frazier
1978; Schwartz et al. 2004). If a grower does not cultivate, this opens the possibility of growing
dry beans in narrow rows since a significant inter-row area in which to cultivate is no longer
needed (Blackshaw et al. 2000). An additional benefit of upright varieties is lower disease
pressure due to a less dense canopy and fewer pods touching the ground (Blackshaw et al. 1999;
Kelly and Adams 1987). Since the canopy structure of upright cultivars makes them less
susceptible to white mold, the increased white mold danger associated with narrow rows is
alleviated (Blackshaw et al. 2000; Kelly and Adams 1987; Saindon et al. 1993).
Preplant incorporated herbicide use, the other component of traditional weed
management in dry bean, also involves soil disturbance and often multiple passes across the
field. Moreover, the current availability of a wide range of preemergence and postemergence
herbicides for dry bean in the United States has reduced the need for reliance on preplant
incorporated herbicides (Sprague and Everman 2011). This suggests the need for research on the
overall effectiveness of herbicide combinations, especially in light of research suggesting that
herbicide combinations work better than single herbicides, even at reduced rates (Blackshaw et
al. 2000).

4

Effects of Row Width and Population on Canopy Development
Numerous studies have shown that decreasing row width, and sometimes increasing
planting population, increases canopy development and consequent light interception in dry bean
(Table 1.1) and other crops (Table 1.2). The increase in canopy development may consist in the
greater development of the canopy during certain parts of the season and/or in greater canopy
closure at peak (Duncan 1986). Increased light interception is often associated with benefits to
crop production, especially higher yields and weed suppression (Andrade et al. 2002;
Norsworthy and Oliver 2001; Peters et al. 1965). In navy bean, Urwin et al. (1996) correlated
several measures of canopy closure, including leaf area index and diffuse non-interception of
light, with reduction in late-season weed emergence in one year of a two-year study; the effect
was non-significant in another year when canopy development was more uniform.
Soybean, another leguminous crop harvested for seed, has been extensively studied with
respect to canopy development. Because its growth habit is similar to that of Type II dry bean,
insights derived from soybean research may prove applicable to dry bean. Harder et al. (2007),
using a factorial design of three row widths and four populations, measured light levels above
and below the soybean canopy at intervals through the growing season and used these values to
calculate leaf area index (LAI). Row widths of 19 and 38 cm resulted in greater LAI mid-season
than 76-cm rows, and the canopy reached the critical LAI that indicated 95% canopy closure 1 to
2 weeks sooner at the two narrow row widths than in the wide row width. Also, in the 19- and
-1

38-cm rows only, the high population (445,000 seeds ha ) had a greater LAI than the low
-1

population (124,000 seeds ha ) on most measurement dates and a greater LAI than the next
lowest population on two dates. Dalley et al. (2004) studied the same row widths (19-, 38-, and
76-cm) at a single population and found large differences in light interception between the 76-cm
5

canopy and those of the narrower row widths; on one date, the 76-cm soybeans had only 18%
interception while the 19-cm had 98% interception, and the 76-cm rows never completely closed
the canopy while the 19- and 38-cm rows always did. On some dates, greater canopy closure
was also evident in the 19-cm than in the 38-cm rows. Taylor (1980) destructively analyzed LAI
at four row widths (25-, 50-, 75-, and 100-cm) at a constant population. Although he did not find
significant differences in LAI among row widths, he did visually observe canopy closure twenty
to sixty days earlier in 25-cm rows than in 100-cm rows; he attributed this to a more uniform
distribution of leaf area.

Yelverton and Coble (1991) compared photosynthetically active

radiation (PAR) interception at 23- and 91- cm row widths in soybean and found that the 23-cm
canopy essentially closed by ten weeks after planting while the 91-cm canopy never reached
greater than 75% interception. Similarly, Légère and Schreiber (1989) found that LAI was
greater through much of the season in 25-cm than in 76-cm rows. Quakenbush and Andersen
(1984) also measured PAR interception in 18- and 76-cm rows and found that when 18-cm rows
reached 95% interception, the 76-cm rows were only at 24% and did not catch up for about three
weeks. Norsworthy and Oliver (2001) conducted a soybean population study in narrow (19-cm)
-1

rows using twelve planting populations ranging from 185,000 to 1,432,000 seeds ha . Light
-1

interception increased by an average of 0.022% for each additional thousand seeds ha ;
however, even the lowest population achieved 89% interceptance, suggesting that narrow rows
may enable efficient light interception at a wide range of populations. Using a four-row-width,
four-population factorial design, Weber et al. (1966) found that LAI increased both with
decreasing row width and increasing population and that whole plant dry weight increased along
with it, but they cautioned that after a certain point, increasing dry weight actually began to result

6

in lower seed yields. Nelson and Renner (1989) found that the soybean canopy closed an
average of 35 days sooner in 19-cm than 76-cm rows.
Some non-leguminous crops have also been shown to respond to narrow row widths with
more complete canopy closure in the early part of the growing season. Dalley et al. (2004)
examined 38- and 76- cm row widths at a constant population in corn (Zea mays L.) and found
that narrow-row corn had higher light interception in the early season, though the advantage was
lost by the time of tassel emergence. Alford et al. (2004) planted corn and sugarbeet (Beta
vulgaris L.) in 38-, 56-, and 76-cm rows; both species demonstrated earlier canopy closure as
measured by light interception in 38-cm rows than in 76-cm rows.
A few similar studies have also been conducted in dry bean with similar results. One of
these was conducted in Alberta by Blackshaw et al. (1999); they studied Type II and III navy
beans at populations of 240,000 or 480,000 ha

-1

and row widths of 23-, 46-, and 69-cm.

Photosynthetically active radiation was measured above and below the canopy several times
during the season and expressed as light penetration. In all three years of the study, PAR
penetration was progressively greater with increasing row width, and 69-cm rows still had 2040% penetration at peak canopy coverage with significantly lower penetration at peak coverage
in the narrower row widths. The higher population resulted in slightly lower penetration in one
year, lower penetration at just one date in another year, and no significant difference in the
remaining year. Blackshaw et al. (2000) also conducted a similar three-year study in Type II
small red beans, growing them in 23- or 69-cm rows at populations of 200,000 or 500,000 plants
-1

ha . They found that PAR interception was usually greater at the higher plant density for most
of the growing season. Also, PAR interception was greater in narrow rows than in wide rows at
many measurement dates, especially early in the season. Working in Queensland, Australia,
7

Redden et al. (1987) conducted seven trials over three years, planting Type II navy beans at
112,500 or 337,500 plants ha

-1

and row widths of 7.8-, 35.6-, 71.1-, and 107-cm. Their method

of canopy analysis relied on visual observation of “the proportion of row width occupied by
plants until rows were closed, after which visible bare ground was estimated by looking directly
down over the canopy” at flowering and maturity. When row width had an effect on ground
cover, increasingly narrow rows resulted in greater ground cover, but this effect did not occur at
all sites. Similarly, increased population often resulted in increased ground cover, but not at all
sites. The actual data from this study was not presented, only regression models of the data. In
Brazil, Ziviani et al. (2009) used digital photography and computer analysis to assess canopy
coverage in both Type II and Type III cultivars and found that in both types, 30- and 40-cm rows
resulted in more complete coverage than 50- and 60-cm rows. They also found that at 36 days
after planting, beans planted at 5 plants per meter of row did not provide as complete coverage as
those planted at 10, 15, or 20 plants m

-1

and that by 49 days, the 10 plants m

-1

treatment was

also lagging in coverage. In another Brazilian study, Vieira et al. (2010) varied within-row
-1

density of Type III dry beans between 4 and 16 plants m , while keeping a constant row width
of 50 cm. Canopy closure was estimated visually twice per year. They reported that canopy
closure increased linearly with increasing within-row density at each date, although it does not
appear that the separation between all treatments was significant at the later estimation dates (62
days after emergence in 2000, 79 days after emergence in 2001), and the precision of their
method, “observing each plot from one end (looking down the rows) and visually estimating the
proportion of soil surface visible between the rows,” seems doubtful. Moreover, the range of
populations used was very low. In Michigan, Xu and Pierce (1998) compared leaf area in

8

‘Mayflower’ navy beans in 56- and 71-cm rows and did not find a significant difference;
however, their decision to destructively sample only three plants in each plot for each
measurement time and the lack of an explanation regarding how they measured leaf area make
this result ambiguous, particularly since greater yield was still observed in narrow rows.
At least two similar studies have also been conducted in snap bean, which, though used in
a very different manner than dry bean and generally considered a horticultural crop, also belong
to Phaseolus vulgaris, and these studies brought up aspects of canopy development and plant
spacing not addressed by agronomic researchers. In Maryland, Teasdale and Frank (1983)
planted snap beans in row widths of 15-, 25-, 36-, 46-, and 91- cm at a constant density of
-1

430,000 plants ha

at two locations in five site-years. Canopy coverage was visually estimated

weekly, and at one location, photon flux density was measured above and below the canopy
weekly. Unlike in the dry bean experiments, canopy coverage and light penetration were treated
separately and related to one another. Canopy cover usually increased and light penetration
usually decreased with decreasing row width. Thus, canopy cover was inversely related to light
penetration, but the relationship was not entirely linear. While it was linear between 15 and 46
cm, greater canopy coverage was required in 91-cm rows to achieve the same reduction in light
penetration. This is likely because canopy coverage was being rated visually, so that only the
outer surface of the canopy was taken into consideration. This suggests that lower leaves play a
role in light absorption and that the canopy was thinner toward the middle of the 91-cm rows
even if the same percent coverage was achieved. Teasdale and Frank also took note of the fact
that in narrowing the rows, plants were increasingly equidistant from one another and were
precisely equidistant at the 15-cm spacing and suggest that equidistance may have played a role
in the quicker canopy development in narrow rows. Another snap bean study by Wahab et al.
9

(1986), conducted in Saskatchewan, suggested that yield advantages could be obtained by
planting in an equidistant pattern. While not specifically addressing canopy development, the
study was specifically designed to look at equidistance by comparing four row widths ranging
from 20- to 125-cm at a within-row spacing of 5-cm, four square designs with equal inter- and
intra- row spacings ranging from 10- to 25-cm, and four honeycombed triangle designs in which
all plants were equidistant from their six nearest neighbors at distances ranging from 10- to 25cm. This two-year experiment was conducted with an eye toward labor-intensive production in
resource- and land-poor nations. Andrade et al. (2002) suggest that some of the advantage
gained by means of equidistant spacing may be a result of more efficient use of water and
nutrients, which would suggest that equidistant spacing improves the development of what might
be termed the “below-ground canopy” - the roots - as well as the above-ground canopy.
However, possibly due to the difficulty of studying roots, this possibility has been little explored
in the literature.

Effects of Planting Practices and Canopy Development on Yield
The principal benefit that growers hope to derive from an increase in canopy closure is an
increase in yield. If decreasing row width or increasing population results in quicker or more
complete canopy closure, more light will be intercepted and thus more energy should be
available to the crop to produce seeds. If the “below-ground canopy” also develops more
efficiently, the crop will have more available water and nutrients, and this may also increase
yield. The other main benefit of greater canopy closure is weed suppression due to reduced light
below the canopy; reduced weed competition could also indirectly result in increased yield by
reducing the amount of light, water, and nutrients made unavailable to the crop by weeds.

10

A large volume of work indicates that narrow rows result in higher yields in soybean and
that planting population should be reasonably high but can safely be varied within wide limits,
although population has also sometimes been varied along with row width, possibly conflating
the effects of the two in certain studies. The principles behind these effects may also apply to
dry beans, especially those with growth habits similar to soybean.
A number of these studies have come in the context of herbicide research, although these have
tended to be somewhat ambiguous in their results. As early as 1965, Peters et al. compared
soybean yield in 51- and 102-cm rows with different cultivation and herbicide program. In two
of three years, 51-cm rows resulted in higher yield than 102-cm rows by 11.1 and 13.8 bu ha

-1

but were approximately equal in the third year. They then compared drilled soybeans in 20-, 41-,
61-, 81-, and 102-cm row widths over three years, using lower populations for wider row widths.
They concluded that widths of 20 to 61 cm resulted in the highest yields in good growing
conditions, but the conclusion was questionable as only a few favorable results were chosen from
the data, much of which was inconclusive. Nelson and Renner (1998), in evaluating the potential
of several acetolactate synthase inhibitor (ALS) herbicides for weed control in soybean, used
-1

both 19- and 76-cm rows, lowering the population by over 100,000 seeds ha in 76-cm rows, at
each of two sites. Yield was greater across both sites in 19-cm rows in all treatments, though it
was only significantly greater in five of the seven treatments; significant differences ranged from
-1

300 to 800 kg ha . Young et al. (2001), in the course of evaluating the effectiveness of
glyphosate with various tank mix partners in glyphosate-resistant soybean, varied soybean row
width using 19-, 38-, and 76-cm widths at three sites over three years. They intended to look at
the effect of row width but in some cases also varied the planting population either between row

11

widths, reducing the population in wider rows. In three site-years, yield was significantly greater
in 19- and 38-cm rows, although in two of these years, population was varied between all row
widths as well. In one site-year, 76-cm rows actually out-yielded 19-cm rows, though the
authors believed that this was due to poor stand in the 19-cm rows. In an evaluation of herbicide
combinations in glufosinate-resistant soybean, Norris et al. (2002) planted soybeans at a constant
population of 480,000 seeds ha

-1

in 38- and 76-cm rows at two locations over two years. Their

results were more equivocal. At one location, no yield differences due to row width were
observed, which the authors attributed to extremely dry conditions in the mid- and late growing
season. At the other site, there were significant differences, and whether narrow row width
resulted in higher yield varied by treatment and year, though the authors do not make it clear
which differences were significant. Wax and Pendleton (1968) used 21-, 51-, 76-, and 102-cm
rows but also reduced the planting populations with increasing row width to the point that 102cm rows had half the population of 21-cm rows. They reported increases of soybean yield
compared with the 102-cm width of 10, 18, and 20% with each decrease in width and concurrent
population increase.
However, a number of other papers have offered clearer data. Légère and Schreiber
(1989) reported that soybean yield was an average of 590 kg ha
rows at a constant population of 39 plants m

-2

-1

higher in 25- than in 76-cm

and that this effect was consistent across all three

years of the study. Lehman and Lambert (1960) compared 51- and 102-cm rows across four
planting densities in two locations and found that at one location, the 51-cm rows had
consistently higher yields across planting densities while at the other location there were not
significant differences; there were no yield differences across planting densities. In an usually

12

lengthy series of studies, Cooper (1977) carried out a six year succession of trials comparing 17-,
50-, and 75-cm rows in three populations in the first year. In 50- and 75-cm rows, there were no
significant differences between populations, while in the 17-cm rows, the high population
-1

(375,000 plants ha ) was optimum and so was adopted as a standard population for the
remainder of the trials. This series of trials, using many different soybean varieties, established
that, assuming good weed control, a yield advantage of 10-20% was usually obtained by 17-cm
rows while early maturing varieties often obtained a 30-40% advantage.

Taylor (1980)

conducted a three year experiment using 25-, 50-, 75-, and 100-cm row widths at a constant
-1

160,000 plant ha

density and experienced one good growing season and two dry years. He

reported that in the year with favorable growing conditions, 25-cm rows resulted in 17% greater
yields than 100-cm rows with the middle row widths having intermediate yields. However, he
also reported no significant differences between row widths in the two dry years, with wide-row
plants outstripping narrow-row plants in leaf area, height, and pod set, allowing them to equal
narrow row yield.
Several studies have treated soybean row width and population (or within-row spacing)
as a factorial design, using every combination of several row widths and planting populations.
An early example of this approach was a 1939 paper by R.G. Wiggans, who used five row
widths ranging from 20 cm to 81 cm, paired with seven set within-row seed spacings ranging
from 1 to 15 cm over a four year period. He discovered a remarkably consistent pattern of yield
increasing with decreasing row width in all years. He also noted a very general trend toward
higher yields at closer within-row spacings but found little of significance up to 8-cm; spacings
above 8-cm began to reduce yield significantly. When Wiggans converted his combinations to
plants per area, he found little benefit in increasing population beyond 650,000 plants ha
13

-1

and

-1

serious yield consequences to populations below 330,000 plants ha . Ethredge et al. (1989)
-1

compared 25-, 51-, and 76-cm rows in soybeans at 260,200, 390,400, and 520,400 plants ha .
In spite of the wide range of populations used, no significant yield differences were found among
populations. Two cultivars were utilized, and the effects of row width within each cultivar were
examined across populations and both years of the study; in one cultivar, the 25-cm rows
outyielded the 76-cm rows by 15% with the 51-cm rows intermediate, while in the other cultivar,
25- and 51-cm rows yielded similarly and outyielded 76-cm rows by 5-7%. Kratochvil et al.
(2004) used standard planting populations for full season and double-crop soybeans in Maryland
(432,250 seeds ha

-1

and 555,750 seeds ha

-1

respectively) along with rates 40% and 20% lower

and 20% higher than standard rates. They planted in 19- and 38-cm rows for three years, using
four cultivars each year in two locations both single and double-cropped for a total of 48
individual cultivar-year-system comparisons. In two years, 19-cm rows averaged significantly
higher yield than 38-cm rows, and while this did not extend to every comparison, in no
comparison in these years did 38-cm rows yield significantly higher. In the other year, there was
no significant yield difference overall between the row widths, but in only one individual
comparison did 38-cm rows yield higher than 19-cm rows. In two years, the 40% reduced rates
were found to reduce yields while the 20% reduced rates did not. The 20% increased rate
increased yield of early maturing cultivars in one year but otherwise had similar yield to the
standard rate. In the third year, under drought conditions, the standard rate and the higher rate
yielded better than the reduced rates at one location but not the other. Harder et al. (2007)
planted soybean in 19-, 38-, and 76-cm rows at five populations ranging from 124,000 to
-1

445,000 plants ha . They did this in three locations over two years in both weedy and weed-free

14

conditions. They found no yield advantage of 445,000 over 296,000 plants ha
yield typically began to drop at 198,000 plants ha

-1

-1

but did find that

and was always less than optimum at

-1

124,000 plants ha . At all populations except the lowest, 19-cm rows yielded higher than 76-cm
rows in both weed control scenarios, and in all but one population-weed control combination, 38cm rows yielded higher than 76-cm rows. In only one case did 19-cm rows yield higher than 38cm rows.
J.K. Norsworthy has also conducted studies in soybean focusing only on the question of
population. Norsworthy and Frederick (2002) noticed that the recommended seeding rate for
-1

narrow row (<76 cm) soybeans (624,000 seeds ha ) was double those for wide row soybeans
-1

(370,000 seeds ha ) in South Carolina and suspected that this was unnecessary. They drilled
narrow-row soybeans of four cultivars at both recommended rates in two years and concluded
that the lower population recommended for wide rows was sufficient to produce yields equal to
the higher population in narrow rows. Earlier, Norsworthy and Oliver (2001) had used twelve
seeding rates ranging from 185,000 to 1,482,000 seeds ha

-1

of the years at two sites. They reported 988,000 seeds ha

-1

in 19-cm rows for two years, in one
to be the optimum seeding rate but

also decided that this was not economically viable. They also found that at one site-year which
experienced very good growing conditions, yield did not vary by seeding rate over the entire
spectrum, suggesting a remarkable ability of narrow-row soybeans to compensate for low
population under ideal conditions.
Studies in non-leguminous crops also suggest the possibility of increasing yields in
narrow rows. For examples, Yonts and Smith (1997) grew sugarbeets in 35-, 56-, 76-, and 9715

cm rows and found that sugar yield was highest in the two lower row widths, and Winter (1989)
compared 76- and 102-cm rows and found sugar quality to be higher in 76-cm rows regardless of
population or irrigation.
However, the work of Sankula et al. (2001) in lima bean (Phaseolus lunatus L.), a close
relative of dry bean, suggested that narrowing row widths may not always increase yield. They
planted lima beans in Maryland and Delaware at a constant population of 175,000 plants ha

-1

at

two sites over two years, with the exception of one site-year in which the narrow rows were
-1

planted to 350,000 plants ha . The two row widths used were 38-cm and 76-cm. In this study,
yield was higher in narrow rows in only in one of the four site-years.
Some work has been conducted regarding similar issues in Phaseolus vulgaris itself, and
the evidence points toward higher yield in narrow rows. However, many of these studies were
conducted under irrigation, which may allow greater productivity in many years than dryland
production and may reduce intraspecific competition, allowing plants to be successfully spaced
more closely. Population effects seem less consistent, and Crothers and Westerman (1976)
found that determinate (i.e., Type I) cultivars tended to response positively to increased
population while indeterminate (i.e., Type II and III) cultivars tend have similar yield over a high
population range due to their greater ability to compensate and fill in gaps, though this work was
carried out in snap beans (also Phaseolus vulgaris L.) rather than dry beans.
Many of these studies have been conducted in Canada rather than in the United States.
Park (1993) worked with dry bean in Ontario but did not specify whether the beans were
irrigated or what class they belonged to, though at least some of the seven varieties used were
navy beans. Both Type I and Type II varieties were planted for two years. Park used 30-, 60-,
and 80-cm row widths with population adjusted up as row width decreased. (The populations
16

-1

stated are 50,000, 33,333, and 26,667 seeds ha , but it seems likely that this is a misstatement
and that the actual populations were ten times that since the stated populations are about 10% of
those used by other studies.)

Average yield increased as row width decreased (with the
-1

associated increase in population) from 2290 to 2882 to 3308 kg ha ; this trend occurred in all
cultivars. Malik et al. (1989) using a factorial design, planted white beans (i.e., navy beans) of
Types I, II, and III in Ontario without irrigation in 23-, 46-, and 69-cm rows combined with
-1

seeding rates of 225,000 and 375,000 seeds ha in both weedy and weed-free conditions. While
beans grown in weed-free conditions exhibited no yield differences by row width, those grown in
weedy conditions exhibited significantly lower yield in 69-cm rows than in 46- or 23-cm rows.
Malik et al. (1990) continued their experiment a second year and reported that in that year, yield
was higher in 46- or 23-cm rows in both the weedy and weed-free plots. In their final paper,
Malik et al. (1993) reported that in both years in 46-cm rows and in one year in 23-cm rows, the
higher planting population resulted in 12-16% higher yields in the weedy plots but in no increase
in yield in the weed-free plots. In Alberta, Blackshaw et al. (1999) also planted both upright and
viny navy beans in 23-, 46-, and 69-cm rows at populations of 240,000 or 480,000 plants ha

-1

under irrigation. This was repeated three years, and in each year, yield was 27-41% higher at the
higher population, and when the high population was used, yield was increased an average of
35% in 23-cm rows compared to 69-cm rows. The narrow row widths also resulted in higher
yields at the low population in one year. While beans were grown both in the presence and in the
absence of hairy nightshade (Solanum sarrachoides Sendtn.), the effects occurred regardless of
this factor except one year when row width was non-significant in the weedy treatment. These
effects occurred regardless of cultivar. Blackshaw et al. (2000) also conducted a similar study in
17

semi-upright small red beans under irrigation in Alberta over three years. They planted at row
-1

widths of 23- and 69-cm and used plant densities of 200,000 or 500,000 plants ha . When
maintained weed-free, the narrow row width had a 19% yield advantage, and the high population
had a 17% advantage; the highest yields in the study were at high population and narrow rows.
However, in one year high population only increased yield in wide rows. In the presence of
weeds, narrow rows and high populations still increased yield except in one year when high
population only increased yield in narrow rows; weedy yields were always less than optimum
weed-free yields.
Similar studies were conducted several decades ago in Australia and New Zealand and
came to similar conclusions. Goulden (1976), in Canterbury on New Zealand’s South Island,
grew Type I ‘Sanilac’ navy beans in 20- and 40-cm rows and within-row spacings of 4.8-, 7.1-,
and 10.2-cm under irrigation. Yield was 57% greater in 20-cm rows than in 40-cm rows. In 20cm rows, yield was greater at the 4.8- and 7.1-cm within-row spacings than in the 10.2-cm
spacing, though within-row spacing had no effect on yield in 40-cm rows. In the Australian state
of Queensland, Redden et al. (1987) planted one of two varieties of Type II navy bean at least
once at six separate sites over three years for a total of eight site-years; some locations were
irrigated while others were not. They were planted at row widths of 17.8-, 35.6-, 71.1-, and 107-1

cm and densities of 112,500 and 337,500 plants ha . In all but one site-year (at which no yield
response was observed), yields decreased with widening rows, especially at low population; a
sometimes lesser response was observed at high population. In two site-years, there was no yield
response to plant density, while at other sites yield was highest at the greater density in at least
some row widths.

18

Few such studies have been conducted in the United States. In North Dakota, Grafton et
al. (1988) planted upright navy beans and vining pinto beans in 19-, 38-, and 76-cm rows and
four populations for each class (99,000 to 321,000 plants ha

-1

for pinto bean, 148,000 to 371,000

-1

plants ha for navy bean) for two years at two sites. Yield decreased with increasing row width,
with the decrease per 25-cm between rows being 565 kg ha

-1

-1

for pinto beans and 435 kg ha for

navy beans. Yield increased with increasing population for navy beans but did not do so for
pintos. In the course of trying to examine the possibility of no-till or ridge-till production of dry
beans, Xu and Pierce (1998), working in Michigan, used both 56- and 71-cm rows for
-1

‘Mayflower’ navy beans for three years. They found yield to be 300 to 600 kg ha higher in 56cm rows than in 71-cm rows. Intended planting population was not specified, and final planting
population was only reported in one year.
Other studies have been conducted in the United States and Canada on snap bean (also
Phaseolus vulgaris). Teasdale and Frank (1983) planted snap beans at row widths of 15-, 25-,
36-, 46-, and 91- cm at a constant density of 430,000 plants ha

-1

for three years at multiple sites

in Maryland. All row widths produced similar yields except the 91-cm width, which yielded less
than the other row widths in four of the five site-years. In Oregon, Peachey et al. (2006) planted
-1

snap beans in 19-, 38-, 75-, 114-, and 150-cm rows in two years at a constant 445,000 seeds ha .
This was a white mold trial, and when fungicide was applied, 19-cm rows produced the highest
yield in one year, while no significant differences were seen among the other row widths. When
fungicide was not applied that year and regardless of fungicide in the following year, 150-cm
rows produced the lowest yield with no significant differences between the other widths. Every

19

-1

significant difference was at least 5000 kg ha . As described in the last section, Wahab et al.
(1986) planted snap beans in Saskatchewan, Canada in three spatial arrangements – rows,
squares, and triangles, in order of increasing equidistance – and four spacings within each
arrangement, observing higher yields in increasingly equidistant designs.

Yield tended to

increase with decreasing row width or plant spacing. Square and triangle designs generally had a
yield advantage over traditional rows, and in one year, the triangular design had an advantage
over the square design. Moore (1991), working with soybeans in Louisiana, added a little
evidence to the idea that equidistant planting increases yields, finding that equidistant within-row
spacing increased yields compared with non-equidistant spacing by 7% and 8% in sequential
years, though neither result was significant at α=0.05.
Some dry bean studies have been conducted in Brazil, and most have had rather different
results than those conducted elsewhere, but this is likely due to the warm, humid nature of
Brazil’s climate, which probably causes high disease pressure when plants are spaced densely,
and also likely increases plant growth compared with beans in cooler climates, thereby
increasing intraspecific competition. These results may have little application to relatively cool
growing regions in the northern United States. The study that would seem more likely to be
climatologically relevant to the northern United States was that conducted by Horn et al. (2000)
in Rio Grande do Sul, the most temperate of Brazil’s states. It was not clear whether the study
was conducted under irrigated conditions. Upright black beans were planted in row widths of
25-, 50-, and 75-cm and paired with populations of 100,000, 200,000, 350,000, and 500,000
-1

plants ha . This study found no yield differences between populations but, contrary to most
studies, found a linear increase in yield as row width increased from 977 kg ha

20

-1

at 25 cm to

1132 kg ha

-1

at 75-cm. In the extremely hot and humid state of Tocantins, Aidar et al. (2001)

decided to conduct row width research due to the low yields being obtained in the region by
growers trying to grow dry bean, a recently introduced crop in Tocantins, and to the observation
that yields were highest in border rows. They used row widths of 45-, 60-, and 75- cm and
-1

within-row densities of 7-, 10-, 13-, and 16-plants m of row and planted Type III carioca beans.
It was determined that row width was the major factor controlling yield and that 60 cm was the
ideal row width for Tocantins, with yield declining again at 75 cm; most growers had been using
45-cm rows. They also determined that yields were highest at about 13 plants m

-1

of row. In

Mato Grosso do Sul, also a state characterized by warm, humid weather, Arf et al. (1996) planted
‘Ouro’ Type III dry bean using within-row densities of 8, 12, and 16 plants m

-1

of row and a

double-row system in which rows were paired with a small space on one side of each row and a
larger space on the other side. They used row widths of 20/80-, 30/80-, 20/70-, and 30/70- cm,
as well as using constant row widths of 60- and 50- cm. They found that yield increased with
increasing in-row density but also, surprisingly, that in no row width combination was yield
significantly different from any other. Perhaps the most interesting Brazilian result was that of
Ziviani et al. (2009), who compared a prostrate variety of carioca bean with an upright variety in
30-, 40-, 50-, and 60-cm rows and 5, 10, 15, and 20 plants m

-1

of row in the tropical highland

state of Goias. They were concerned that upright varieties were being unjustly rejected as lowyielding due to being grown at low densities and wide row widths and found that, indeed, the
upright variety demonstrated the highest yields at 30- and 40-cm and at 20 plants m
while the prostrate variety yielded similarly across all treatments.

21

-1

of row

Effects of Planting Practices and Canopy Development on Weed Suppression
Given the strong evidence that narrow rows in many circumstances and high planting
populations in some circumstances can improve canopy development and increase yield, it
would be expected that they also improve weed suppression by allowing less light to penetrate
the canopy and be utilized by emerging weeds and by making the crop more competitive for
water and nutrients.
A number of studies have addressed the effect of planting practices on weed populations
in soybean, so once again the information from these studies may give an indication of what to
expect from dry bean. Harder et al. (2007) observed weed density and biomass following
glyphosate application and in a weedy control using three row widths and four populations.
They had found that 19-cm and 38-cm rows increased canopy closure compared with 76-cm
rows. Weed populations were counted and weed biomass measured weekly two through five
weeks following glyphosate application; 19-cm rows had lower weed density and biomass across
all timings than 76-cm rows, and 38-cm rows had lower weed density and biomass than 76-cm
rows in weeks three through five. Thus, narrow rows suppressed weed resurgence following
herbicide application. Although high populations tended to improve canopy closure in narrow
rows, weed density and biomass following glyphosate application did not respond to planting
population.

In the weedy control, weed density was unaffected by row width or planting

population. Weed biomass was higher in the lowest population than in the highest but was
otherwise unaffected by population. At the two lower populations, (124,000 and about 190,000
-1

plants ha ), row width also had no effect on weed biomass. However, at the moderate and high
-1

populations (about 300,000 and 445,000 plants ha ), weed biomass was lower in 19-cm rows

22

than in 38- and 76-cm rows. There was not a clear correlation between weed suppression and
yield.
Wax and Pendleton (1968) planted soybeans in 10-, 20-, 30-, and 40-cm rows but greatly
reduced population with increasing row width; they noted that with decreasing row width (and
increasing population), grass control improved, and that biomass of some, but not all, broadleaf
weeds was reduced at 10- and 20-cm compared with the wider spacings. Peters et al. (1965)
planted soybeans in 20-, 41-, 61-, 81-, and 102-cm row widths and found that the three lower
row widths reduced the need for cultivation for good weed control; unfortunately, the three
narrower widths were also at the same high population while population was lowered for the
wider widths, possibly calling into question which was the effective cause.
Nelson and Renner (1998) found that planting soybeans in 19-cm rows resulted in lower
weed biomass in most herbicide combinations than planting in 76-cm rows, especially when an
herbicide treatment resulted in only partial weed control.

This advantage coincided with

observed yield advantages in 19-cm rows in all treatments except the weed-free treatment and
the most effective herbicide combination. It also agreed with their observation of faster canopy
closure in 19-cm rows. Yelverton and Coble (1991) planted soybeans in 91-, 46-, and 23-cm
rows in two years. Herbicides were applied and weed resurgence measured in terms of weed
density. Resurgence increased linearly with increasing row width, which agrees with their
observation that canopy cover was much greater in 23-cm rows than in 91-cm rows.
Some researchers have looked at the control of specific weeds as affected by row width.
Légère and Schreiber (1989) intentionally sowed redroot pigweed in soybean in 25- and 76-cm
rows.

They observed canopy geometry to determine whether soybean and pigweed were

competing and measured leaf area and biomass of both the crop and the weed. They found that

23

pigweed was competing with soybean for light even though soybean always produced more leaf
area. Pigweed was 43% of biomass and 29% of leaf area on average in 76-cm rows but biomass
was reduced to 24% and leaf area to 15% in 25-cm rows. This was consistent with their findings
that narrow rows increased canopy closure and yield. Young et al. (2001) examined the control
of velvetleaf (Abutilon theophrasti Medik.), giant foxtail (Setaria faberi Herrm.), and common
waterhemp (Amaranthus rudis Sauer) in 19-, 38-, and 76-cm rows in soybean using a variety of
herbicide combinations in eight site-years. At some sites, within some herbicide treatments,
giant foxtail control was improved in 19-cm rows compared with 76-cm rows, but control was
good with most herbicide combinations, regardless of row width. However, common waterhemp
control was better across all treatments and years in 19- or 38-cm rows than in 76-cm rows, and
velvetleaf control was better across all treatments in 19- and 38-cm rows in four site-years. This
was consistent with their reports of higher yields in 19- and 38-cm rows in some site-years.
Norris et al. (2002) planted 38- and 76-cm soybean rows at a constant population and used a
variety of herbicide combinations on fields with high populations of barnyardgrass (Echinoloa
crus-galli Beauv.), hemp sesbania (Sesbania herbacea Mill. (McVaugh)), pitted morningglory
(Ipomoea lacunose L.), large crabgrass (Digitaria sanguinalis L. (Scop.)) and sicklepod (Senna
obtusifolia (L.) H.S. Irwin and Barneby). In all weeds and nearly all herbicide treatments,
control was improved, often substantially, in 38-cm rows; in some cases, 100% control was
achieved in 38-cm rows. However, there was no clear trend with regard to yield, which the
authors attribute to drought and poor stand establishment.
Chandler et al. (2001) looked at weed seed return in addition to weed biomass. They
planted soybean in 76- and 38-cm rows as well as in twin rows 19 cm apart with each pair 76 cm
from the next. Seed return was estimated by collecting seeds on or just under the soil surface

24

2

within two 400 cm quadrats of each plot. While weed biomass was only reduced in the 38-cm
rows, yield was higher in both the 38-cm and twin-row treatments. Weed seed return was also
diminished in both 38-cm and twin row treatments.
In addition to the research that has been done in soybean, a few studies have also been
conducted relating planting patterns and growth habit to weed control in dry bean. Malik et al.
(1993) planted three navy bean varieties of Types I, II, and III in 69-, 46-, and 23-cm rows and
-1

populations of 250,000 or 375,000 plants ha , though only at 250,000 for the widest row width.
This research was conducted in two years in Ontario. Weed biomass was estimated by handharvesting six square meters of each plot several times over the growing season.

The

indeterminate vines (Types II and III) were found to result in less weed biomass than the Type I
determinate bush variety, probably due to the fact that determinate cultivars stop growing after
flowering while indeterminate cultivars, like weeds, continue growing. While weed biomass
before flowering was similar across row widths, and population did not affect weed biomass,
weed biomass after flowering was reduced in 23- or 46-cm rows compared with 69-cm rows;
reductions in weed biomass ranged from 15-21% in both years. A negative correlation was
observed between weed biomass and yield, with 1 kg ha
about 0.38 kg ha

-1

-1

of weed biomass corresponding to

yield loss. Uncontrolled weed populations reduced yield as much as 70%.

Blackshaw et al. (2000) in Alberta planted semi-upright small red beans for three years in 23-1

and 69-cm rows at populations of 200,000 and 500,000 plants ha . Weed control was evaluated
visually. Narrow (23-cm) rows did not improve weed control over that in wide (69-cm) rows at
200,000 plants ha

-1

-1

but did improve weed control at 500,000 plants ha . In 69-cm rows, the

25

effect of population was inconsistent, but in 23-cm rows, high population resulted in improved
weed control. Thus, weed control was maximized at high population and narrow row width.
This strongly correlated with observed increases in yield and also agreed with observations of
improved interception of photosynthetically-active radiation in narrow rows and at high
populations.
Blackshaw et al. (1999) looked specifically at the interaction between navy beans and
hairy nightshade (Solanum sarrachoides Sendtn.) over three years. Both an upright Type II
cultivar and a prostrate Type III cultivar were planted in 23-, 46-, and 69-cm rows at 240,000 and
-1

480,000 plants ha . There were few treatment differences in hairy nightshade biomass at three
or six weeks after planting, but nine weeks after planting, biomass was reduced with decreasing
row width in all three years, and the higher planting density decreased weed biomass in two of
three years. Little difference was noted in weed suppression between cultivars. In all years,
weedy plots had much lower yields than weed-free plots, demonstrating that hairy nightshade
competition affected yield, and hairy nightshade was also found to reduce light available to the
canopy. While yield increases often occurred in narrow rows in the weed-free plots, they always
occurred in the weedy plots, suggesting that weed suppression caused by narrow rows may
sometimes translate into yield benefits.
While Urwin et al. (1996) in Nebraska did not specifically look at planting practices, they
did compare canopy closure in twenty dry bean varieties of different growth forms and relate it
to late-season weed emergence. The cultivars were a mix of Types I, II, and III and of many
market classes. Canopy closure was measured in terms of diffuse non-interceptance of light, leaf
area index, plant canopy volume, and projected canopy cover. Plots were cultivated and kept
weed-free through the last cultivation, and late emerging weeds were counted. In one unusually
26

cool year, correlation was observed between weed emergence and many measures of canopy
closure, along with a general trend toward better closure in more viny cultivars. However, the
following year, growing conditions were better, making canopy closure more uniform and
largely eliminating the correlation.
In addition, the Teasdale and Frank (1983) study in snap bean addressed the effect of row
spacing on weed competition. They found that row widths of 15-, 25-, and 36- cm reduced weed
biomass an average of 18% compared with 91-cm rows when weeds were allowed to emerge
with the crop and that they had 82% less biomass than 91-cm rows when weeds were controlled
for the first half of the season. This was consistent with their finding that narrow rows increased
canopy closure. Yield increases were also observed in narrow rows but extended to 46-cm rows
in which weed control benefits were inconsistently observed.

Disease Considerations with Regard to Planting Practices and Canopy Development
Despite the studies suggesting benefits to planting in narrow rows at moderate to high
populations, concern has been expressed that planting beans more densely may increase disease
pressure, especially white mold, due to reduced air flow through a thicker canopy. However,
studies have shown that upright varieties are better at avoiding white mold infection (Kelly et al.
2009a; Saindon et al. 1995). Kelly et al. (2010) found that ‘Santa Fe,’ an upright pinto bean
variety, had higher yield and lower levels of infection under intentional white mold inoculation
than a traditional prostrate variety. Park (1993) in Ontario found in one year of a two year study
that white mold incidence was 34% greater in 30-cm rows than in 60-cm rows and 56% greater
than in 80-cm rows, although in a year of lower white mold pressure there were no significant
differences between row widths. However, three Type I bush varieties and four upright vine
Type II varieties were planted, and both years, the disease severity was found to be much lower,
27

regardless of row width, in the Type II varieties for which narrow row recommendations are
primarily being developed. Park, along with Saindon et al. (1995), suggested that the erect,
narrow canopy allows better airflow and less plant contact with the ground than other types.
White mold incidence was positively correlated with lodging in one year, suggesting that podsoil contact is a significant factor in initial infection, and Type I beans exhibited more lodging,
especially in the sense of lower branches coming into contact with the ground. Furthermore,
despite the higher white mold incidence in narrow rows, yield continued to increase with
decreasing row width even in Type I cultivars, suggesting that the increase in white mold
infection was not economically significant.

In addition to narrow rows, dense within-row

spacing (and therefore high population) has also been implicated in increasing white mold.
Vieira et al. (2010) in Minas Gerais state, Brazil, planted the Type III Pérola cultivar at 15, 7.5,
or 5 seeds m

-1

of row in constant 50-cm rows in one year and at 16, 12, 8 and 4 seeds m

-1

in a

second year. White mold incidence increased linearly with increasing within-row density in the
first year, but fungicide application was found to be much more effective than planting at a low
density. Disease severity index increased with increasing density, and yield was also reduced
with increasing density. In the second year, disease severity index, disease incidence, and
sclerotia weight all increased linearly with increasing within-row density, while yield was not
significantly affected. However, it should be remembered that this research was conducted in
Brazil and may not be applicable to the climate of the northern United States.
Saindon et al. (1993) in Alberta had more encouraging results with regard to white mold.
They planted Type II navy beans, along with Type III small red beans as a viny control. In the
first year, they used a factorial design with row widths of 30-, 45-, and 60-cm combined with
within-row spacings of 4.0-5.0-, 5.5-6.9-, and 7.0-9.5- cm under irrigation. The following year,
28

they dropped the 4.0-5.0-cm spacing was dropped and replaced with a 10.0-13.0- cm spacing,
and the viny control was established only at 7.0-9.5-cm. Plots were intentionally inoculated with
white mold, yet the disease difference between the viny control and the upright navy beans was
large. In the first year, 73% of viny plants were infected, and 45.6% were killed while only
18.5% of upright plants were infected, and almost none were killed. In the second year, 45.8%
of viny plants were infected and 11.2% killed while the upright plants were not infected by white
mold at all. This suggests that the adoption of upright cultivars reduces white mold infection,
independent of row width or population. Furthermore, row width did not affect white mold
infection in upright cultivars, and within-row spacing did not affect white mold development at
all. Moreover, in both years, yield increased linearly with decreasing row width in the absence
of white mold and did so even in the presence of white mold in one year. Saindon et al. (1995)
went on to conduct another four-year study specifically on planting density, using populations of
250,000, 350,000, 500,000, and 600,000 plants ha-1 in 23-cm rows. Two upright navy bean
cultivars, an upright dark red kidney bean cultivar, and an upright black bean cultivar were
grown in addition to a viny small red cultivar. The field was intentionally inoculated and
irrigated. In this study, planting density did affect white mold infection in all but one year in
which white mold was particularly severe; in general, higher planting densities led some
cultivars, including upright cultivars, to develop more severe disease. However, disease severity
remained low in all upright cultivars except in the year with the highest white mold pressure, and
in every year, the viny cultivar had the highest rates of infection and death at every planting
density. Moreover, yield of upright cultivars increased with higher planting densities in three of
the four years.

29

In Oregon, Peachey et al. (2006) also looked at white mold infection and row width,
though in snap beans rather than dry beans. White mold infestation in the field was naturally
-1

occurring. Row widths used were 19-, 38-, 75-, 114-, and 150- cm at 445,000 seeds ha . The
authors found that white mold severity decreased 3-5% for each 10 cm increase in row width in
the absence of fungicide but that disease levels in the absence of fungicide were still
unacceptable even at the widest row width since snap beans are grown for direct human
consumption, and diseased pods are not considered acceptable. When fungicide was applied,
yield was highest in 19-cm rows, but in the absence of fungicide, it was highest in either 38-cm
or 75-cm rows, depending on year, suggesting that row width-aggravated white mold can cause
yield losses in snap bean. However, the growth form of the snap bean cultivar used was not
stated and may not resemble Type II dry beans.
While concern about row width-aggravated disease has centered on white mold, one
paper by Conner et al. (2006) in Manitoba investigated the influence of growth habit and row
width on bean anthracnose (Colletotrichum lindemuthianum (Sacc. & Magnus) Briosi & Cavara)
development.

Concern regarding this disease may be lower in part because sources of

anthracnose resistance have been identified and have begun to be bred into new varieties.
However, new varieties are not necessarily anthracnose-resistant and even those that are may not
be resistant to all races (Hosfield et al. 2004; Hosfield et al. 2005; Kelly et al. 2009b). Conner et
al. used four navy and two pinto bean varieties at two sites in two years; three of the six varieties
were viny Type III beans, two were upright Type II, and one was a bushy Type I. In some plots
seeds were intentionally infected with anthracnose while others were left uninfected.
Anthracnose infection was visually assessed in the canopy twice, and percentage of pod tissue
covered in lesions was assessed before harvest. Row widths of 30- and 60-cm were used. Pinto
30

beans were shown to be more susceptible than navy beans. Neither row width nor growth habit
had any consistent effect on bean anthracnose development, but 30-cm rows resulted in
significantly higher yield in three of four site-years.

Effects of Planting Practices on Yield Components and Morphological Characteristics
Underlying the effects of planting patterns on canopy closure, yield, weed control, and
disease must be some changes in basic morphological or physiological realities. A handful of
studies have attempted to elucidate these realities. In bean crops, the components of per-plant
yield are seed weight, number of seeds per pod, and number of pods per plant (Grafton et al.
1988; Lehman and Lambert 1960).

Branching patterns and the relative allocation of

reproductive resources between branches and the main stem may also change in response to
planting patterns (Norsworthy and Shipe 2004). Three important characteristics not directly
related to but potentially affecting yield are height, lodging, and maturity rate (Lee et al. 1996).
Several papers have examined these traits in soybean, a crop in which yield is determined
by the same components as dry bean yield and in which basic morphology is similar. Pederson
and Lauer (2004) noted that soybeans cannot usually be bred to enhance a specific yield
component because yield-determining factors are largely outside the reproductive parts, and the
plant tends to respond to environmental conditions by increasing and decreasing yield
components proportionally to produce the same final yield.
Weber et al. (1966) planted soybeans in 13-, 25-, 51-, and 102-cm rows at populations of
-1

64,200, 128,500, 257,000, and 516,400 plants ha . Yield was maximized at 128,500 plants ha

-

1

, and it was found that in the higher populations, especially the highest, plants tended to be

taller, to lodge more, to mature later, and to set fewer pods per plant, all of which were
31

suggestive of intra-specific competition at high density. Row width had little effect on the
characteristics that do not directly affect yield, but number of pods per plant was maximized in
-1

25-cm rows and at 64,200 plants ha ; it was reduced with increasing population, presumably
because soybeans compensate for lower populations by growing larger, enabling them to
produce more seed. Number of seeds per pod was maximized in 51-cm rows and intermediate
populations while seed weight was independent of row width and also maximized in
intermediate populations. Ethredge et al. (1989) grew two varieties of determinate soybeans in
-1

25-, 51-, and 76-cm rows in populations of 260,200, 390,000, and 520,400 plants ha . In one
variety, seed weight was lower in 25-cm rows, but this was apparently the result of greater seed
numbers such that yield was inversely related to seed weight. Lehman and Lambert (1960)
planted two varieties of soybean in 51- and 102-cm rows and four within-row spacings. They
found that the number of seeds per plant declined steeply with decreasing within-row spacing
and that the 102-cm rows produced more seeds per plant than the 51-cm rows. Number of seeds
per pod was also greatest in 102-cm rows, and number of seeds per pod declined along with
number of seeds per plant at lower within-row spacings. Seed weight was greater in 51-cm rows
at one site, and at that site yield was also greater in 51-cm rows; at the other site, seed weight
was greater in 102-cm rows, and yield was not significantly different between row widths,
though it seems that 102-cm rows ought to have demonstrated higher yields than 51-cm at that
site, having the advantage in both seed number and seed weight.
Board (1987) planted eight determinate soybean cultivars in order to compare yields and
yield components and determine what yield components were most tightly correlated to yield.
He found that pods per plant as well as seeds per plant were closely correlated to yield while
seed size and seeds per pod were very weakly correlated. Similarly, Carpenter and Board (1997)
32

planted soybeans in 75-cm rows at populations of 70,000, 164,000, 189,000, and 234,000 plants
ha-1 and found that 164,000 plants ha

-1

was the optimal population for yield. They determined

that seed size and seeds per pod were not affecting yield across populations, and thus concluded
that pods per plant must be the primary factor determining yield in various populations. Branch
pods made up 75-87% of the yield in this study, so number of branch pods was strongly
correlated to yield.
Norsworthy and Shipe (2004) were interested in the partition of seed yield between the
main stem and branches in wide and narrow rows. They planted six varieties of soybean in 19and 97-cm rows at populations of 432,000 and 272,000 seeds ha

-1

respectively. They found that

averaged over two years and the six varieties, main stem yield accounted for 69% of 19-cm yield
but only 45% of 97-cm yield. In 19-cm rows, branch yield ranged from 14-57% of total yield,
while in 97-cm rows it ranged from 47-74% of total yield. The authors noted that reduced
branching has also been reported in high populations, which in this case were concurrent with
narrow rows, but the authors assumed that reduced branching was due to narrow rows and
suggested that varieties be examined for good main stem yield for narrow rows and good branch
yield for wide rows. This suggests that when soybean plants have significant space between one
another, they partially compensate by branching more than they would with less space. This also
suggests that main stem yield is enhanced when there is less branching and that, given the
tendency toward higher yields in conditions that might discourage branching, the main stem of
soybean is more efficient at producing seed than are branches. This is supported by Ethredge et
al. (1989), who found that main stem yield decreased with increasing row width as did overall
yield. However, despite Norsworthy and Shipe’s attribution of branch yield reduction to narrow
rows, the Ethredge et al. study saw that branch yield was lowest at the highest population but
33

was not consistently affected by row width. On the other hand, Lehman and Lambert (1960) did
report more branches and a greater proportion of pods on branches in 102-cm rows than in 51-cm
rows as well as more branches at wider within-row spacings.

They also noted that seeds on

branches tended to be smaller than those on the main stem.
In contrast to the large body of work in soybean, a rather limited amount of work has
been done looking at the morphological responses of dry bean. Crothers and Westerman (1976)
found that in two semi-upright, Type II dry bean varieties, the optimum planting population was
about 300,000 plants ha

-1

and that yields decreased at higher populations due to intraspecific

competition manifesting itself in the death of older branches and the production of pods mainly
on upper nodes. Consistent with this, Bennett et al. (1977) noted that when yield is reduced at
overly high populations, the most sensitive yield component is number of pods per plant. They
found that the causes of this were decreases in branching, racemes per node, and pods per node
with increased population. They found that number of branches and pods per branch were
inversely related and thus concluded that breeders ought not to focus on either element since
increasing one decreases the other. Grafton et al. (1988) planted a determinate navy bean
cultivar and an indeterminate Type III pinto bean cultivar in four row widths and four
populations. The determinate cultivar showed yield increases with increased population while
the indeterminate cultivar did not; in the determinate cultivar, pods per plant decreased with
increasing population while seed weight remained constant, but number of seeds per pod
increased, apparently accounting for part of the yield advantage at high population. In the
indeterminate cultivar, pods per plant and seeds per pod both decreased with increasing
population, eliminating the advantage that might have been expected from more plants per area,
and suggesting that indeterminate cultivars are better able to compensate for low populations
34

than indeterminate cultivars.

The cause for increasing yields with decreasing row width

remained obscure since no yield component appeared to vary significantly with row width.
Goulden et al. (1976) found that seed weight and number of seeds per pod were
unaffected by row width in Type I navy beans. However, the two were found to be inversely
related to one another. Variation in yield per plant was explained entirely by number of pods per
plant, but pods per plant and yield per plant were inversely related to yield ha

-1

as well as to

plant density. The inverse relationship of yield components to one another is consistent with the
finding of Adams (1967) that number of pods per plant is often inversely related to the other two
components of yield. He postulated that the three components are genetically independent of
one another and that the plant first initially focuses on allocating resources to pod formation, then
to seed formation, then to increasing seed size. He further postulated that nutrient supply
oscillates during the growing season and that if fewer resources are available during the
determination of one yield component, the plant will compensate by allocating more resources to
later-developing yield component. Later, Duarte and Adams (1972) noted that number of pods
per plant was correlated with number of leaves since dry bean flowers form in leaf axils. They
also noted that seed size correlated with leaf size and suggested that the two are genetically
linked.
Horn et al. (2000) observed lodging in three row widths of Type II dry bean and found
that lodging increased with wider rows and lower populations. The 25-cm rows were also found
to reduce plant height compared to wider rows and to reduce the number of pods touching the
ground.

35

Weed Control Systems in Dry Bean
While planting density or row width may affect many aspects of plant development and
thereby affect weed management, changes in these factors alone will not normally result in
sufficient weed control, as evidenced by the much lower yields obtained in soybean regardless of
row width or population (Harder et al. 2007). Blackshaw et al. (2000) also observed that dry
bean yield remained low in the absence of herbicide application regardless of row width or
population.

As such, some form of chemical or mechanical weed control is likely to be

necessary for effective weed control.
Two studies have tried to establish a critical time of weed control in dry beans; that is, a
period beginning at the time when weeds that emerge with the crop begin to impact yield and
ending at the time when newly emerging weeds cease to impact yield (Burnside et al. 1998;
Woolley et al. 1993). This is the period during which weeds must be controlled to avoid yield
loss. Burnside et al. (1998) in Minnesota reported that yield was not affected by waiting four
weeks after planting to control weeds but was affected by waiting six weeks. They reported that
yield was unaffected if weed control ended at six weeks; however, it was affected if weed control
ended at two weeks and was affected at one site if weed control ended at four weeks. This
suggests that the critical period for weed control is four to six weeks after planting, though the
authors suggest three to six weeks. Wooley et al. (1993) in Ontario planted two navy bean
varieties in two years and calculated the critical period of weed control based on growth stage of
the plant and either allowed weeds to grow up to a certain stage, afterward keeping the plot
weed-free, or kept the plot weed-free up to a certain stage, then allowed weeds to grow. Three of
the variety-years had a critical period from second trifoliate to first flower; the other variety-year
had a longer period from second trifoliate to mid-flower. Arnold et al. (1993) in New Mexico

36

assert without explanation that five to seven weeks after planting is the period in which
competition from late-season weeds is most intense.
Burnside et al. (1994), working in Minnesota, demonstrated that there were numerous
combinations of herbicides and mechanical control treatments available to dry bean growers that
resulted in sufficient weed control. However, every treatment involved cultivation alone or
cultivation plus rotary hoeing, and every herbicide combination except the two least effective
involved pre-plant incorporated herbicides (clomazone, ethalfluralin, EPTC, trifluralin, and
alachlor).

In some treatments, all herbicides were pre-plant incorporated (PPI).

Unless

mechanical control was used as the sole method of control, a single cultivation was sufficient to
produce effective weed control, but both cultivation and PPI herbicides involve soil disturbance
and additional fuel use, and cultivation is impossible in narrow rows, so these combinations are
not desirable for many modern dry bean growers. The only postemergence (POST) herbicides
used by Burnside et al. were acifluorfen, bentazon, imazethapyr, sethoxydim, and an unnamed
experimental herbicide, and no pre-emergence (PRE) herbicides were used. All herbicides used
in this study are still registered for dry beans in the United States, except acifluorfen (Sprague
and Everman 2011).
Hekmat et al. (2008) were working with dry beans in Ontario, where herbicide options
for dry bean have been even more limited than for U.S. growers due to Canadian policy, having
only bentazon and fomesafen available as POST herbicides. As such, they experimented with
combining bentazon with imazamox and examined crop safety in a variety of market classes,
which they thought would be acceptable due to evidence that bentazon can act as a safener for
imadizolinone herbicides such as imazamox and imazethapyr. This supposition was correct, as
the combination was found to result in minimal, transient crop injury. Imazamox is registered

37

for dry beans in the United States (Sprague and Everman 2011). They did not comment on this
combination’s effectiveness.
Wall (1995) also examined bentazon tank-mixes for control of redroot pigweed and
common lambsquarters in navy bean in Manitoba. His concern was that bentazon does not
always fully control these weeds; tank-mixes of bentazon and imazethapyr applied POST were
found to give adequate control of both. While such mixes usually injured dry beans, they always
recovered by four weeks after treatment.
Arnold et al. (1993) expressed dissatisfaction with the state of weed control among New
Mexico pinto bean growers, who were largely relying on PPI herbicides, and so explored
imazethapyr applied PPI, PRE, or POST, sometimes in combination with other herbicides. They
found that imazethapyr applied as a PRE with either metolachlor or pendimethalin was highly
effective at controlling the major weeds in the study. As a POST application alone, imazethapyr
did not provide good control of barnyardgrass but did control many other weeds. All treatments
resulted in yields that did not differ from the weed-free, suggesting that imazethapyr suppressed
weeds well overall.
Blackshaw and Esau (1991) in Alberta investigated herbicides suitable for control of
hairy nightshade, redroot pigweed, and common lambsquarters in pinto beans.

Fomesafen

applied POST was found to be insufficient for control of these weeds, while imazethapyr applied
POST provided good control, though not at reduced rates; bentazon alone applied POST was
inconsistent in its control of hairy nightshade but gave fairly good control of the other weeds.
Sikkema et al. (2009) in Ontario used combinations of S-metolachlor and fomesafen
applied both PPI and PRE to control a variety of weeds in black, navy, kidney, and cranberry
bean. Both herbicides are currently registered for use on dry bean in the United States (Sprague

38

and Everman 2011). They concluded that this combination is safe for all four classes but did not
comment on the effectiveness of weed control.
Blackshaw et al. (2000) found ethalfluralin, applied PPI followed by bentazon or
imazethapyr, to be an effective herbicide combination.

In general, they concluded that

combinations of herbicides, even at reduced rates, controlled weeds as well or better than the full
rate of individual herbicides. They found that imazethapyr alone applied POST failed to control
barnyardgrass and exhibited fairly low control of some other species but that tank-mixing it with
bentazon improved control of some weeds. However, they suggested that imazamox has the
potential to control weeds effectively applied alone as a POST treatment.
Aside from ringing endorsements of various imazethapyr tank mixes, very little
information seems to exist on the effectiveness of available PRE or POST herbicide
combinations in dry beans. However, at least six PRE herbicides and seven POST herbicides are
available for dry bean in the United States (Sprague and Everman 2011), suggesting the need for
additional research into which combinations are effective.

39

Table 1.1. Summary of studies examining row width and population in dry bean.
Author/Year

Location

Classes;
Types

Widths (cm);
Results
Populations
-1
(thousand ha )
or within row
densities
45, 60, 75;
Yield optimized at 60-cm,
-1
7, 10, 13, and
13 m
-1
16 plants m of
row

Aidar et al.
2001

Tocantins,
Brazil

Carioca; II, III

Alford et al.
2004

Wyoming

Not specified;
not specified

38, 56, 76;
not specified

No impact of row width on
yield

Arf et al.
1996

Mato Grosso
do Sul,
Brazil

‘Ouro’;
not specified

50, 60, and four
split-row
(20x80, 30x80,
20x70, 30x70);
8, 12, and 16
-1
plants m of
row

No impact of row width on
yield; increased yield with
increasing within-row
density

Blackshaw et
al. 1999

Alberta

Navy; II/III

23, 46, 69;
240, 480

- Increased PAR
interception with narrowing
rows; incomplete canopy
closure in 69-cm rows
- Inconsistent increase in
PAR interception at 480 ha
1

- Higher yields in 23- than
in 69-cm rows but not
-1
always at 240 ha ; highest
-1

yields in 480 ha
- Most consistent yield
increases in weedy plots
- Hairy nightshade biomass
reduced with narrowing
-1
rows, at 480 ha

40

Table 1.1 (cont’d)
Blackshaw et
al. 2000

Alberta

Small red; II

23, 69;
200, 500.

- PAR interception often
higher in 23-cm, especially
early; also higher in 500 ha
1

during much of the season
- Yield optimized at 69-cm
-1
and 500 ha
- Weed control maximized
-1
at 23-cm and 500 ha
Conner et al.
2006

Manitoba

Navy, pinto;
II/III

30, 60;
not specified

Higher yields in 30-cm in
three of four site-years

Crothers and
Westerman
1976

Idaho

Snap bean (I),
Pinto (III),
Small red (III)

Not specified;
ranged from
107.6 to 969.7

-Increased yield with higher
populations in determinate
cultivars but not usually in
indeterminate cultivars

Goulden 1976 Canterbury,
New
Zealand

Navy; I

20, 40;
Within-row
spacings of 4.8,
7.1, and 10.2
cm

Much higher yield in 20-cm
than 40-cm; within 20-cm,
yield higher in 4.8- and 7.1cm within row spacings

Grafton et al.
1988

North
Dakota

Navy, Pinto;
I/II

25, 50, 75, 100;
99, 173, 247,
321

Increased yield with
narrowing rows; increased
yield with higher population
in navy beans but not in
pinto beans

Horn et al.
2000

Rio Grande
do Sul,
Brazil

‘Pampa’; II

25, 50, 75;
100, 200, 350,
500

Yield increased with
widening rows; no impact of
population on yield

Malik et al.
1989, 1990,
1993

Ontario

Navy; I/II/III

23, 46, 69;
250, 375

- Generally Higher yields in
23- and 46-cm
-1
- Higher yield at 375 ha in
23- or 46-cm rows
- Weed biomass highest in
69-cm
- Negative correlation
between weed biomass and
yield
- Better weed suppression in
indeterminate cultivars

41

Table 1.1 (cont’d)
Park 1993

Ontario

Not specified*; 30 (500), 60
I/II
(333.3), 80
(266.7);

Increased yield with
narrowing rows

Redden 1987

Queensland,
Australia

Navy; II

8, 36, 71, 107;
112.5, 337.5

- Inconsistent canopy
ground cover increase with
narrowing rows; also
sometimes greater ground
-1
cover at 337.5 ha
- Increased yield with
narrowing rows in all but
one site-year; often
increased yield with
increased population

Saindon et al.
1995

Alberta

Dark red
kidney (I),
navy (II),
black (II),
small red (III)

23;
250, 350, 500,
600

-Increased yield with
increasing population in
three of four years

Vieira et al.
2010

Minas
Gerais,
Brazil

Carioca; III

50;
4, 5, 7.5, 8, 15,
and 16 plants
-1
m of row

Increased canopy closure
with higher within-row
density

Welacky and
Park 1987

Ontario

Navy; II

30 (412.4),
60 (275)

Higher yields in 60-cm

Xu and Pierce
1998

Michigan

Navy; not
specified

56, 71;
not stated

- Leaf area unaffected by
row width
- Yield higher in 56-cm

Ziviani et al.
2009

Goias,
Brazil

Carioca;
II/III

30, 40, 50, 60;
5, 10, 15, and
-1
20 plants m of
row

- Canopy ground cover
greater at 30- and 40-cm
than at 50- and 60-cm; lower
-1
ground cover at 5 plants m
throughout the season and in
-1
10 plants m by mid-season
- Yield optimized at 30- or
-1
40-cm, 20 m

42

Table 1.2. Summary of studies examining row width and population in crops other than dry
bean.
Author/Year

Location

Crop

Widths (cm);
Populations
-1
(thousand ha )
or within-row
density

Results

Alford et al.
2004

Wyoming

Sugarbeet, Corn

38, 56, 76;
not stated

Reached 95%
light interception
earlier in 38-cm
than in 76-cm

Andrade et al.
2002

Buenos Aires,
Argentina

Soybean

19, 30, 52, 57,
60,70;
150, 300, 350,
400, 450, 460,
520

Increased light
interception and
yield in narrow
rows; light
interception
increases
correlated with
yield increases

Carpenter and
Board 1997

Louisiana

Soybean

Not stated;
- Yield reduced
70, 164, 189, 234 at 70 ha-1
- Pods per plant
increased with
decreasing
population,
largely on
branches

Chandler et al.
2001

Ontario

Soybean

19/76 (twin row), - Yield higher in
38, 76;
38-cm, twin row
200
- Weed biomass
lower in 38-cm

Cooper 1977

Illinois

Soybean

17, 50, 75;
188, 281, 375

43

Highest yields in
-1
17-cm, 375 ha
combination

Table 1.2 (cont’d)
Dalley et al.
2004

Michigan

Soybean, Corn

Soy: 19, 38, 76;
422
Corn: 38, 76
77

- Soy: Much
greater daily and
peak light
interception in
19- and 38-cm;
yield sometimes
greater in 19than 38-cm
- Corn: Higher
early-season light
interception in
38-cm

Ethredge et al.
1989

Georgia

Soybean

25, 51, 76;
260.2, 390.4,
520.4

- Increased yield
with narrow rows
- No yield
differences
between
populations

Harder et al.
2007

Michigan

Soybean

19, 38, 76;
124, 190,
296, 445

- Greater LAI,
earlier closure in
19- and 38-cm;
greater LAI in
-1
445 ha than
-1

124 or 190 ha
- Higher yields
with narrowing
rows; yield
higher at 296 and
-1
445 ha
- Decreased
weed density and
biomass with
narrow rows
after herbicide
application; no
impact of
population on
weeds
- Weed biomass
higher in 124 ha
1

, 38- or 76-cm

44

Table 1.2 (cont’d)
Soybean
19, 38;
259.35, 333.45,
345.8, 432.25,
444.6, 518.7,
555.75, 666.9

Kratochvil et al.
2004

Maryland

Légère and
Schreiber 1989

Indiana

Soybean

25, 76;
385.5

- Greater LAI in
26-cm through
much of the
season
- Greater yield in
26-cm
- Reduced
percentage of
pigweed in
biomass and leaf
area of plot in
76-cm

Lehman and
Lambert 1960

Minnesota

Soybean

51, 102;
13, 26, 52, and
-1
79 plants m of
row

Greater yield in
51-cm at one of
two sites; no
yield differences
between withinrow densities

Moore 1991

Louisiana

Soybean

97;
Increased yield
3/6, 26/19, 48/28 with equidistant
-1
plants m of row within-row
spacing

Nelson and
Renner 1998

Michigan

Soybean

19 (358/371), 76
(508/469)

45

- Higher yields in
19-cm in five of
six site-years
- Yield lower in
lowest two
populations;
occasionally
higher in higher
populations

- Canopy closed
35 d sooner in
19-cm
- Higher yields in
19-cm
- Lower weed
biomass
following partial
herbicide in 19cm rows

Norris et al. 2002 Mississippi

Table 1.2 (cont’d)
Soybean
38, 76;
480

- Occasionally
higher yields in
38-cm
- Better
suppression all
weed species
after herbicide in
38-cm

Norsworthy and
Oliver 2001

Arkansas

Soybean

19;
185, 247, 371,
494, 618, 741,
865, 988, 1112,
1235,1359, 1482

- 0.022%
increase in light
interception per
thousand plants
-1
ha , but still
89% at 185
-1
ha
- Little variation
in yield across
populations; 988
-1
ha optimum for
yield

Norsworthy and
Frederick 2002

South Carolina

Soybean

Drilled;
370, 620

No effect of
population on
yield

Peachey et al.
2006

Oregon

Snap bean

19, 38, 75, 114,
150;
445

- With fungicide,
highest yield at
19-cm, all others
similar
- Without
fungicide, lowest
yields at 150-cm,
highest at 38- or
75-cm, all other
yields similar

46

Table 1.2 (cont’d)
Peters et al. 1965

Missouri

Soybean

20, 41, 51, 61,
81, 102 (not all
in same trial);
not stated

- Higher yields
with narrowing
rows
- Less cultivation
needed for
adequate weed
control with
narrowing rows

Quakenbush and
Andersen 1984

Minnesota

Soybean

18 (556),
76 (395)

95% PAR
interception 3 wk
earlier in 18-cm

Sankula et al.
2001

Delaware and
Maryland

Lima bean

38, 76;
175, 350

Yield increase in
38-cm only in
one of four siteyears

Taylor 1980

Iowa

Soybean

25, 50, 75, 100;
160

- No LAI
differences;
earlier
appearance of
canopy closure
with narrowing
rows
- Yield higher
with narrowing
rows in good
growing
conditions, not in
drought

Teasdale and
Frank 1983

Maryland

Snap bean

15, 25, 36, 46,
91;
430

- Increased
canopy cover and
decreased light
penetration with
narrowing rows.
- Lower yield in
91-cm rows;
otherwise similar
across row
widths
- Weed biomass
reduced in 15-,
25-, and 36-cm

47

Table 1.2 (cont’d)
Snap bean
Row widths: 20
(1000), 40 (440),
80 (250), 125
(160)
Square designs:
10 (1000), 15
(440), 20 (250),
25 (125)
Triangle designs:
10 (1160), 15
(510), 20 (290),
25 (180)

Wahab et al.
1986

Saskatchewan

Wax and
Pendleton 1968

Illinois

Soybean

25, 51, 76, 102;
not specified but
decreased with
increasing row
width

- Increased yield
with narrowing
rows
- Weed biomass
decreased with
narrowing rows

Weber et al.
1966

Iowa

Soybean

13, 25, 38, 51;
64.5, 129, 258,
516.5

- Increased LAI
with narrowing
rows and
increasing
population
- Increased dry
matter with
increasing LAI,
but decreased
yield at very high
dry matter

Wiggans 1939

New York

Soybean

20, 31, 41, 61,
81;
Within-row
spacings of 1.3,
2.5, 3.8, 5.1, 7.6,
10.2, and 15.2
cm

Increased yield
with narrowing
rows; slight
yield increases
with decreasing
within-row
spacings; large
yield declines
above 7.6 cm

48

- Increased yield
with increasing
equidistance.
- Generally
increased yield
with decreasing
between-plant
spacing
- Yield highest in
triangle and
square designs;
one year triangle
higher than
square

Winter 1989

Texas

Table 1.2 (cont’d)
Sugarbeet
76, 102;
43, 86

Yelverton and
Coble 1991

North Carolina

Soybean

23, 46, 91;
not stated

- Earlier and
more complete
PAR interception
in 23- than in 91cm
- Lower weed
density following
herbicide in 23than in 91-cm

Yonts and Smith
1997

Nebraska

Sugarbeet

35, 56, 76, 97;
25, 40, 65, 100,
150

Sugar yield
highest in 35and 56-cm

Young et al.
2001

Illinois

Soybean

19, 38, 76;
260, 321, 346,
420, 445, 457
(varied with row
width and not all
present in any
site-year)

- Yield usually
higher in
narrower row
widths; higher in
one site-year at
76-cm
- Suppression of
several weeds
enhanced
following
herbicide in
narrow rows

49

Sugar quality
highest in 76-cm

LITERATURE CITED

50

LITERATURE CITED

Aidar, H., J. Kluthcouski, M. Thung, I.P. de. Oliveira, and F.J.P. Zimmerman. 2001. Effect of
spacing and number of plants in the row on bean (Phaseolus vulgaris L.) production in tropical
lowland of Brazil. Annual Report of the Bean Improvement Cooperative 44:63-64.
Alford, C.M., S.D. Miller, and J.T. Cecil. 2004. Using row spacing to increase crop competition
with weeds. Proceedings of the 4th International Crop Science Conference. Website:
http://www.cropscience.org.au/icsc2004/poster/2/4/1/412_alfordcm.htm.
Accessed April 4,
2011.
Adams, M.W. 1967. Basis of yield component compensation in crop plants with special
reference to the field bean, Phaseolus vulgaris. Crop Sci. 7:505-510.
Adams, M.W. 1995. An historical perspective on significant accomplishments in dry bean
research – the first hundred years. Bean Improvement Cooperative Review Article. Website:
http://www.css.msu.edu/bic/PDF/BeanHistory.pdf, accessed April 20, 2011.
Andrade, F.H., P. Calvino, A. Cirilo, and P. Barbieri. 2002. Yield responses to narrow rows
depend on increased radiation interception. Agron. J. 94:975-980.
Arf, O., M.E. DeSa, C.S. Okita, M.A. Tiba, G.G. Neto, and F.Y. Ogassawara. 1996. Effect of
different spacing and densities of sowing on the development of beans (Phaseolus vulgaris L.).
Pesq. agropec. bras. 31:629-634.
Arnold, R.N., M.W. Murray, E.J. Gregory, and D. Smeal. 1993. Weed control in pinto beans
(Phaseolus vulgaris) with imazethapyr combinations. Weed Technol. 7:361-364.
Bennett, J.P., M.W. Adams, and C. Burga. 1977. Pod yield component variation and
intercorrelation in Phaseolus vulgaris L. as affected by planting density. Crop Sci. 17:73-75.
Blackshaw, R.E. and R. Esau. 1991. Control of annual broadleaf weeds in pinto beans
(Phaseolus vulgaris). Weed Technol. 5:532-538.
Blackshaw, R.E., L.J. Molnar, H.H. Müendel, G. Saindon, and X. Li. 2000. Integration of
cropping practices and herbicides improves weed management in dry bean (Phaseolus vulgaris).
Weed Technol. 14:327-336.
Blackshaw, R.E., H.H. Müendel, and G. Saindon. 1999. Canopy architecture, row spacing and
plant density effects on yield of dry bean (Phaseolus vulgaris) in the absence and presence of
hairy nightshade (Solanum sarrachoides). Can. J. Plant Sci. 79:663-669.
Board, J.E. 1987. Yield components related to seed yield in determinate soybean. Crop Sci.
27:1296-1297.
51

Burnside, O.C., W.H. Ahrens, B.J. Holder, M.J. Wiens, M.M. Johnson, and E.A. Ristau. 1994.
Efficacy and economics of various mechanical plus chemical weed control systems in dry beans
(Phaseolus vulgaris). Weed Technol. 8:238-244.
Burnside, O.C., M.J. Wiens, B.J. Holder, S. Weisberg, and E.A. Ristau. 1998. Critical periods
for weed control in dry beans (Phaseolus vulgaris). Weed Sci. 46:301-306.
Carpenter, A.C. and J.E. Board. 1997. Branch yield components controlling soybean yield
stability across plant populations. Crop Sci. 37:885-891.
Chandler, K., A. Shreshtha, and C.J. Swanton. 2001. Weed seed return as influenced by the
critical weed-free period and row spacing of the no-till glyphosate-resistant soybean. Can. J.
Plant Sci. 81:877-880.
Chikoye, D., S.F. Weise, and C.J. Swanton. 1995. Influence of common ragweed (Ambrosia
artemisiifolia) time of emergence and density on white bean (Phaseolus vulgaris). Weed Sci.
43:375-380.
Conner R.L., D.W. McAndrew, P. Balasubramanian, F.A. Kiehn, and Y. Dongfang. 2006.
Influence of growth habit, row spacing, and seed infection on bean anthracnose development.
Can. J. Plant Pathol. 28:411-418.
Cooper, R.L. 1977. Response of soybean cultivars to narrow rows and planting rates under
weed-free conditions. Agron. J. 69:89-92.
Crothers, S.E. and D.T. Westerman. 1976. Plant population effects on the seed yield of
Phaseolus vulgaris L. Agron. J. 68:958-960.
Dalley, C.D., J.J. Kells, and K.A. Renner. 2004. Effect of glyphosate application timing and
row spacing on weed growth in corn (Zea mays) and soybean (Glycine max). Weed Technol.
18:177-182.
Duarte, R.A., M.W. Adams. 1972. A path coefficient analysis of some yield component
interrelations in field beans (Phaseolus vulgaris L.). Crop Sci. 12:579-582.
Duncan, W.G. 1986. Planting patterns and soybean yields. Crop Sci. 26:584-588.
Ethredge, W.J., D.A. Ashley, and J.M. Woodruff. 1989. Row spacing and plant population
effects on yield components of soybean. Agron. J. 81:947-951.
Goulden, D.S. 1976. Effect of plant population and row spacing on yield and components of
yield in navy beans (Phaseolus vulgaris). New Zeal. J. Exp. Agr. 4:177-180.

52

Grafton, K.F., A.A. Schneiter, and B.J. Nagle. 1988. Row spacing, plant population, and
genotype x row spacing interaction effects on yield and yield components of dry bean. Agron. J.
80:631-634.
Harder, D.B., C.L. Sprague, and K.A. Renner. 2007. Effect of soybean row width and
population on weeds, crop yield, and economic return. Weed Technol. 21:744-752.
Hekmat, S., N. Soltani, C. Shropshire, and P.H. Sikkema. 2008. Effect of imazamox plus
bentazon on dry bean (Phaseolus vulgaris L.). Crop Prot. 27:1491-1494.
Horn, F.L., L.O. Braga Schuch, E.P. Silveira, I.F. Antunes, J.C. Vieira, G. Marchioro, D.F.
Medeiros, and J.E. Schwengber. 2000. Spacing and plant population for dry bean harvest
mechanization. Pesq. agropec. bras. 35:41-46.
Hosfield, G.L., S.M. Bushey, and J.D. Kelly. 2005. Merlot: a new small red dry bean for
Michigan. Michigan State University Extension Bulletin E-2933, East Lansing, MI.
Hosfield, G.L., G.V. Varner, M.A. Uebersax, and J.D. Kelly. 2004. Registration of ‘Merlot’
small red bean. Crop Sci. 44:351-352.
Kelly, J.D. 1994. Traditional and novel approaches to bean improvement. Mich. Bean Digest
19(1):4-7.
Kelly, J.D. 2001. Remaking bean plant architecture for efficient production. Advances in
Agronomy 71:109-143.
Kelly, J.D. 2010. The story of bean breeding. 2010. Michigan State University publication
prepared for BeanCAP and PBG Works, June 2010.
http://www.css.msu.edu/Bean/_pdf/Story_of_Bean_Breeding_in_the_US.pdf. Accessed March
28, 2012.
Kelly, J.D. and M.W. Adams. 1987. Phenotypic recurrent selection in ideotype breeding of
pinto beans. Euphytica 36:69-80.
Kelly, J.D., G.V. Varner, and B. Long. 2010. Registration of ‘Santa Fe’ pinto bean. J. Plant
Registrations 4(1):12-16.
Kelly, J.D., G.V. Varner, P. O’Boyle, and B. Long. 2009a. Registration of ‘Zorro’ black bean.
J. Plant Registrations 3:226-230.
Kelly, J.D., G.V. Varner, and C.L. Sprague. 2009b. Zorro: a new black bean for Michigan.
Michigan State University Extension Bulletin E-3069, East Lansing, MI.

53

Kratochvil, R.J., J.T. Pearce, and M.R. Harrison. 2004. Row-spacing and seeding rate effects on
glyphosate-resistant soybean for mid-Atlantic production systems. Agron. J. 96:1029-1038.
Lamey, H.A., R.K. Zollinger, D.K. McBride, R.C. Venette, and J.R. Venette. 1991. Production
problems and practices of Northarvest dry bean growers in 1989. North Dakota Farm Res.
49:17-24.
Lee, S.H., M.A. Bailey, M.A.R. Mian, E.R. Shipe, D.A. Ashley, W.A. Parrott, R.S Hussey, and
H.R. Boerma. 1996. Identification of quantitative trait loci for plant height, lodging, and
maturity in a soybean population segregating for growth habit. Theor. Appl. Genet. 92:516-523.
Légère, A. and M.M. Schreiber. 1989. Competition and canopy architecture as affected by
soybean (Glycine max) row width and density of redroot pigweed (Amaranthus retroflexus).
Weed Sci. 37:84-92.
Lehman, W.F. and J.W. Lambert. 1960. Effects of spacing of soybean plants between and
within rows on yield and its components. Agron. J. 52:84-86.
Malik, V.S., C.J. Swanton, and T.E. Michaels. 1989. Impact of white bean (Phaseolus vulgaris
L.) cultivars, row-spacing and seeding rate on establishment and suppression of annual weeds.
Annual Report of the Bean Improvement Cooperative 32:81-82.
Malik, V.S., C.J. Swanton, and T.E. Michaels. 1990. Impact of white bean (Phaseolus vulgaris
L.) cultivars, row-spacing and seeding rate on annual weeds interference. Annual Report of the
Bean Improvement Cooperative. 33:25-26.
Malik, V.S., C.J. Swanton, and T.E. Michaels. 1993. Interaction of white bean (Phaseolus
vulgaris L.) cultivars, row spacing, and seeding density with annual weeds. Weed Sci. 41(1):6268.
Michigan Bean Commission. 2011. Annual summary (1.12.11).
http://www.michiganbean.org/december-crop-production-12-10-1/. Accessed April 4, 2011.
Mooers, C.A. 1909. The soy-bean as a farm crop. Agron. J. 1:153-158.
Moore, S.H. 1991. Uniformity of plant spacing effect on soybean population parameters. Crop
Sci. 31:1049-1051.
Nelson, K.A. and K.A. Renner. 1998. Weed control in wide- and narrow- row soybean (Glycine
max) with imazamox, imazethapyr, and CGA-277476 plus quizalofop. Weed Technol. 12:137144.
Norris, J.L., Shaw, D.R., and Snipes, C.E. 2002. Influence of row spacing and residual
herbicides on weed control in glufosinate-resistant soybean (Glycine max). Weed Technol.
16:319-325.
54

Norsworthy, J.K. and J.R. Frederick. 2002. Reduced seeding rate for glyphosate-resistant,
drilled soybean on the southeastern Coastal Plain. Agron. J. 94:1282-1288.
Norsworthy, J.K. and L.R. Oliver. 2001. Effect of seeding rate of drilled glyphosate-resistant
soybean (Glycine max) on seed yield and gross profit margin. Weed Technol. 15:284-292.
Norsworthy, J.K. and E.R. Shipe. 2005. Effect of row spacing and soybean genotype on
mainstem and branch yield. Agron. J. 97:919-923.
Park, S.J. 1993. Response of bush and upright plant type selections to white mold and seed
yield of common beans grown in various row widths in southern Ontario. Can. J. Plant Sci.
73:265-272.
Peachey E., R.L. Ludy, M.L. Powelson, and D.M. McGrath. 2006. Modification of plant
arrangement suppresses white mold of snap beans. HortScience 41:1298-1302.
Pederson, P. and J.G. Lauer. 2004. Response of soybean yield components to management
system and planting date. Agron. J. 96:1372-1381.
Peters, E.J., M.R. Gebhardt, and J.F. Stritzke. 1965. Interrelations of row spacings, cultivations,
and herbicides for weed control in soybeans. Weeds 13:285-289.
Quakenbush, L.S. and R.N. Andersen. 1984. Effect of soybean (Glycine max) interference on
eastern black nightshade (Solanum ptycanthum). Weed Sci. 32:638-645.
Redden, R.J. 1987. Response of navy beans to row width and plant population density in
Queensland. Aust. J. Exp. Agric. 27:455-463.
Robertson, L.S. and R.D. Frazier, eds. 1978. Dry bean production – principles and practices.
Michigan State University Extension Bulletin E-1251, East Lansing, MI.
Saindon, G., H.C. Huang, and G.C. Kozub. 1995. White-mold avoidance and agronomic
attributes of upright common beans grown at multiple planting densities in narrow rows. J.
Amer. Soc. Hort. Sci. 120:843-847.
Saindon, G., H.C. Huang, G.C. Kozub, H.H. Mundel, and G.A. Kemp. 1993. Incidence of white
mold and yield of upright bean grown in different planting patterns. Phytopathology 137:118124.
Sankula, S., M.J. VanGessel, W.E. Kee, C.E. Beste, and K.L. Everts. 2001. Narrow row
spacing does not affect lima bean yield or management of weeds and other pests. HortScience
36:884-888.

55

Schwartz, H.F., M.A. Brick, R.M. Harveson, and G.D. Franc, eds. 2004. Dry bean production
& integrated pest management – 2nd edition. Regional publication by Central High Plains Dry
Bean and Beet Group, Colorado State University, University of Nebraska, and University of
Wyoming: Bulletin 562A.
Sikkema, P.H., C. Shropshire, N. Soltani. 2009. Response of dry bean to pre-plant incorporated
and pre-emergence applications of S-metolachlor and fomesafen. Crop Prot. 28:744-748.
Sprague, C.L., and W.J. Everman. 2011. Weed control guide for field crops. Michigan State
Extension Bulletin E-434.
Taylor, H.M. 1980. Soybean growth and yield as affected by row spacing and by seasonal water
supply. Agron. J. 72:543-547.
Teasdale, J.R. and J.R. Frank. 1983. Effect of row spacing on weed competition with snap
beans (Phaseolus vulgaris). Weed Sci. 31:81-85.
United States Department of Agriculture – National Agricultural Statistics Service. 2011.
National statistics for beans.
http://www.nass.usda.gov/Statistics_by_Subject/index.php?sector=CROPS. Accessed April 4,
2011.
Urwin, C.P., R.G. Wilson, and D.A. Mortensen. 1996. Late season weed suppression from dry
bean (Phaseolus vulgaris) cultivars. Weed Technol. 10:699-704.
Vieira, R.F., T.J. Paula, H. Texeira, and J.E. de San Carneiro. 2010. White mold management in
common bean by increasing within-row distance between plants. Plant Dis. 94:361-367.
Wahab, M.N.J., D.H. Dabbs, and R.J. Baker. 1986. Effects of planting density and design on
pod yield of bush snap bean (Phaseolus vulgaris L.). Can. J. Plant Sci. 66:669-675.
Wall, D.A. 1995. Bentazon tank-mixtures for improved redroot pigweed (Amaranthus
retroflexus) and common lambsquarters (Chenopodium album) control in navy bean. Weed
Technol. 9:610-616.
Wax, L.M. and J.W. Pendleton. 1968. Effect of row spacing on weed control in soybeans.
Weed Sci. 16:462-465.
Weber, C.R., R.M. Shibles, and D.E. Byth. 1966. Effect of plant population and row spacing on
soybean development and production. Agron. J. 58:99-102.
Welacky, T.W. and S.J. Park. 1987. Response for row width of white bean lines selected for
upright plant types. Annual Report of the Bean Improvement Cooperative 30:54.

56

Wiggans, R.G. 1939. The influence of space and arrangement on the production of soybean
plants. Agron. J. 31:314-321.
Winter, S.R. 1989. Sugarbeet yield and quality response to irrigation, row width, and stand
density. J. Sugar Beet Res. 26:26-33.
Woolley, B.L., T.E. Michaels, M.R. Hall, C.J. Swanton. 1993. The critical period of weed
control in white bean (Phaseolus vulgaris). Weed Sci. 41:180-184.
Xu, C. and F.J. Pierce. 1998. Dry bean and soil response to tillage and row spacing. Agron. J.
90:393-399.
Yelverton, F.H. and H.D. Coble. 1991. Narrow row spacing and canopy formation reduces
weed resurgence in soybeans (Glycine max). Weed Technol. 5:169-174.
Yonts, C.D., and J.A. Smith. 1997. Effects of plant population and row width on yield of
sugarbeet. J. Sugar Beet Res. 34:21-30.
Young, B.G., J.M. Young, L.C. Gonzini, S.E. Hart, and L.M. Wax. 2001. Weed management in
narrow- and wide-row glyphosate-resistant soybean (Glycine max). Weed Technol. 15:112-121.
Ziviani, A.C., W.Q Ribeiro, M.L.G. Ramos, M. Barbosa, A. Cordeiro, and L.V. Franca. 2009.
Spatial arrangements for common beans with different architecture and their effects on yield and
plant covering. Biosci. Journal. 25:1-9.

57

CHAPTER 2
EFFECTS OF ROW WIDTH AND PLANTING POPULATION ON CROP AND WEED
DEVELOPMENT IN TYPE II VARIETIES OF BLACK BEAN AND SMALL RED BEAN

Abstract
The development of upright black and small red bean varieties gives dry bean growers
the opportunity to plant in narrower rows, which has been associated with yield and weed control
benefits in many crops. Field studies were conducted in 2010 and 2011 at two locations in
Michigan to examine the effect of row width and population on canopy closure, weed
suppression, pest management, and yield in two upright varieties, ‘Zorro’ black and ‘Merlot’
small red beans. Each variety was planted in 38- and 76-cm rows, as well as 51-cm rows at one
location. Populations were 196500, 262000, and 327500 plants ha-1 in black bean and 148000,
196500, and 262000 plants ha

-1

in small red bean. Canopy closure was assessed by repeatedly

measuring light above and below the canopy. Emerging weeds were counted in set quadrats, and
weed biomass was harvested and weighed at the end of the season. Dry bean plants were
examined for disease and insect feeding during the season and for lodging and maturity shortly
before harvest. Yield was obtained using direct harvest by combine. In many cases, narrow
rows increased canopy closure, especially in the mid- to late-season, while population had little
effect on canopy closure. Narrow rows decreased weed biomass in three of four years and
occasionally decreased weed emergence; population had no effect on weed suppression. Row
width and population had little effect on disease pressure. Narrow rows sometimes decreased
insect feeding while population did not affect it. In some cases, narrow rows reduced lodging
and increased maturity while population had little effect. Narrow rows increased yield in four of

58

eight site-years; population did not affect yield.

Narrow rows generally improved canopy

closure and thus improved weed control, yield, and other variables, while planting population
had minimal effects on these variables.
Nomenclature: Dry bean, Phaseolus vulgaris.

Introduction
Quicker and more complete crop canopy development has beneficial effects, including
increased crop yield and greater weed suppression (Andrade et al. 2002; Norsworthy and Oliver
2001; Peters et al. 1965). Narrowing row width increases canopy development in several crops,
including dry bean (Phaseolus vulgaris L.) (Blackshaw et al. 1999; Blackshaw et al. 2000;
Harder et al. 2007; Redden et al. 1987; Yelverton and Coble 2001).

However, dry bean

producers traditionally plant in 71- or 76-cm rows to allow for weed control and harvest
preparation by inter-row cultivation (Blackshaw et al. 2000; Goulden 1976; Malik et al. 1993).
Through the 1970s, most commercial dry bean varieties were prostrate and vining, known as
Type III, or bushy with low branches and determinate growth, known as Type I (Adams 1995;
Kelly et al. 2009; Schwartz et al. 2004). Type I and Type III dry bean varieties must be
harvested using a two-pass system in which plants are first pulled and windrowed, then harvested
and threshed with a specially equipped combine (Robertson and Frazier 1978; Schwartz et al.
2004). However, in the late twentieth century, breeding efforts produced commercially viable
Type II dry bean varieties with an upright habit suitable for direct harvest by combine in one
pass (Hosfield et al. 2004; Hosfield et al. 2005; Kelly 1994, 2001, 2010; Welacky and Park
1987). Compared with the windrow system, this method improves harvest efficiency and may
reduce harvest losses (Kelly 2010; Robertson and Frazier 1978; Schwartz et al. 2004). It also

59

eliminates the need for cultivation as a hilling method to prepare beans for harvest (Robertson
and Frazier 1978; Schwartz et al. 2004). In fact, direct harvest precludes cultivation as a weed
control method since it requires a level surface free of rocks and soil clods (Schwartz et al.
2004).

This inability to cultivate may no longer be a serious problem in most dry bean

production systems due to the availability of effective modern herbicides (Sprague and Everman
2011). If no cultivation is planned, growers have the option of growing dry beans in narrower
rows than in the past (Blackshaw et al. 2000), which could lead to higher yields and better weed
suppression. However, row width cannot be considered independent of planting population since
changing row width while keeping population constant changes within-row plant spacing. In
addition, planting at an unnecessarily high population is a cost that growers should avoid.
In soybean (Glycine max L.), a related crop with a similar growth habit to Type II dry
beans, studies indicate that narrower row widths increase leaf area index (LAI) and
photosynthetically active radiation (PAR) interception (Dalley et al. 2004; Harder et al. 2007;
Weber et al. 1966), indicating an increase in canopy closure. While some studies show a canopy
closure benefit at higher populations, others suggest that soybean is able to compensate over a
wide range of populations (Harder et al. 2007; Norsworthy and Oliver 2001; Weber et al., 1966).
Very little work has been published on canopy closure in dry bean; however, Blackshaw et al.
(1999, 2000) studied PAR interception in Type II and Type III navy beans and Type II small red
beans and concluded that interception was greater in narrow rows and, in some cases, at higher
populations.
Increased light interception due to greater canopy closure should suppress weeds since
this denies weeds below the canopy a critical resource. Malik et al. (1993) found that planting
navy beans in 23- or 46-cm rows decreased weed biomass 15-21% compared with 69-cm rows,

60

but that planting population had no impact. Blackshaw et al. (2000) evaluated weed control in
small red beans at various row width and plant population combinations and found that weed
control was optimized by narrow rows with high populations.

Similar effects have been

observed in soybean with regard to row width, including by Harder et al. (2007), who found an
effect of row width but not population, Nelson and Renner (1998), whose observation of
decreased weed biomass in narrow rows correlated with increased canopy closure, and Légère
and Schreiber (1989), who observed decreases in both weed biomass and leaf area of redroot
pigweed (Amaranthus retroflexus L.) in narrow rows.
Weed suppression could indirectly improve yield, but enhanced canopy development
could also improve yield by increasing the amount of total light energy captured and utilized by
the crop.

Many studies exist in soybean showing increased yields in narrow rows, with

population able to be varied within wide limits to achieve optimum yield (Harder et al. 2007;
Légère and Schreiber 1989; Lehman and Lambert 1960; Norsworthy and Frederick 2002;
Norsworthy and Oliver 2001; Young et al. 2001). However, Norsworthy and Oliver (2001)
showed that populations can be so high that yield begins to be reduced. In dry bean, Park (1993)
found that simultaneously increasing population and decreasing row width resulted in increasing
yields in both Type I and Type II navy beans. Malik et al. (1993) reported that 23-cm rows
resulted in yield increases in navy beans of all three types compared with 46-cm rows and
-1

375,000 plants ha

-1

resulting in higher yields than 225,000 plants ha , but only in the presence

of weeds. In Type II small red beans, Blackshaw et al. (2000) found a 19% yield advantage of
23-cm rows over 69-cm rows and a 17% yield advantage of 500,000 plants ha
-1

plants ha . Grafton et al. (1988) found that yield increased by 400-600 kg ha

61

-1

-1

over 200,000

for each 25-cm

decrease in row width in Type III pinto bean and Type I navy bean; they also found that yield
increased with increasing population in pinto beans, but not in navy beans.
A key reason for enhanced canopy development and its accompanying effects in narrow
rows may be that narrowing row width typically results in an increasingly equidistant plant
arrangement. As the distance between rows decreases, it becomes increasingly similar to the
smaller distance between plants within the row, improving the ability of the crop canopy to fill in
the inter-row space. Wahab et al. (1986) compared a variety of planting designs in snap bean
(which is the same species as dry bean) that achieved different levels of equidistance. They
found that yield was maximized with an equidistant planting design, presumably because this
allows each plant to reach its full potential and minimizes gaps in the canopy. Andrade et al.
(2002) suggested that part of the advantage of equidistance is found in the below-ground canopy
of the roots; that is, equidistance may improve the efficiency with which water and nutrients are
gathered by the roots, as well as the efficiency with which light is captured by the leaves.
Presumably, any positive effects of increasing population on canopy development and the
benefits associated with it are due to the fact that dropping the population below some critical
level will prevent the crop from producing enough leaves or roots to take advantage of the
available resources. However, very high populations could result in intra-specific competition
that causes excessive vegetative growth at the expense of seed yield (Bennett et al. 1977;
Crothers and Westerman 1976).
Despite the apparent advantages of narrow rows and enhanced canopies, concern has
been expressed that this would also provide an enhanced environment for disease development.
Park (1993) found that under high disease pressure, white mold (Sclerotinia sclerotiorum (Lib.)
de Bary ) incidence was 34% greater in 30-cm rows than in 60-cm rows and 56% greater than in

62

80-cm rows, though in a year of lower white mold pressure, there were no significant
differences. However, Park also found that white mold development was much lower in Type II
than Type I navy beans, and despite higher white mold incidence, yield continued to increase
with decreasing row width even in Type I beans. Saindon et al. (1995) and Park suggested that
Type II beans are less susceptible to disease than other types due to better airflow in the canopy
and fewer pods touching the ground. Saindon et al. found no effect of row width on white mold
infection of Type II plants and much less severe infection on Type II plants than on Type III
plants. Conner et al. (2006) found no impact of row width on the severity of anthracnose
(Colletotrichum lindemuthianum (Sacc. & Magnus) Briosi & Cavara) in any navy or pinto bean
type. In addition to disease pests, a key insect pest in dry bean is western bean cutworm,
Striacosta albicosta (Smith), the larvae of which bore into developing bean pods and damage the
seeds (Michel et al. 2010). There is no published research on the effect of dry bean row width or
planting population on this insect, so there is a need for research to determine whether changing
these variables can increase or decrease its severity.
Although new Type II varieties of black and small red bean are widely used and
economically important in Michigan, there is a notable lack of research regarding the agronomic
properties of these varieties. Most of what exists focuses on navy or pinto beans, and very little
research here has been done on black beans or small red beans. In addition, most dry bean
research has occurred in climates dissimilar to that of Michigan. Thus, the objectives of this
study were to determine the effect of row width and planting population of Type II black and
small red bean on 1) canopy closure, 2) weed biomass and emergence, 3) disease and insect
damage, and 4) seed yield and gross margin.

63

Materials and Methods
Field studies were conducted in 2010 and 2011 at the Michigan State University Harry
and Hazel Box Farm near East Lansing, Michigan and at the Michigan State University Saginaw
Valley Research and Extension Center near Richville, Michigan. The soil at the East Lansing
site was a Capac loam (fine-loamy, mixed, mesic, Aeric Ochraqualfs) with pH 6.6 and 4%
organic matter in 2010 and pH 6.5 and 2.6% organic matter in 2011. The soil at the Richville
site was a Tappan-Londo loam (fine-loamy, mixed, active, calcareous, mesic Typic Endoaquolls
and fine-loamy, mixed, mesic, Aeric Glossaqualfs) with pH 7.8. Soil organic matter was 2.7%
and 2.6% in 2010 and 2011, respectively. The crops prior to these studies were winter wheat
(Triticum aestivum L.) at East Lansing and corn (Zea mays L.) at Richville. Field preparation
consisted of either fall moldboard or chisel plow followed by one to two field cultivations in the
spring. Prior to planting, 336 kg of 17-8-15 (N-P-K) fertilizer containing 1.5% manganese and
1.5% zinc was broadcast onto the soil surface and incorporated.
Two Type II (upright indeterminate vine) varieties, ‘Zorro’ black beans and ‘Merlot’
small red beans (Michigan Crop Improvement Association, Okemos, MI) were planted in 38and 76-cm row widths at East Lansing on June 16, 2010 and June 8, 2011 and in 38-, 51-, and
76-cm row widths at Richville on June 10, 2010 and June 6, 2011. Beans were planted using a
John Deere split-row planter (Deere and Company, Moline, IL) at East Lansing and two
Monosem planters (Monosem Incorporated, Edwardsville, KS) at Richville. Planter settings
were adjusted to attain three target populations for each variety-row width combination. Target
populations were 196,500, 262,000, and 327,500 plants ha

-1

for black beans and 148,000,

-1

196,500, and 262,000 plants ha for small red beans. Populations were held constant across row
widths, leading to differences in within-row plant spacing for each population (Table 2.1). Each
64

combination of row width and population was repeated with two weed control systems: a total
postemergence (POST) herbicide treatment and a weed-free control. The POST treatment was
applied at the V2 stage of dry bean and consisted of imazamox (Raptor, BASF Corporation,
Research Triangle Park, NC) at 35 g ai ha

-1

Research Triangle Park, NC) at 28 g ai ha

-1

plus bentazon (Basagran, BASF Corporation,
applied with 1% v/v of crop oil concentrate

(Herbimax, Loveland Products Inc., Loveland, CO) and 2.8 kg ha
(Actamaster, Loveland Products Inc., Loveland, CO).

-1

ammonium sulfate

The PRE treatment consisted of S-

metolachlor (Dual II Magnum, Syngenta Corporation, Wilmington, DE) at 1.4 kg ai ha
halosulfuron (Permit, Gowan Company, Yuma, AZ) at 35 g ai ha

-1

-1

plus

applied preemergence (PRE)

within 24 h of planting followed by hand-weeding to maintain weed-free conditions. The POST
treatment was used to assess the effect of row width and population on weed dynamics, while the
weed-free treatment was used to assess the effect of row width and population on the crop,
independent of weed competition.

At East Lansing, clethodim (Select Max, Valent USA

Corporation, Walnut Creek, CA) at 77 g ai ha

-1

with crop oil concentrate at 1% v/v was also

applied POST to all plots to relieve high grass (largely Setaria spp.) pressure. Herbicides were
applied using a custom-built, tractor-mounted compressed-air sprayer calibrated to deliver 177 L
-1

ha at 193 kPa using AirMix 11003 nozzles (Greenleaf Technologies, Covington, LA).
Precipitation data was obtained throughout the growing season from the Michigan
Automated Weather Network (http://www.agweather.geo.msu.edu/mawn/, Michigan State
University, East Lansing, MI), which maintains stations on the Michigan State University
Horticulture Farm less than a mile from both field locations. In addition, automated humidity
65

and temperature sensors (HOBO U23 Pro v2 Temperature/Relative Humidity Data Logger,
Onset Computer Corporation, Pocasset, MA) were attached to stakes and placed beneath the
canopy 8 cm above the ground in two replications of the 38- and 76-cm row widths at the high
and low black bean populations. These sensors recorded readings at 1-hr intervals.
The experimental design was a randomized complete block with regard to variety, row
width, and planting population. Each combination of these three factors was treated as a split
plot with regard to weed control system. Each treatment was replicated four times. Plot width
was 3 m, with seven, five, and four rows per plot for the 38-, 51-, and 76-cm row widths,
respectively. Plot length ranged from 9.1 to 10.7 m, depending on location. After the crop fully
emerged, stand counts were taken in two 3-m sections of row in each plot to determine actual
emerged plant populations.

Dry bean populations. In many cases, actual early-season stand counts fell short of (or, less
commonly, exceeded) the target populations (Table 2.2). However, within each row width at
each site, the highest target population had the highest stand count and the lowest population in
the lowest stand count. In fifty-four of the sixty site-year-class-width-population combinations,
stand count was similar enough to target population to essentially maintain the proper
relationship of populations among treatments; in six cases stand count was dramatically lower
than target planting population, although still higher than the next lowest population within the
row width. Four of these six cases occurred at Richville in 2010 and may have been the result of
poor emergence in dry soil (Table 2.3). The small number of interactions between population
and row width suggests that the effect of most of these variations from target population had
limited impact on the results of the studies.

66

Crop canopy development. Measurements of canopy light interception were taken in all weedfree plots at 1-2 wk intervals from approximately 3 wk after planting (WAP) through canopy
senescence. Light measurements were taken from three locations in each plot above and below
the canopy using a SunScan Probe (SunScan Canopy Analysis System, Dynamax Inc., Houston,
TX), a 1- x 0.013-m wand containing 64 light sensors, connected to a hand-held computer
(Recon, Trimble Navigation Limited, Sunnyvale, CA). The wand was placed perpendicular to
the rows above and below the canopy of the center rows. The percentage of light intercepted by
the canopy was calculated by dividing the below-canopy photosynthetically active radiation
(PAR) by the above-canopy PAR and averaging the three measurements from each plot.
Measurements were taken around solar noon in full or nearly full sun, or under uniformly
overcast conditions.

2

Weed emergence and biomass. Two 0.25-m quadrats were established in each POST-treated
plot after the herbicide application controlled most of the early-emerging weeds. These quadrats
were cleared of any surviving weeds, and newly emerging weeds were counted and recorded two
to three times during the growing season. Prior to harvest, above-ground weed biomass in each
POST-treated plot was harvested and dry weight was recorded.
In addition to counts of emerging weeds, weed biomass samples were taken shortly
before harvest. Depending on weed population, area for the biomass samples ranged from four
2

0.25-m subsamples to a swath 1 m wide running the entire length of the plot. The species of
harvested weeds was noted. The predominant weeds at Richville were common lambsquarters
(Chenopodium album L.) and pigweed (Amaranthus spp.). The predominant weeds at East

67

Lansing were common lambsquarters, pigweed, velvetleaf (Abutilon theophrasti Medik.), and
common ragweed (Ambrosia artemisiifolia L.)

Insect monitoring and feeding. Western bean cutworm (WBC) populations were monitored
using milk jug pheromone traps at each location. Moths trapped were counted and removed at
least once weekly, and the number caught was reported to Michigan State University
entomologists maintaining a statewide trap network, which allowed the monitoring of regional
WBC populations. At the R5-R6 stages of bean development, two 1-m sections of the outside
rows were randomly sampled for WBC damage in all POST-treated plots. The total number of
pods and number of pods with apparent WBC damage were noted in order to calculate the
percentage of pods with WBC damage.

Disease monitoring.

Each year, all studies were repeatedly examined for white mold

(Sclerotinia sclerotiorum (Lib.) de Bary) infection, but no white mold infection was observed in
any site-year. However, at East Lansing in both years and at Richville in 2011, a pod spot
fungus (Alternaria alternata (Fr.) Keissl. (1912)) was noted on bean pods prior to harvest.
Alternaria was evaluated visually on a 0-100% scale in weed-free plots and by randomly
sampling 50 pods from the border rows of each of the POST-treated plots. Pods were evaluated
individually on a scale of 0-3: “0” represented virtually no infection, “1” infection on less than
half of the pod surface, “2” infection over the majority of the pod surface, and “3” infection of
the entire pod surface or infection that resulted in significant deformation.
infection was then calculated for each plot.

68

Average pod

Maturity, lodging, and yield. At East Lansing only, bean plants failed to senesce evenly, and
many plots contained green plants at harvest. Immediately before harvest, all plots were visually
evaluated for maturity on a scale of 0-100 with “0” representing no observable yellowing and
“100” representing all plants fully senesced.

At East Lansing in 2011, maturity was also

evaluated about a month before harvest.
Immediately prior to harvest, all plots were visually evaluated for lodging.

The

evaluation system was a 0-5 scale in which “0” indicated no lodging, and “5” indicated that all
plants were prostrate. A preharvest desiccation treatment of glyphosate (Roundup PowerMax,
Monsanto Company, Saint Louis, MO) at 0.84 kg ae/ha was applied to all plots at both locations
in 2011. Dry bean plots were direct harvested with a small plot combine (Massey-Ferguson
8XP, AGCO, Duluth, GA) with a 1.5-m header. The center two, three, and five rows were
harvested in the 76-, 51-, and 38-cm row width plots, respectively. Dry bean yield was adjusted
to 18% moisture. Harvest dates were September 6 at Richville and September 29 at East
Lansing in 2010 and October 6 at Richville and October 18 at East Lansing in 2011.

Economic analysis. Gross margins were calculated at grower-received prices of $0.66, $0.88,
and $1.10 kg-1 for black beans and $0.66, $0.99, and $1.21 kg-1 for small red beans paired with
dealer seed prices of $0.77 and $1.25 kg-1 for black beans and $0.77 and $1.43 kg-1 for small red
beans (USDA-AMS 2011). The low and high prices were approximately the lowest and highest
prices occurring in Michigan during the last five years, while the middle grower-received prices
were those currently being contracted by Michigan growers (Varner 2012 pers. corr.). Gross
margin was calculated using the formula:
Gross margin = (Y × Pgrower) – ((SR*Pdealer) + Pherbicide)

69

Where Y was yield, Pgrower was the bean price received by growers, SR was the seeding rate,
Pdealer was the seed price paid by growers, and Pherbicide was the cost of the POST herbicide
treatment (Local herbicide prices 2011). Costs other than seed and weed control were not
considered. There is no cost associated with changing row widths unless a new planter is
purchased in order to plant at the desired row width.

Statistical analysis. Data were subjected to ANOVA using PROC MIXED in SAS 9.2 (SAS
Institute Inc., Cary, NC), and treatment means were compared using Fisher’s protected LSD at
the α < 0.05 level of significance. Some data were log or square root transformed for analysis,
but all data are presented untransformed. Data were combined over population, row width, and
herbicide when no significant interactions were present. Data could not be combined across
years due to significant interactions in almost all cases. A regression analysis relating yield to
observed final dry bean population was performed in SigmaPlot 12 (Systat Software Inc.,
2

Chicago, IL), but no model with a significance level above R =0.06 could be generated.

Results and Discussion
Growing conditions. Rainfall from June through September in 2010 was lower than the 30-yr
average at both locations (Table 2.3). At East Lansing, above-average rainfall in June and nearaverage rainfall in July and September allowed dry bean plants to develop normally despite a dry
August. At Richville, rainfall was well below average in every month, and much of the June
rainfall came prior to planting; by the beginning of September, most bean plants had senesced,
rendering September rainfall irrelevant. Only 11 cm of rain fell at Richville between planting
and harvest, resulting in stunted crop plants and little weed growth.
70

In 2011, rainfall from June through September was slightly above the 30-yr average at
East Lansing, while at Richville it was once again below the 30-yr average. At East Lansing,
rainfall was slightly to moderately below average in three months, but in late July a single
weather system brought 16 cm of rain over a three-day period, making that month’s total
dramatically above average and supplying ample water for pod formation, especially with
moderate rainfall continuing through September. At Richville, rainfall was again below average
in every month, but all rainfall in June occurred after planting, and monthly totals were higher
than those of the previous year in July and August. Rainfall was sufficient for normal plant
development, and pod filling continued into September.

Crop canopy development. Dry bean canopy development progressed similarly in three of the
four site-years; at Richville in 2010 under drought conditions (Table 2.3), canopy development
was slower and was less complete at its peak. In 2010, canopy closure (as measured by percent
interception of photosynthetically active radiation) was generally greater in small red bean than
in black bean at every measurement time at both sites. In 2011, canopy closure at East Lansing
was greater in small red bean from 3-6 weeks after plantings (WAP), while at Richville, canopy
closure was greater in small red bean from 3-4 WAP and again from 9-12 WAP.
Row width effects on canopy development.

In seven of eight site-year-class combinations,

narrow rows improved canopy closure during a portion of the growing season; the sole exception
occurred at Richville in 2010 in black bean.
At East Lansing in 2010, black bean canopy closure was much greater in 38-cm rows
than in 76-cm rows from 5 to 10 WAP, with difference ranging from 9 to 21 percentage points
(Figure 1). Peak canopy closure occurred 8 WAP in 38-cm rows at 94%, while in 76-cm rows,

71

canopy closure peaked a week later at 77%.

At East Lansing in 2011, 76-cm rows actually

provided a canopy closure advantage at 6 WAP; however, from 8-12 WAP, canopy closure was
again greater in 38-cm rows than in 76-cm rows (Figure 2a). Peak canopy closure was 96% in
38-cm rows and only 92% in 76-cm rows.

At Richville in 2011, black bean canopy closure

showed no clear trend with regard to row width from 3-9 WAP; however, from 10-11 WAP,
canopy closure was greater in 38- or 51-cm rows than in 76-cm rows by 5-9%, and at 12 WAP,
closure in 38-cm rows was still greater than in 76-cm rows (Figure 3a). Canopy closure peaked
at 83% in 38-cm rows, 80% in 51-cm rows, and 75% in 76-cm rows.
Differences in small red bean canopy closure between 38- and 76-cm rows were
generally not as great as the differences between wide and narrow rows for black beans.
However, at East Lansing in 2010, canopy closure was significantly greater in 38-cm rows than
in 76-cm rows at 5, 6, and 10 WAP (Figure 4a). At East Lansing in 2011, canopy closure was
greater in 38-cm rows than in 76-cm rows by 1-3% from 8-12 WAP, although this difference
was only significant at 10 WAP (Figure 5a). At Richville in 2011, canopy closure at 5 WAP was
greater in 76-cm rows than in either 38- and 51-cm rows; however, at 9 WAP, canopy closure
was greater in 38-cm rows than in 51- or 76-cm rows, and at 10 and 12 WAP, canopy closure
was still greater in 38-cm rows than in 76-cm rows (Figure 6a).
At Richville in 2010, black bean canopy closure was higher in 38-cm rows than in 51- or
76-cm rows at 3 WAP; however, after 4 WAP, canopy closure was not different between 38-cm
rows and other row widths (Figure 7a). Peak canopy closure was only 60%. In small red bean,
the effect of row width on canopy closure showed no clear trend at any time during the growing
season, and canopy closure peaked between 65 and 70% (Figure 8a).

72

Early in the growing season when canopy closure was less than 50% and plants were in
the vegetative or early flowering stages, canopy closure was sometimes erratic and was
occasionally greater in wide rows, possibly due to the closer within-row spacing at a given
population in wide rows. However, except under serious drought conditions (at Richville in
2010), canopy closure was improved by narrow rows as canopy closure exceeded 50% in mid- to
late-season, which corresponds to full flowering and pod filling; this is likely the time when dry
beans can benefit the most from increased light interception. This agrees with the results of
Blackshaw et al. (1999) and Ziviani et al. (2009) who found that canopy closure in navy and
carioca bean increased with decreasing row width. However, Blackshaw et al. (2000) reported
earlier development of the small red bean canopy in narrow rows, while these studies found most
of the canopy closure benefit to occur in mid- to late-season. Peak canopy closure was higher in
narrow rows in three site-year-class combinations. The effect of greater canopy closure in
narrow rows tended to be more pronounced in black beans than in small red beans, probably due
to the larger size of individual small red bean plants.
Planting population effects on canopy development. The effect of planting population on canopy
closure was less pronounced than that of row width. In seven of the eight site-year-class
combinations, the high population resulted in greater canopy closure than the low and/or medium
populations in at least one circumstance, but these effects were rarely large, and were never
consistent across a large portion of the growing season.
At East Lansing in 2010, there was a population interaction with row width with regard to
black bean canopy closure, with the effect of population differing between 38- and 76-cm rows.
In 38-cm rows, canopy closure was greater at 327,500 seeds ha

-1

than at 196,500 seeds ha

-1

at 6

WAP (Figure 9a). However, in 76-cm rows, canopy closure was greater at 196,500 seeds ha
73

-1

than at 327,500 seeds ha

-1

at 6 WAP, and from 9-10 WAP, canopy closure at both 196,500 and

-1

262,000 seeds ha was greater than at 327,500 seeds ha

-1

(Figure 9b). At East Lansing in 2011,

planting population had little effect on black bean canopy closure, with the 195,500 and 262,000
seeds ha

-1

populations occasionally lagging slightly behind the 327,500 seeds ha

-1

population

(Figure 2b). At Richville in 2010, black bean canopy closure at 3 and 7 WAP was greater at
327,500 seeds ha

-1

-1

than at 196,500 seeds ha (Figure 7b). No other differences were observed.

At Richville in 2011, black bean canopy closure at 3 WAP was greater at 327,500 seeds ha

-1

-1

than at 262,000 seeds ha , after which population did not affect canopy closure (Figure 3b).
In small red bean, canopy closure tended to lag at the lowest population of 148,000 seeds
-1

ha . At East Lansing in 2010, this population lagged behind canopy closure of the 196,500 and
262,000 seeds ha

-1

planting populations from 5-6 WAP; differences ranged from 13-18%

(Figure 4b). At East Lansing in 2011, canopy closure at 148,000 seeds ha

-1

was lower than at

the two higher populations at 3, 4, and 6 WAP (Figure 5b). At Richville in 2010, closure at
-1

148,000 seeds ha lagged behind 262,000 seeds ha-1 only at 3 WAP (Figure 8b). At Richville in
2011, closure at 148,000 seeds ha

-1

was lower than at 262,000 seeds ha

-1

only at 3-4 WAP

(Figure 6b).
Canopy closure tended to increase with increasing population when there was an effect of
population, but this effect occurred mainly in the early season and was neither consistent nor
-1

clearly of practical significance. The low population in small red beans (148,000 seeds ha )

74

lagged in canopy closure enough to be of some concern; otherwise, dry beans had the ability to
compensate for a wide range of populations, intercepting more light per plant at lower
populations. Although Blackshaw et al. (2000), reported an in light interception at high small
red bean population, this may be due to that study’s use of a much higher population range, with
-1

their high population being 500,000 seeds ha .

Weed suppression.

Weed pressure was greater at both sites in 2011 than in 2010, with

aboveground weed biomass in the weediest treatments being over 400 kg ha
compared with 123 kg ha

-1

-1

at both sites

-1

at East Lansing in 2010 and only 78 kg ha at Richville in 2010. At

East Lansing in 2010, black bean plots had about 3.5 times more weed biomass than small red
bean plots, but class did not have a significant effect on weed biomass at Richville or in 2011.
Weed biomass. In six of the eight class-site-year combinations, narrow rows reduced weed
biomass, with both exceptions being at Richville in 2010 (Table 2.4).
At East Lansing in 2010, weed suppression in black beans was greatly enhanced in
narrow rows; weed biomass was 81% lower in 38-cm rows than in 76-cm rows.

Weed

suppression in small red beans was also strongly enhanced in narrow rows, being reduced by
75% in 38-cm rows compared with 76-cm rows. These results corresponded with increases in
canopy closure in 38-cm rows during much of the growing season. At East Lansing in 2011,
weed suppression in black beans was again enhanced by narrow rows; in 38-cm rows, weed
biomass was 61% lower than in 76-cm rows. Again, this corresponded to greater canopy closure
in 38-cm rows over much of the growing season.

In small red beans, there was a row width

interaction with planting population with regard to weed biomass (Table 2.5). Weed biomass

75

differences at 148,000 and 196,500 seeds ha

-1

-1

were non-significant, but at 262,000 seeds ha ,

weed biomass was 81% lower in 38-cm rows than in 76-cm rows. This corresponded to a small
increase in canopy closure in narrow rows.
At Richville in 2011, the effect of row width on weed biomass in black beans was nonsignificant at a significance level of α < 0.05; however, this seems to have been largely due to
excessive variance, and at a significance level of α < 0.1, weed suppression improved with
narrowing rows. Weed biomass averaged 430 kg ha
greater than the 181 kg ha

-1

-1

in 76-cm rows, which was significantly

in 38-cm rows, with 51-cm rows intermediate at 302 kg ha

-1

(Table

2.4). This corresponded with the late season improvement in canopy closure with narrower
rows. In small red bean, 38-cm rows improved weed suppression relative to both 51-cm and 76cm rows by 76% and 71%, respectively. This corresponded to the late-season canopy closure
advantage in 38-cm rows over 76-cm rows and the early- to mid- season lag in canopy closure in
51-cm rows.
At Richville in 2010, row width had no effect on weed biomass in either class. This
corresponded to the presence of only small and erratic early-season differences in canopy closure
(Table 2.4).
Although planting population impacted canopy closure at some measurement dates,
planting population did not affect weed biomass in any site-year-class combination (Table 2.6).
This was similar to the findings of Malik et al. (1993) that dry bean population did not alter weed
biomass. Blackshaw et al. (1999, 2000) found that weed biomass was reduced at high dry bean
populations, but their higher populations were much higher than the high populations used in
these studies.

76

Weed emergence. Weeds that emerged following the postemergence (POST) herbicide
application tended to stay small and suppressed beneath the crop canopy for the duration of the
growing season, so it seems unlikely that these weeds significantly affected yield. As such, weed
biomass is a more relevant measure of the actual weed control benefit derived from optimum
row width than weed emergence. However, late-emerging weeds could still produce seed late in
the year after crop removal.
At East Lansing in 2010, weed emergence following the POST application was very low,
consisted largely of annual grass, and virtually ended by early July. At Richville in 2010, no
more than two emerging weeds were observed in any plot at any counting time. Neither row
width nor planting population significantly impacted weed emergence at either site in 2010.
At East Lansing in 2011, weed emergence was much higher, and when surviving weeds
that had emerged since the POST application were counted in September, narrow rows reduced
either overall weed emergence or survival of emerged weeds in black beans. In 76-cm rows,
-1

there were 346,000 recently emerged weeds ha , while in 38-cm rows, there were only 48,000
-1

recently emerged weeds ha . This was consistent with the increase in canopy closure observed
over much of the season in 38-cm rows. Row width did not significantly affect weed emergence
in small red bean.
At Richville in 2011, weed emergence was again low but was much higher than in 2010.
Across both classes, 76-cm rows reduced weed emergence compared with 51-cm rows with 38cm rows intermediate. Weed emergence averaged 68,000 weeds ha
-1

with 31,000 weeds ha in 76-cm rows and 46,000 weeds ha

-1

-1

in 51-cm rows compared

in 38-cm rows. This is consistent

with the early season advantage of 76-cm rows and disadvantage in 51-cm rows observed in this
77

study rather than with the late-season canopy closure that apparently affected overall weed
biomass.
Planting population was never found to affect weed emergence in any site-year-class
combination.
Summary. Narrow rows enhance the crop’s ability to suppress weeds under normal conditions
by improving canopy closure. The increased mid- to late-season canopy closure often observed
in narrow rows would be expected to result in increased weed suppression, and this did occur;
weed biomass was lower in narrow rows in every case when narrow rows resulted in increased
canopy closure. The exception occurred when narrow rows did not result in an increase in midto late-season canopy closure due to drought. Weed emergence following the POST application
was less influenced by row width than was weed biomass, but when it was influenced, it
appeared to respond to early- to mid-season canopy closure enhancement. Although planting
population had sporadic impacts on canopy closure, these impacts were too small to significantly
affect weed suppression; in no case did planting population have any significant impact on weed
biomass or emergence.

Disease development and humidity.
White mold. The chief disease on which concern has focused with regard to narrow rows in dry
beans is white mold; frequent scouting was undertaken for this disease during every site-year,
but none was ever observed.
Alternaria infection. However, in three of the four site-years, Alternaria pod spot infection
occurred; infection levels were similar in all three of these site-years. Virtually no infection was
observed at Richville in 2010, presumably due to inadequate moisture for disease development,

78

so disease was not evaluated for this site-year. In every site-year in which Alternaria was
evaluated, infection was much more severe in small red beans than in black beans, with average
disease scores (on a 0-4 scale) ranging from 1.2-1.6 in black beans and 2.0-2.4 in small red beans
and average disease ratings (on a 0-100% scale) ranging from 28-39% in black beans and 7075% in small red beans. While most Alternaria infection occurred on the surface of the pods,
moderately to severely infected pods often contained shrunken or aborted beans, and some beans
in severely infected pods showed visible signs of infection, so Alternaria did appear to be having
at least a small impact on yield and on bean quality.
Although Alternaria development began late in the season as pods began to fill, at which
time canopy closure tended to be more complete in narrow rows, Alternaria infection was not
exacerbated by narrow rows in any site-year. Planting population also did not have a large or
consistent effect on Alternaria severity.
Humidity. Because white mold and some other diseases benefit from moist conditions in the
canopy, an increase in humidity within the canopy in a given treatment may indicate a potential
disease risk in that treatment even if no disease exacerbation was observed in these studies. In
the early season prior to extensive canopy development and very late season after the beginning
of senescence, there was very little difference between the treatments, so average humidity
during the period of significant canopy development was used to determine whether there were
differences between treatments. Average humidity was similar at East Lansing in 2010 (85%)
and at Richville in 2011(89%); it was higher at East Lansing in 2011 (94%) and much lower at
Richville in 2010 (75%). In two of the four site-years, humidity was lower in wide rows at high
populations than in narrow rows or than in wide rows with low population (Table 2.7).

79

At East Lansing in both 2010 and 2011, average humidity between July 25 and
September 1 was 6-9 percentage points less in 76-cm rows at 327,000 seeds ha

-1

than at other

treatments. At Richville in both 2010 and 2011, humidity did not differ among row widths or
populations.
These data suggest that narrow rows may have the potential to increase humidity
compared with wide rows, but the reason wide-row humidity was only lower at high population
is not obvious, and the increase in humidity in narrow rows was inconsistent between sites.
However, the possibility that these 6-9% increases in average humidity in narrow rows may
sometimes be biologically significant, potentially creating a disease-enhancing environment,
cannot be ruled out. In fact, an increase in humidity may be the underlying reason for the higher
levels of white mold observed in narrow rows by Park (1993) and Peachey et al. (2006).
According to Hagedorn and Inglis (1986), Alternaria alternata may also cause a destructive leaf
spot disease in dry beans in cool, wet conditions; it is possible that if this occurs, narrow rows
could increase disease development, but no such increase was observed for Alternaria as a pod
rot. It was somewhat surprising not to observe higher humidity at higher populations, especially
in light of the finding of Saindon et al. (1995) and Vieira et al (2010) that white mold infection
increases with plant density.

Insect feeding. One insect pest affecting dry bean in Michigan is western bean cutworm.
Number of male WBC moths trapped varied by site and year. At East Lansing in 2010, 421
moths were caught in the two traps; similarly 412 moths were caught in 2011. At Richville, trap
catch was 226 and 118 moths in 2010 and 2011 respectively. Peak WBC flight occurred during
the week of July 18 in 2010; it occurred a week later in 2011 during the week of July 24. Bean
80

leaf beetle (Cerotoma trifurcata Forster) was also present, but the presence of boring in pods was
considered diagnostic of western bean cutworm. Due to the superficial nature of bean leaf beetle
feeding on pods, only suspected western bean cutworm feeding was evaluated. Suspected
western bean cutworm (WBC) feeding varied between site-years. In 2010, 2.5% of pods at East
Lansing were damaged by WBC, while at Richville 1.5% of pods were damaged. In 2011, 3.3%
of pods at East Lansing were damaged by WBC, while at Richville, feeding was higher than at
other site-years but differed by class, with 5.2% of black bean pods being damaged compared
with 8.2% of small red bean pods. In four of eight site-year-class combinations, WBC feeding
was reduced in narrow rows within at least one population (see Table 2.8 for main effect and
Tables 2.9 and 2.10 for interactions). Planting population had no significant main effect on
WBC feeding in any site-year (Table 2.11).
At East Lansing in 2010, there was an interaction between row width and planting
-1

population with regard to WBC damage in black bean (Table 2.9) at 327,500 seeds ha , WBC
damage was higher in 76-cm rows than in 38-cm rows, while row width was non-significant at
lower populations. There was also an interaction with regard to WBC damage in black bean at
-1

Richville in 2011 (Table 2.10); within 196,500 and 262,000 seeds ha , WBC damage was
-1

greater in 76-cm rows than in 38-cm rows with 51-cm rows intermediate; at 327,000 seeds ha ,
row width was not significant. In 38-cm rows, WBC damage was higher at 327,000 seeds ha

-1

than at either of the lower populations; in the other row widths, planting populations had no
effect.

81

At East Lansing in 2011, WBC damage was greater in 76-cm rows than in 38-cm rows
across both classes and all populations. In 76-cm rows, 4.2% of pods were damaged while in 38cm rows, only 2.4% were damaged.
At Richville in 2010, WBC feeding was very low, and neither row width nor population
had a significant effect on WBC damage in either class. Except at East Lansing in 2011, neither
row width nor planting population affected WBC feeding in small red bean.
In every site-year except Richville in 2010, which experienced low insect pressure,
narrow rows reduced WBC damage in black bean in at least some populations, and this effect
also occurred in one site-year in small red bean; narrow rows never increased WBC damage. In
each case of reduced WBC damage, canopy closure was increased in narrow rows during at least
part of the WBC active period from mid-July to mid-August. Moreover, in wide rows, there was
typically a gap between rows for much, if not all, of the WBC active period whereas the gap was
smaller or non-existent in narrow rows. If WBC moths prefer a more open canopy for easier
flight or typically enter the canopy from the side rather from above, this may explain the
apparent reduction of WBC damage in narrow rows. The effect may be less pronounced in small
red beans because of their more sprawling growth habit and less distinct rows. It is also possible
that those plots that demonstrated reduced WBC feeding had higher below-canopy humidity,
which could encourage the development of fungal pathogens that infect WBC larvae (Sprague et
al. 2010). The effect of planting population was erratic and conflicting, and the reason for the
observed effects is not clear. This appears to be the first study that has related row width and
planting population to insect damage in dry bean.

82

Lodging. Lodging was always at least 80% greater in small red bean than in black bean in every
site-year based on a 0-5 scale. In four of the eight site-year-class combinations, narrow rows
reduced lodging (Table 2.12) while planting population had very little effect on lodging (Table
2.13).
At East Lansing in 2010, neither row width nor population significantly affected lodging
in either class. At Richville in 2010, almost no lodging was observed in black bean, and
treatments did not significantly differ; however, in small red bean, 38-cm rows reduced lodging
compared with both 51-cm and 76-cm rows. Planting population never affected lodging in 2010.
At East Lansing in 2011, row width again did not have a significant effect on lodging in
black bean. In small red bean, lodging was again lower in 38-cm rows than in 76-cm rows,
averaging 3.25 in 38-cm rows and 4.2 in 76-cm rows. Planting population did not significantly
impact lodging in either class. At Richville in 2011, black bean lodging was reduced in both 38and 51-cm rows compared with 76-cm rows. Planting population did not affect lodging. In
small red bean, lodging was also reduced in both 38- and 51-cm rows compared with 76-cm
rows. Lodging was also reduced at medium population compared with low population, with
medium population intermediate; this was the only significant effect of population on lodging.
In half of the site-year-class combinations, 38-cm rows reduced lodging compared with
76-cm rows, and in half of the cases that included 51-cm rows, these also reduced lodging
compared with 76-cm rows; one of the exceptions to both effects was a case in which virtually
no lodging was observed, and 76-cm rows never produced an advantage with regard to lodging.
Significant differences ranged from 0.41 to 0.95 on the 0-5 scale. This advantage of narrow
rows with regard to lodging may be attributed to the mutual support of intertwining rows, which
presumably prevents individual plants from leaning into the inter-row space. It confirms the

83

observation of Horn et al. (2000) that lodging increases with increasing row width. Planting
population had no clear effect on lodging, with lodging being minimized at medium population
in the one case in which a significant effect was detected.

Maturity. At East Lansing, but not at Richville, beans failed to mature at an even rate across the
field. In 2010, maturity was rated once 1 wk prior to harvest; in 2011, it was rated twice, 7 wk
and 2 wk prior to harvest. Black beans generally matured more quickly than small red beans. In
three of four site-classes, plants growing in narrow rows were more mature shortly before
harvest in at least one herbicide treatment, while 7 wk before harvest, plants growing in wide
rows were more mature in both classes (Table 2.14). Neither planting population (Table 2.15)
nor herbicide treatment affected maturity.
At East Lansing in 2010, black bean maturity 1 wk prior to harvest was greater in 38-cm
rows than in 76-cm rows. In small red bean, treatments did not significantly differ from one
another.
In 2011, black bean maturity 2 wk prior to harvest was again greater in 38-cm rows than
in 76-cm rows. In small red bean in the weed-free treatments only, average maturity was greater
in 38-cm rows than in 76-cm rows (Table 2.16). At 7 wk prior to harvest in 2011, maturity in
both classes was greater in 76-cm rows than in 38-cm rows.
As dry beans began to mature, maturation was initially more rapid in 76-cm rows, but as
physiological maturity approached, maturation became more rapid in 38-cm rows, to the point
that 38-cm treatments caught up to or passed 76-cm treatments in maturity.

84

Dry bean yield. Crop yield was similar at East Lansing in 2010 and Richville in 2011, both of
which were moderately dry site-years. At East Lansing in 2010, black bean yield averaged 3003
-1

-1

kg ha , while small red bean yield averaged 2519 kg ha . At Richville in 2011, black bean
yield averaged 2893 kg ha

-1

-1

while small red bean yield averaged 2461 kg ha . At East Lansing

in 2011, abundant rainfall led to higher yields, with black bean averaging 4198 kg ha
-1

red bean averaging 3080 kg ha .

-1

and small

In all three of these site-years, black bean yield was

significantly higher than small red bean yield. At Richville in 2010, dry conditions led to
extremely low yields that were similar between classes: 1694 kg ha
ha

-1

in small red bean.

-1

in black bean and 1683 kg

Herbicide treatment did not interact with row width or planting

population, so yields were combined across herbicide treatments.
In four of the eight site-year-class combinations, yield was higher in narrow rows than in
wide rows (Table 2.17). At East Lansing in 2010 and 2011, black bean yield was 17% and 15%
higher in 38-cm rows than in 76-cm rows. Small red bean yield was also 18% higher in 38-cm
rows than in 76-cm rows in 2011. In each of these cases, the increase in yield corresponded with
both a mid- to late-season increase in canopy closure and an overall decrease in weed pressure in
38-cm rows. At East Lansing in 2011, in both classes, yield was 7-11% higher in weed-free
treatments than in POST treatments. At Richville in 2011, small red bean yield was 11% higher
in 38-cm rows and 9% higher in 51-cm rows than in 76-cm rows. In 38-cm rows, canopy closure
was significantly greater than in 76-cm rows during late-season, and canopy closure in 51-cm
rows was often not statistically different from that in 38-cm rows. However, 51-cm rows never
resulted in a clear advantage in canopy closure over 76-cm rows, and 38-cm rows reduced weed

85

biomass relative to both 51- and 76-cm rows, so the reason for the increase in yield in 51-cm
rows compared with 76-cm rows is obscure; it may be that the canopy closure advantages in 51cm rows compared with 76-cm rows are real even if non-significant or that the advantage was
due to more efficient development of the root system.
At East Lansing in 2010, small red bean yield was not different between treatments
despite a significant increase in mid- to late-season canopy closure in 38-cm rows (Tables 2.17,
18). At Richville in 2011, black bean yield also was not affected by row width or population
despite a late season canopy closure advantage in 38-cm rows and several other narrow row
advantages. However, this may have been due to the erratic differences in canopy closure
observed in the early- to mid-season.
At Richville in 2010, black bean yield was marginally higher (α ≤ 0.1) in 76-cm rows
than in 51-cm or 38-cm rows, while in small red bean, yield was higher in 38- and 76-cm rows
than in 51-cm rows at medium population in weed-free treatments only. Black bean yield was
also 9% higher in POST than in weed-free across all row widths and populations. Small red
bean yield was 12% higher at 196,500 seeds ha

-1

than at 262,500 seeds ha

-1

across all

combinations of row width and herbicide. These unusual results were presumably the result of
drought-induced randomness in which yield was determined more by small differences in soil
moisture than by the factors studied.
Planting population never had a significant overall effect on yield in black bean (Table
2.18). In small red bean at Richville in 2010 under drought conditions, small red bean yield was
higher at medium population than at high populations, while at East Lansing in 2011, yield was
higher at high population than at low population. In the other two site-years, planting population
did not affect small red bean yield.
86

In four of the eight site-year-class combinations, yield was higher in 38-cm rows than in
76-cm rows (Table 2.17). In the only one of these four combinations that included 51-cm rows,
yield was higher in 51-cm rows than in 76-cm rows. Two of the four exceptions occurred at
Richville in 2010 when drought resulted in low yields and seemingly arbitrary differences
between treatments. In all but one of these cases (the yield increase in 51-cm rows), the yield
advantage of narrow rows is readily explained by advantages in canopy closure in narrow rows,
possibly augmented by other observed benefits of narrow rows such as weed suppression and
reduced insect feeding. It is also possible that some of the yield benefit of narrow rows resulted
from undocumented improvements in root system efficiency. In no site-year-class combination
did wide rows increase yield at the α ≤ 0.05 level, and only in one class at Richville in 2010 did
wide rows have any positive effect on yield. This trend toward higher yield in narrow rows
confirms the findings of Blackshaw et al. (1999, 2000), Redden (1987), and Malik et al. (1993)
that yield is often, though not always, higher in narrow rows; Grafton et al. (1988) and Goulden
(1976) found even more consistent yield advantages in narrow rows.
Only under drought conditions at Richville in 2010 did population affect yield, and this
effect was not consistent between classes or row widths. This contrasted with the findings of
Blackshaw et al. (1999, 2000) that increasing population increases yield; however, Blackshaw et
al. much higher population ranges and used irrigation. Other studies (Malik et al. 1993, Goulden
1976, Redden et al. 1987) reported a more sporadic positive effect of increased population.
However, planting population had virtually no effect on yield in these studies.
In one site-year, weed-free yield was higher than POST yield in both classes, which was
likely the result of reduced weed competition; only in one class at Richville in 2010 did POST
treatments have higher yields than weed-free treatments.

87

Economic Returns. In each of the four site-year-class combinations in which yield increased in
38-cm rows and in the one case in which yield increased in 51-cm rows compared with 76-cm
rows, economic return also increased (Table 2.19). Under drought conditions at Richville in
-1

2010 in small red beans, high population (262,000 seeds ha ) reduced economic returns
compared with the 196,500 or 148,000 seeds ha

-1

planting populations (Table 2.20); in this case,

the high population also reduced yield compared with the low population. Otherwise, planting
population did not significantly affect economic returns (Table 2.21). Neither seed price paid by
the grower (the dealer price) nor bean price received by the grower (the grower price) affected
which differences were significant within the range of prices that occurred from September 2007
to February 2012. Dealer prices ranged from $0.77-$1.25 kg
1

-1

in small red bean. Grower prices ranged from $0.66-1.10 kg

in black bean and $0.77-1.43 kg
-1

-

in black bean and from $0.66-

-1

$1.21 kg in small red bean (USDA-AMS 2011).

Conclusions. At Richville in 2010, drought conditions were so extreme that virtually every data
set behaved differently than those from most or all of the other site-years. If this site-year is
discounted, narrow rows provided an advantage in the majority of site-year-class combinations
analyzed in six different variables: canopy closure, weed biomass, western bean cutworm
feeding, lodging, pre-harvest maturity, and yield. Weed density and emergence were only rarely
affected by row width, but when they were affected, narrow rows again provided an advantage.
Narrow rows appear to provide significant advantages under favorable environmental conditions.
The only variable in which wide rows offered a possible advantage was humidity; within the
88

high population, humidity was higher in narrow rows in half of the site-years observed, and high
humidity could be conducive to disease development.

However, no increase in disease

development was observed in narrow rows.
Since two of the four site-year-class combinations in which 51-cm rows were observed
were at Richville in 2010, it is difficult to draw conclusions about whether 51-cm rows function
more similarly to 38-cm rows or 76-cm rows, or are intermediate. With regard to yield, 51-cm
rows appear to provide the same advantages as 38-cm rows. No clear conclusions could be
drawn regarding the effect of 51-cm rows on lodging. With regard to canopy closure, 51-cm
rows varied from intermediate between 38-cm and 76-cm rows to similar to 38-cm rows, to
lower than either row width in the early season. With regard to weed suppression, 51-cm rows
were either intermediate between 38- and 76-cm rows or resulted in equally poor weed
suppression as 76-cm rows.
Planting population had virtually no effect on any variable except canopy closure, in
which high and medium populations occasionally resulted in higher canopy closure than low
populations, especially in small red beans.
Because 38-cm rows clearly confer advantages with regard to canopy closure, weed
suppression, lodging, and yield in many cases with few disadvantages relative to 76-cm rows,
and because the effect of 51-cm rows is less consistent, 38-cm rows appears to be the optimum
row width among those studied for Type II black and small red bean. Since planting at the high
population confers little or no advantage over the medium population, and the low populations
sometimes resulted in a lag in canopy closure, the medium populations should be the target
populations for growers of Type II black and small red beans regardless of row width.

89

Table 2.1. Target within-row plant spacings for Type II black and small red beans planted in
three row widths and at three planting populations.
Target black bean population ha
Row width

196,500

-1

262,000
327,500
______________
cm
10
8
8
6
5
4

______________

38 cm
51 cm
76 cm

13
10
7

90

Target small red bean population ha
148,000

196,500
262,000
______________
cm
13
10
10
8
7
5

______________

18
13
9

-1

Table 2.2. Early season stand counts in 38- and 76-cm rows at East Lansing and 38-, 51-, and 76-cm rows at Richville compared with
target planting populations.
East Lansing
Richville
Target planting
Populations (seeds
-1
ha )
Black beans
196,500
262,000
327,500

2010
38 cm

76 cm

________________

197,500
252,000
313,500

2011
38 cm

2010

76 cm

38 cm

-1_________________

187,500 168,000
222,500 226,000
286,500 263,000

Small red beans
148,000
149,500 162,000
141,000 139,500
196,500
211,000 207,500
175,500 173,000
262,000
289,000 267,000
246,000 236,000
* Stand count was at or below the next lowest target population.

76 cm

______________________

plants ha

175,500
246,000
305,000

51 cm

2011
38-cm

plants ha

51-cm

76-cm

-1 ______________________

174,000
226,000
327,500

163,000
210,000
274,000

165,500
196,500*
253,000*

183,000 183,000 131,000*
233,500 246,000 218,500
295,000 300,000 228,500*

141,000
187,500
233,500

131,000
148,000*
153,000*

138,500
179,000
222,000

142,000 148,000 135,000
175,500 186,500 165,500
238,500 221,000 205,000

91

Table 2.3. Monthly and 30-yr average precipitation at the Harry and Hazel Box Farm in East
Lansing, MI and the Saginaw Valley Research and Extension Center in Richville, MI in 2010
a
and 2011 .
Month

East Lansing
2011
30-yr avg.

2010

_______________

June

b

10.6 (4.1)

July
August
September

6.4
3.4
9.2

2010

cm _______________
4.4
8.4
17.8
6.4
6.7

Richville
2011

_______________

b

6.9 (2.8)

8.2
8.3
9.3

2.3
3.2
d

2.9

30-yr
c
avg.

cm _______________
3.8
9.3
3.4
7.6
5.8

7.6
8.0
10.7

Total
29.6
35.3
34.2
15.3
20.6
35.6
a
Michigan Automated Weather Network, http://www.agweather.geo.msu.edu/mawn/, Michigan
State University, East Lansing, MI
b
Precipitation data in parentheses is from the time of planting.
c
d

30-yr average for Caro, MI; none was available for Richville
Does not include September rainfall after harvest

Table 2.4. Main effect of row width on weed biomass in black and small red bean plots at East
a
Lansing and Richville in 2010 and 2011 .
East Lansing
2010
Row
width

Black

Small red

Richville
2011

2010

Black

Black

____________________________________

kg ha

Small red

2011
Black

Small red

-1 ____________________________________

38 cm
24 b
9b
174 b
102 a
33 a
181 b
111 b
51 cm
50 a
56 a
302 ab
470 a
76 cm
130 a
36 a
422 a
57 a
21 a
430 a
385 a
Mean
77
22
298
70
37
304
322
a
Means within each column that the same letter are statistically similar to one another at the α ≤
0.05 level of significance.

92

Table 2.5. Interaction of row width and planting population with regard to weed biomass in
a
small red bean at East Lansing in 2011 .
-1

Planting population (plants ha )
Row width

148,000

196,000

_________________________________

38 cm
76 cm

kg ha

344 ab
397 ab

262,000

-1 _________________________________

406 ab
308 ab

107 b
562 a

a

Means throughout the table that have the same lower-case letter are statistically similar to one
another at the α ≤ 0.05 level of significance.

Table 2.6. Main effect of planting population on weed biomass in black and small red bean plots
a
at East Lansing and Richville in 2010 and 2011 .
Planting
b
population

East Lansing
2010
Black

Richville
2011

Small red

Black

2010
Black

________________________________________

kg ha

Small red

2011
Black

Small red

-1 ________________________________________

Low
62 a
27 a
378 a
49 a
29 a
235 a
423 a
Medium
75 a
23 a
213 a
55 a
42 a
218 a
355 a
High
93 a
35 a
347 a
104 a
39 a
324 a
188 a
a
Means within each column that the same letter are statistically similar to one another at the α ≤
0.05 level of significance.
b
-1
-1
In black bean, Low = 196,500 seeds ha , Medium = 262,000 seeds ha , and High = 327,500
-1

-1

-1

seeds ha ; in small red bean, Low = 148,000 seeds ha , Medium = 196,500 seeds ha , and
-1

High = 262,000 seeds ha .

93

Table 2.7. Effects of row width and planting population on average humidity under the black
bean canopy during the period in which the canopy was well-developed at East Lansing and
a
Richville in 2010 and 2011 .
East Lansing
Planting
population
-1
(seeds ha )

Row width

2010

2011

Richville
2010

2011

------------------------ humidity (%) ----------------------196,500

76 cm

85.3 a

94.8 a

74.5 a

86.9 a

196,500
38 cm
85.7 a
95.1 a
74.8 a
88.3 a
327,500
76 cm
79.3 b
89.1 b
74.3 a
88.7 a
327,500
38 cm
87.9 a
96.0 a
75.2 a
87.2 a
a
Means within each column that the same letter are statistically similar to one another at the α ≤
0.05 level of significance.

Table 2.8. Main effect of row width on western bean cutworm damage in black and small red
a
bean in at East Lansing and Richville in 2010 and 2011 .
East Lansing
Row width

2010
(Small red only)

Richville

2011
(Both classes)

_____________________________

2010
(Both classes)

2011
b
(Small red only)

% of pods damaged _____________________________
38 cm
2.3 a
2.4 b
2.4 a
4.5 b
51 cm
1.4 a
5.1 ab
76 cm
2.7 a
4.2 a
1.7 a
6.0 b
a
Means within each column that the same letter are statistically similar to one another at the α ≤
0.05 level of significance.
b
Significance level of α ≤ 0.1

94

Table 2.9. Interaction of row width and population with regard to western bean cutworm
a
damage in black bean at East Lansing in 2010 .
-1

Planting population (seeds ha )
Row width

196,500

262,000
327,500
_________________________
% of pods damaged
38 cm
3.9 ab
1.4 bc
0.1 c
76 cm
1.1 bc
2.7 abc
5.9 a
a
Means throughout the table that have the same lower-case letter are statistically similar to one
another at the α ≤ 0.05 level of significance.
_________________________

Table 2.10. Interaction of row width and planting population with regard to western bean
a
cutworm damage in black bean at Richville in 2011 .
-1

Planting population (seeds ha )
Row width

196,500
__________________________

38 cm
51 cm
76 cm

3.6 c
4.4 bc
6.4 ab

262,000
327,500
__________________________
% of pods damaged
3.6 c
6.4 ab
5.2 abc
5.7 ab
6.7 a
4.9 abc

a

Means throughout the table that have the same lower-case letter are statistically similar to one
another at the α ≤ 0.05 level of significance.

Table 2.11. Main effect of planting population on western bean cutworm damage by planting
a
population in black and small red bean at East Lansing and Richville in 2010 and 2011 .
East Lansing
Planting
b
population

2010
(Small red only)

Richville

2011
(Both classes)

2010
(Both classes)

2011
(Small red only)

_______________________________

% of pods damaged _______________________________
Low
3.6 a
3.0 a
1.7 a
4.8 a
Medium
1.7 a
3.7 a
2.4 a
5.2 a
High
2.2 a
3.0 a
1.4 a
5.7 a
a
Means within each column that have the same letter are statistically similar to one another at
the α ≤ 0.05 level of significance.
b
-1
-1
In black bean, Low = 196,500 seeds ha , Medium = 262,000 seeds ha , and High = 327,500
-1

-1

-1

seeds ha ; in small red bean, Low = 148,000 seeds ha , Medium = 196,500 seeds ha , and
-1

High = 262,000 seeds ha .
95

Table 2.12. Main effect of row width on lodging in black and small red bean at East Lansing and
a
Richville in 2010 and 2011 .
East Lansing
2010
Row
width

Richville
2011

2011

Small
Small
Small
Black
Black
red
red
red
___________________________________________
___________________________________________
0-5 scale
38 cm
0.38 a
3.92 a
2.12 a
3.25 b
0a
0.62 b
1.83 b
2.83 b
51 cm
0a
1.17 a
1.83 b
2.96 b
76 cm
0.71 a
4.00 a
1.77 a
4.21 a
0.21 a
1.29 a
2.5 a
3.44 a
Mean
0.54
3.96
1.94
3.73
0.07
1.03
2.05
3.08
a
Means within each column that the same letter are statistically similar to one another at the α ≤
0.05 level of significance.
Black

Small
red

2010

Black

Table 2.13. Main effect of planting population on lodging in black and small red bean at East
a
Lansing and Richville in 2010 and 2011 .
East Lansing

Population
Low
Medium
High

b

Richville

2010
Small
Black
red

2011
2010
2011
Small
Small
Small
Black
Black
Black
red
red
red
_________________________________________
_________________________________________
0-5 scale
0.56 a
3.75 a
1.97 a
4.0 a
0a
1.00 a
1.81 a
3.25 a
0.56 a
3.94 a
1.81 a
3.62 a
0a
1.08 a
1.94 a
3.04 ab
0.50 a
4.19 a
2.06 a
3.56 a
0.21 a
1.00 a
2.42 a
2.93 b

a

Means within each column that the same letter are statistically similar to one another at the α ≤
0.05 level of significance.
b
-1
-1
In black bean, Low = 196,500 seeds ha , Medium = 262,000 seeds ha , and High = 327,500
-1

-1

-1

seeds ha ; in small red bean, Low = 148,000 seeds ha , Medium = 196,500 seeds ha , and
-1

High = 262,000 seeds ha .

96

Table 2.14. Main effect of row width on maturity in black and small red bean at East Lansing in
a
2010 and 2011 .
2010

2011

Late-season rating
Row width

Black

Mid-season rating

Small red

Late-season
rating
Black

Black
Small red
____________________________________________
%
30.1 b
16.6 b
92.0 a
46.9 a
22.5 a
81.6 b
38.5
19.6
86.8

____________________________________________

38 cm
76 cm
Mean

88.0 a
72.2 b
80.1

78.5 a
75.8 a
77.2

a

Means within each column that the same letter are statistically similar to one another at the α ≤
0.05 level of significance.

Table 2.15. Main effect of planting population on maturity in black and small red bean at East
a
Lansing in 2010 and 2011 .
2010

2011

Late-season rating
Planting
b
populations

Black

Mid-season rating

Small red

Black

Small red

Late-season
rating
Black

_____________________________________________

% _____________________________________________
Low
79.2 a
73.1 a
41.1 a
20.4 a
88.4 a
Medium
81.7 a
79.1 a
38.5 a
19.3 a
86.4 a
High
79.4 a
79.3 a
35.9 a
19.6 a
85.6 a
a
Means within each column that have the same letter are statistically similar to one another at
the α ≤ 0.05 level of significance.
b
-1
-1
In black bean, Low = 196,500 seeds ha , Medium = 262,000 seeds ha , and High = 327,500
-1

-1

-1

seeds ha ; in small red bean, Low = 148,000 seeds ha , Medium = 196,500 seeds ha , and
-1

High = 262,000 seeds ha .

97

Table 2.16. Interaction of row width and planting population with regard to late-season maturity
a
in small red bean at East Lansing in 2011 .
-1

Planting population (seeds ha )
Row width

148,000

196,500
262,000
____________________________________________
%
38 cm
60.2 ab
57.8 ab
64.5 a
76 cm
59.4 ab
55.2 ab
50.9 b
a
Means throughout the table that have the same lower-case letter are statistically similar to one
another at the α ≤ 0.05 level of significance.
____________________________________________

Table 2.17. Main effect of row width on yield of black bean and small red bean at East Lansing
a
and Richville in 2010 and 2011 .
East Lansing
2010
Row
width

Black

Richville
2011

Small
red

Black

2010
Small
red

______________________________________________

b

Black

2011
Small
red

Black

Small
red

-1 ______________________________________________

kg ha

38 cm
3257 a
2553 a
4494 a
3329 a
1488 b
1612 a
3040 a
2592 a
51 cm
1553 b
1674 a
2924 a
2550 a
76 cm
2785 b
2486 a
3903 b
2830 b
1814 a
1797 a
2896 a
2339 b
Mean
3021
2520
4198
3080
1618
1694
2953
2494
a
Means within each column that the same letter are statistically similar to one another at the α ≤
0.05 level of significance unless otherwise noted.
b
Significance level of α ≤ 0.1

98

Table 2.18. Main effect of planting population on yield of black and small red bean at East
a
Lansing and Richville in 2010 and 2011 .
East Lansing
2010
Planting
a
population

Black

Richville
2011

Small
red

Black

2010

Small
redb

__________________________________________

Black

kg ha

2011

Small
red

Black

Small
red

-1 __________________________________________

Low
3020 a
2425 a
4232 a
2958 b
1674 a 1694 ab 2915 a
2427 a
Medium
3138 a
2554 a
4103 a
3059 ab 1656 a 1774 a
2984 a
2523 a
High
2902 a
2577 a
4258 a
3222 a
1751 a 1582 b
2956 a
2619 a
a
Means within each column that the same letter are statistically similar to one another at the α ≤
0.05 level of significance unless otherwise noted.
b
Significance level of α ≤ 0.1

Table 2.19. Main effect of row width on economic return of black and small red bean at East
a

Lansing and Richville in 2010 and 2011 .
East Lansing
2010
Row
width

Black

Small red

Richville
2010

2011
Black

Small red

__________________________________

c

Black

2011

Small red

Black

-1 b __________________________________

$ ha

38 cm
1110 a
626 a
1552 a
1278 a
526 a
643 a
1015 a
51 cm
546 ab
592 a
971 a
76 cm
941 b
609 a
1341 b
1079 b
591 b
615 a
960 a
Mean
1026
618
1446
1178
554
617
982
a
Means within each column that the same letter are statistically similar to one another at the α ≤
0.05 level of significance unless otherwise noted.
b
Assumes black bean dealer price of $1.25 kg-1 and grower price of $0.88 kg-1, small red bean
dealer price of $1.43 kg-1 and grower price of $0.99 kg-1 (USDA-AMS 2011, Varner 2012) and
June 2011 herbicide prices for POST treatment (Local herbicide prices 2011)
c
Significance level of α ≤ 0.1

99

Table 2.20. Interaction of row width and planting population with regard to economic return in
a
small red bean at Richville in 2011 .
-1

Planting population (seeds ha )
Row width

1

148,000 seeds ha-

196,500 seeds ha

____________________________

38 cm
51 cm
76 cm

940 abcd
901 cd
910 bcd

-1

262,000 seeds ha

-1

-1b ____________________________

$ ha

1007 ab
982 abcd
782 e

1003 abc
1014 a
888 de

a

Means throughout the table that have the same lower-case letter are statistically similar to one
another at the α ≤ 0.05 level of significance.
b
-1
-1
Assumes small red bean dealer price of $1.43 kg and grower price of $0.99 kg (USDAAMS 2011, Varner 2012) and June 2011 herbicide prices for POST treatment (Local herbicide
prices 2011)

Table 2.21. Main effect of planting population on economic return of black and small red bean
at East Lansing and Richville in 2010 and 2011

a, b

.

East Lansing
2010
Planting
b
population

Black

Small red

Richville
2010

2011
Black

Small red

____________________________________________

Black

2011

Small red

Black

-1c ____________________________________________

$ ha

Low
1030 a
599 a
1463 a
1137 a
553 a
630 a
967 a
Medium
1067 a
628 a
1413 a
1171 a
541 a
656 a
998 a
High
979 a
625 a
1463 a
1228 a
569 a
564 a
980 a
a
Means within each column that the same letter are statistically similar to one another one
another at the α ≤ 0.05 level of significance.
b
-1
-1
In black bean, Low = 196,500 seeds ha , Medium = 262,000 seeds ha , and High = 327,500
-1

-1

-1

seeds ha ; in small red bean, Low = 148,000 seeds ha , Medium = 196,500 seeds ha , and
-1

High = 262,000 seeds ha .
c

-1

-1

Assumes black bean dealer price of $1.25 kg and grower price of $0.88 kg , small red bean
-1

-1

dealer price of $1.43 kg and grower price of $0.99 kg (USDA-AMS 2011, Varner 2012) and
June 2011 herbicide prices for POST treatment (Local herbicide prices 2011)

100

100

Canopy closure (%)

80

60

40

20

38-cm rows
38-cm rows
76-cm rows
76-cm rows
0
2

4

6

8

10

12

Weeks after planting
Figure 1. Main effect of row width on canopy closure for black beans planted at East Lansing,
MI in 2010. Vertical bars represent Fisher’s protected LSD at the α < 0.05 level of significance.

101

100

(a)

Canopy closure (%)

80

60

40

20

38-cm rows
38-cm rows
76-cm rows
76-cm rows
0
2

4

6

8

10

12

100

(b)

Canopy closure (%)

80

60

40

-1

20

196,500 seeds ha-1
196,500 seeds ha
-1
262,000 seeds ha
262,000 seeds ha-1
-1
327,500 seeds ha-1
327,500 seeds ha

0
2

4

6

8

10

12

Weeks after planting

Figure 2. Main effects of (a) row width and (b) planting population on canopy closure for black
beans at East Lansing, MI in 2011. Vertical bars represent Fisher’s protected LSD at the α <
0.05 level of significance.
102

100

(a)

Canopy closure (%)

80

60

40

20

38-cm rows
38-cm rows
51-cm rows
51-cm rows
76-cm rows
76-cm rows

0
2

4

6

8

10

12

100

(b)

Canopy closure (%)

80

60

40

-1

20

196,500
196,500 seeds ha-1
-1
262,000
262,000 seeds ha-1
-1
327,500 seeds ha
327,500 seeds ha -1

0
2

4

6

8

10

12

Weeks after planting

Figure 3. Main effects of (a) row width and (b) planting population on canopy closure for black
beans at Richville, MI in 2011. Vertical bars represent Fisher’s protected LSD at the α < 0.05
level of significance.

103

100

(a)

Canopy closure (%)

80

60

40

20

38-cm rows
38-cm rows
76-cm rows
76-cm rows
0
2

4

6

8

10

12

100

(b)

Canopy closure (%)

80

60

40

20

-1

148,000 seeds ha-1
148,000 seeds ha-1
196,500 seeds ha-1
196,500 seeds ha-1
262,000 seeds ha -1
262,000 seeds ha

0
2

4

6

8

10

12

Weeks after planting

Figure 4. Main effects of (a) row width and (b) planting population on canopy closure for small
red beans at East Lansing, MI in 2010. Vertical bars represent Fisher’s protected LSD at the α <
0.05 level of significance.
104

100

(a)

Canopy closure (%)

80

60

40

20

38-cm rows
38-cm rows
76-cm
76-cm rows
0
2

4

6

8

10

12

100

(b)

Canopy closure (%)

80

60

40

-1

20

196,500 seeds ha-1
148,000 seeds ha-1
262,000 seeds ha -1
196,500 seeds ha
-1
262,000 seeds ha
327,500 seeds ha -1

0
2

4

6

8

10

12

Weeks after planting

Figure 5. Main effects of (a) row width and (b) planting population on canopy closure for small
red beans at East Lansing, MI in 2011. Vertical bars represent Fisher’s protected LSD at the α <
0.05 level of significance.
105

100

(a)

Canopy closure (%)

80

60

40

20

38-cm rows
38-cm
51-cm
51-cm rows
76-cm rows
76-cm rows
0
2

4

6

8

10

12

100

(b)

Canopy closure (%)

80

60

40

-1

20

148,000 seeds ha -1
148,000 seeds ha
-1
196,500 seeds ha-1
196,500 seeds ha-1
262,000 seeds ha -1
262,000 seeds ha

0
2

4

6

8

10

12

Weeks after planting
Figure 6. Main effects of (a) row width and (b) planting population on canopy closure for small
red beans at Richville, MI in 2010. Vertical bars represent Fisher’s protected LSD at the α <
0.05 level of significance.
106

100

(a)

Canopy closure (%)

80

60

40

20

38-cm rows
38-cm rows
51-cm rows
51-cm rows
76-cm rows
76-cm rows

0

2

4

6

8

10

12

100

(b)

Canopy closure (%)

80

60

40

20

-1

196,500 seeds ha-1
196,500 seeds ha-1
262,000 seeds ha-1
262,000 seeds ha-1
327,500 seeds ha
327,500 seeds ha -1

0

2

4

6

8

10

12

Weeks after planting

Figure 7. Main effects of (a) row width and (b) planting population on canopy closure for black
beans at Richville, MI in 2010. Vertical bars represent Fisher’s protected LSD at the α < 0.05
level of significance.
107

100

(a)

Canopy closure (%)

80

60

40

20

38-cm rows
38-cm rows
51-cm rows
51-cm rows
76-cm
76-cm rows
rows

0
2

4

6

8

10

12

100

(b)

Canopy closure (%)

80

60

40

20

-1
-1

148,000 seeds ha
148,000 seeds ha-1
196,500 seeds ha-1
196,500 seeds ha -1
262,000 seeds ha-1
262,000 seeds ha

0
2

4

6

8

10

12

Weeks after planting

Figure 8. Main effects of (a) row width and (b) planting population on canopy closure for small
red beans at Richville, MI in 2011. Vertical bars represent Fisher’s protected LSD at the α <
0.05 level of significance.
108

100

(a)

Canopy closure (%)

80

60

40

-1

20

196,500 seeds ha -1
196,500 seeds -1
262,000 seeds ha -1
262,000
-1
327,500 seeds
327,500 seeds ha -1

0
2

4

6

8

10

12

100

(b)

Canopy closure (%)

80

60

40

-1

20

196,500 seeds ha-1
196,500 seeds ha -1
262,000 seeds ha -1
262,000 seeds ha
-1
327,500 seeds ha
327,500 seeds ha -1

0
2

4

6

8

10

12

Weeks after planting
Figure 9. Effect of planting population on canopy closure for black beans planted in (a) 38- and
(b) 76-cm rows at three populations at East Lansing, MI in 2010. Vertical bars represent Fisher’s
protected LSD at the α < 0.05 level of significance.

109

APPENDIX

110

Table A2.1. Schedule of activities in the field for population-row width studies.
Event
Planting
Preemergence application
Cutworm trap establishment
Stand counts
Bean thinning
Clethodim application
First light meter reading
Postemergence application
Humidity meter placement
Emergence quadrat placed
First emergence count
Cutworm scouting
Plot length reduction
Early maturity rating
Glyphosate application
Weed harvest
Disease rating
Lodging rating
Final Maturity Rating
Disease pod harvest
Yield component harvest
Final harvest

2010
East Lansing
Richville
June 16
June 10
June 17
June 11
June 22
June 23
June 28
June 23
June 28
June 24
July 1
none
July 6
June 29
July 6
July 1
July 6
July 7
July 12
July 13
July 25
July 29
August 19
August 20
September 27
August 31
none
none
none
August 31
September 21 September 4
September 23
none
September 23 September 7
September 23 September 7
September 23
none
September 30 September 7
September 29 September 7

111

2011
East Lansing
June 8
June 8
June 21
June 21
none
July 12
June 29
July 5
July 15
July 13
July 22
August 29
none
September 2
October 10
September 22
October 4
October 12
October 3
October 2
October 11
October 18

Richville
June 6
June 6
June 20
June 20
none
none
July 6
June 30
July 6
July 6
July 14
August 30
none
none
September 12
September 9/10
October 6
September 23
none
October 6
September 23
October 6

Table A2.2. Dates of light meter readings and weed emergence counts in population-weed
control studies.
Site-year
Light meter readings
Weed emergence counts
East Lansing 2010
July 6, 14, 25, 30
July 25, 30
August 10*, 13*, 18, 27
August 10
Richville 2010
June 29
July 23, 29
July 7, 13, 26
August 17
August 7, 17
East Lansing 2011
June 29
July 22
July 8, 15, 21, 31
August 8
August 10, 19, 26
September 20
September 1
Richville 2011
June 30
July 14
July 6, 12, 25
September 2
August 11, 16, 22, 28
*Replications 1 and 2 only were measured on August 10, while replications 3 and 4 were
measured on August 13.

112

LITERATURE CITED

113

LITERATURE CITED

Adams, M.W. 1995. An historical perspective on significant accomplishments in dry bean
research – the first hundred years. Bean Improvement Cooperative Review Article. Website:
http://www.css.msu.edu/bic/PDF/BeanHistory.pdf, accessed April 20, 2011.
Andrade, F.H., P. Calvino, A. Cirilo, and P. Barbieri. 2002. Yield responses to narrow rows
depend on increased radiation interception. Agron. J. 94:975-980.
Bennett, J.P., M.W. Adams, and C. Burga. 1977. Pod yield component variation and
intercorrelation in Phaseolus vulgaris L. as affected by planting density. Crop Sci. 17:73-75.
Blackshaw, R.E., L.J. Molnar, H.H. Müendel, G. Saindon, and X. Li. 2000. Integration of
cropping practices and herbicides improves weed management in dry bean (Phaseolus vulgaris).
Weed Technol. 14:327-336.
Blackshaw, R.E., H.H. Müendel, and G. Saindon. 1999. Canopy architecture, row spacing and
plant density effects on yield of dry bean (Phaseolus vulgaris) in the absence and presence of
hairy nightshade (Solanum sarrachoides). Can. J. Plant Sci. 79:663-669.
Conner R.L., D.W. McAndrew, P. Balasubramanian, F.A. Kiehn, and Y. Dongfang. 2006.
Influence of growth habit, row spacing, and seed infection on bean anthracnose development.
Can J. of Plant Path. 28:411-418.
Crothers, S.E. and D.T. Westerman. 1976. Plant population effects on the seed yield of
Phaseolus vulgaris L. Agron. J. 68:958-960.
Dalley, C.D., J.J. Kells, and K.A. Renner. 2004. Effect of glyphosate application timing and
row spacing on weed growth in corn (Zea mays) and soybean (Glycine max). Weed Technol.
18:177-182.
Goulden, D.S. 1976. Effect of plant population and row spacing on yield and components of
yield in navy beans (Phaseolus vulgaris). New Zeal. J. Exp. Agr. 4:177-180.
Grafton, K.F., A.A. Schneiter, and B.J. Nagle. 1988. Row spacing, plant population, and
genotype x row spacing interaction effects on yield and yield components of dry bean. Agron. J.
80:631-634.
Hagedorn, D.J. and D.A. Inglis. 1986. Handbook of bean diseases. University of Wisconsin
Extension Publication A3374, Madison, WI.
Harder, D.B., C.L. Sprague, and K.A. Renner. 2007. Effect of soybean row width and
population on weeds, crop yield, and economic return. Weed Technol. 21:744-752.
114

Horn, F.L. L.O. Braga Schuch, E.P. Silveira, I.F. Antunes, J.C. Vieira, G. Marchioro, D.F.
Medeiros, and J.E. Schwengber. 2000. Spacing and plant population for dry bean harvest
mechanization. Pesq. agropec. bras. 35:41-46.
Hosfield, G.L., S.M. Bushey, and J.D. Kelly. 2005. Merlot: a new small red dry bean for
Michigan. Michigan State University Extension Bulletin E-2933, East Lansing, MI.
Hosfield, G.L., G.V. Varner, M.A. Uebersax, and J.D. Kelly. 2004. Registration of ‘Merlot’
small red bean. Crop Sci. 44:351-352.
Kelly, J.D. 1994. Traditional and novel approaches to bean improvement. Mich. Bean Digest
19(1):4-7.
Kelly, J.D. 2001. Remaking bean plant architecture for efficient production. Advances in
Agronomy 71:109-143.
Kelly, J.D. 2010. The story of bean breeding. 2010. Michigan State University publication
prepared for BeanCAP and PBG Works, June 2010.
http://www.css.msu.edu/Bean/_pdf/Story_of_Bean_Breeding_in_the_US.pdf. Accessed March
28, 2012.
Kelly, J.D., G.V. Varner, P. O’Boyle, and B. Long. 2009. Registration of ‘Zorro’ black bean. J.
of Plant Registrations 3:226-230.
Légère, A. and M.M. Schreiber. 1989. Competition and canopy architecture as affected by
soybean (Glycine max) row width and density of redroot pigweed (Amaranthus retroflexus).
Weed Sci. 37:84-92.
Lehman, W.F. and J.W. Lambert. 1960. Effects of spacing of soybean plants between and
within rows on yield and its components. Agron. J. 52:84-86.
Local herbicide prices. 2011. Herbicide price sheet. Effective Friday, June 3, 2011.
Malik, V.S., C.J. Swanton, and T.E. Michaels. 1993. Interaction of white bean (Phaseolus
vulgaris L.) cultivars, row spacing, and seeding density with annual weeds. Weed Sci. 41(1):6268.
Michel, A.P., C.H. Krupke, T.S. Baute, and C.D. DiFonzo. 2010. Ecology and management of
the western bean cutworm (Lepidoptera: Noctuidae) in corn and dry beans. J. Integ. Pest Mngnt.
1(1):1-10.

115

Nelson, K.A. and K.A. Renner. 1998. Weed control in wide- and narrow- row soybean (Glycine
max) with imazamox, imazethapyr, and CGA-277476 plus quizalofop. Weed Technol. 12:137144.
Norsworthy, J.K. and J.R. Frederick. 2002. Reduced seeding rate for glyphosate-resistant,
drilled soybean on the southeastern Coastal Plain. Agron. J. 94:1282-1288.
Norsworthy, J.K. and L.R. Oliver. 2001. Effect of seeding rate of drilled glyphosate-resistant
soybean (Glycine max) on seed yield and gross profit margin. Weed Technol. 15:284-292.
Peachey E., R.L. Ludy, M.L. Powelson, and D.M. McGrath. 2006. Modification of plant
arrangement suppresses white mold of snap beans. HortScience 41:1298-1302.
Park, S.J. 1993. Response of bush and upright plant type selections to white mold and seed
yield of common beans grown in various row widths in southern Ontario. Can. J. Plant Sci.
73:265-272.
Peters, E.J., M.R. Gebhardt, and J.F. Stritzke. 1965. Interrelations of row spacings, cultivations,
and herbicides for weed control in soybeans. Weeds 13:285-289.
Redden, R.J. 1987. Response of navy beans to row width and plant population density in
Queensland. Aust. J. Exp. Agric. 27:455-463.
Robertson, L.S. and R.D. Frazier, eds. 1978. Dry bean production – principles and practices.
Michigan State University Extension Bulletin E-1251, East Lansing, MI.
Saindon, G., H.C. Huang, and G.C. Kozub. 1995. White-mold avoidance and agronomic
attributes of upright common beans grown at multiple planting densities in narrow rows. J.
Amer. Soc. Hort. Sci. 120:843-847.
Schwartz, H.F., M.A. Brick, R.M. Harveson, and G.D. Franc, eds. 2004. Dry bean production
& integrated pest management – 2nd edition. Regional publication by Central High Plains Dry
Bean and Beet Group, Colorado State University, University of Nebraska, and University of
Wyoming: Bulletin 562A.
Sprague, C.L., C.D. DiFonzo, and L.E. Hanson. 2010. Optimizing row width and plant
populations to improve weed management and yield in Michigan dry bean production systems.
Project Greeen grant proposal.
Sprague, C.L., and W.J. Everman. 2011. Weed control guide for field crops. Michigan State
Extension Bulletin E-434.
United States Department of Agriculture, Agricultural Marketing Service. 2011. Bean market
news 2011 summary. USDA Market News.
116

Varner, P. 2012. Personal correspondence. By e-mail.
Vieira, R.F., T.J. Paula, H. Texeira, and J.E. de San Carneiro. 2010. White mold management in
common bean by increasing within-row distance between plants. Plant Dis. 94:361-367.
Wahab, M.N.J., D.H. Dabbs, and R.J. Baker. 1986. Effects of planting density and design on
pod yield of bush snap bean (Phaseolus vulgaris L.). Can. J. Plant Sci. 66:669-675.
Weber, C.R., R.M. Shibles, and D.E. Byth. 1966. Effect of plant population and row spacing on
soybean development and production. Agron. J. 58:99-102.
Welacky, T.W. and S.J. Park. 1987. Response for row width of white bean lines selected for
upright plant types. Annual Report of the Bean Improvement Cooperative 30:54.
Yelverton, F.H. and H.D. Coble. 1991. Narrow row spacing and canopy formation reduces
weed resurgence in soybeans (Glycine max). Weed Technol. 5:169-174.
Young, B.G., J.M. Young, L.C. Gonzini, S.E. Hart, and L.M. Wax. 2001. Weed management in
narrow- and wide-row glyphosate-resistant soybean (Glycine max). Weed Technol. 15:112-121.
Ziviani, A.C., W.Q Ribeiro, M.L.G. Ramos, M. Barbosa, A. Cordeiro, and L.V. Franca. 2009.
Spatial arrangements for common beans with different architecture and their effects on yield and
plant covering. Biosci. Journal. 25:1-9.

117

CHAPTER 3
WEED CONTROL IN WIDE- AND NARROW-ROW BLACK BEANS

Abstract
Dry beans are traditionally grown in wide rows with a heavy reliance on pre-plant
incorporated herbicides and inter-row cultivation for weed control. Narrow rows have been
shown to increase weed suppression in other crops, and preemergence and postemergence
herbicides are a key component of most contemporary weed management systems, so various
herbicide combinations were examined in wide and narrow rows. Field studies were conducted
in 2010 and 2011 at two locations to examine the effect of row width and herbicide combination
on weed suppression and yield in ‘Zorro’ black beans, a new upright variety. Dry beans were
planted in 38- and 76-cm rows. Six weed control strategies were examined: S-metolachlor +
halosulfuron (PRE), S-metolachlor (PRE) fb. bentazon + fomesafen (POST), halosulfuron (PRE)
+ clethodim (+ fomesafen in one site-year) (POST), imazamox + bentazon (POST), weed-free,
and no weed control. Weed control and crop injury were three to four times beginning at the last
POST herbicide application. In addition, weeds were counted by species in late July, and weed
biomass was harvested and weighed at the end of the season. Yield was obtained by direct
harvest. Narrow rows reduced weed population in two of the four site-years, reduced weed
biomass in three of four site-years, and often improved control of upright broadleaf weeds. All
herbicide combinations generally reduced weed populations and biomass, but control of specific
weeds was variable. Crop injury was generally slight and transient. Yield was greater in narrow
rows in three of four site-years. All herbicide combinations increased yield compared with the

118

untreated and resulted in similar yields to one another.

Yield and weed suppression was

maximized in narrow rows, while herbicide performance varied by year and weed spectrum
Nomenclature: Dry bean, Phaseolus vulgaris; preplant incorporated; preemergence;
postemergence.

Introduction
Weed management in dry bean production has traditionally relied heavily on pre-plant
incorporated (PPI) herbicides and between-row cultivation after beans have emerged (Arnold et
al. 1993, Blackshaw et al. 2000; Burnside et al. 1994; Burnside 1998; Robertson and Frazier
1978).

Traditional harvest methods involve pulling and windrowing mature plants and

harvesting them in a second pass (Robertson and Frazier 1978; Schwartz 2004). However,
development of new upright varieties of many dry bean classes, including black bean, has led
many growers to adopt single-pass direct harvest with a typical combine (Hosfield et al. 2005;
Kelly 1994, 2001, 2010; Schwartz et al. 2004). This method of harvest strongly discourages
between-row cultivation due to the need for a level soil surface free of stones at harvest
(Schwartz et al. 2004). However, the need for cultivation has been the primary limiting factor
requiring beans to be grown in wide rows, typically 76-cm; if cultivation is not performed, row
width can be narrowed (Blackshaw et al. 2000). Many studies in dry bean and other crops have
shown that narrow rows can improve weed suppression, thereby at least partly offsetting the lack
of cultivation (Blackshaw et al. 1999; Blackshaw 2000; Harder et al. 2007; Malik et al. 1993;
Nelson and Renner 1998; Peters et al. 1965). In addition, PPI herbicides have also become less
popular due to fuel and time costs and soil disturbance associated with incorporation. However,
a number of preemergence (PRE) and postemergence (POST) herbicides are now available for

119

use in dry bean in the United States, which may be able to maintain weed control in the absence
of cultivation or PPI herbicides (Sprague and Everman 2011). Therefore, research is warranted
to determine effective herbicide combinations for modern dry bean production and whether
narrow rows can play a role in current weed management programs.
Many studies have shown that decreasing row width increases canopy development in
dry bean and other crops (Blackshaw et al. 1999; Blackshaw et al. 2000; Harder et al. 2007;
Redden et al. 1997; Yelverton and Coble 2001). Quicker or more complete canopy development
may have beneficial effects, including weed suppression and increased yield (Andrade et al.
2002; Norsworthy and Oliver 2001; Peters et al. 1965). Malik et al. (1993) found that planting
navy beans in 23- or 46-cm rows reduced weed biomass 15-21% compared with beans planted in
69-cm rows. In small red beans, Blackshaw et al. (2000) evaluated weed control at various row
width and population combinations and found that control was optimized by narrow rows and
high populations. Blackshaw et al. (1999) also found that peak canopy closure was increasingly
close to 100% as rows were narrowed. Similar effects have been observed in soybean with
regard to row width. Planting soybean in row widths of 25-cm or less has resulted in lower weed
emergence, biomass and leaf area compared with 76-cm rows (Harder et al. 2007; Légère and
Schreiber 1989; Nelson and Renner 1998). These observations often correlated with increased
canopy closure.
In addition to improving weed suppression, growing dry beans in narrow rows could also
result in higher yields. Weed suppression could indirectly result in improved yield, but enhanced
canopy development could also improve yield by increasing the amount of total light energy
captured and utilized by the crop. Although very little research has been published on canopy
development in dry beans, abundant studies exist in soybean showing increased yields in narrow

120

rows (Harder et al. 2007; Légère and Schreiber 1989; Lehman and Lambert 1960; Young et al.
2001). For instance, Cooper (1977) compared three row widths and determined that 17-cm rows
obtained a yield advantage over wider rows ranging from 10-40%, and Ethredge (1989) analyzed
the effect of row width in two soybean cultivars at many populations and found that in both
cultivars, yield was greater in narrow rows. Some work has also been carried out in dry beans;
Park (1993) simultaneously increased population and decreased row width and found that this
resulted in increasing yields in both Type I and Type II navy bean cultivars. Malik et al. (1993)
reported that 23-cm rows resulted in yield increases in three types of navy beans compared with
46-cm rows. In Type II small red beans, Blackshaw et al. (2000) found a 19% yield advantage
of 23-cm rows over 69-cm rows. Grafton et al. (1988) found that yield increased an average of
565 kg ha

-1

for each 25-cm decrease in row width in Type III pinto beans and an average of 435

-1

kg ha for each such decrease in Type I navy bean.
In spite of the possible weed suppression and yield increases gained by planting in
narrow rows, Blackshaw et al. (2000) observed that small red bean yield remained depressed
regardless of row width in the absence of chemical or mechanical weed control. Harder et al.
(2007) made similar observations in soybean, obtaining low yield in the absence of weed control,
regardless of row width or population. Thus, effective herbicide combinations are needed to give
narrow-row dry beans an initial competitive advantage over weeds in order to maximize yield.
In 1994, Burnside et al. demonstrated that there were numerous combinations of
herbicides and mechanical control treatments available to dry bean growers that resulted in
sufficient weed control; however, every treatment involved cultivation or rotary hoeing, and
every herbicide combination except the two least effective involved PPI herbicides. In 1993,
Arnold et al. (1993) had expressed dissatisfaction with the state of weed control among New
121

Mexico pinto bean growers, who largely relied on PPI herbicides, and so explored imazethapyr
applied PPI, PRE, or POST; they found that it was highly effective applied PRE in combination
with metolachlor and pendimethalin and was generally effective at increasing yields above the
untreated control in any treatment. Blackshaw and Esau (1991) in Alberta had investigated the
potential of imazethapyr, bentazon, and fomesafen applied POST and found generally good
results for imazethapyr but inconsistent results for bentazon and fomesafen alone. Blackshaw et
al. (2000) investigated several PPI and POST herbicides and concluded that herbicide
combinations were better than any one herbicide applied alone, though they indicated that
imazamox may be somewhat effective alone. While these investigations suggest promise for
PRE and POST herbicide combinations in dry bean, little research is available. Much of what is
available focuses on imazethapyr, and none has explored the relationship of herbicide
combinations to row width or investigated herbicide combinations in Type II black beans.
Therefore, the objectives of this research were to investigate the combined effect of row width
and herbicide combinations in the absence of both cultivation and PPI herbicides on: 1) weed
control in Type II black bean, 2) Type II black bean yield, and 3) economic returns.

Materials and Methods
Field trials were conducted in 2010 and 2011 at the Michigan State University Harry and
Hazel Box Farm near East Lansing, Michigan and at the Michigan State University Saginaw
Valley Research and Extension Center near Richville, Michigan. The soil at the East Lansing
site was a Capac loam (fine-loamy, mixed, mesic, Aeric Ochraqualfs) with pH 6.5 and 3.3%
organic matter in 2010 and pH 6.5 and 2.6% organic matter in 2011. The soil at the Richville
site was a Tappan-Londo loam (fine-loamy, mixed, active, calcareous, mesic Typic Endoaquolls

122

and fine-loamy, mixed, mesic, Aeric Glossaqualfs) with pH 7.8. Soil organic matter was 2.7%
and 2.6% in 2010 and 2011, respectively. The crops prior to these studies were winter wheat
(Triticum aestivum L.) at East Lansing and corn (Zea mays L.) at Richville. Field preparation
consisted of either fall moldboard or chisel plow followed by one to two field cultivations in the
spring. Prior to planting, 336 kg of 17-8-15 (N-P-K) fertilizer containing 1.5% manganese and
1.5% zinc was broadcast onto the soil surface and incorporated.
The variety used at both locations was ‘Zorro’ black bean (Michigan Crop Improvement
Association, Okemos, MI), which is a Type II (upright indeterminate vine) variety. Beans were
-1

planted in 38- and 76-cm rows at a target population of 262,000 plants ha on June 16, 2010 and
June 8, 2011 at East Lansing using a John Deere split-row planter (Deere and Company, Moline,
IL) and on June 10, 2010 and June 6, 2011 at Richville using a Monosem split-row planter
(Monosem Incorporated, Edwardsville, KS). Six weed control systems were examined in each
row width (38- and 76-cm rows), resulting in a total of twelve treatments. The different weed
control systems are summarized in Table 3.1. PRE herbicide applications were made within 24
hours of planting, and POST applications were made when the majority of weeds were 5-10 cm
tall. In 2010, the POST portion of the PRE fb. POST treatments were applied 4-6 d after the
total POST treatment; in 2011, they were applied at the same time. Herbicides were applied
1

using a tractor-mounted, compressed air sprayer designed to deliver 177 l ha- at 193 kPa using
AirMix 11003 nozzles (Greenleaf Technologies, Covington, LA).
The experimental design was a randomized complete block with regard to herbicide
treatment; each herbicide treatment was treated as a split-plot between the two row widths. Each
treatment was replicated four times. Plot width was 3 m, with seven rows per plot for the 38-cm

123

row width and 4 rows per plot for the 76-cm row width. Plot length ranged from 9.1 to 10.4 m,
depending on location.
Precipitation data was obtained throughout the growing season from the Michigan
Automated Weather Network (http://www.agweather.geo.msu.edu/mawn/, Michigan State
University, East Lansing, MI), which maintains stations on the Michigan State University
Horticulture Farm less than a mile from both field locations.

Crop injury and weed control, density, and biomass. Weed control by species and crop injury
were both visually evaluated throughout the growing season on a scale of 0-99%, with 0
indicating no injury and 99 indicating that all plants were dead. Weed control is presented for
the 28 d after postemergence treatment (DAT) evaluation unless otherwise noted. Evaluations
were also performed at 14 and 7 DAT and at the time of POST, as well as at 45 DAT at Richville
in 2010 and at 3 DAT at East Lansing in 2011. At the time of the 28 DAT evaluations, the
number of weeds of each species in each plot was counted. These numbers were added to find
total weed densities for each plot. Prior to harvest, above-ground weed biomass in each plot was
harvested, and dry weights were recorded. Depending on weed population, area for the biomass
2

samples ranged from four 0.25-m subsamples to the entire length of the plot. The predominant
weeds at Richville were common lambsquarters (Chenopodium album L.) and pigweed
(Amaranthus spp.). Common lambsquarters, pigweed, velvetleaf (Abutilon theophrasti Medik.),
common ragweed (Ambrosia artemisiifolia L.), eastern black nightshade (Solanum ptycanthum
Dun.), and annual grasses, chiefly foxtail (Setaria spp.) were among the predominant weeds at
East Lansing in 2010; in 2011, these were joined by common purslane (Portulaca oleracea L.).

124

Yield. A preharvest treatment of glyphosate (Roundup PowerMax, Monsanto Company, Saint
Louis, MO) at 0.84 kg ae ha

-1

was applied to all plots at both locations in 2011 to hasten weed

and bean desiccation. Dry beans were direct harvested with a small plot combine (MasseyFerguson 8XP, AGCO, Duluth Georgia) with a 1.5-m header. The middle two and four rows
were harvested in the 76- and 38-cm row width plots, respectively. Dry bean yield was adjusted
to 18% moisture. Harvest occurred on September 6 at Richville and September 29 at East
Lansing in 2010 and on October 6 at Richville and on October 18 at East Lansing in 2011.

Economic return. Gross margins were calculated at grower-received prices of $0.66, $0.88, and
-1

$1.10 kg

paired with dealer seed prices of $0.77 and $1.25 kg

-1

(USDA-AMS 2011). The low

and high prices were approximately the lowest and highest prices occurring in Michigan during
the last five years, while the middle grower-received prices were those currently being contracted
by Michigan growers (Varner 2012 pers. corr.). This calculation was performed using the
formula:
Gross margin = (Y × Pgrower) – (C + Pherbicide)
Where Y is yield, Pgrower is the bean price received by growers, C the cost of seed, and
Pherbicide is the cost of the weed control treatment at June 2011 prices (Local herbicide costs
2011). The cost of weed control for the untreated was assumed to be $0 ha
weed control for the weed-free was assumed to be $323.57 ha

-1
.

-1

while the cost of

Costs other than seed and weed

control were not considered. There is no cost associated with changing row widths except the
upfront cost of a new planter if one is not available for the desired row width.
125

Data analysis. Data were subjected to ANOVA using PROC MIXED in SAS 9.2 (SAS Institute
Inc., Cary, NC) and treatment means were compared using Fisher’s protected LSD at the α <
0.05 level of significance. Some data were log or square root transformed for analysis, but all
data is presented untransformed. Data were combined over row width and weed management
system when no significant interactions were present.

Results and Discussion
Growing conditions. Rainfall from June through September in 2010 was lower than the 30-yr
average at both locations (Table 3.2). At East Lansing, above-average rainfall in June and nearaverage rainfall in July and September allowed dry bean plants to develop normally despite a dry
August. At Richville, rainfall was well below average in every month, and much of the June
rainfall came prior to planting; by the beginning of September, most bean plants had senesced,
rendering September rainfall irrelevant. Only 11 cm of rain fell at Richville between planting
and harvest, resulting in stunted crop plants and little weed growth.
In 2011, rainfall from June through September was slightly above the 30-yr average at
East Lansing, while at Richville it was once again below the 30-yr average. At East Lansing,
rainfall was slightly to moderately below average in three months, but in late July a single
weather system brought 16 cm of rainfall over three days, making that month’s total dramatically
above average and supplying ample water for pod formation, especially with moderate rainfall
continuing through September. At Richville, rainfall was below average in every month, but all
rainfall in June occurred after planting, and monthly totals were higher than those of the previous

126

year in July and August. Rainfall was sufficient for normal plant development, and pod filling
continued into September.

Crop injury. The severity of dry bean injury from herbicide treatments differed among siteyears. Crop injury never exceeded 17%, and by 14 DAT, injury symptoms nearly always
disappeared (Table 3.3).
At East Lansing in 2010, crop injury was not greater than 5% in any treatment at the time
of POST application, and no injury was observed afterward.
At Richville in 2010, all herbicide treatments initially caused 6-8% injury, but 7 days
after the POST treatment (DAT), S-metolachlor fb. bentazon + fomesafen had 16% injury while
other treatments had less than 2% injury. At 14 DAT, injury in this treatment was only 4%, and
injury in other treatments had disappeared.
At Richville in 2011, all treatments containing S-metolachlor caused 7-12% injury at
POST and 7 DAT while treatments not containing S-metolachlor had less than 5% injury.
However, all injury symptoms were gone at 14 DAT. Similarly, at East Lansing in 2011, all
treatments containing S-metolachlor had injury ranging from 10-15% at POST, while other
treatments had less than 5% injury, but by 7 DAT, injury in all treatments was less than 5%, and
at 14 DAT, no injury was observed.
The highest injury consistently occurred in treatments involving S-metolachlor.
Treatments containing halosulfuron often resulted in minor injury, as did imazamox + bentazon,
which is the same result seen by Hekmat et al. (2008). Fomesafen resulted in more pronounced
injury when applied with bentazon following S-metolachlor under drought conditions at
Richville in 2010 but did not injure dry bean in other site-years. At East Lansing in 2011, injury

127

was slightly higher in 38- than in 76-cm rows, but at other site-years, row width was nonsignificant. In every case, injury was transient and never resulted in any obvious effects beyond
14 DAT.

Weed control by species. The effectiveness of each herbicide combination and of narrow rows
for the control of each major weed observed in these studies is presented in Tables 3.4-3.9 and
summarized in Table 3.9. While weed control was evaluated several times during the growing
season, control data is presented from the evaluation 28 days after the POST application unless
otherwise noted.
Common lambsquarters. Common lambsquarters occurred in significant numbers in every siteyear and was the dominant weed at Richville in both years. Common lambsquarters population
-1

was extremely high (167,300 weeds ha ) at Richville in 2011 and relatively low at East Lansing
-1

in 2010 (8900 weeds ha ).
At East Lansing in 2010, when weed pressure was low and PRE herbicides were
incorporated by rainfall, all herbicide combinations provided excellent (≥ 94%) control of
common lambsquarters regardless of row width (Table 3.4).

Common lambsquarters

populations were reduced to <15% of the untreated in all treatments and to levels similar to that
of the weed-free free in all treatments except imazamox + bentazon (POST) (data not shown).
At Richville in both years, PRE herbicides were not effectively incorporated by rainfall,
and all treatments involving PRE herbicides provided only 63-73% control of common
lambsquarters and failed to significantly reduce common lambsquarters populations compared
with the untreated. Imazamox + bentazon (POST) was better, providing an average of 89%

128

control and reduced population to a level similar to the weed-free. At Richville, common
lambsquarters control averaged 78% in 38-cm rows and only 70% in 76-cm rows (Table 3.5).
At East Lansing in 2011, there was an interaction between row width and herbicide
treatment with regard to common lambsquarters control (Table 3.6).

S-metolachlor +

halosulfuron (PRE) and halosulfuron (PRE) fb. clethodim + fomesafen (POST) again provided
excellent (≥ 97%) control of common lambsquarters regardless of row width. Imazamox +
bentazon (POST) provided 98% control in 38-cm rows; however, in 76-cm rows, it only
provided 90% control. S-metolachlor (PRE) fb. bentazon + fomesafen (POST) provided only
89% control in 38-cm rows and even lower control (80%) in 76-cm rows. The two best
treatments reduced common lambsquarters population to levels similar to that of the weed free,
while the other two reduced populations to 37-48% of the untreated (data not shown).
When PRE herbicides were incorporated by timely rainfall, the two treatments containing
halosulfuron provided excellent control of common lambsquarters, even reducing populations to
levels similar to the weed-free; however, the efficacy of the halosulfuron treatments was
dramatically reduced when poorly incorporated.

Imazamox + bentazon (POST) was the

treatment that was most consistent over all site-years and provided the best control when PRE
herbicides were poorly incorporated. S-metolachlor fb. bentazon + fomesafen provided good
control when the PRE was well-incorporated but sometimes less than the halosulfuroncontaining treatments, and control was greatly reduced when the PRE was not well-incorporated.
Imazamox appears to be a better partner for bentazon than fomesafen for common lambsquarters
control. Narrow rows often improved control of common lambsquarters when herbicide control
was less than complete; however, row width did not significantly affect mid-season common
lambsquarters populations.

129

Pigweed. Pigweed (mostly Amaranthus retroflexus L. and Amaranthus powellii S. Wats., with
some Amaranthus hybridus L.) was present at significant numbers in all four site-years but were
more numerous at Richville. Populations ranged from 7200 (East Lansing 2011) to 33,200
-1

weeds ha (Richville 2010.)
At East Lansing in both years and at Richville in 2011, all herbicide treatments provided
excellent (≥ 97%) control of pigweed and reduced populations to levels similar to the weed-free
(Table 3.4). In 38-cm rows, control averaged 99% across these three site-years, while in 76-cm
rows, control was reduced to 97%.
At Richville in 2010, when PRE herbicides were poorly incorporated and drought
conditions prevailed, pigweed control was substantially reduced in all treatments containing PRE
applications, ranging from 74-79%. These treatments still reduced pigweed populations to 1628% of the untreated. Imazamox + bentazon (POST) still provided excellent (94%) control and
reduced pigweed population to a level similar to that of the weed-free. Overall, pigweed control
averaged 75% in 38-cm rows compared with just 67% in 76-cm rows (Table 3.5), and pigweed
population was higher in 76-cm rows with 11,500 weeds ha

-1

than in 38-cm rows with 7700

-1

weeds ha .
All herbicide treatments effectively controlled pigweed under normal environmental
conditions.

When PRE herbicides were not properly incorporated and drought conditions

prevailed afterward, the effectiveness of all treatments involving PRE applications was reduced,
although all still provided significant control. Imazamox + bentazon provided excellent control
of pigweed regardless of environmental conditions. Imazamox appears to be a better partner

130

with bentazon than fomesafen for control of pigweed. Narrow rows improved pigweed control
and reduced mid-season population compared to wide rows.
Velvetleaf. Velvetleaf (Abutilon theophrasti Medik.) was present only at East Lansing. In 2010,
velvetleaf was present in the untreated at a density of only 2000 weeds ha

-1

but in 2011 at a

-1

density of 17,600 weeds ha . Populations in 2010 were too low and sporadic to reliably detect
differences in weed density between treatments.
In 2010, imazamox + bentazon (POST) provided virtually total (99%) control of
velvetleaf (Table 3.4). S-metolachlor (PRE) fb. bentazon + fomesafen (POST) and halosulfuron
(PRE) fb. clethodim (POST) also provided excellent (94-96%) control of velvetleaf. Control
with S-metolachlor + halosulfuron (PRE) at 88% was significantly lower than imazamox +
bentazon but statistically similar to the other two treatments.

Row width did not have a

significant effect on velvetleaf control (Table 3.5).
In 2011, when velvetleaf pressure was much higher, imazamox + bentazon continued to
provide excellent (97%) control, and halosulfuron fb. clethodim + fomesafen also provided
excellent (92%) control, and both reduced velvetleaf populations to levels similar to that of the
weed-free. However, velvetleaf control with S-metolachlor fb. bentazon + fomesafen (69%) and
S-metolachlor + halosulfuron (56%) was greatly reduced, and velvetleaf populations were only
reduced to 54-60% of the untreated. Velvetleaf control in 38-cm rows averaged 84% compared
with just 72% in 76-cm rows.
When velvetleaf pressure was low in 2010, all herbicide treatments were reasonably
effective. However, in 2011, treatments involving S-metolachlor paired either with halosulfuron
or with bentazon and fomesafen were unsatisfactory.

131

However, halosulfuron paired with

fomesafen or bentazon paired with imazamox were effective for velvetleaf control. In 2011,
velvetleaf control was greater in narrow rows than in wide rows.
Eastern black nightshade. Eastern black nightshade (Solanum ptycanthum Dun.) occurred in
significant numbers only at East Lansing; density was more than ten times greater in 2011 than
in 2010.
At East Lansing in 2010, S-metolachlor + halosulfuron (PRE), S-metolachlor (PRE) fb.
bentazon + fomesafen (POST), and imazamox + bentazon (POST) all provided very good to
excellent (88-99%) control of eastern black nightshade and reduced nightshade populations to
levels similar to the weed-free regardless of row width. However, halosulfuron (PRE) fb.
clethodim (POST) provided very poor (39%) control of eastern black nightshade, and nightshade
density was similar to that of the untreated (Table 3.7).
At East Lansing in 2011, the three treatments that previously provided very good to
excellent control of eastern black nightshade all provided virtually complete (99%) control
regardless of row width (Table 3.8). Halosulfuron fb. clethodim + fomesafen provided much
lower control that differed by row width, with 81% control in 38-cm rows and 72% control in
76-cm rows.

-1

Moreover, it increased eastern black nightshade population to 85,500 ha
-1

compared with just 14,400 ha in the untreated.
Imazamox + bentazon and the two treatments that involved S-metolachlor provided
excellent control of eastern black nightshade. Halosulfuron fb. clethodim, with or without
fomesafen, failed to adequately control eastern black nightshade; in fact, this treatment increased
eastern black nightshade population, presumably by removing other weeds that had been
suppressing eastern black nightshade. The effect of row width was not assessed due to the
sporadic nature of eastern black nightshade populations.
132

Common ragweed. Common ragweed was present in significant numbers only at East Lansing.
-1

Population in the untreated in 2010 was 10,200 weeds ha . In 2011, there were only 1100
-1

weeds ha in the untreated.
In 2010, all treatments involving a PRE application provided excellent (94-99%) control
of common ragweed (Table 3.7) and reduced weed populations to levels similar to the weed-free
(data not shown). Imazamox + bentazon (POST) provided lesser but still good (87%) control but
did not reduce weed population compared with the untreated. Common ragweed control was
significantly greater (96%) in 38-cm rows than in 76-cm rows (91%) (Table 3.9), but row width
did not affect common ragweed population.
-1

In 2011, common ragweed control was not rated due to low density (1100 ha ) in the
untreated; however, common ragweed population was ten times higher in the S-metolachlor fb.
bentazon + fomesafen and imazamox + bentazon treatments than in the untreated, suggesting
that common ragweed was being suppressed by other weeds and was released by these
treatments. While there were few common ragweed plants in the halosulfuron fb. clethodim +
fomesafen treatment mid-season, the reason fomesafen was applied in this site-year was the
presence of a high density of ALS-resistant common ragweed following halosulfuron
application; fomesafen effectively controlled this population. Common ragweed density was
-1

-1

more than twice as high in 76-cm rows (5200 ha ) as in 38-cm rows (2500 ha ).
Treatments containing halosulfuron or fomesafen were generally effective at controlling
common ragweed, but one exception was observed in both cases; in halosulfuron, this was due to
an ALS-resistant population.

The Total POST provided less effective control of common

ragweed in one year and appeared to provide no control in the other year; it cannot be
133

recommended for fields with high common ragweed populations or those with ALS-resistant
common ragweed. Narrow rows improved common ragweed control in 2010 and decreased
common ragweed population in 2011.
Annual grass. Annual grass was present in numbers significant enough to evaluate only at East
Lansing in 2011. The predominant grasses were foxtails (Setaria faberi Herrm., Setaria glauca
P.Beauv. and Setaria pumila (Poir.) Roem. & Schult.), but many other species were also present,
including large crabgrass, Digitaria sanguinalis (L.) Scop., and stinkgrass, Eragrostis cilinensis
(All.) Vign. Ex Janchen, among others. Annual grass population in the untreated was 42,200
weeds ha

-1

.

S-metolachlor (PRE) fb. bentazon + fomesafen (POST), and halosulfuron (PRE) fb.
clethodim + fomesafen (POST) both provided excellent (94%) control of annual grass (Table
3.7). Control in the latter combination was slightly delayed due to the slowness (1-2 wk) of
clethodim activity but was more effective at actually killing, rather than merely suppressing,
annual grass than the former treatment. However, halosulfuron fb. clethodim + fomesafen
demonstrated no control of stinkgrass. S-metolachlor + halosulfuron (PRE) provided good
(87%) control of annual grass and resulted in populations similar to the weed-free but also in
some uninjured escapes. Imazamox + bentazon provided some suppression (77%) of annual
grass, but it failed to actually kill it, resulting in a 273% increase in annual grass population
compared to the untreated, presumably due to release from broadleaf weed competition.
Narrow rows were generally not effective at suppressing annual grass (Table 3.9).
Annual grass control averaged 90% in 38-cm rows and 86% in 76-cm rows; this was only
significant at α=0.1. Row width did not affect mid-season annual grass population.

134

Common purslane.

Common purslane (Portulaca oleracea L.) was present in significant

numbers only at East Lansing in 2011. Common purslane population in the untreated was
-1

92,400 weeds ha .
S-metolachlor (PRE) fb. bentazon + fomesafen (POST) and halosulfuron (PRE) fb.
clethodim + fomesafen both provided excellent (98-99%) control of common purslane (Table
3.7) and reduced populations to levels similar to the weed-free. S-metolachlor + halosulfuron
(PRE) also reduced common purslane population to a level similar to the weed-free but provided
somewhat less control (88%) overall because it allowed a number of uninjured escapes.
Imazamox + bentazon (POST) provided good suppression (82%) of common purslane but failed
to actually kill the plants, with common purslane population being virtually identical to the
weed-free. Row width had no significant effect on common purslane control or population
(Table 3.9).
Summary.

The effect of each herbicide combination and of row width on each weed is

summarized in Table 3.3.
S-metolachlor + halosulfuron was the most consistently effective treatment overall when
incorporated by rainfall but was much less effective when it was not incorporated and was
ineffective against a high population of velvetleaf even when incorporated.
The performance of both PRE fb. POST combinations was also significantly decreased if
not well incorporated. Otherwise, S-metolachlor fb. clethodim, with or without fomesafen,
provided consistently effective control of most weeds but provided virtually no control of eastern
black nightshade; and fomesafen was needed to control ALS-resistant common ragweed.
Halosulfuron fb. bentazon + fomesafen provided good but not excellent control of common
lambsquarters and poor control of velvetleaf when velvetleaf population was high; it suppressed
135

annual grass but did not reduce its population. It provided good to excellent control of other
weeds. Imazamox + bentazon had the advantage of being little affected by environmental
conditions and provided consistently excellent control of the most prominent weeds as well as
providing good control of velvetleaf, a weed two other treatments controlled inadequately, along
with eastern black nightshade; its disadvantages were merely fair control of common ragweed,
and suppression without population reduction in annual grass and common purslane. There was
no perfect herbicide combination, so the optimal combination for a given situation should be
chosen based on weeds known to be present in the field and expected environmental conditions.
Narrow rows can play a supplementary role in the suppression of upright broadleaf weeds but
appear to be largely ineffective at suppressing annual grass and common purslane.

Mid-season weed populations. Weed populations shortly before peak canopy closure (41-44
days after planting) were much higher in 2011 than in 2010 at both sites and were slightly higher
in at East Lansing than at Richville in both years. In every site-year, differences in weed
population were observed among the weed control treatments and in two site-years between row
widths. Weed populations by row width and herbicide combination are presented in Tables 3.10
and 3.11.
At East Lansing in 2010, all treatments with a PRE component were similar to the weedfree. Overall weed population significantly differed from the weed-free only in the imazamox +
-1

bentazon (POST) treatment, which still had 68% fewer weeds (19,950 weeds ha ) than the
-1

untreated (62,140 weeds ha .) At East Lansing in 2011, under higher weed pressure, overall
weed populations in both treatments that included halosulfuron, S-metolachlor + halosulfuron
(PRE) and halosulfuron (PRE) fb. clethodim (+ fomesafen at this site-year only) (POST), were
136

again similar to the weed-free. However, overall weed populations in imazamox + bentazon and
S-metolachlor (PRE) fb. bentazon + fomesafen (POST) were similar to the untreated.
At Richville in both years, all herbicide treatments reduced overall weed population
compared with the untreated. In 2010, the three treatments that included a PRE component were
similar to one another and had higher weed populations than the weed-free. Imazamox +
bentazon had fewer weeds and was similar to the weed-free. In 2011, all herbicide treatments
were statistically similar to one another, but only the halosulfuron fb. clethodim and imazamox +
bentazon treatments were statistically similar to the weed-free.
In 2010, there were fewer weeds in 38-cm rows than in 76-cm rows. At East Lansing 76cm rows averaged 21,000 weeds ha-1 and 38-cm rows just 10,900 weeds ha

-1

across all

-1

treatments. At Richville, 76-cm rows averaged 22,500 weeds ha , while 38-cm rows averaged
-1

16,600 weeds ha . In 2011, under higher weed pressure, row width did not have a significant
effect on overall weed population at either site.
In fourteen of sixteen cases, herbicide treatment reduced weed population compared to
the untreated. However, no treatment was consistently best. At East Lansing, where weed
diversity was much greater and the PRE was incorporated by rainfall, S-metolachlor +
halosulfuron (PRE) was among the most effective treatments at reducing weed population in
both years, while imazamox + bentazon (POST) was less effective and did not differ from the
untreated in 2011. At Richville, where common lambsquarters (Chenopodium album L.) and
redroot pigweed (Amaranthus retroflexus L.) dominated and the PRE applications were not
effectively incorporated by rainfall, imazamox + bentazon was most effective in both years while
S-metolachlor + halosulfuron allowed significantly more weeds than the weed-free.

137

Halosulfuron fb. clethodim was the most consistent, being among the most effective at reducing
weed population in every site-year except Richville 2010, when all treatments involving PRE
herbicides compared unfavorably to the weed free; however, it should be noted that in the siteyear of highest weed pressure, East Lansing 2011, fomesafen was added to this treatment POST
due to the failure of halosulfuron to control some broadleaf weeds, especially ALS-resistant
common ragweed (Ambrosia artemisiifolia L.) S-metolachlor fb. bentazon + fomesafen was
only among the most effective at reducing weed populations at East Lansing in 2010, and at East
Lansing in 2011 it did not differ from the untreated. In 2010, when weed populations were lower
overall, 38-cm rows reduced weed populations by 44% and 26% compared with 76-cm rows at
East Lansing and Richville respectively, but in 2011 with higher weed populations, narrow rows
failed to significantly reduce weed populations. The observed differences were large enough to
suggest that under some circumstances, narrow rows can reduce weed populations, which agrees
with the finding of Harder et al. (2007) that weed populations are reduced in narrow row
soybean.

Final weed biomass. Weed biomass was much greater in 2011 than in 2010 at both sites, and
although weeds were more numerous mid-season at East Lansing, final weed biomass was higher
at Richville within each year. At East Lansing in 2010, weed biomass was much lower than in
other site-years. In every site-year, most or all herbicide treatments reduced weed biomass
compared with the untreated (Table 3.12), and in three of the four site-years, weed biomass was
reduced in 38-cm rows compared to 76-cm rows (Table 3.13).
At East Lansing in 2010, under low weed pressure, all herbicide treatments reduced weed
biomass to levels similar to the weed-free. At East Lansing in 2011, imazamox + bentazon

138

(POST) and halosulfuron (PRE) fb. clethodim + fomesafen (POST) again reduced weed biomass
to levels similar to the weed-free despite imazamox + bentazon having failed to reduce the midseason weed population. S-metolachlor + halosulfuron (PRE) and S-metolachlor (PRE) fb.
bentazon + fomesafen (POST) reduced weed biomass compared with the untreated by 86-91%
but both still had more weed biomass than the weed-free.
At Richville in both years, only imazamox + bentazon reduced weed biomass to a level
similar to the weed-free. The effectiveness of the three treatments that involved PRE applications
was reduced by poor PRE incorporation, but in 2010, all three were similar to one another and
still reduced weed biomass by 77-82% compared to the untreated. In 2011, S-metolachlor +
halosulfuron and S-metolachlor fb. bentazon + fomesafen reduced weed biomass by 48-56%
compared with the untreated, but weed biomass was still much higher than in the weed-free, and
halosulfuron + clethodim failed to significantly reduce weed biomass.
Weed biomass was significantly reduced in 38-cm rows compared to 76-cm rows in three
of the four site years. This reduction in weed biomass was 53% at East Lansing in 2010, 30% at
East Lansing in 2011, and 51% at Richville in 2011. At Richville in 2010, when the canopy
never fully developed, row width did not affect weed biomass.
In fifteen of sixteen cases, herbicide treatments reduced pre-harvest weed biomass
compared with the untreated. Imazamox + bentazon was the most consistent weed control
treatment, reducing weed biomass to a level similar to the weed-free in every site-year.

S-

metolachlor + halosulfuron and S-metolachlor fb. bentazon + fomesafen resulted in weed
biomass similar to the weed-free only once but always reduced it relative to the untreated.
Halosulfuron fb. clethodim twice resulted in weed biomass similar to that of the weed-free but

139

once resulted in weed biomass that did not significantly differ from the untreated, and one of the
cases in which it was similar to the weed-free was the case in which fomesafen was added POST.
Imazamox + bentazon was affected only minimally by environmental conditions, while the
efficacy of those involving PRE herbicides was reduced in years when little rainfall followed
PRE application. Treatments involving PRE applications usually performed similarly; however,
halosulfuron fb. clethodim was the least consistent of the three, demonstrating that relying solely
on a single PRE herbicide for broadleaf control is a poor strategy in many situations.
Narrow rows decreased weed biomass across all treatments except under drought
conditions, which led to incomplete canopy closure even in narrow rows. As Blackshaw et al.
(2000) noted, narrow rows cannot replace herbicides; in no case did narrow rows reduce weed
biomass in the untreated to levels similar to those of the most effective herbicide treatments.
However, narrow rows can play a significant supplementary role in weed control. This is similar
to the finding of Teasdale and Frank (1983) that narrow-row snap bean (Phaseolus vulgaris L.)
reduces weed biomass compared to wide rows.

Dry bean yield. Site and year had a strong influence on yield. Yield was highest at East
Lansing in 2011, the only site-year with above average rainfall; yield in the weed-free averaged
-1

4829 kg ha . Yield was lowest at Richville in 2010 under drought conditions; yield in the weed-1

free averaged 2083 kg ha . Yield was intermediate at East Lansing in 2010 and Richville in
2011 under slightly dry conditions, with weed-free yield averaging 3242 and 3082 kg ha
respectively.

140

-1

Yield was increased by herbicide application in every site-year, but yield was nearly
always similar across herbicide treatments, and in only one site-year did the weed-free have
significantly higher yield than some of the herbicide treatments (Table 3.14). In two of the four
site-years, narrow rows increased yield compared to wide rows (Table 3.15).
At East Lansing in 2011, yield was similar across all herbicide treatments. All herbicide
treatments increased yield by 70-83% compared with the untreated.

Weed-free yield was

significantly higher than any of the herbicide treatments except the halosulfuron + clethodim (+
fomesafen at this site-year only). At Richville in both years, yield was similar across all
herbicide treatments and the weed-free, and all of these resulted in higher yield than the
untreated. Yield increases compared with the untreated ranged from 14-19% in 2010 and from
30-48% in 2011.
At East Lansing in 2010, yield was similar across all herbicide treatments and the weedfree in 76-cm rows and in 38-cm rows within three herbicide treatments; yields ranged from
-1

-1

3135 kg ha to 3408 kg ha . In S-metolachlor (PRE) fb. bentazon + fomesafen (POST) and the
weed-free, yield within 38-cm rows was similar to that of the untreated (Table 3.16).
At East Lansing and Richville in 2011, yield was higher across all treatments in 38-cm
rows than in 76-cm rows. At East Lansing in 2011, yield was 19% higher in 38-cm rows, and at
Richville in 2011, it was 8% higher.
Under drought conditions at Richville in 2010, yield did not significantly differ by row
width.
All herbicide combinations used in this study in this study consistently improved yield
relative to the untreated, and none increased yield more than any other. This indicates the
importance of weed control for obtaining optimum dry bean yield and also suggests that there are
141

many good options available for weed control in dry bean. In three of the four years, all
herbicide treatments resulted in yields similar to the weed-free. However, in the site-year with
both the highest yields and the highest weed pressure, three of the herbicide combinations failed
to provide yields similar to the weed-free. Narrow rows improved yield in two of the four siteyears, with the magnitude of the yield improvement being higher at higher yield. Narrow rows
were not capable of replacing herbicide applications, but they did provide a supplemental yield
increase. In two site-years, yield did not increase in narrow rows; in one site-year, it actually
increased in wide rows in two herbicide treatments; the reason for this was unclear. In the other
site-year, drought conditions severely reduced canopy development and yield, so yield did not
significantly differ by row width. This trend toward higher yield in narrow rows confirms the
findings of Blackshaw et al. (1999, 2000), Redden et al. (1987), and Malik et al. (1993) that yield
is often, though not always, higher in narrow rows; Grafton et al. (1988) and Goulden (1976)
found even more consistent yield advantages in narrow rows.

Economic returns. In both of the site-years in which yield was higher in 38-cm rows than in
76-cm rows, gross margin was likewise higher in 38-cm rows (Table 3.17). Bean price received
by the grower (grower price) did not affect the relationship among treatments within the range of
prices occurring from September 2007 to February 2012 except under drought conditions at
Richville in 2010 (USDA-AMS 2011). At Richville in 2010, yields was low, and all herbicide
treatments resulted in similar economic returns to the untreated at the lowest grower price in the
-1

-1

five year period ($0.66 kg ) while at the highest grower price in the period ($1.10 kg ),
imazamox + bentazon (POST) and S-metolachlor + halosulfuron (PRE) both increased economic

142

-1

return compared to the untreated; at the medium grower price ($0.88 kg ), only imazamox +
bentazon increased economic return.
Weed control treatment affected economic return in every site-year. The S-metolachlor +
halosulfuron (PRE) treatment was also always among those with the highest gross margins
(Table 3.18). Halosulfuron (PRE) fb. clethodim (POST) was among the treatments with the
highest gross margins in every site-year, including East Lansing 2011 when fomesafen was
included POST. Imazamox + bentazon (POST) was among the treatments with the highest gross
margins in every site-year except when a grower-received bean price of $1.10 kg

-1

(the highest

price) was used at East Lansing in 2011. Since S-metolachlor (PRE) fb. bentazon + fomesafen
(POST) was the most expensive herbicide combination but did not significantly increase yield
compared to other herbicides, it was not among the treatments with the highest gross margins at
East Lansing except in wide rows in 2010 (Table 3.19), although it was among them at Richville.
In 2010, when weed pressure was relatively low, the weed-free treatment was not among those
with the highest gross margins except in wide rows at East Lansing; however, under heavier
weed pressure in 2011, the weed-free was among them. The untreated control was never among
the treatments with the highest gross margins in any site-year except in narrow rows at East
Lansing in 2010 and at low grower price at Richville in 2010, indicating that weed control is
necessary to maximize profit in dry bean.

Conclusions.

Imazamox + bentazon (POST) treatment was the most effective treatment

investigated at reducing both weed biomass and populations over various environmental
conditions. However, it provides weak control of annual grass and common ragweed and fails to
reduce common purslane populations.
143

S-metolachlor often results in mild early crop injury. Halosulfuron and the Total POST
treatment result in very minor early crop injury. Injury symptoms for all of these normally
disappeared by 14 DAT and totally disappeared by 28 DAT. It seems likely that these injury
symptoms result in little or no yield loss.
The effectiveness of all three treatments involving PRE applications is compromised
when significant rainfall does not soon follow the PRE application; this was especially evident
with regard to common lambsquarters control.
The S-metolachlor + halosulfuron (PRE) treatment was the most effective treatment for
reducing overall weed populations when incorporated by rainfall. It consistently reduced weed
biomass in various environmental conditions. It allows some annual grass and common purslane
escapes but is otherwise effective at reducing the population of these weeds. However, it
provides poor control of velvetleaf.
The S-metolachlor (PRE) fb. bentazon + fomesafen (POST) treatment can reduce weed
biomass across various environmental conditions but only reduced weed populations in three of
four site-years. It provides weaker control of common lambsquarters than the other treatments,
control of common ragweed is inconsistent, and it does not reduce annual grass populations.
Halosulfuron (PRE) fb. clethodim (POST) was among the most effective treatments at
reducing weed population in three of the four site-years but required the addition of fomesafen at
East Lansing in 2011 to control ALS-resistant common ragweed escapes from the halosulfuron.
Whether this treatment was among the best at reducing weed biomass, reduced biomass to a
lesser extent than other treatments, or failed to reduce weed biomass at all depended on site-year.
When fomesafen was added, it was also among the best at reducing weed biomass. POST annual
grass control is excellent but is delayed due to poor PRE control and the slow action of

144

clethodim. The addition of fomesafen is needed to control common purslane. This treatment
provides no control of eastern black nightshade.
Narrow rows reduced weed population in one of two years. They reduced weed biomass
in all site-years except the one in which a severe drought occurred. Narrow rows usually
improve control of upright broadleaf weeds unless herbicide control is excellent but generally
fail to improve control of annual grass and common purslane. Narrow-row weed suppression is
not strong enough to replace herbicide application but is a significant supplement to herbicide
application.
All herbicide combinations improved yield compared with the untreated and had yields
that were similar to one another in all site-years. All herbicide combinations resulted in yields
similar to the weed-free except at East Lansing in 2011 when only halosulfuron fb. clethodim +
fomesafen did so. Narrow rows often result in an increase in yield compared to wide rows.
Because 38-cm rows often result in better weed suppression and higher yield than 76-cm rows,
this row width should be adopted as an effective supplement to herbicide application. However,
in order to maximize weed control and yield in the absence of mechanical control, herbicide
application is necessary. Imazamox + bentazon (POST) is the most consistent treatment of those
investigated for overall weed suppression.

145

Table 3.1. Weed control systems, rates, and application timings in black beans planted in 38- and 76-cm rows.
Dry bean growth stage at application
East Lansing
Richville
-1
Weed control system
Timing
2010
2011
2010
2011
Rate ha
a

S-metolachlor + halosulfuron
b

S-metolachlor fb.
fomesafen + bentazon + COC
halosulfuron fb.
c
clethodim + fomesafen + COC
imazamox + bentazon + COC + AMS
d

Weed-free

1.4 kg + 35 g

PRE

--

--

--

--

1.4 kg fb.
0.28 kg + 0.56 kg + 1% v/v

PRE fb.
POST

-- fb. V4

-- fb. V3

-- fb. V6

-- fb. V2

35 g fb.
77 g + 0.28 kg + 1% v/v

PRE fb.
POST

-- fb. V4

-- fb. V3

-- fb. V6

-- fb. V2

35 g + 0.56 kg + 1% v/v + 2.8
kg
--

POST

V2

V3

V2

V2

--

--

--

--

--

Untreated
------a
S-metolachlor (Dual II Magnum, Syngenta Corporation, Wilmington, DE); halosulfuron (Permit, Gowan Company, Yuma, AZ);
fomesafen (Reflex, Syngenta Corporation, Wilmington, DE); bentazon (Basagran, BASF Corporation, Research Triangle Park, NC);
clethodim (Select Max, Valent U.S.A. Corporation, Walnut Creek, CA); imazamox (Raptor, BASF Corporation, Research Triangle
Park, NC); crop oil concentrate (Herbimax, Loveland Products, Loveland, CO); ammonium sulfate (Actamaster, Loveland Products
Inc., Loveland, CO).
b
Abbreviations: fb., followed by; COC, crop oil concentrate; AMS, ammonium sulfate; PRE, preemergence; POST, postemergence.
c
d

Fomesafen was only included in this treatment at East Lansing in 2011.
-1

-1

Weed-free treatment consisted of s-metolachlor (1.4 kg ha ) + halosulfuron (35 g ha ) applied PRE supplemented with handweeding.

146

Table 3.2. Monthly and 30-yr average precipitation at the Harry and Hazel Box Farm in East
Lansing, MI and at the Saginaw Valley Research and Extension Center in Richville, MI in 2010
a
and 2011 .
Month

East Lansing
2011
30-yr avg.

2010

_________________

June
July
August
September
Total

b

10.6 (4.1)
6.4
3.4
9.2
29.6

cm _________________
4.4
8.4
17.8
6.4
6.7
35.3

8.2
8.3
9.3
34.2

a

Richville
2011

2010

_________________

b

6.9 (2.8)
2.3
3.2
2.9
15.3

30-yr
c
avg.

cm _________________
3.8
9.3

3.4
7.6
5.8
20.6

7.6
8.0
10.7
35.6

Michigan Automated Weather Network, http://www.agweather.geo.msu.edu/mawn/, Michigan
State University, East Lansing, MI
b
Precipitation data in parenthesis is from the time of planting.
c

30-yr average for Caro, MI; none available for Richville

147

Table 3.3. Summary of the effect of herbicide treatments and of narrow rows on crop injury and control of various weed species.
E. black
Common
Annual
Weed control
Common
a Common
Crop injury
Pigweed
a
a
b
b
Velvetleaf
treatment
lambsquarters
ragweed
nightshade
purslane
grass
S-metolachlor Very slight
Excellent if
Excellent Good if
Excellent
Excellent
Good
Good
+ halosulfuron to moderate PRE well
if normal
population
(PRE)
incorporated; moisture; is low;
poor to fair
good in
poor to fair
otherwise
drought
if high
S-metolachlor Very slight
Good to
Excellent Excellent if Fair to
Excellent
Excellent
Excellent but
(PRE) fb.
to moderate excellent if
if normal
population excellent
with some
bentazon +
PRE well
moisture; is low;
escapes
fomesafen
incorporated; fair to
poor to fair
(POST)
poor to good
good in
if high
otherwise
drought
halosulfuron
Very slight
Excellent if
Excellent Excellent if Excellent
Poor
Excellent if Excellent but
(PRE) fb.
to slight
PRE
if normal
population with
fomesafen delayed on
clethodim (+
incorporated
moisture; is low or
fomesafen,
is included most species;
fomesafen at
by rainfall;
fair to
fomesafen fair to
poor on
East Lansing
poor to fair
good in
is included excellent
stinkgrass
in 2011)
otherwise
drought
without it
(POST)
imazamox +
bentazon
(POST)
Narrow rows

a
b

Very slight
to slight

Good to
excellent

Excellent

Excellent

Poor to
good

Excellent

Fair to
good

Usually no
effect; once
slightly
increased
injury

Consistently
improve
control

Improve
control
and
reduce
density

Improve
control if
population
is high

Sometimes
improve
control and
reduce
density

No effect on
weed
density

Rarely
improve
control

Only present in significant numbers at East Lansing
Only present in significant numbers at East Lansing in 2011
148

Fair but
usually nonlethal
Little
improvement
in control

Table 3.4. Main effect of herbicide treatment on control of common lambsquarters, pigweed, and velvetleaf at 28 DAT at East
a
Lansing and Richville in 2010 and 2011 .
Common lambsquarters
Pigweed
Velvetleaf
Herbicide
East
Richville
East Lansing
Richville 2010
East Lansing
East Lansing
b
d
Lansing
2010/2011,
2010
2011
combination
2010 /2011
2010
Richville 2011
____________________________________________________
Control (%) ____________________________________________________
S-metolachlor +
98 a
63 b
97 a
79 b
88 b
56 b
halosulfuron
(PRE)
S-metolachlor
97 a
73 b
98 a
74 b
96 ab
69 b
(PRE) fb.
bentazon +
fomesafen
(POST)
halosulfuron (PRE)
98 a
71 b
98 a
74 b
94 ab
92 a
fb. clethodim
c
(POST)
Imazamox+
94 a
89 a
99 a
94 a
99 a
97 a
Bentazon (POST)
a
Means within each column that the same letter are statistically similar to one another at the α < 0.05 level.
b
c
d

Herbicide rates, application timings, and additives are presented in Table 3.1.
Treatment included fomesafen applied POST at East Lansing in 2011.
Data used were from the 45 DAT evaluation.

149

Table 3.5. Main effect of row width on control of common lambsquarters, pigweed, and velvetleaf at 28 DAT at East Lansing and
a
Richville in 2010 and 2011 .
Common lambsquarters
Pigweed
Velvetleaf
East Lansing
Richville
East Lansing
East Lansing
East Lansing
Row width
2010/2011,
Richville 2010
b
2010
2010
2011
2010 /2011
Richville 2011
___________________________________________
Control (%) ___________________________________________
38 cm
98 a
78 a
99 a
75 a
96 a
84 a
76 cm
96 a
70 b
97 b
67 b
93 a
72 b
a
b

Means within each column that the same letter are statistically similar to one another at the α < 0.05 level.
Data used were from the 45 DAT evaluation.

Table 3.6. Interaction of row width and herbicide combination with regard to control of
a
common lambsquarters at 28 DAT at East Lansing in 2011 .
Row width
b
38 cm
76 cm
Herbicide combination
______________________
______________________
Control (%)
S-metolachlor + halosulfuron
97 a
98 a
(PRE)
S-metolachlor (PRE) fb.
89 b
80 c
bentazon + fomesafen (POST)
halosulfuron (PRE) fb.
99 a
98 a
clethodim + fomesafen (POST)
Imazamox + bentazon (POST)
98 a
90 b
a
b

Means throughout the table that have the same lower-case letter are statistically similar to one another at the α < 0.05 level.
Herbicide rates, application timings, and additives are presented in Table 3.1.

150

Table 3.7. Main effect of herbicide treatment on control at 28 DAT of eastern black nightshade and common ragweed at East Lansing
a
in 2010 and of annual grass and common purslane at East Lansing in 2011 .
Herbicide combination

b

Eastern black
nightshade:
d
East Lansing 2010

Common ragweed:
East Lansing 2010

__________________________________

Annual grass:
East Lansing 2011

Common purslane:
e
2011

S-metolachlor +
halosulfuron (PRE)
S-metolachlor (PRE) fb.
bentazon + fomesafen
(POST)

99 a

96 a

Control (%) ___________________________________
87 b

96 a

99 a

94 a

99 a

halosulfuron (PRE) fb.
c
clethodim (POST)

39 b

94 a

94 a

98 a

Imazamox + bentazon
88 a
87 b
77 c
(POST)
a
Means within each column that the same letter are statistically similar to one another at the α < 0.05 level.
b
c

88 b

82 b

Herbicide rates, application timings, and additives are presented in Table 3.1.
Treatment included fomesafen applied POST at East Lansing in 2011.

d

Data used were from POST evaluation as eastern black nightshade was largely confined to the halosulfuron fb. clethodim plots
throughout the growing season.
e
Data used were from the 14 DAT evaluation; since common purslane is a low-growing weed, canopy closure made it difficult to
evaluate at 28 DAT.

151

Table 3.8. Interaction of row width and herbicide treatment at 28 DAT with regard to control of
a
eastern black nightshade at East Lansing in 2011 .
Row width
b
38 cm
76 cm
Herbicide combination
______________________

S-metolachlor + halosulfuron
(PRE)
S-metolachlor (PRE) fb.
bentazon + fomesafen
(POST)
halosulfuron (PRE) fb.
clethodim + fomesafen
(POST)
Imazamox+ bentazon (POST)

99 a

Control (%) ______________________
99 a

99 a

99 a

81 b

72 c

99 a

99 a

a

Means throughout the table that have the same lower-case letter are statistically similar to one
another at the α < 0.05 level.
b
Herbicide rates, application timings, and additives are presented in Table 3.1.

Table 3.9. Main effect of row width on control of eastern black nightshade and common
ragweed at 28 DAT at East Lansing in 2010 and of annual grass and common purslane at East
a
Lansing in 2011 .
Row width
Eastern black
Common
Annual grass:
Common
d
nightshade:
ragweed: East
East Lansing
purslane: 2011
c
East Lansing
Lansing 2010
2011
b
2010
_____________________________________
Control (%) _____________________________________
38 cm
78 a
96 a
90 a
93 a
76 cm
83 a
91 b
86 b
91 a
a

Means within each column that the same letter are statistically similar to one another at the α <
0.05 level.
b
Data used were from POST evaluation as eastern black nightshade was largely confined to the
halosulfuron fb. clethodim plots throughout the growing season.
c
Significant level of α ≤ 0.1
d

Data used were from the 14 DAT evaluation; since common purslane is a low-growing weed,
canopy closure made it difficult to evaluate at 28 DAT.

152

Table 3.10. Main effect of row width on mid-season weed populations at East Lansing and
a
Richville in 2010 and 2011 .
East Lansing
2010
2011

Row width

__________

weeds m

Richville
2010

2011

__________

-2 __________

-2 __________

weeds m

38-cm
1.1 b
15.5 a
1.7 b
7.8 a
76-cm
2.1 a
14.5 a
2.3 a
8.6 a
a
Means within each column that the same letter are statistically similar to one another at the α <
0.05 level.

Table 3.11. Main effect of weed control system on mid-season weed populations at East Lansing
a
and Richville in 2010 and 2011 .
Weed control system

b

East Lansing
2010
2011
________

S-metolachlor + halosulfuron
(PRE)
S-metolachlor (PRE) fb.
fomesafen + bentazon (POST)
halosulfuron (PRE) fb.
c
clethodim (POST)

weeds m

-2 _______

Richville
2010
2011
________

weeds m

-2 _______

0.3 c

5.2 b

1.6 b

9.2 b

0.2 c

16.5 ab

1.7 b

10.4 b

0.7 c

5.8 b

1.8 b

4.0 b

imazamox + bentazon (POST)
2.0 b
25.2 a
0.5 c
5.4 b
Weed-free
0.1 c
1.9 b
0.2 c
0.5 c
Untreated
6.3 a
22.4 a
4.5 a
20.1 a
a
Means within each column that the same letter are statistically similar to one another at the α <
0.05 level.
b
Herbicide rates, application timings, and additives are presented in Table 3.1.
c

Fomesafen applied POST was included in this treatment at East Lansing in 2011.

153

Table 3.12. Main effect of weed control system on end-of-season weed biomass at East Lansing
b
and Richville in 2010 and 2011 .
Weed control system

East Lansing
2010
2011

a

________

S-metolachlor + halosulfuron
(PRE)
S-metolachlor (PRE) fb.
fomesafen + bentazon (POST)
halosulfuron (PRE) fb.
c
clethodim (POST)

Richville
2010
2011

-1 _______

________

kg ha

kg ha

-1 _______

0.4 b

337 b

431 b

1468 bc

4.9 b

212 bc

517 b

1755 b

2.6 b

97 cd

551 b

2306 ab

imazamox + bentazon (POST)
3.6 b
82 cd
27 c
Weed-free
0.1 b
17 d
4.6 d
Untreated
196 a
2361 a
2375 a
a
Herbicide rates, application timings, and additives are presented in Table 3.1.

376 cd
25 d
3375 a

b

Means within each column that the same letter are statistically similar to one another at the α <
0.05 level.
c
Fomesafen applied POST was included in this treatment at East Lansing in 2011.

Table 3.13. Main effect of row width on end-of-season weed biomass at East Lansing and
a
Richville in 2010 and 2011 .
Row width

East Lansing
2010
2011
__________

kg ha

-1 __________

Richville
2010
__________

2011
kg ha

-1 __________

38-cm
22 b
427 b
595 a
1020 b
76-cm
47 a
609 a
706 a
2082 a
a
Means within each column that the same letter are statistically similar to one another at the α <
0.05 level.

154

Table 3.14. Main effect of weed control system on bean yield at East Lansing in 2011 and
b
Richville in 2010 and 2011 .
Weed control system

East Lansing
2011

a

________

S-metolachlor + halosulfuron (PRE)
S-metolachlor (PRE) fb.
fomesafen + bentazon (POST)
halosulfuron (PRE) fb.
c
clethodim (POST)

Richville
2010

-1 _______

2011

________

kg ha

kg ha

-1 _______

4175 b
3953 b

2129 a
2114 a

2703 a
2796 a

4260 ab

2087 a

2732 a

imazamox + bentazon (POST)
4143 b
2166 a
Weed-free
4828 a
2171 a
Untreated
2333 c
1832 b
a
Herbicide rates, application timings, and additives are presented in Table 3.1.

3067 a
3081 a
2086 b

b

Means within each column that the same letter are statistically similar to one another at the α <
0.05 level.
c
Fomesafen applied POST was included in this treatment at East Lansing in 2011.

Table 3.15. Main effect of row width on bean yield at East Lansing in 2011 and Richville in
a
2010 and 2011 .
Row width

East Lansing
2011
__________

38-cm
76-cm

kg ha

Richville
2010

-1 __________

4294 a
3607 b

__________

2126 a
2040 a

a

2011
kg ha

-1 __________

2857 a
2634 b

Means within each column that the same letter are statistically similar to one another at the α <
0.05 level.

155

Table 3.16. Interaction of row width and weed control system with regard to bean yield at East
b

Lansing in 2010 .
Row width (cm)
Weed control system

a

38-cm

76-cm

_______________________

kg ha

-1 _______________________

S-metolachlor + halosulfuron
3162 abc
3343 ab
(PRE)
S-metolachlor (PRE) fb.
2816 d
3133 abc
fomesafen + bentazon (POST)
halosulfuron (PRE) fb.
3337 ab
3167 abc
clethodim + fomesafen (POST)
imazamox + bentazon (POST)
3210 ab
3169 abc
Weed-free
3073 bcd
3406 a
Untreated
3046 bcd
2865 cd
a
Herbicide rates, application timings, and additives are presented in Table 3.1.
b
Means throughout the table that have the same lower-case letter are statistically similar to one
another at the α < 0.05 level.

Table 3.17. Main effect of row width on economic return at East Lansing in 2011 and Richville
a
in 2010 and 2011 .
Row width

East Lansing
2011
__________

-1c

$ ha

Richville
2010

__________

__________

2011
-1c __________

$ ha

38-cm
1850 a
698 a
1235 a
76-cm
1544 b
667 a
1097 b
a
Means within each column that the same letter are statistically similar to one another at the α <
0.05 level.
c
-1
-1
Assumes black bean dealer price of $1.25 kg and grower price of $0.88 kg (USDA-AMS
2011, Varner 2012) and herbicide and adjuvant costs as of June 2011 (Local herbicide costs
2011)

156

Table 3.18. Main effect of weed control system on economic return at East Lansing in 2011 and
b
Richville in 2010 and 2011 .
Weed control system

a

East Lansing
2011
________

S-metolachlor + halosulfuron (PRE)
S-metolachlor (PRE) fb.
fomesafen + bentazon (POST)
halosulfuron (PRE) fb.
c
clethodim (POST)

-1d _______

$ ha

Richville
2010

2011

________

-1d _______

$ ha

1812 ab
1698 b

711 ab
690 ab

1145 a
1182 a

1846 ab

708 ab

1180 a

imazamox + bentazon (POST)
1798 b
724 a
Weed-free
2002 a
626 c
Untreated
1022 c
637 bc
a
Herbicide rates, application timings, and additives are presented in Table 3.1.

1382 a
1225 a
946 b

b

Means within each column that the same letter are statistically similar to one another at the α <
0.05 level.
c
Fomesafen applied POST was included in this treatment at East Lansing in 2011.
d

-1

-1

Assumes black bean dealer price of $1.25 kg and grower price of $0.88 kg (USDA-AMS
2011, Varner 2012) and herbicide and adjuvant costs as of June 2011 (Local herbicide costs
2011)

157

Table 3.19. Interaction of row width and weed control system with regard to economic return at
a
East Lansing in 2010 .
Row width (cm)
Weed control system

b

38-cm
_______________________

S-metolachlor + halosulfuron
(PRE)
S-metolachlor (PRE) fb.
fomesafen + bentazon (POST)
halosulfuron (PRE) fb.
clethodim + fomesafen (POST)
imazamox + bentazon (POST)
Weed-free
Untreated

76-cm
-1c _______________________

$ ha

1361 ab

1442 a

1191 d

1332 abc

1469 a

1374 ab

1382 ab
1220 cd
1340 abc

1364 ab
1369 ab
1260 bcd

a

Means throughout the table that have the same lower-case letter are statistically similar to one
another at the α < 0.05 level.
b
Herbicide rates, application timings, and additives are presented in Table 3.1.
c

-1

-1

Assumes black bean dealer price of $1.25 kg and grower price of $0.88 kg (USDA-AMS
2011, Varner 2012) and herbicide and adjuvant costs as of June 2011 (Local herbicide costs
2011)

158

APPENDIX

159

Table A3.1. Schedule of activities in the field for weed control-row width studies.
Event
Planting
Preemergence application
Cutworm trap establishment
Stand counts
Bean thinning
Early postemergence application
Late postemergence application
3-day rating
7-day rating
14-day rating
Total weed count
28-day rating
45-day rating
Weed harvest
Final harvest

2010
East Lansing
Richville
June 16
June 10
June 17
June 11
June 22
June 23
June 28
June 23
June 28
June 24
July 8
July 1
July 12
July 7
none
none
July 23
July 13
July 27
July 21
July 27
July 23
August 10
August 6
none
August 23
September 21 September 6
September 29 September 7

160

2011
East Lansing
Richville
June 8
June 6
June 8
June 6
June 21
June 20
June 21
June 20
none
none
July 5
June 30
July 5
June 30
July 8
None
July 13
July 6
July 21
July 14
July 21
July 20
July 25
August 8
none
none
September 22
September 9
October 18
October 6

LITERATURE CITED

161

LITERATURE CITED

Andrade, F.H., P. Calvino, A. Cirilo, and P. Barbieri. 2002. Yield responses to narrow rows
depend on increased radiation interception. Agron. J. 94:975-980.
Arnold, R.N., M.W. Murray, E.J. Gregory, and D. Smeal. 1993. Weed control in pinto beans
(Phaseolus vulgaris) with imazethapyr combinations. Weed Technol. 7:361-364.
Blackshaw, R.E. and R. Esau. 1991. Control of annual broadleaf weeds in pinto beans
(Phaseolus vulgaris). Weed Technol. 5:532-538.
Blackshaw, R.E., L.J. Molnar, H.H. Müendel, G. Saindon, and X. Li. 2000. Integration of
cropping practices and herbicides improves weed management in dry bean (Phaseolus vulgaris).
Weed Technol. 14:327-336.
Blackshaw, R.E., H.H. Müendel, and G. Saindon. 1999. Canopy architecture, row spacing and
plant density effects on yield of dry bean (Phaseolus vulgaris) in the absence and presence of
hairy nightshade (Solanum sarrachoides). Can. J. Plant Sci. 79:663-669.
Burnside, O.C., W.H. Ahrens, B.J. Holder, M.J. Wiens, M.M. Johnson, and E.A. Ristau. 1994.
Efficacy and economics of various mechanical plus chemical weed control systems in dry beans
(Phaseolus vulgaris). Weed Technol. 8:238-244.
Burnside, O.C., M.J. Wiens, B.J. Holder, S. Weisberg, and E.A. Ristau. 1998. Critical periods
for weed control in dry beans (Phaseolus vulgaris). Weed Sci. 46:301-306.
Cooper, R.L. 1977. Response of soybean cultivars to narrow rows and planting rates under
weed-free conditions. Agron. J. 69:89-92.
Ethredge, W.J., D.A. Ashley, and J.M. Woodruff. 1989. Row spacing and plant population
effects on yield components of soybean. Agron. J. 81:947-951.
Goulden, D.S. 1976. Effect of plant population and row spacing on yield and components of
yield in navy beans (Phaseolus vulgaris). New Zeal. J. Exp. Agr. 4:177-180.
Grafton, K.F., A.A. Schneiter, and B.J. Nagle. 1988. Row spacing, plant population, and
genotype x row spacing interaction effects on yield and yield components of dry bean. Agron. J.
80:631-634.
Harder, D.B., C.L. Sprague, and K.A. Renner. 2007. Effect of soybean row width and
population on weeds, crop yield, and economic return. Weed Technol. 21:744-752.
Hekmat, S., N. Soltani, C. Shropshire, and P.H. Sikkema. 2008. Effect of imazamox plus
bentazon on dry bean (Phaseolus vulgaris L.). Crop Prot. 27:1491-1494.
162

Hosfield, G.L., S.M. Bushey, and J.D. Kelly. 2005. Merlot: a new small red dry bean for
Michigan. Michigan State University Extension Bulletin E-2933, East Lansing, MI.
Kelly, J.D. 1994. Traditional and novel approaches to bean improvement. Mich. Bean Digest
19(1):4-7.
Kelly, J.D. 2001. Remaking bean plant architecture for efficient production. Advances in
Agronomy 71:109-143.
Kelly, J.D. 2010. The story of bean breeding. 2010. Michigan State University publication
prepared for BeanCAP and PBG Works, June 2010.
http://www.css.msu.edu/Bean/_pdf/Story_of_Bean_Breeding_in_the_US.pdf. Accessed March
28, 2012.
Légère, A. and M.M. Schreiber. 1989. Competition and canopy architecture as affected by
soybean (Glycine max) row width and density of redroot pigweed (Amaranthus retroflexus).
Weed Sci. 37:84-92.
Lehman, W.F. and J.W. Lambert. 1960. Effects of spacing of soybean plants between and
within rows on yield and its components. Agron. J. 52:84-86.
Local herbicide prices. 2011. Herbicide price sheet. Effective Friday, June 3, 2011.
Malik, V.S., C.J. Swanton, and T.E. Michaels. 1993. Interaction of white bean (Phaseolus
vulgaris L.) cultivars, row spacing, and seeding density with annual weeds. Weed Sci. 41(1):6268.
Nelson, K.A. and K.A. Renner. 1998. Weed control in wide- and narrow- row soybean (Glycine
max) with imazamox, imazethapyr, and CGA-277476 plus quizalofop. Weed Technol. 12:137144.
Norsworthy, J.K. and L.R. Oliver. 2001. Effect of seeding rate of drilled glyphosate-resistant
soybean (Glycine max) on seed yield and gross profit margin. Weed Technol. 15:284-292.
Park, S.J. 1993. Response of bush and upright plant type selections to white mold and seed
yield of common beans grown in various row widths in southern Ontario. Can. J. Plant Sci.
73:265-272.
Peters, E.J., M.R. Gebhardt, and J.F. Stritzke. 1965. Interrelations of row spacings, cultivations,
and herbicides for weed control in soybeans. Weeds 13:285-289.
Redden, R.J. 1987. Response of navy beans to row width and plant population density in
Queensland. Aust. J. Exp. Agric. 27:455-463.

163

Robertson, L.S. and R.D. Frazier, eds. 1978. Dry bean production – principles and practices.
Michigan State University Extension Bulletin E-1251, East Lansing, MI.
Schwartz, H.F., M.A. Brick, R.M. Harveson, and G.D. Franc, eds. 2004. Dry bean production
& integrated pest management – 2nd edition. Regional publication by Central High Plains Dry
Bean and Beet Group, Colorado State University, University of Nebraska, and University of
Wyoming: Bulletin 562A.
Sprague, C.L., and W.J. Everman. 2011. Weed control guide for field crops. Michigan State
Extension Bulletin E-434, East Lansing, MI.
Teasdale, J.R. and J.R. Frank. 1983. Effect of row spacing on weed competition with snap
beans (Phaseolus vulgaris). Weed Sci. 31:81-85.
United States Department of Agriculture, Agricultural Marketing Service. 2011. Bean market
news 2011 summary. USDA Market News.
Varner, P. 2012. Personal correspondence. By e-mail.
Yelverton, F.H. and H.D. Coble. 1991. Narrow row spacing and canopy formation reduces
weed resurgence in soybeans (Glycine max). Weed Technol. 5:169-174.
Young, B.G., J.M. Young, L.C. Gonzini, S.E. Hart, and L.M. Wax. 2001. Weed management in
narrow- and wide-row glyphosate-resistant soybean (Glycine max). Weed Technol. 15:112-121.

164

CHAPTER 4
EFFECTS OF ROW WIDTH AND PLANTING POPULATION ON TYPE II BLACK
BEAN AND SMALL RED BEAN ARCHITECTURE AND YIELD COMPONENTS

Abstract
The development of upright black and small red bean varieties gives dry bean growers
the opportunity to plant in narrower rows, which has been associated with yield benefits in many
crops. Yield differences between row widths or lack of yield differences between planting
populations must be explained by a change in one or more yield components in individual plants:
-1

-1

pods plant , seeds pod , or seed weight, possibly coupled with changes in plant architecture.
Field studies were conducted in 2010 and 2011 at two locations in Michigan to examine the
effect of row width and plant population on yield, yield components, and plant architecture in
two upright varieties: ‘Zorro’ black and ‘Merlot’ small red beans. Each variety was planted in
38- and 76-cm rows, as well as 51-cm rows at one location. Populations were 196500, 262000,
and 327000 plants ha

-1

in black bean and 148000, 196500, and 262000 plants ha

-1

in small red

bean. Yield was obtained using direct harvest by combine. Plants were also hand-harvested, and
on each plant, height was measured, branches and pods were counted, and pods were separated
-1

into main-stem and branch fractions and by number of seeds pod . One hundred seeds from
each plot were weighed. Narrow rows resulted in higher yields than wide rows in two of four
site-years in both dry bean classes. Planting population did not influence yield. The component
-1

most sensitive to both row width and population was number of pods plant ; branch pods were
more responsive than main-stem pods. Number of beans pod

165

-1

was slightly higher at the low

population but rarely affected by row width. Seed weight differences between treatments were
rare and small. Branching increased at lower populations and in narrow rows. Row width and
population had little effect on plant height. The most sensitive yield component to both row
width and population was number of pods per plant, with most of the increase occurring on
branches, although number of seeds per pod also increased at low population.
Nomenclature: Dry bean, Phaseolus vulgaris.

Introduction
Dry bean yield is a function of plant population per area and the three components of
individual plant yield: number of pods per plant, number of seeds per pod, and seed size (Adams
1967; Lehman and Lambert 1960; Grafton et al. 1988). Plant population is a function of plant
spacing within and between rows and is therefore largely dependent on the planter and planter
settings used by the grower. However, dry bean plants are capable of compensating for lower
populations through increases in one or more of the three yield components, while extremely
high plant populations may reduce yield due to intraspecific competition (Bennett et al. 1977).
The objective of this study was to observe the response of the three plant-level yield components
and other aspects of plant architecture to changes in population and row width.
One morphological trait that has been associated with yield in dry bean and observed to
vary with plant population and row width is branching; more branching has been reported to
accompany higher yield (Bennett et al. 1977). Plant height and lodging have also been observed
to vary with row width or population, and both could have an indirect impact on yield, especially
since low-hanging pods are associated with yield loss in direct harvest systems (Horn et al.
2000).

166

Little research exists regarding dry bean yield components; however, there is a great deal
of published literature with regard to yield components and architecture in soybean. Since the
growth habit of Type II dry bean is similar to that of soybean, and the components of yield are
the same, these results may be suggestive of the behavior of Type II dry bean. Weber et al.
(1966) planted soybeans in row widths ranging from 13 to 102 cm and populations ranging from
-1

64,200 to 516,400 plants ha . They found that yield was maximized at 128,500 plants ha

-1

or

higher populations and that with increasing population, height and lodging increased while pods
per plant decreased. With regard to row width, yield and number of pods per plant were
maximized in 25-cm rows, though the number of seeds per pod was maximized in 51-cm rows.
The number of pods per plant was the primary factor controlling yield, and it was greatest at
intermediate row widths and populations. Norsworthy and Shipe (2004) found that the main
stem contributed the majority of the yield in narrow row soybean while the branches contributed
the majority of the yield in wide rows. In contrast, Ethredge et al. (1989) found that main stem
yield decreased in wide rows while branch yield was unaffected. They also found that branch
yield decreased as population increased.
In dry beans, a few studies have touched on the subject of yield components. Crothers
and Westerman (1976) found that yield decreased above 300,000 plants ha

-1

due to decreased

pod formation on lower nodes and death of branches. Bennett et al. (1977) found that the yield
component that changed the most when yields were reduced at high populations was pods per
plant. Grafton et al. (1988) found that in determinate Type I dry beans, yield tended to rise with
population; however, indeterminate Type III pinto beans maintained similar yields over a wide
range of populations due to decreases in both pods per plant and seeds per pod with increasing

167

population. In a review paper, Adams (1967) noted that dry bean yield components were often
inversely related to one another.
Few, if any, studies have been conducted comparing the development of yield and
morphology across row widths and populations in Type II dry bean. Understanding how Type II
dry beans compensate, or fail to compensate, for low populations and wide rows and how
crowding in high populations or narrow rows affects the structure and reproductive allocation of
individual plants will be useful for determining the optimum planting population for various row
widths and for evaluating varieties for their potential in wide- or narrow-row systems.
Therefore, the objectives of this study were to determine how row width and population in Type
II black and small red bean affect: 1) the ability of plants to grow tall yet remain upright, 2) the
three yield components of individual plant yield, and 3) branching patterns and 4) the allocation
of yield between branches and main stem.

Materials and Methods
Field experiments were conducted in 2010 and 2011 at the Michigan State University
Harry and Hazel Box Farm near East Lansing, Michigan and at the Michigan State University
Saginaw Valley Research and Extension Center near Richville, Michigan. Two Type II (upright
indeterminate vine) varieties were planted at both sites: ‘Zorro’ black beans and ‘Merlot’ small
red beans (Michigan Crop Improvement Association, Okemos, MI). At East Lansing, both
varieties were planted in 38- and 76-cm rows using a John Deere split-row planter (Deere and
Company, Moline, IL) on June 16, 2010 and June 8, 2011. At Richville, both varieties were also
planted in 51-cm rows; two separate Monosem planters (Monosem Incorporated, Edwardsville,
KS) were used to plant the three different row widths on June 10, 2010 and June 6, 2011. Planter
settings were adjusted to attain three target populations for each variety-row width combination.
168

Target populations were 196,500, 262,000, and 327,500 plants ha
196,500, and 262,000 plants ha

-1

-1

for black bean and 148,000,

for small red bean. Populations were held constant across row

widths, leading to differences in within-row plant spacing for each population (Table 4.1). When
the crop fully emerged, the number of bean plants in two 3-m sections of row in each plot was
counted to obtain actual emerged plant populations. Actual early-season stand counts following
emergence showed populations that often fell short of (or, less commonly, exceeded) the desired
populations. However, within each row width at each site, the highest target population did
result in the highest stand count and the lowest population in the lowest stand count (Table 4.2).
The experimental design was a randomized complete block with regard to variety, row width,
and planting population. Each treatment was replicated four times. All plots were maintained
weed-free throughout the growing season with a preemergence (PRE) application of Smetolachlor (Dual II Magnum, Syngenta Corporation, Wilmington, DE) at 1.4 kg ai ha
halosulfuron (Permit, Gowan Company, Yuma, AZ) at 35 g ai ha

-1

-1

plus

supplemented with hand

weeding. At East Lansing, clethodim (Select Max, Valent U.S.A. Corporation, Walnut Creek,
CA) at 77 g ha

-1

with crop oil concentrate at 1% v/v was applied postemergence (POST) to all

plots to control excess grasses (mainly Setaria spp.). Plot width was 3 m; with seven, five, and
four rows per plot for the 38-, 51-, and 76-cm row widths, respectively. Plot length ranged from
9.1 to 10.7 m, depending on location.
After the crop fully emerged, stand counts were taken in two 3-m sections of row in each plot to
determine actual emerged plant populations. In many cases, actual early-season stand counts
following emergence fell short of (or, less commonly, exceeded the target populations (Table
4.2). However, within each row width at each site, the highest population resulted in the lowest
169

stand count and the lowest population in the lowest stand count. In fifty-four of sixty site-yearclass combinations, stand count was similar enough to target population to essentially maintain
the proper relationship of populations among treatments.

In six cases, stand count was

dramatically lower than target planting population, although still higher than the next lowest
population within the row width. Four of these six cases occurred at Richville in 2010 and may
have been the result of poor emergence in dry soil (Table 4.3). The small number of interactions
between population and row width suggests that the effect of most of these variations from target
population had limited impact on the results of these studies.
Precipitation data was obtained throughout the growing season from the Michigan Automated
Weather Network (http://www.agweather.geo.msu.edu/mawn/, Michigan State University, East
Lansing, MI), which maintains stations on the Michigan State University Horticulture Farm less
than a mile from both field locations.
2

At physiological maturity, dry bean plants were sampled from three 0.25 m areas in each
plot. For each plant, the following data were taken: height, number of branches per plant,
number of pods on branches, and pods on the main stem. Branch and main-stem pods were
individually separated into categories of 1 to 9 seeds per pod. Beans were then threshed, and one
hundred seeds were counted and weighed to obtain the weight per hundred seeds. The data was
aggregated to obtain average number of pods per plant and average number of seeds per pod for
each sample. The ratio of the number of pods on the branches to the number of pods on the main
stem and the ratio of the average number of seeds per pod on the branches to the average number
of seeds per pod on the main stem were also calculated. In 2011, instead of examining all plants
2

per sample, a subsample of five plants was examined from each 0.25 m area.

170

Immediately prior to harvest, all plots were visually evaluated for lodging of bean plants. The
evaluation system was a 0-5 scale in which 0 indicated no lodging, and 5 indicated that all plants
were prostrate.

A preharvest treatment of glyphosate (Roundup PowerMax, Monsanto

Company, Saint Louis, MO) at 0.84 kg ae ha

-1

was applied to all plots at both locations in 2011

to expedite bean desiccation. Dry bean plots were direct harvested with a small plot combine
(Massey-Ferguson 8XP, AGCO, Duluth, GA) with a 1.5-m header. The center two-, three-, and
five-rows were harvested in the 76-, 51-, and 38-cm row width plots, respectively. Dry bean
yield was adjusted to 18% moisture.
Data were subjected to ANOVA using PROC MIXED in SAS 9.2 (SAS Institute Inc.,
Cary, NC) and treatment means were compared using Fisher’s protected LSD at the α < 0.05
level of significance. Some data were log or square root transformed for analysis, but all data is
presented untransformed. Data were combined over population, row width, year, and location
when no significant interactions were present.

Results and Discussion
Growing conditions. Rainfall from June through September in 2010 was lower than the 30-yr
average at both locations (Table 4.3). At East Lansing, above-average rainfall in June and nearaverage rainfall in July and September allowed dry bean plants to develop normally despite a dry
August. At Richville, rainfall was well below average in every month, and much of the June
rainfall came prior to planting; by the beginning of September, most bean plants had senesced,
rendering September rainfall irrelevant. Only 11 cm of rain fell at Richville between planting
and harvest, resulting in stunted crop plants.

171

In 2011, rainfall from June through September was slightly above the 30-yr average at
East Lansing, while at Richville it was once again below the 30-yr average. At East Lansing,
rainfall was slightly to moderately below average in three months, but in late July a single
weather system brought 16 cm of rainfall over three days, making that month’s total dramatically
above average and supplying ample water for pod formation, especially with moderate rainfall
continuing through September. At Richville, rainfall was below average in every month, but all
rainfall in June occurred after planting, and monthly totals were higher than those of the previous
year in July and August. Rainfall was sufficient for normal plant development, and pod filling
continued into September.

Yield. Yield was significantly greater in 38-cm rows than in 76-cm rows in four of the eight
site-year-class combinations: East Lansing 2010 black beans, East Lansing 2011 black beans,
East Lansing 2011 small red beans, and Richville 2011 small red beans (Table 4.4). In only one
site-year-class combination (Richville 2010 black beans, which experienced severe drought) did
the opposite effect occur, and this was not significant at α < 0.05. In addition, in a companion set
of studies in black bean (presented in chapter 3), yield was higher in 38-cm rows than in 76-cm
rows in three of four site-years. Yield rarely differed by planting population (Table 4.5).

Lodging. Lodging was always at least 80% greater in small red bean than in black bean in every
site-year on a 0-5 evaluation scale (Table 4.6). At Richville in 2010, virtually no lodging was
observed in black bean. In four of the remaining seven site-year-class combinations, 38-cm rows
significantly reduced lodging compared with 76-cm rows, and in two of four combinations, 51cm rows resulted in a similar lodging reduction.

172

In only one combination, narrow rows

marginally increased lodging (α ≤ 0.1). Planting population did not affect lodging in any siteyear (Table 4.7).

Plant height. Average plant height was consistently greater in small red beans than in black
beans in all site-years. At East Lansing in 2010, there was a row width by planting population
interaction: black bean height was 6.6 cm greater in 38-cm rows than in 76-cm rows at 327,000
seeds ha

-1

only. At Richville in 2010, black bean height was 4.2 cm greater in 51-cm rows than

in 38-cm rows across all populations; height was intermediate in 76-cm rows (Table 4.8).
Planting population did not affect height at either site (Table 4.9). In 2011, plant height was not
affected by row width or planting population in either class at either site-year. Small red bean
height was unaffected by row width or planting population in all site-years.
Planting population had no effect on plant height in any circumstance, while the effect of
row width was inconsistent and often not present at all. Plant height appears to be essentially
independent of the factors examined in these studies. This contrasts with the work of Weber et
al. (1966) in soybean, who found that soybean height increased with increasing populations, and
Horn et al. (2000) in dry bean, who found that dry bean height was lower in narrow rows.

Pods per plant. Black beans had significantly more pods per plant than small red beans except
at East Lansing in 2010. Average number of pods per plant in black bean varied widely among
site-years from 13.2 at Richville in 2010 to 19.3 at Richville in 2011. Average number of pods
per plant in small red beans ranged from 10.4 at Richville in 2010 to 14.4 at East Lansing in
2011.

173

At East Lansing, there were more pods per plant in 38-cm rows than in 76-cm rows in
every site-class combination. In 2010, the number of pods per plant in 38-cm rows was 17%
greater in black bean and 18% greater in small red bean than in 76-cm rows. In 2011, number of
pods per plant was 36% and 40% greater in 38-cm than in 76-cm rows respectively (Table 4.10).
At Richville, the effect of row width was less consistent. In 2010, there was a row width
by population interaction: there were fewer black bean pods per plant in 38-cm rows (8.5 pods
-1

-1

plant ) than in 51- (11.4) or 76-cm rows (10.8) at 327,500 seeds ha only. Across populations,
there were more small red bean pods per plant in 51-cm rows (11.8) than in 38-cm rows (10.5),
and more in either narrow row width than in 76-cm rows (8.8) (Table 4.10). At Richville in
2011, there were more black bean pods per plant in 76-cm rows (21.0) than in 51-cm rows
(17.8), with 38-cm rows intermediate (19.4). In small red beans, row width did not have a
significant effect on number of pods per plant.
The effect of planting population on number of pods per plant was consistent. In every
site-year-class combination, number of pods per plant increased as population decreased. Across
all sites, years, and classes, number of pods per plant averaged 11.8 at high populations, 14.3 at
medium populations, and 17.7 at low populations (Table 4.9).
In five of the eight site-year-class combinations, there were more pods per plant in 38-cm
rows than in 76-cm rows. In one of the four combinations in which 51-cm rows were used, there
were more pods per plant in 51-cm rows than in 76-cm rows. In only one case did 76-cm rows
result in more pods per plant than 51-cm rows, and in no case did 76-cm rows result in more
pods per plant than 38-cm rows across all populations. Thus, narrow rows tended to result in
more pods per plant, probably because of the increase in light interception often observed in
narrow rows. When combined across site-years and classes, number of beans per pod increased
174

with decreasing population; number of pods per plant was greater at low populations than at
either medium or high populations and greater in medium populations than in high populations
(Table 4.9). Thus, dry bean plants are able to compensate for low population, producing more
pods per plant when fewer plants are present; this should enable them to produce similar yields
across a range of populations. This confirms the findings of Goulden et al. (1976), Bennett et al.
(1977), and Grafton et al. (1988) that number of pods per plant in dry bean is lower at higher
populations. Change in number of pods is an important mechanism controlling yield changes in
individual plants.

Seeds per pod. Black beans had significantly more seeds per pod than small red beans in every
site-year. Average number of black bean seeds per pod in various site-years ranged from 3.99 to
5.39, while average number of small red bean seeds per pod ranged from 3.07 to 3.65.
The effect of row width on seeds per pod was inconsistent (Table 4.11). At East Lansing
in 2010 and Richville in 2011, row width did not have a significant effect on number of seeds per
pod. However, at East Lansing in 2011, number of black bean seeds per pod averaged 4.99
seeds pod

-1

in 38-cm rows compared with just 4.60 seeds pod

-1

in 76-cm rows; row width did

not affect number of seeds per pod in small red bean.
Under drought conditions at Richville in 2010, number of seeds per pod significantly
increased with widening row width across classes. In black bean, the average number of seeds
per pod was 3.85 in 38-cm rows, 4.04 in 51-cm rows, and 4.43 in 76-cm rows. In small red
beans, the average number of beans per pod was 3.15 in 38-cm rows, 3.49 in 51-cm rows, and
3.55 in 76-cm rows.

175

The fact that the site-year in which number of seeds per pod increased in wide rows was a
drought year suggests that number of seeds per pod may have been reduced in narrow rows by
intraspecific competition for water between the rows. This may also explain the apparent yield
reduction in narrow rows in this site-year-class combination. Number of pods per plant is more
likely to play a role in yield differences due to row width than number of seeds per pod, but it
appears that seeds per pod may also sometimes be responsive to intraspecific competition. The
general lack of difference in number of seeds per pod between different row widths is consistent
with the work of Goulden et al. (1976) who saw no differences in seeds per pod between row
widths.
Although within most site-year-class combinations the effect of planting population on
number of seeds per pod was non-significant, when analyzed across all site-year-class
combinations, there were significantly more seeds per pod at low populations than at high
-1

-1

populations; low populations averaged 3.99 seeds pod , medium populations 3.92 seeds pod ,
and high populations 3.85 seeds pod

-1

(Table 4.9). This suggests that an increase in the number

of seeds per pod plays a role in dry bean compensation for low population. This confirms the
finding of Grafton et al. (1988) that number of seeds per pod decreases with increasing
population. While this effect was less pronounced than the increase in number of pods per plant
with decreasing population, it does appear that an increase in number of seeds per pod plays a
role in dry bean compensation for low population.

Weight per hundred seeds. Dry bean seed weight was consistently higher in small red bean
than in black bean as both are bred to maintain a relatively uniform seed size. Weight per

176

hundred seeds varied among site-years. Black bean weight per hundred seeds ranged from 15.4
g to 24.0 g, and small red bean weight per hundred seeds ranged from 29.5 g to 41.8 g.
At East Lansing in both years and at Richville in 2010, weight per hundred seeds within
each class was not significantly different between row widths (Table 4.12).

However, at

Richville in 2011, weight per hundred seeds was 11% and 12% greater in 76-cm rows than in 51or 38-cm rows respectively in black bean and 5% and 8% greater respectively in small red bean.
Across all site-years in small red bean seed weight was higher at 148,000 seeds ha-1 than
-1

at 262,000 seeds ha at the α=0.1 level of significance (Table 4.9). However, population had no
effect on seed weight in black bean.
In one site-year, seed weight was greater in wide rows than in narrow rows; the effect
was small and did not occur in other site-years. It does appear that small red bean weight was
slightly higher at low population than at high population, but this effect was very slight. Seed
weight generally appears to be insensitive to row width and population; this is a positive finding
because growers would not want substantial variation in bean size from the industry standard for
a given class. This is also consistent with the finding of Goulden et al. (1976) that seed weight
did not differ between row widths. The reason for the increase in seed weight in wide rows at
Richville in 2011 is not obvious, particularly since in one of the classes at that site-year, yield
was actually higher in narrow rows. That seed weight should increase at low population does
seem to fall into the pattern of plants compensating for low plant density, but the small size of
the increase makes it questionable whether this is really the correct explanation.

Branching. Number of branches per plant was unaffected by class except at Richville in 2011
-1

when black beans averaged 0.35 more branches plant than small red beans, and at all site-years,
177

the classes were affected in the same manner by planting population and row width, so data is
presented across classes. Number of branches per plant was similar at Richville in 2010, at East
Lansing in 2011, and in small red beans at Richville in 2011, averaging 1.83-1.95 branches plant

-

1

. Number of branches per plant was higher in black bean at Richville in 2011 at 2.25 branches
-1

-1

plant and was lower at East Lansing in 2010 at 1.56 branches plant .
Narrow rows increased branching in all planting populations in three of the four siteyears (Table 4.13). At East Lansing in 2010, plants in 38-cm rows averaged 23% more branches
per plant than those in 76-cm rows. At East Lansing in 2011, plants in 38-cm rows averaged
47% more branches than those in 76-cm rows. At Richville in 2011, branching was 18% and
12% greater in 38-cm rows than in 51- or 76-cm rows respectively.
At Richville in 2010 under drought conditions, the effect of row width on branching
depended on planting population (Table 4.14). At low populations, branching was 18% higher in
38-cm rows than in 76-cm rows with 51-cm rows intermediate. At medium populations, the
effect of row width was non-significant. At high populations, branching was 39% greater in 51cm rows than in 38-cm rows with 76-cm rows intermediate.
The effect of planting population on branching was consistent across all site-year-class
combinations (Table 4.9). Across all combinations, number of branches per plant increased as
population decreased, averaging 1.52 branches plant

-1

at high populations, 1.80 branches plant

-1

-1

at medium populations, and 2.28 branches plant at low populations.
In most cases, plants had a higher number of branches in 38-cm rows than in 76-cm rows;
51-cm rows were either intermediate or similar to 76-cm rows. It appears that narrow rows tend
to encourage branching, which may account for a portion of the ability of narrow- row dry beans
178

to develop a more complete canopy than in wide rows; this differs from findings in soybean that
row width does not affect branching (Ethredge et al. 1989) or that branching increases in wide
rows (Lehman and Lambert 1960). However, an increase in branching in narrow rows makes
sense in light of the much greater within-row plant spacing in narrow rows within a given
population; presumably, narrow rows allow the plants more room to grow toward one another
without coming into competition with one another. Since Bennett et al. (1977) found that
branching was associated with higher yields in dry bean, this increase in branching may
contribute to observed increases in yield in narrow rows.
Branching increased with decreasing population. This confirms the finding of Bennett et
al. (1977) that dry bean plants produce fewer branches with increasing population. Dry bean
plants clearly compensate for lower populations by branching, allowing individual plants to grow
larger and fill in the gaps between plants; this behavior presumably accounts for the limited
effect of planting population on canopy closure, yield, and other parameters observed in other
portions of these studies. It is also similar to the findings of Ethredge et al. and Lehman and
Lambert that branching increases with decreasing population in soybean.

Pod distribution between main stem and branch. The average number of pods on the main
stem was somewhat similar across site-years, ranging from 6.8 to 8.7 in black beans and from 5.6
to 6.0 in small red beans. There were more pods on the main stem in black beans than in small
red beans in every site-year. Average number of pods on branches per plant was more variable
between site-years, ranging from 5.4 to 10.3 in black bean and from 4.4 to 7.4 in small red bean.
The lowest average numbers of branch pods per plant occurred at Richville in 2010, suggesting

179

that drought conditions reduced the number of branch pods despite only a marginal reduction in
number of branches.
Main-stem pods. At East Lansing in both years and in small red bean at Richville in 2011, row
width did not affect the number of pods on the main stem. In the other cases, the effect was
inconsistent (Table 4.15). At Richville in 2010 across classes, main-stem pods per plant was
higher in 51-cm rows than in 38-cm rows, but higher in both 51 and 38-cm rows than in 76-cm
rows. In contrast, at Richville in 2011, there were more small red bean main-stem pods per plant
at 76- or 51-cm than at 38-cm.
In six out of eight site-year-class combinations, lower populations led to more main-stem
pods per plant (Table 4.16). At East Lansing in 2010 across classes, there were more main-stem
pods per plant at low and medium populations than at high populations; differences were 24%
and 15% respectively. At East Lansing in 2011 in black bean, there were more main-stem pods
per plant at low population than at medium population or high population; differences were 25%
and 33%. At Richville in 2010, main-stem pods per plant decreased with increasing population
across classes, with differences ranging from 6-17%. At Richville in 2011, there were more
-1

black bean main-stem pods per plant at 196,500 seeds ha than at 262,000 or 327,500 seeds ha

-

1

, differences of 14% and 17%. However, in 2011 at both sites, main-stem pods per plant in

small red bean was not affected by planting population.
Number of pods on the main stem was often unaffected by row width, but when it was
affected, it was maximized in 51-cm rows; this occurred in all three site-year-class combinations
in which row width had a significant effect on main-stem pods per plant. In six of eight siteyear-class combinations, lower populations increased main-stem pods per plant. It appears that

180

pod development on the main stem is a part of compensation for lower populations in many
cases.
Branch pods. In seven of eight site-year-class combinations, narrower rows led to more branch
pods per plant in at least some circumstances (Table 4.17 for main effects, Tables 4.18 and 4.19
for interactions).
At East Lansing in both years, number of branch pods per plant was higher in 38-cm
rows than in 76-cm rows across both classes. In 2010, there were 33% more branch pods per
plant in 38-cm rows than in 76-cm rows. In 2011, there were 73% more branch pods per plant in
38-cm rows than in 76-cm rows (Table 4.17).
At Richville in 2010, there was an interaction of row width with planting population
across classes with regard to branch pods (Table 4.18). At low populations, there were more
branch pods per plant in 38-cm rows than in 76-cm rows, with 51-cm rows intermediate, whereas
at high populations, there were more branch pods per plant in 51-cm rows than in 38-cm rows,
with 76-cm rows intermediate. At medium populations, row width did not have a significant
effect on branch pods. At Richville in 2011 in small red bean, there was also an interaction of
-1

row width with planting population (Table 4.19). At 196,500 and 327,000 seeds ha , row width
-1

did not have a significant effect on branch pods; at 262,000 seeds ha , branch pods per plant
was higher in 38-cm rows than in 76-cm rows, with 51-cm rows intermediate.
At Richville in 2011 in black bean, there were 25% more branch pods per plant in 76-cm
rows than in 51-cm rows, with 38-cm rows were intermediate.
In all year-class combinations at East Lansing, there were more branch pods per plant in
narrow rows than in wide rows. At Richville, there were interactions in two of the four yearclass combinations in which branch pods per plant were higher in narrow rows than wide rows in
181

certain cases, and in an additional combination, the intermediate row width had more branch
pods per plant than the wide row width. In only one year-class combination were there more
branch pods per plant in wide rows, and this occurred during the 2010 Richville drought.
Across all site-year-class combinations, number of branch pods plant

-1

increased with

decreasing population (Table 4.9, footnote c.). At high populations, number of branch pods per
plant averaged 5.22, at medium populations 6.81, and at low populations 9.47.
Branch pods are highly responsive to planting population and likely play a major role in
allowing dry bean plants to compensate for low population. They are also often responsive to
narrow rows and likely play a major role in the increase in yield often observed in narrow rows.
Ratio of branch pods to main-stem pods. The ratio of pods on the branches to pods on the main
stem (branch-stem ratio) was similar between classes except at East Lansing in 2010 when it was
1.15 in black beans and 0.71 in small red beans. In black beans, the branch-stem ratio was
similar across site-years, ranging from 1.12 to 1.18 except at Richville in 2010 when it was 0.70.
The branch-stem ratio in small red beans differed by year, being 0.68-0.71 in 2010 compared to
1.03-1.11 in 2011. The low ratios at Richville in 2010 suggest that in drought conditions, a
higher percentage of pods develop on the main stem than is typical in a year of adequate
moisture.
In six of the eight site-year-class combinations, the branch-stem ratio increased with
narrowing rows (Table 4.20). Across all site-year-class combinations, the branch-stem ratio
increased with reduced populations (Table 4.9).
In three site-years, the branch-stem ratio was lower in wide rows than in narrow rows. At
East Lansing in both years, the black bean branch-stem ratio was lower in 76-cm rows than in
38-cm rows in both classes. In 2010, the ratios in 76-cm and 38-cm rows were 0.62 and 0.82 for
182

black bean and 0.98 and 1.39 for small red bean; in 2011, across both classes, they were 0.89 and
1.52. At Richville in 2011, the black bean branch-stem ratio was marginally lower (α=0.1) in
51-cm rows at 1.02 than in 38-cm rows at 1.22 with 76-cm rows intermediate at 1.14. The small
red bean branch-stem ratio was lower in 51- and 76-cm rows at 0.98 and 0.84 respectively than
in 38-cm rows at 1.25. Under drought conditions at Richville in 2010, row width did not affect
the branch-stem ratio in either class.
The branch-stem ratio was higher in 38-cm rows than in 76-cm rows in five of the eight
site-year-class combinations and was higher in 38-cm rows than in 51-cm rows in one of these
five combinations plus an additional one (Table 4.20).

The only exceptions to higher

branch-stem ratios in narrow rows occurred at Richville in 2010 under drought conditions, and
the reverse effect never occurred. This indicates that in narrow rows, a higher percentage of
pods produced are often borne on the branches. It appears that the increased formation of branch
pods is a major mechanism by which yield may be increased in narrow rows. This is probably
due at least partly to the greater within-row spacing between seeds in narrow rows for a given
population. Each plant has more room to spread out toward neighboring plants in narrow rows,
thus encouraging branching parallel to the row.
Across all site-year-class combinations, the branch-stem ratio increased with decreasing
population (Table 4.9). At high populations, the branch-stem ratio averaged 0.71, at medium
populations 0.87, and at low populations 1.25. This indicates that forming branch pods is a
major means by which dry beans compensate for low populations and that more branch pods are
formed due to this compensation than main-stem pods.

183

Distribution of seeds per pod between main stem and branch. Average number of seeds per
pod on the main stem in black bean ranged from 3.98 to 5.44 and in small red bean from 2.99 to
3.7. Average number of seeds per pod on the branches in black beans ranged from 3.94 to 5.34
and in small red beans from 3.15 to 3.6. In every site-year-class combination, there more seeds
per pod both on the main stem and on the branches in black bean than in small red bean.
Seeds per main-stem pod.

At East Lansing in 2010, there were more black bean seeds per

main stem pod in 38- than in 76-cm rows; row width did not affect seeds per main-stem pod in
small red beans. Conversely, at Richville in 2010 under drought conditions, number of seeds per
main-stem pod increased with increasing row width across classes. In 2011, row width did not
affect seeds per main-stem pod at either site, and planting population did not affect seeds per
main-stem pod in any site-year.
One case in which narrow rows increased the number of seeds per pod on the main stem
was observed, suggesting the possibility that this may occasionally play a role in increasing
yields, but this was hardly definitive. The observation of reduced seeds per main-stem pod in
narrow rows under drought conditions at Richville in 2010 corresponded to the overall reduction
in seeds per pod in that site-year, which may indicate intra-specific competition for water
between the rows in narrow rows.
Seeds per branch pod. At East Lansing in 2011, number of seeds per branch pod was higher in
38- than in 76-cm rows. Conversely, at Richville in 2010, number of seeds per branch pod
increased with increasing row width across classes. At East Lansing in 2010 and at Richville in
2011, row width did not affect seeds per branch pod in either class, and planting population
never significantly affected seeds per branch pod.

184

The increase in seeds per branch pod at one site-year once again suggests the possibility
that an increase in the number of seeds per pod may sometimes play a small role in positive yield
responses to narrow rows. Once again, the increase in seeds per branch pod in wide rows at
Richville in 2010 suggests the possibility of intra-specific competition for water reducing
number of seeds per pod.
Ratio of branch seeds per pod to main-stem seeds per pod. The ratio of the average number of
seeds per pod on the branches to the average number of seeds per pod on the main stem (branchstem seed ratio) tended to be close to 1.0, indicating that the number of seeds per pod does not
strongly differ between the main stem and the branches. In black bean, the stem-branch seed
ratio ranged from 0.93 to 1.03 while in small red beans, it ranged from 0.91 to 1.05. The branchstem seed ratio differed very little or not at all between classes. The lowest branch-stem seed
ratios occurred at Richville in 2010, suggesting that drought conditions may cause bean plants to
form more seeds per pod on the main stem than on the branches. The next lowest branch-stem
seed ratios were at East Lansing in 2010 and were 0.97 in small red beans and 0.98 in black
beans; at the other two site-years, ratios were 1.0 or higher.
The branch-stem seed ratio remained near 1.0 regardless of row width. At East Lansing
in both years and at Richville in 2011, row width did not affect the branch-stem seed ratio (Table
4.21). However, under drought conditions at Richville in 2010, the branch-stem seed ratio was
lower across both classes in 38-cm rows than in 51- or 76-cm rows; in 38-cm rows, the branchstem seed ratio was 0.89, while in both wider row widths, it was 0.93, another possible indication
of drought conditions favoring seed production on the main-stem and of drought stress being
increased by intraspecific competition in narrow rows.

185

The branch-stem seed ratio was also remained near 1.0 regardless of population. In seven
of the eight site-years, planting population did not affect the branch-stem seed ratio (Table 4.22).
-1

The only exception was at East Lansing in 2010 in small red bean; at 262,000 seeds ha , branchstem seed ratio was just 0.93 compared to 1.01 at 196,500 seeds ha

-1

and 0.99 at 148,000 seeds

-1

ha .
The branch-stem seed ratio is fairly stable; the decrease in the ratio in narrow rows under
drought conditions and the decrease at high population in one other site-year-class combination
suggest that intra-specific competition can result in disequilibrium in the number of seeds per
pod between the main stem and branches, with the main stem being favored; however, this effect
is not large. It does not appear that any increases in yield with branching are due a higher
number of seeds per pod in branch pods than in main-stem pods.

Conclusions. Neither row width nor planting population were found to consistently affect plant
height. However, lodging was often reduced in narrow rows, probably due to the mutual support
of the rows across the smaller inter-row space. Planting population did not have a pronounced
effect on lodging. Branching was often increased in narrow rows and always increased with
increasing population.
The yield component that was most sensitive to row width and planting population was
number of pods per plant, which was often higher in narrow rows and always increased with
decreasing population.
Number of pods on the main stem was often unaffected by row width but was maximized
in 51-cm (intermediate width) rows when it was affected. Number of pods on the main stem was

186

often higher at lower populations, but this did not always occur. However, just as branching was
highly sensitive to row width and population, number of branch pods was highly sensitive to row
width and population. Branch pods often increased in response to reduced row width and always
increased in response to reduced population. In addition, percentage of pods on the branches
was higher in both narrow rows and at low populations, indicating that much of the increase in
number of pods per plant was occurring on the branches.
Number of seeds per pod was less sensitive to changes in row width and population than
number of pods. In only one site-year-class combination did narrow rows have a positive effect
on average number of seeds per pod; however, low populations had an average of 0.14 more
seeds per pod than high populations across all site-years. This suggests that increases in seeds
per pod may at times play a role in yield responses to row width and do play a role in yield
compensation for low population, but it also suggests that this role is relatively small. Under
drought conditions at Richville in 2010, number of seeds per pod increased with increasing row
width, an increase observed in both the main-stem and branch fractions, and at low populations
the number of branch seeds per pod was marginally higher than at high population. This
suggests that narrow rows and high populations may lead to intraspecific competition for water
in drought conditions, which can be expressed in a reduction in the number of seeds per pod.
Differences in bean seed weights between treatments were few and small, with two
instances of average weight being reduced in narrow rows and a marginally significant increase
in small red bean weight at low population compared to high population overall. It seems
unlikely that these differences reflect a significant contribution of seed weight variation toward
yield differences between row widths, although it is possible that seed variation does play a small
role in small red bean compensation for low population.

187

Thus, the primary driver of yield increases in narrow rows and yield compensation for
low populations seems to be increasing numbers of branch pods. While main-stem pods did
sometimes change in response to the treatments, especially planting population, and number of
seeds per pod showed some response to them, branch pods demonstrated the most
responsiveness. It appears that increased branching is key for increasing plant yield; this is
because yield increases seem to be largely the result of adding more pods on the branches and
because branches enable the plants to more efficiently fill the canopy, thereby capturing more
light energy that can be used to form more pods.
The often-observed yield advantage in narrow rows presumably occurs because a higher
percentage of available resources may be appropriated to the crop plants in narrow rows. While
this is likely due in part to the greater efficiency of narrow rows – the distance between rows is
more similar to the difference between plants within the row – it may also be partially the result
of increased branching in narrow rows. Resources are then allocated to form a greater number
pods per plant, especially on the branches. This may at times be supplemented by a small
advantage in the average number of seeds per pod. In addition, the reduction in lodging in
narrow rows may play a small role in causing higher yields by reducing yield loss at harvest.
Similarly, the often-observed failure of high populations to increase yield apparently
occurs because more branching occurs at low populations. This allows plants to compensate for
lower populations by growing larger and branching, thus filling the canopy as well or nearly as
well as they would at higher populations. There may also be a corresponding increase in the
extent of the root system, though this has not been confirmed. Resources are again allocated to
form a greater number of pods per plant, especially on the branches. This may typically be

188

supplemented by a small advantage in the average number of seeds per pod and, in small red
bean, by a small increase in seed size.

189

Table 4.1. Within-row plant spacing for Type II black and small red beans planted in three row
widths and at three planting populations.
Target black bean population
Target small red bean population
Row width 196,500 ha-1 262,000 ha-1 327,500 ha-1 148,000 ha-1 196,500 ha-1 262,000 ha-1
____________________

38 cm
51 cm
76 cm

13
10
7

cm ____________________
10
8
8
6
5
4

190

____________________

18
13
9

cm ____________________
13
10
10
8
7
5

Table 4.2. Early-season stand counts of black and small red beans compared with target planting populations at East Lansing and
Richville in 2010 and 2011.
East Lansing
Richville
Target planting
Populations (seeds
-1
ha )
Black beans
196,500
262,000
327,500

2010
38-cm

76-cm

_______________

197,500
252,000
313,500

2011
38-cm

2010

76-cm

38-cm

-1

plants ha _______________
175,500
187,500 168,000
246,000
222,500 226,000
305,000
286,500 263,000

51-cm

2011
76-cm

___________________________

38-cm

51-cm

76-cm

-1

174,000
226,000
327,500

141,000
187,500
233,500

Small red beans
148,000
149,500 162,000
141,000 139,500
196,500
211,000 207,500
175,500 173,000
262,000
289,000 267,000
246,000 236,000
*Stand count was at or below the next lowest target population.

191

163,000
210,000
274,000

plants ha ___________________________
165,500
183,000 183,000 131,000*
196,500*
233,500 246,000 218,500
253,000*
295,000 300,000 228,500*

131,000
148,000*
153,000*

138,500
179,000
222,000

142,000 148,000
175,500 186,500
238,500 221,000

135,000
165,500
205,000

Table 4.3. Monthly and 30-yr average precipitation at the Harry and Hazel Box Farm in East
Lansing, MI and at the Saginaw Valley Research and Extension Center in Richville, MI in 2010
a
and 2011.
Month

East Lansing
2011
30-yr avg.

2010

_________________

b

June

Richville
2011

2010

cm _________________

_________________

b

30-yr
c
avg.

cm _________________

10.6 (4.1)

4.4

8.4

6.9 (2.8)

3.8

9.3

6.4
3.4
9.2
29.6

17.8
6.4
6.7
35.3

8.2
8.3
9.3
34.2

2.3
3.2
2.9
15.3

3.4
7.6
5.8
20.6

7.6
8.0
10.7
35.6

July
August
September
Total
a

Michigan Automated Weather Network, http://www.agweather.geo.msu.edu/mawn/, Michigan
State University, East Lansing, MI
b
Precipitation data in parenthesis is from the time of planting.
c

30-yr average for Caro, MI; none was available for Richville

Table 4.4. Main effect of row width on yield of black and small red bean at East Lansing and
a
Richville in 2010 and 2011 ..
East Lansing
2010
Row
width

Black

Richville
2011

Small
red

Black

2010
Small
red

________________________________

38 cm
51 cm
76 cm
Mean

3257 a
2785 b
3021

2553 a
2486 a
2520

4494 a
3903 b
4198

kg ha

3329 a
2830 b
3080

a

Black

b

2011
Small
red

Black

Small
red

-2 _________________________________

1488 b
1553 b
1814 a
1618

1612 a
1674 a
1797 a
1694

3040 a
2924 a
2896 a
2953

2592 a
2550 a
2339 b
2494

Means within each column that the same letter are statistically similar to one another at the α ≤
0.05 level of significance unless otherwise noted.
b
Significance level of α ≤ 0.1

192

Table 4.5. Main effect of planting population on yield of black and small red bean at East
a
Lansing and Richville in 2010 and 2011 .
East Lansing
2010
Planting
b
population

Black

Small
red

Richville

2011
Small
Black
c
red

_______________________________

kg ha

2010
Black

Small
red

2011
Black

Small
red

-1 _________________________________

Low
3020 a
2425 a
4232 a
2958 b
1674 a 1694 ab 2915 a
2427 a
Medium
3138 a
2554 a
4103 a
3059 ab 1656 a 1774 a
2984 a
2523 a
High
2902 a
2577 a
4258 a
3222 a
1751 a 1582 b
2956 a
2619 a
a
Means within each column that the same letter are statistically similar to one another at the α ≤
0.05 level of significance unless otherwise noted.
b
-1
Low population targets were 148,000 seeds ha in small red beans and 196,500 in black beans;
medium population targets were 196,500 seeds ha

-1

in small red beans and 262,000 in black
-1

beans high population targets were 262,000 seeds ha in small red beans and 327,500 in black
beans.
c
Significance level of α ≤ 0.1

Table 4.6. Main effect of row width on lodging in black and small red bean at East Lansing and
a
Richville in 2010 and 2011 .
East Lansing
Richville
2010
2011
2010
2011
Row
Small
Small
Small
Small
Black
Black
Black
Black
width
red
red
red
red
_________________________________
(0-5 scale) _________________________________
38 cm
0.38 a
3.92 a
2.12 a
3.25 b
0a
0.62 b
1.83 b
2.83 b
51 cm
0a
1.17 a
1.83 b
2.96 b
76 cm
0.71 a
4.00 a
1.77 a
4.21 a
0.21 a
1.29 a
2.5 a
3.44 a
Mean
0.54
3.96
1.94
3.73
0.07
1.03
2.05
3.08
a
Means within each column that the same letter are statistically similar to one another at the α ≤
0.05 level of significance.

193

Table 4.7. Main effect of population on lodging in black and small red bean at East Lansing and
a
Richville in 2010 and 2011 .
East Lansing
Richville
2010
2011
2010
2011
Planting
Small
Small
Small
Small
Black
Black
Black
Black
population
red
red
red
red
___________________________________________
___________________________________________
(0-5 scale)
Low
0.56 a
3.75 a
1.97 a
4.0 a
0a
1.00 a
1.81 a
3.25 a
Medium
0.56 a
3.94 a
1.81 a
3.62 a
0a
1.08 a
1.94 a
3.04 ab
High
0.50 a
4.19 a
2.06 a
3.56 a
0.21 a
1.00 a
2.42 a
2.93 b
a
Means within each column that the same letter are statistically similar to one another at the α ≤
0.05 level of significance.

Table 4.8. Main effect of row width on plant height in black and small red bean at East Lansing
a

and Richville in 2010 and 2011 .
East Lansing
Richville
2010
2011
2010
2011
Row
Small
Small
Small
Small
b
Black
Black
Black
Black
width
red
red
red
red
_____________________________________
Plant height (cm) __________________________________
38 cm
47 a
62 a
55 a
73 a
35 b
54 a
56 a
73 a
51 cm
39 a
56 a
55 a
72 a
76 cm
47 a
61 a
53 a
69 a
38 ab
55 a
56 a
78 a
Mean
47
61.5
54
71
37.3
55
55.7
74.3
a
Means within each column that the same letter are statistically similar to one another at the α ≤
0.05 level.
b
-1
-1
196,500 and 262,000 seeds ha only; within 327,500 seeds ha , plant height in 38-cm rows
averaged 51 cm, which was statistically different from 76-cm rows, in which plant height
averaged 44 cm.

194

Table 4.9. Main effect of planting population on yield components and plant architecture in
a
black and small red beans at East Lansing and Richville in 2010 and 2011 .
Planting
Pods per
Seeds
Weight per hundred
Plant
Branches
Ratio of
b
c
per pod
beans (g)
height
per plant
branch
population
plant
(cm)
pods to
Black
Small
d
main-stem
Red
pods
Low
17.68 a
3.99 a
19.47 a
36.61 a
57 a
2.28 a
1.25 a
Medium
14.27 b
3.92 ab
19.72 a 35.79 ab
57 a
1.80 b
0.87 b
High
11.84 c
3.85 b
19.27 a
35.34 b
57 a
1.52 c
0.71 c
a

Means within each column that the same letter are statistically similar to one another at the α ≤
0.05 level unless otherwise noted.
b
-1
Low population targets were 148,000 seeds ha in small red beans and 196,500 in black beans.
-1

Medium population targets were 196,500 seeds ha in small red beans and 262,000 in black
-1

beans. High population targets were 262,000 seeds ha in small red beans and 327,500 in black
beans.
c
-1
Pods plant on branches also increased with decreasing row width across all site-year-class
-1

-1

combinations, averaging 5.22 plant at high populations, 6.81 plant at medium populations,
-1

-1

and 9.47 plant at low populations. Pods plant on the main stem did so in six of eight
combinations (Table 4.16)
d
Significance level of α ≤ 0.1

Table 4.10. Main effect of row width on number of pods per plant in black and small red bean at
a

East Lansing and Richville in 2010 and 2011 .
East Lansing
Richville
2010
2011
2010
2011
Row
Small
Small
Small
Small
b
Black
Black
Black
Black
width
red
red
red
red
____________________________
Number of pods per plant ____________________________
38-cm
16.2 a
15.2 a
19.7 a
16.8 a
15.2 a
10.5 b
19.4 ab
12.8 a
51-cm
14.3 a
11.8 a
17.8 b
13.4 a
76-cm
13.8 b
12.9 b
14.5 b
12.0 b
14.6 a
8.8 c
21.0 a
13.3 a
Mean
15.0
14.0
17.1
14.4
14.7
10.4
19.4
13.2
a

Means within each column that the same letter are statistically similar to one another at the α ≤
0.05 level unless otherwise indicated.
b
-1
-1
196,500 and 262,000 seeds ha only; within 327,500 seeds ha , there were more pods per
plant in 76-cm (10.8) or 51-cm rows (11.4) than in 38-cm rows (8.5).

195

Table 4.11. Main effect of row width on number of seeds per pod in black and small red bean at
a
East Lansing and Richville in 2010 and 2011 .
East Lansing
Richville
2010
2011
2010
2011
Row
Black
Small red
Black
Small red Combined
Black
Small red
width
________________________________
Seeds per pod________________________________
38 cm
5.48 a
3.59 a
4.99 a
3.11 a
3.47 c
3.99 a
3.02 a
51 cm
3.74 b
3.96 a
3.09 a
76 cm
5.29 a
3.71 a
4.60 b
3.09 a
3.95 a
4.02 a
3.10 a
Mean
5.38
3.65
4.80
3.10
3.72
3.99
3.07
a
Means within each column that the same letter are statistically similar to one another at the α ≤
0.05 level.

Table 4.12. Main effect of row width on weight per hundred seeds in black and small red bean at
a
East Lansing and Richville in 2010 and 2011 .
East Lansing
Richville
2010
2011
2010
2011
Row
Small
Small
Small
Small
Black
Black
Black
Black
width
red
red
red
red
___________________________
Weight per hundred seeds (g) ___________________________
38 cm
15.67 a 28.72 a 24.06 a 42.47 a 18.37 a 34.58 a 19.34 b 35.26 b
51 cm
18.44 a 33.93 a 19.40 b 36.51 b
76 cm
15.21 a 30.25 a 24.00 a 41.04 a 19.55 a
35.5 a
21.07 a 38.70 a
Mean
15.44
29.48
24.03
41.76
18.79
34.67
19.94
36.82
a
Means within each column that the same letter are statistically similar to one another at the α ≤
0.05 level.

Table 4.13. Main effect of row width on number of branches per plant combined across black
a
and small red bean at East Lansing and Richville in 2010 and 2011 .
East Lansing
Richville
Row width
2010
2011
2011
_____________________
Branches per plant ______________________
38 cm
1.72 a
2.32 a
2.27 a
51 cm
1.92 b
76 cm
1.40 b
1.58 b
2.03 b
a

Means within each column that the same letter are statistically similar to one another at the α ≤
0.05 level.

196

Table 4.14. Interaction between row width and planting population with regard to number of
a
branches per plant in black and small red bean at Richville in 2010 .
Planting population
Row width

b

Low

Medium
High
___________________________
Branches per plant
38 cm
2.51 a
1.86 cd
1.13 f
51 cm
2.29 ab
1.78 cde
1.56 de
76 cm
2.12 bc
1.77 cde
1.44 ef
a
Means throughout the table that have the same lower-case letter are statistically similar to one
another at the α ≤ 0.05 level.
b
-1
Low population targets were 148,000 seeds ha in small red beans and 196,500 in black beans.
___________________________

-1

Medium population targets were 196,500 seeds ha in small red beans and 262,000 in black
-1

beans. High population targets were 262,000 seeds ha in small red beans and 327,500 in black
beans.

Table 4.15. Main effect of row width on number of pods on the main stem in black and small
a
red beans at East Lansing and Richville in 2010 and 2011 .
East Lansing
2010
Row
width

Black

2011
Small
red

Black

____________________________

Richville
2011

2010
Small
red

Combined

Main-stem pods plant

-1

Black

Small
red

____________________________

38 cm
51 cm
76 cm

7.6 a
7.5 a

5.7 a
5.8 a

6.7 a
6.9 a

5.8 a
5.4 a

6.2 b
6.6 a
5.8 c

8.6 a
8.5 a
9.0 a

5.3 b
6.3 a
6.6 a

Mean

7.55

5.75

6.80

5.60

6.27

8.7

6.07

a

Means within each column that the same letter are statistically similar to one another at the α ≤
0.05 level.

197

Table 4.16. Main effect of planting population on number of pods on the main stem in black and
a
small red beans at East Lansing and Richville in 2010 and 2011 .
2010
Planting
b
population

East Lansing
2011

2010

Combined

Black

Combined

_______________________

Low
Medium
High

7.3 a
6.8 a
5.9 b

8.0 a
6.4 b
6.0 b

Richville

Small red

Main-stem pods plant
5.4 a
6.1 a
5.4 a

-1

6.8 a
6.4 b
5.8 c

2011
Black

Small red

______________________

9.6 a
8.4 b
8.2 b

6.4 a
5.9 a
5.9 a

a

Means within each column that the same letter are statistically similar to one another at the α ≤
0.05 level.
b
-1
Low population targets were 148,000 seeds ha in small red beans and 196,500 in black beans.
-1

Medium population targets were 196,500 seeds ha in small red beans and 262,000 in black
-1

beans. High population targets were 262,000 seeds ha in small red beans and 327,500 in black
beans.

Table 4.17. Main effect of row width on number of pods on the branches in black and small red
a
beans at East Lansing in 2010 and 2011 and in black beans at Richville in 2011 .
East Lansing
Richville
2010
2011
2011
Row width
Combined
Combined
Black
____________________
____________________
Branch pods per plant
38 cm
51 cm
76 cm

7.7 a
5.8 b

10.7 a
6.2 b

a

10.6 ab
9.2 b
11.5 a

Means within each column that the same letter are statistically similar to one another at the α ≤
0.05 level.

198

Table 4.18. Interaction of row width and planting population with regard to number of pods on
a
the branches in black and small red beans at Richville in 2010 .
Planting population

b

38-cm rows

51-cm rows

______________

Low
Medium
High

7.5 a
4.7 c
2.5 e

76-cm rows

Branch pods per plant ______________
6.8 ab
6.3 b
5.0 c
4.4 c
4.0 cd
3.0 de

a

Means throughout the table that have the same lower-case letter are statistically similar to one
another at the α ≤ 0.05 level.
b
-1
Low population targets were 148,000 seeds ha in small red beans and 196,500 in black beans.
-1

Medium population targets were 196,500 seeds ha in small red beans and 262,000 in black
-1

beans. High population targets were 262,000 seeds ha in small red beans and 327,500 in black
beans.

Table 4.19. Interaction of row width and planting population with regard to number of pods on
a
the branches in small red bean at Richville in 2011 .
Planting population
38-cm rows
51-cm rows
76-cm rows
-1
(seeds ha )
____________________
Branch pods per plant ____________________
148,000
8.4 a
7.9 ab
8.8 a
196,500
8.0 a
6.9 abc
5.4 bc
262,000
4.8 c
5.3 bc
4.5 c
a

Means throughout the table that have the same lower-case letter are statistically similar to one
another at the α ≤ 0.05 level.

199

Table 4.20 . Main effect of row width on the ratio of branch pods to main-stem pods in black
a
and small red beans at East Lansing and Richville .
East Lansing
Richville
2010
2011
2010
2011
Row
Small
Small
Small
Small
b
Black
Black
Black
Black
width
Red
Red
Red
Red
___________________
___________________
Ratio of branch pods to main stem pods
38 cm
0.82 a
1.39 a
1.75 a
1.44 a
0.67 a
0.63 a
1.22 a
1.25 a
51 cm
0.70 a
0.77 a
1.02 b
0.98 b
76 cm
0.62 b
0.98 b
0.89 b
0.90 b
0.72 a
0.64 a
1.14 ab
0.84 b
Mean
0.72
1.18
1.32
1.17
0.70
0.68
1.13
1.02
a

Means within each column that the same letter are statistically similar to one another at the α ≤
0.05 level unless otherwise indicated.
b
Significant at α ≤ 0.1

Table 4.21. Main effect of row width on the ratio of number of seeds in branch pods to number
of seeds in main-stem pods in black and small red beans at East Lansing and Richville in 2010
a
and 2011 .
East Lansing
Richville
Row width
2010
2011
2010
2011
Ratio of seed number per pod on branch pods to seed number per pod on the
_______________________________
main stem _______________________________
38-cm
0.97 a
1.02 a
0.89 b
1.02 a
51-cm
0.93 a
1.03 a
76-cm
0.98 a
1.01 a
0.93 a
1.04 a
a
Means within each column that the same letter are statistically similar to one another at the α ≤
0.05 level.

200

Table 4.22. Main effect of planting population on the ratio of number of seeds in branch pods to
number of seeds in main-stem pods in black and small red beans at East Lansing and Richville in
a
2010 and 2011 .
East Lansing
Richville
2010
2011
2010
2011
Planting
Small
Small
Small
Small
Black
Black
Black
b Black
Red
Red
Red
Red
population
Ratio of seed number per pod on branch pods to seed number per pod on the
_______________________________
main stem _______________________________
Low
0.99 a 0.99 a
1.05 a
1.00 a
0.93 a
0.92 a
1.00 a
1.06 a
Medium
1.03 a 1.01 a
1.00 a
1.00 a
0.94 a
0.92 a
1.02 a
1.05 a
High
0.99 a 0.93 b
1.03 a
1.01 a
0.93 a
0.90 a
1.00 a
1.04 a
Mean
1.00
0.98
1.03
1.00
0.93
0.91
1.01
1.05
a

Means within each column that the same letter are statistically similar to one another at the α ≤
0.05 level.
b
-1
Low population targets were 148,000 seeds ha in small red beans and 196,500 in black beans.
-1

Medium population targets were 196,500 seeds ha in small red beans and 262,000 in black
-1

beans. High population targets were 262,000 seeds ha in small red beans and 327,500 in black
beans.

201

LITERATURE CITED

202

LITERATURE CITED

Adams, M.W. 1967. Basis of yield component compensation in crop plants with special
reference to the field bean, Phaseolus vulgaris. Crop Sci. 7:505-510.
Bennett, J.P., M.W. Adams, and C. Burga. 1977. Pod yield component variation and
intercorrelation in Phaseolus vulgaris L. as affected by planting density. Crop Sci. 17:73-75.
Crothers, S.E. and D.T. Westerman. 1976. Plant population effects on the seed yield of
Phaseolus vulgaris L. Agron. J. 68:958-960.
Ethredge, W.J., D.A. Ashley, and J.M. Woodruff. 1989. Row spacing and plant population
effects on yield components of soybean. Agron. J. 81:947-951.
Goulden, D.S. 1976. Effect of plant population and row spacing on yield and components of
yield in navy beans (Phaseolus vulgaris). New Zeal. J. Exp. Agr. 4:177-180.
Grafton, K.F., A.A. Schneiter, and B.J. Nagle. 1988. Row spacing, plant population, and
genotype x row spacing interaction effects on yield and yield components of dry bean. Agron. J.
80:631-634.
Horn, F.L., L.O. Braga Schuch, E.P. Silveira, I.F. Antunes, J.C. Vieira, G. Marchioro, D.F.
Medeiros, and J.E. Schwengber. 2000. Spacing and plant population for dry bean harvest
mechanization. Pesq. agropec. bras. 35.
Lehman, W.F. and J.W. Lambert. 1960. Effects of spacing of soybean plants between and
within rows on yield and its components. Agron. J. 52:84-86.
Norsworthy, J.K. and E.R. Shipe. 2005. Effect of row spacing and soybean genotype on
mainstem and branch yield. Agron. J. 97:919-923.
Weber, C.R., R.M. Shibles, and D.E. Byth. 1966. Effect of plant population and row spacing on
soybean development and production. Agron. J. 58:99-102.

203