!
!

OPTIMIZING RYE-VETCH COVER CROP MIXTURE MANAGEMENT IN VEGETABLE
CROPPING SYSTEMS: OPPORTUNITIES AND TRADEOFFS
By
Zachary D. Hayden

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Horticulture - Doctor of Philosophy
2014

!
!
ABSTRACT
OPTIMIZING RYE-VETCH COVER CROP MIXTURE MANAGEMENT IN VEGETABLE
CROPPING SYSTEMS: OPPORTUNITIES AND TRADEOFFS
By
Zachary D. Hayden
Cereal-legume cover crop mixtures composed of cereal rye (Secale cereale L.) and hairy
vetch (Vicia villosa Roth) have the potential to combine the unique strengths of the component
species while taking advantage of interspecific synergies. However, management practices and
environmental conditions can influence both species interactions and residue decomposition.
The overarching objective of this research was to evaluate strategies for optimizing rye-vetch
mixture performance within vegetable cropping systems.
The main component of this research evaluated 1) how the relative proportions of rye and
vetch sown in mixtures influenced cover crop biomass production, winter annual weed
suppression, vetch winter survival, and vetch N fixation; and 2) the interactive effects of ryevetch residue characteristics and the use of black polyethylene mulch (PM) on soil N dynamics,
microbial communities, and vegetable yields during production of bell pepper (Capsicum
annuum L.) and slicing cucumber (Cucumis sativus L.). Following a replacement series design,
main plot treatments consisted of a gradient of seven rye-vetch mixture proportions ranging from
100% vetch to 100% rye, in addition to a no-cover-crop control. Following cover crop
termination, subplots consisted of pepper and cucumber grown either with or without PM.
Density and biomass composition in the mixtures were highly correlated with rye and vetch
seeding rates, with little evidence of substantial interspecific interference. Total shoot biomass
in all mixtures was equal to or greater than that of either monoculture, but no differences were
detected in vetch winter survival or the efficiency of N fixation. Changing the proportions of rye

!
!
and vetch in the mixtures resulted in important tradeoffs among some of the services provided by
the cover crop. Increasing vetch in mixtures generally led to greater fixed N accumulation, soil
inorganic N levels, and vegetable yields, but also led to increased seed costs, pore water NO3concentrations below the crop root zone, and reduced winter annual weed suppression. Yields
and soil N levels were generally higher when PM was used, but the differences between PM and
bare ground treatments varied between the two years due in part to differences in the magnitude
and timing of precipitation events. Lower microbial biomass was observed under PM compared
to bare ground 3 wk after plastic application, but cover crop and mulching treatments did not
explain the majority of variation observed in patterns of carbon substrate utilization from BiologEcoplateTM data.
In a separate field experiment, we also investigated how fall planting dates influenced
rye-vetch cover crop biomass quantity and quality in the spring, and evaluated whether
staggering (delaying) rye seeding could improve vetch performance in mixtures. Treatments
consisted of a two-way factorial of three vetch planting dates (late August, mid September, and
late September) and three lengths of rye seeding stagger (co-seeded, short stagger, and long
stagger). Later planting of co-seeded mixtures generally led to reduced total shoot biomass and
lower proportions of vetch biomass, resulting in cover crop residues with less fixed N and a
higher total C:N. For earlier planting dates, delaying rye seeding until vetch emergence (short
stagger) increased vetch shoot biomass by 760 - 1,060 kg ha-1 (30-36 kg vetch N ha-1) relative to
co-seeding. Staggered seeding provided no benefit to vetch biomass at later planting dates, and
delaying rye seeding until the vetch 3-4 leaf stage (long stagger) resulted in significant
reductions in vetch winter survival compared with co-seeding.

!
!
ACKNOWLEDGEMENTS

I would like to express my appreciation to my major advisors, Dr. Mathieu Ngouajio and
Dr. Dan Brainard. I am indebted to you both for the sound guidance you have provided me over
the years, and for giving me the freedom to develop a research program that was truly my own.
Your mentorship has had a profound impact on my growth as a scientist, and I am fortunate to
have had the opportunity to work closely with two such exceptional minds in the field of
sustainable agriculture.
I also want to recognize the contributions of my guidance committee members, Dr. John
Biernbaum and Dr. Alexandra Kravchenko, whose thoughtful direction and input throughout my
graduate program challenged me to think in different ways— greatly improving the quality of
this work and that of future research to come.
I would like to thank Bill Chase and the staff of the MSU Horticulture Teaching and
Research Center, Jon Dahl and the staff of the MSU Plant and Soil Nutrient Laboratory, and all
the past and present colleagues who contributed to this research in the field and otherwise,
including Dr. Ajay Nair, Ben Henshaw, Drey Clark, Damen Kurzer, Erin Haramoto, Carolyn
Lowry, Rebekah Faivor, Aaron Yoder, and Buck Counts. Many thanks also to Dr. Greg Lang
and the Department of Horticulture Graduate Programs Committee for bringing me to MSU in
the first place, and who along with the exceptional Department of Horticulture administrative
staff, supported me throughout my graduate program.
Finally, none of this would have been possible without the incredible support of my
family, and especially that of my wife Ashley. Thank you for your love, your confidence, and

iv

!
!
for giving us our handsome baby boy. Looks like we survived the Michigan winters together—
couldn’t have done it without you.

v

!
!
TABLE OF CONTENTS

LIST OF TABLES .......................................................................................................................ix
LIST OF FIGURES .....................................................................................................................xi
INTRODUCTION .......................................................................................................................1
CHAPTER I
WINTER ANNUAL WEED SUPPRESSION IN RYE-VETCH COVER CROP
MIXTURES .....................................................................................................................5
Abstract....................................................................................................................7
Introduction .............................................................................................................7
Materials and Methods ............................................................................................11
Results and Discussion ............................................................................................14
APPENDIX .............................................................................................................22
REFERENCES ........................................................................................................29
CHAPTER II
RYE-VETCH MIXTURE PROPORTION TRADEOFFS: COVER CROP PRODUCTIVITY,
NITROGEN ACCUMULATION, AND WEED SUPPRESSION .................................34
Abstract....................................................................................................................36
Introduction .............................................................................................................36
Materials and Methods ............................................................................................41
Site Description ...............................................................................................41
Treatment and Experimental Design...............................................................41
Field Management and Data Collection..........................................................42
Corrections for Soil Contamination on Vetch Shoots .....................................43
Land Equivalent Ratios ...................................................................................44
Vetch N Fixation Estimates ............................................................................45
Cost of Nitrogen Fixed ....................................................................................46
Statistical Analysis ..........................................................................................47
Results and Discussion ............................................................................................48
Weather Conditions.........................................................................................48
Cover Crop Density and Vetch Winter Survival ............................................48
Cover Crop Biomass Production and Land Equivalent Ratios .......................49
Winter Annual Weed Biomass and PAR ........................................................52
Cover Crop Nitrogen Content .........................................................................53
Vetch Nitrogen Fixation..................................................................................54
Conclusions .............................................................................................................56
APPENDIX .............................................................................................................58
REFERENCES ........................................................................................................71

vi

!
!
CHAPTER III
RYE-VETCH COVER CROP RESIDUE QUALITY AND PLASTIC MULCH AFFECT SOIL
NITROGEN DYNAMICS, MICROBIAL COMMUNITIES, AND YIELDS IN
VEGETABLE PRODUCTION SYSTEMS ....................................................................77
Abstract....................................................................................................................79
Introduction .............................................................................................................80
Materials and Methods ............................................................................................83
Site Description ...............................................................................................83
Treatment and Experimental Design...............................................................84
Field Management...........................................................................................84
Data Collection and Analysis ..........................................................................86
Cover Crop Residue Quantity and Quality ..............................................86
Vegetable Crop Yields and Fruit Quality ................................................86
Soil Inorganic Nitrogen Dynamics ..........................................................87
Soil Microbial Communities ....................................................................89
Statistical Analysis ..........................................................................................90
Results .....................................................................................................................92
Weather Conditions.........................................................................................92
Rye-vetch Mixture Residues ...........................................................................92
Vegetable Yields and Fruit Quality ................................................................93
Soil Inorganic Nitrogen ...................................................................................94
Lysimeter Pore Water Nitrate Concentrations ................................................95
Microbial Communities ..................................................................................95
Discussion................................................................................................................96
Rye-Vetch Residue Effects on N Dynamics and Vegetable Yields ................96
Plastic Mulch Effects ......................................................................................99
Responses of Soil Microbial Communities .....................................................102
Management Implications ...............................................................................104
APPENDIX .............................................................................................................106
REFERENCES ........................................................................................................119
CHAPTER IV
PLANTING DATE AND STAGGERED SEEDING OF RYE-VETCH MIXTURES:
BIOMASS, NITROGEN, AND LEGUME WINTER SURVIVAL ...............................126
Abstract....................................................................................................................128
Introduction .............................................................................................................128
Materials and Methods ............................................................................................131
Statistical Analysis and Data Presentation ......................................................134
Results and Discussion ............................................................................................135
Weather and GDD Accumulation ...................................................................135
Effect of Planting Date on Co-seeded Mixture Biomass ................................136
Staggered Seeding Effects ..............................................................................138
Vetch Winter Survival ....................................................................................140
Conclusions .............................................................................................................141
APPENDIX .............................................................................................................144
REFERENCES ........................................................................................................151

vii

!
!
CHAPTER V
CONCLUSIONS .........................................................................................................................156
!

viii

!
!
LIST OF TABLES

Table 1.1. Dates of key field operations and data collection at College Rd and Jolly Rd. .........23
Table 1.2. Dominant weed species present at College Rd and Jolly Rd in 2010 and 2011,
including percent composition based on density and biomass calculated from control
treatments. ..................................................................................................................24
Table 1.3. Effect of cover crop treatment on (A) total shoot dry weight, (B) density, and (C)
shoot dry weight per plant of dominant winter annual weeds present in the College Rd
experiment in 2010 and 2011. ....................................................................................25
Table 1.4. Effect of cover crop treatment on density of dominant winter annual weeds present in
the Jolly Rd experiment in 2010. ...............................................................................26
Table 2.1. Rye (R) and vetch (V) seeding rates and estimated seed costs across cover crop
mixture proportions....................................................................................................59
Table 2.2. Dates of key field activities and data collection.........................................................60
Table 2.3. Monthly growing degree day (GDD) and rainfall totals in Holt, MI during cover crop
growth, 2009-2011. ....................................................................................................61
Table 2.4. Vetch apparent overwinter survival (population change) across cover crop mixture
proportions. ................................................................................................................62
Table 2.5. Partial land equivalent ratios for vetch (Lv) and rye (Lr) and total land equivalent
ratios (LER) in rye-vetch cover crop mixtures. .........................................................63
Table 2.6. Vetch shoot N concentration, percentage of N derived from the atmosphere (Ndfa),
estimated shoot N fixed (SNF) and cost of N fixed (CNF) across cover crop mixture
proportions. ................................................................................................................64
Table 3.1. Dates of key field activities and data collection.........................................................107
Table 3.2. Shoot residue dry weight of vetch, rye, and vetch+rye (total) across cover crop
mixture proportions in 2010 and 2011. ......................................................................108
Table 3.3. Bell pepper total number of marketable fruit harvested, average weight of marketable
fruit, and percentage non-marketable yield by weight when grown either on black
plastic mulch or bare ground (no plastic) following 7 rye-vetch cover crop mixture
proportions or a no cover crop control in 2010 and 2011. .........................................109

ix

!
!
Table 3.4. Slicing cucumber total number of marketable fruit harvested, average weight of
marketable fruit, and percentage non-marketable yield by weight when grown either
on black plastic mulch or bare ground (no plastic) following 7 rye-vetch cover crop
mixture proportions or a no cover crop control in 2010 and 2011. ...........................110
Table 3.5. Microbial biomass carbon (MBC), soil respiration, and metabolic quotient in soils
sampled from pepper rows grown either on black plastic mulch or bare ground (no
plastic) following select rye-vetch cover crop mixture proportions or a no cover crop
control. Soils were sampled on 21 and 20 June in 2010 and 2011, respectively. .....111
Table 4.1. Seeding schedule and accumulated growing degree days (GDD) during rye and vetch
growth in cover crop mixture treatments. ..................................................................145
Table 4.2. Percentage vetch shoot N derived from the atmosphere (Ndfa), vetch and rye shoot N
content, and total (vetch+rye) shoot N content and C:N ratio in co-seeded, short
stagger, and long stagger mixtures sown in late August, Mid September, and Late
September in 2009 and 2010......................................................................................146

x

!
!
LIST OF FIGURES

Figure 1.1. Mean (±SE) shoot dry weight of vetch (V), rye-vetch mixture (RV), and rye (R)
winter cover crops in the spring at (A) College Rd in 2010 and 2011, and (B) Jolly Rd
in 2009 and 2010. For RV treatments, the height of the stacked bar represents total
cover crop biomass (rye+vetch), while depicted standard errors correspond to the mean
biomass of each component species in the mixture. Within a given site and year, total
cover crop biomass is not significantly different for treatments labeled with the same
letter (α=0.05). ...............................................................................................................27
Figure 1.2. Mean (±SE) dry weight per plant of transplanted field pennycress (Thlaspi arvense
L.) in control (C), vetch (V), rye-vetch mixture (RV), and rye (R) treatments at Jolly Rd
in 2009 and 2010. Within a given year, means labeled with the same letter are not
significantly different (α=0.05). ....................................................................................28
Figure 2.1. Spring density of vetch (gray), rye (white), and vetch+rye (black) across cover crop
mixture proportions in 2010 and 2011. 2010 responses of vetch: y=1.34x, r2=0.81***;
rye: y=2.44x, r2=0.87***; and vetch+rye: y=1.14x+136.23, r2=0.73***. 2011 responses
of vetch: y=1.56x, r2=0.84***; rye: y=0.006x2+0.46x, r2=0.80***; and vetch+rye:
y=1.12x+182.26, r2=0.47***. Means (+/- SE). Single, double, and triple asterisks
following r2 values denote regression model significance at the 0.05, 0.01, and 0.001
levels, respectively. .......................................................................................................65
Figure 2.2. Spring shoot biomass dry weights of vetch (gray), rye (white), and vetch+rye (black)
across cover crop mixture proportions in 2010 and 2011. 2010 responses of vetch:
y=5.38x, r2=0.71***; rye: y=-0.036x2+7.67x, r2=0.50***; and vetch+rye: y=1.46x+607.32, r2=0.16*. 2011 responses of vetch: y=-0.020x2+5.07x, r2=0.69***; rye:
y=-0.025x2+5.62x, r2=0.76***; and vetch+rye: y=-0.038x2+3.62x+332.52, r2=0.23*.
Means (+/- SE). Single, double, and triple asterisks following r2 values denote
regression model significance at the 0.05, 0.01, and 0.001 levels, respectively.
Asterisks in the figure indicate where total biomass (vetch+rye) in a mixture was
significantly greater than in either monoculture in that year at the 0.05 probability level.
.......................................................................................................................................66
Figure 2.3. Percentage of photosynthetically active radiation (PAR) penetrating the cover crop
stand canopy across mixture proportions. Means are for 2010 and 2011 data combined
(n=8), presented +/- SE. The triple asterisks following the r2 value in the figure denote
regression model significance at the 0.001 level. ..........................................................67
Figure 2.4. Winter annual weed biomass (dry weight) across cover crop mixture proportions in
2010 and 2011, expressed as mean percentage of the no-cover-crop control treatment
(+/- SE). 2010 response: y=0.089x+4.18, r2=0.18*. 2011 response: y=0.35x+2.27,
r2=0.40***. Single, double, and triple asterisks following r2 values denote regression
model significance at the 0.05, 0.01, and 0.001 levels, respectively.............................68
xi

!
!
Figure 2.5. Shoot N accumulated in vetch (gray), rye (white), and vetch+rye (black) across cover
crop mixture proportions in 2010 and 2011. 2010 responses of vetch: y=0.92x,
r2=0.66***; rye: y=-0.00048x2+0.086x, r2=0.36***; and vetch+rye: y=-0.13x+17.83,
r2=0.67***. 2011 responses of vetch: y=-0.00066x2+0.17x, r2=0.66***; rye: y=0.00044x2+0.084x, r2=0.61***; and vetch+rye: y=-0.00097x2+0.032x+10.69,
r2=0.58***. Means (+/- SE). Single, double, and triple asterisks following r2 values
denote regression model significance at the 0.05, 0.01, and 0.001 levels, respectively.
.......................................................................................................................................69
Figure 2.6. Rye shoot N concentration across rye-vetch cover crop mixture proportions in 2010
and 2011. The slope of the response to rye-vetch proportion was equal in both years
and significant overall (analysis of covariance, P<0.001). 2010 equation:
y=0.0018x+0.96, r2=0.15. 2011 equation: y=0.0018x+1.31, r2=0.13. Means (+/- SE).
.......................................................................................................................................70
Figure 3.1. Daily rainfall amounts (bars) and minimum, maximum, and average daily air
temperatures (lines) at the Michigan State University Horticulture Teaching and
Research Center (Holt, Michigan) from April through September in 2010 and 2011.
The first and second dotted lines denote the dates of cover crop termination and plastic
mulch application, respectively, in each year. ...............................................................112
Figure 3.2. Total (vetch+rye) shoot residue N content (a) and C:N ratio (b) across cover crop
mixture proportions in 2010 (open circles) and 2011 (closed circles). The horizontal
dotted line references a C:N value of 25, a common estimate for the cutoff between
expectations of net N mineralization and immobilization in soils following residue
incorporation. Means +/- SE. The responses of total N content in 2010: y=-1.3x+178.3,
r2=0.67, P<0.001; and 2011: y=-0.0097x2+0.32x+106.9, r2=0.58, P<0.001. The
responses of C:N in 2010: 0.0038x2-0.13x+13.73, r2=0.90, P<0.001; and 2011:
0.0019x2-0.01x+12.77, r2=0.85, P<0.001. ....................................................................113
Figure 3.3. Total marketable yields of bell pepper (a, c) and slicing cucumber (b, d) when grown
either on black plastic mulch (black squares) or bare ground (white squares) following 7
rye-vetch cover crop mixture proportions or a no cover crop control (0:0) in 2010 (a, b)
and 2011 (c, d). Means +/- SE. ....................................................................................114
Figure 3.4. Correlations between the average soil inorganic N concentration in bell pepper rows
during the summer and a) total bell pepper marketable yield and b) bell pepper SPAD
meter leaf absorbance readings (dimensionless units) for 2010 and 2011 data combined.
Rye-vetch cover crop mixture treatments are designated by grayscale gradient, while
black plastic mulch and bare ground treatments are differentiated by shape. Pearson
correlation coefficients (R) and their significance are indicated on the plots. ..............115
Figure 3.5. Total soil inorganic N concentrations during the summer in bell pepper rows grown
either on black plastic mulch (a, c) or bare ground (b, d) in 2010 (a, b) and 2011 (c, d)
following 7 rye-vetch cover crop mixture proportions or a no cover crop control.
xii

!
!
Means +/- SE. For clarity, error bars are only presented for select cover crop treatments
(Control, 100 V: 0 R, 50 V: 50 R, and 0 V: 100 R). .....................................................116
Figure 3.6. NO3--N concentrations of pore water extracted during the summer from suction
lysimeters installed to a depth of 61 cm below bell pepper rows grown either on black
plastic mulch (a, c) or bare ground (b, d) in 2010 (a, b) and 2011 (c, d) following select
rye-vetch cover crop mixture proportions or a no-cover-crop control. Means +/- SE. .117
Figure 3.7. Principal component analysis of soil microbial substrate utilization data from BiologEcoPlatesTM in 2010 (a, b) and 2011 (c, d). Component scores for black plastic mulch
(black) and bare ground (white) management following select rye-vetch cover crop
mixture treatments (shapes) are plotted for the first 2 principal components (a, c), as
well as the correlations (loadings) between the 2 principal components and the original
6 substrate class variables (b, d). The percentages of the total variance accounted for by
each principal component in a given year are indicated in parentheses in axis titles. ..118
Figure 4.1. Vetch (black), rye (white), and total (vetch+rye, gray) shoot dry biomass (A), total
residue N content (B), and total residue carbon:nitrogen ratio (C) in co-seeded mixtures
sown in late August, Mid September, and Late September in 2009 (circles) and 2010
(triangles). Data from the two years are plotted based on total growing degree days
(GDD) base 4 ºC accumulated during vetch growth on a reversed scale, with planting
date treatments indicated above the x-axis. Means +/- SE. Responses to GDD
accumulation (years combined) of vetch shoot biomass: y=0.0008x2-0.95x+298.53, r2=
0.82***; rye shoot biomass: y=-0.0016x2+2.99x-1122.92, r2= 0.26*; total shoot
biomass: y=0.69x-216.93, r2= 0.56***; total N content: y=0.02x-10.32, r2= 0.71***;
and total C:N: y=-0.0001x2+0.10x-21.08, r2= 0.36**. Single, double, and triple
asterisks following r2 values denote regression model significance at the 0.05, 0.01, and
0.001 levels, respectively. .............................................................................................147
Figure 4.2. Vetch (A), rye (B), and total (vetch+rye, C) shoot dry biomass in short rye stagger
(gray) and long rye stagger (white) mixtures relative to co-seeded mixtures (black)
sown in late August, Mid September, and Late September in 2009 (circles) and 2010
(triangles). Data from the two years are plotted based on total growing degree days
(GDD) base 4 ºC accumulated during vetch growth on a reversed scale, with planting
date treatments indicated above the x-axis. Actual GDD accumulation for rye is less
than that for vetch in staggered mixtures (Table 1). Within a given year and planting
date treatment, levels of stagger are not significantly different when means (+/- SE) are
followed by the same letter (α =0.05). Responses to GDD accumulation (years
combined) of vetch co-seeded: y=0.0008x2-0.95x+298.53, r2= 0.82***; vetch short
stagger: y=0.0011x2-1.22x+342.41, r2= 0.89***; vetch long stagger: y=0.79x-531.85,
r2= 0.71***; rye co-seeded: y=-0.0016x2+2.99x-1122.92, r2= 0.26*; rye short stagger:
y=-0.0012x2+2.51x-1048.44, r2= 0.62***; rye long stagger: y=0.38x-232.79, r2=
0.81***; total co-seeded: y=0.69x-216.93, r2= 0.56***; total short stagger: y=1.13x630.00, r2= 0.85***; total long stagger: y=1.17x-764.63, r2= 0.80***. Single, double,
and triple asterisks following r2 values denote regression model significance at the 0.05,

xiii

!
!
0.01, and 0.001 levels, respectively. ..............................................................................148
Figure 4.3. Percentage vetch winter survival in co-seeded (black), short rye stagger (gray), and
long rye stagger (white) mixtures sown in late August, Mid September, and Late
September in 2009 (circles) and 2010 (triangles). Data from the two years are plotted
based on total growing degree days (GDD) base 4 ºC accumulated during vetch growth
on a reversed scale, with planting date treatments indicated above the x-axis. Within a
given year and planting date treatment, levels of stagger are not significantly different
when means (+/- SE) are followed by the same letter (α=0.05). Responses to GDD
accumulation (years combined) of co-seeded: y=-0.0023x+92.10, r2=0.0011; short
stagger: y=-0.00021x2+0.41x-107.20, r2=0.29*; and long stagger: y=-0.00034x2+0.71x277.40, r2=0.77***. Single, double, and triple asterisks following r2 values denote
regression model significance at the 0.05, 0.01, and 0.001 levels, respectively. ..........150
!

xiv

!
!

INTRODUCTION

1

!
!
INTRODUCTION

Cereal-legume cover crop mixtures can combine the unique functional strengths of the
component species while also taking advantage of potential interspecific synergies. Winter
annual mixtures composed of cereal rye (Secale cereale L.) and the legume hairy vetch (Vicia
villosa Roth) have been well-studied for their potential to accumulate significant fixed N while
also providing services more characteristic of rye, including more effective weed suppression,
erosion control, scavenging of residual soil N, greater biomass production, and lower overall
seed costs than a vetch monoculture. In addition, cereal-legume complementarity may contribute
to advantages in rye-vetch mixtures, including greater resource-use efficiency, improved N
availability for rye, and enhanced N fixation, light interception, and winter survival for vetch.
Furthermore, the moderation of total residue quality in mixtures may alter patterns of net N
mineralization following cover crop termination— alleviating challenges with N tie-up
associated with rye residues, while potentially improving the synchrony of N release with crop
demand relative to monoculture vetch. The winter hardiness and grower familiarity of the two
species make them well suited for broad applicability in American agriculture, supporting the
value of additional research to better understand species interactions and improve mixture
management.
The services provided by rye-vetch cover crops and the extent to which possible
facilitative benefits are realized depend on both environmental conditions and management
factors that influence species interactions during cover crop growth, the final composition of
mixture stands, and the decomposition of residues following cover crop termination. Evaluating
cover crop performance requires an understanding of how the cover crops influence multiple

2

!
!
components of the agroecosystem, from weed suppression to N availability, so that management
decisions can take into account a farmer’s principal goals and crop/soil management practices.
The unifying objective of this work was to document the performance of rye-vetch mixtures
relative to rye and vetch monocultures while leveraging our understanding of ecological
principles to develop and evaluate strategies for improving rye-vetch mixture management
within vegetable cropping systems.
Although generally considered less of a management priority than summer annual weeds,
winter annual weeds can serve as alternative hosts for important pests and diseases, and can
interfere directly with summer cash crops in reduced tillage systems. In Chapter I, we report
results from two studies evaluating the effectiveness of rye and vetch cover crops in suppressing
the biomass and density of components of winter annual weed communities, comparing the
performance of a rye-vetch mixture with rye and vetch monocultures.
Recognizing that seeding rates were a logical tool for managing rye-vetch mixture
composition, the next two chapters present results from a large systems-level experiment
evaluating how the proportion of rye and vetch sown in a mixture influences species interactions
and cover crop performance. Chapter II describes the effects of species proportions on the
living cover crops, focusing on stand establishment, biomass productivity, vetch winter survival
and N fixation, and tradeoffs among seed costs and cover crop services like fixed N
accumulation and winter annual weed suppression. Chapter III follows with the impact of
incorporated residues on vegetable production after cover crop termination, evaluating the
interactive effects of rye-vetch residue quality across mixture proportions and the use of black
polyethylene mulch on cash crop yields and quality, soil N dynamics, and microbial
communities during production of both bell pepper (Capsicum annuum L.) and slicing cucumber

3

!
!
(Cucumis sativus L.).
Finally, inclement weather and late harvests of previous cash crops can force delays in
planting of rye and vetch cover crops past recommended dates in some years, and fall
establishment is an important determinant of spring biomass production. Furthermore, cereals
tend to be the more-competitive component in cereal-legume mixtures, and depending on
conditions, interference from rye can suppress vetch productivity in mixtures. Chapter IV
presents results from an additional experiment that investigated how fall planting dates
influenced rye-vetch cover crop biomass quantity and quality in the spring, and tested a unique
approach for improving vetch productivity and N fixation in mixtures by staggering (delaying)
the seeding of rye after vetch in the fall. This research also provided the opportunity to explore
interactions between vetch developmental stage and the presence of rye on vetch overwinter
survival in the Great Lakes region.

4

!
!

CHAPTER I

WINTER ANNUAL WEED SUPPRESSION IN RYE-VETCH COVER CROP MIXTURES

5

!
!
Winter Annual Weed Suppression in Rye-Vetch Cover Crop Mixtures

Zachary D. Hayden*, Daniel C. Brainard, Ben Henshaw, and Mathieu Ngouajio

Department of Horticulture, Michigan State University, Plant and Soil Sciences Building, 1066
Bogue St., East Lansing, MI 48824.
*Corresponding author (haydenza@msu.edu).

This research received financial support from the U.S. Department of Agriculture-National
Institute of Food and Agriculture (USDA-NIFA) Sustainable Agriculture Research and
Education (SARE) program (project no. GNC09-108), from USDA/CSREES Special Research
Grant no. 03162 (Sustainable Agriculture 2008: Developing Sustainable Agriculture and Food
Systems), from MSU AgBioResearch Project GREEEN (project no. GR09-068), and from The
Ceres Trust Organic Research Initiative. Additional support came from the C.S. Mott Group for
Sustainable Food Systems and the Ecological Food and Farming Systems Program at Michigan
State University. Special thanks to Ajay Nair, Drey Clark, Corey Noyes, Damen Kurzer, and
Daryl Lederle for their assistance with field management and data collection.

Manuscript published in WEED TECHNOLOGY, 26:818–825. 2012.

6

!
!
ABSTRACT
Winter annual weeds can interfere directly with crops and serve as alternative hosts for
important pests, particularly in reduced tillage systems. Field experiments were conducted on
loamy sand soils at two sites in Holt, Michigan between 2008 and 2011 to evaluate the relative
effects of cereal rye, hairy vetch, and rye-vetch mixture cover crops on the biomass and density
of winter annual weed communities. All cover crop treatments significantly reduced total weed
biomass compared to a no cover crop control, with suppression ranging from 71 - 91% for vetch,
to 95 - 98% for rye. In all trials, the density of non-mustard family broadleaf weeds was either
not suppressed or suppressed equally by all cover crop treatments. In contrast, the density of
mustard family weed species was suppressed more by rye and rye-vetch mixtures than by vetch.
Cover crops were more consistently suppressive of weed dry weight per plant than of weed
density, with rye-containing cover crops generally more suppressive than vetch. Overall, rye
was most effective at suppressing winter annual weeds; however, rye-vetch mixtures can match
the level of control achieved by rye, in addition to providing a potential source of fixed nitrogen
for subsequent cash crops.

INTRODUCTION
Crop production losses in the United States due to agricultural weeds may be as high as
$33 billion annually, and American farmers spend an estimated $6 billion each year on
herbicides, tillage, and cultivation for weed control (Liebman et al., 2001; Pimentel et al., 2005).
Summer annual weed species are undoubtedly the greatest contributors to these costs, because
they have life cycles that facilitate direct interference with most agronomic and vegetable crops.
As a result, comparatively little research in summer annual cropping systems has focused on

7

!
!
winter annual weeds, which establish prior to warm-season production windows and are
typically controlled by pre-planting tillage or herbicide applications. However, due in part to
increasing adoption of reduced tillage systems (CTIC, 2008), reductions in residual herbicide
usage (Young, 2006), and trends toward milder winter temperatures in North America (Hayhoe
et al., 2006), winter annuals may become a greater management priority in the future (Creech et
al., 2008).
In the absence of an established forage or cover crop, winter annual weeds may provide
valuable services in fields during the off-season, including erosion control and recycling of
residual nutrients (Jordan and Vatovec, 2004). However, the significant challenges posed by
increasingly prevalent winter annual populations likely outweigh any potential benefits. In notill and other reduced tillage systems, winter annual and perennial weeds not controlled in the fall
or spring can interfere directly with early-season or summer cash crops, and their previous
establishment can make them highly competitive (Liebman et al., 2001; Brainard et al., 2012b).
This is particularly problematic in organic reduced tillage systems, where the prohibition of
synthetic herbicides leaves few options for effective pre-planting control. Failure to effectively
control winter annual weeds will also increase soil seedbanks over time, which can lead to costly
infestations in winter annual cash crops (e.g., winter cereals) grown later in a rotation (Mirsky et
al., 2010). Moreover, many “winter annual” weed species are actually facultative winter
annuals, which can germinate in both the fall and spring (Cici and Van Acker, 2009), and can
reduce yields, interfere with harvest, and serve as costly contaminants in summer annual as well
as winter annual crops. For example, mayweed chamomile (Anthemis cotula L.) is a weed in
winter wheat and can also interfere with combining during pea (Pisum sativum L.) harvest in
Washington and Northern Idaho (Ogg et al., 1993). Furthermore, as alternative hosts for

8

!
!
economically important pests and diseases, overwintering weeds can also serve as “biological
bridges” from one growing season to the next (Wisler and Norris, 2005; Norris and Kogan,
2005), fostering plant parasitic nematodes and plant pathogenic viruses, as well as their insect
vectors in some cases (Creech et al., 2007; Duffus 1971; Groves et al., 2001). Since many
common winter annuals belong to the Brassicaceae (mustard family), vegetable crops in this
family may be particularly vulnerable to pests and diseases where winter annuals are not
controlled (Schaad and Dianese, 1981; Chen et al., 2009).
Cover cropping is an important component of integrated approaches to weed
management. Residues from winter cover crops are well-studied for their potential to suppress
summer weed populations, particularly when maintained as a surface mulch in reduced tillage
systems (Carrerra et al., 2004; Teasdale, 1996; Teasdale and Mohler, 1993). However, in
addition to the provision of agroecosystem services like erosion control, organic matter addition,
nutrient recycling, and nitrogen fixation, winter cover crops also have significant potential for
managing winter annual weeds. Displacing winter annual populations with cover crop species
that are unsuitable or less-suitable alternative hosts may reduce overwintering reservoirs of
important pests and diseases, as well as draw down weed seedbanks over time. Cover crops are
more likely to be adopted, however, if they can combine effective weed suppression with other
desirable services.
Winter cover crops composed of mixtures of cereal and legume species, such as cereal
rye and hairy vetch, have been studied for their potential to provide significant fixed nitrogen
with greater weed suppression and lower overall seed costs than monoculture legumes (Brainard
et al., 2012a). Mixtures are often more efficient than monocultures in the capture of light, water,
and nutrients (Liebman and Dyck, 1993), which may contribute to greater biomass productivity

9

!
!
in cover crop mixtures vs. monocultures, and suggests that mixtures may be more competitive
with weeds. In practice, weed suppression in cereal-legume mixtures may be more closely
related to the presence of the competitively dominant cereal species, rather than on the diversity
of the mixture, per se (Liebman and Dyck, 1993). Accordingly, many studies suggest that
cereal-legume mixtures often suppress weeds better than a monoculture of the legume, but less
than or equivalent to a monoculture of the cereal (Akemo et al., 2000; Brainard et al., 2011;
Brennan and Smith, 2005; Mohler and Liebman, 1987; Poggio, 2005).
While cereal rye and hairy vetch are well-researched both as winter cover crop
monocultures and in mixture, few studies have documented the effects of rye-vetch mixtures on
winter annual weeds. In monoculture, the notable ability of rye to suppress weeds, reduce nitrate
leaching, and control erosion is often tempered by the high C:N ratio of its residues and the
threat of subsequent yield losses due to nitrogen immobilization (McCracken et al., 1994;
Shipley et al., 1992; Wagger et al., 1998). Vetch, on the other hand, can fix large amounts of
nitrogen, but generally provides less effective weed suppression than rye (Clark, 2007; Clark et
al., 2007; Mennan et al., 2009). In mixture, rye-vetch stand characteristics vary among studies,
but total N release from mixture residues can approach the amount released from vetch
monocultures, and total dry matter yields of rye-vetch cover crops can be greater than yields of
either species in monoculture (Ranells and Wagger, 1996; Sainju et al., 2005). These qualities
suggest that rye-vetch mixtures could be effective cover crops for controlling winter annual
weeds, in balance with providing other important services.
The objective of this research was to evaluate the relative effects of rye, hairy vetch, and
rye-vetch mixture cover crops on the biomass and density of winter annual weed communities.

10

!
!
MATERIALS AND METHODS
Two separate experiments were conducted at the Michigan State University (MSU)
Horticulture Teaching and Research Center (HTRC) in Holt, MI (42°40” N, 84°28” W) at two
sites within 2 km of each other (hereafter referred to as College Rd and Jolly Rd), both on Spinks
loamy sand soil (sandy, mixed, mesic Lamellic Hapludalf). Initial soil chemical characteristics
at College Rd included pH 6.6; CEC 7.7 cmol kg-1; and P, K, and Mg levels of 117, 146, and 42
mg kg-1, respectively. Initial soil chemical characteristics at Jolly Rd included pH 7.9; CEC 7.5
cmol kg-1; and P, K, and Mg levels of 71, 83, and 266 mg kg-1, respectively. Experiments at both
sites were repeated for two seasons, alternating between adjacent fields, from 2008-2011.
Summer cover crops of sorghum sudangrass (Sorghum bicolor x S. bicolor var. sudanense) were
grown on each field prior to seeding winter cover crops in the fall.
Experiments at both sites investigated the effects of rye and hairy vetch cover crops on
winter annual weed populations using a randomized complete block design with four
replications. Cover crop treatments common to the College Rd and Jolly Rd experiments
included: hairy vetch in monoculture (V), cereal rye in monoculture (R), a rye-vetch mixture
(RV), and a control treatment with no winter cover crop (C). Vetch seeding rates were 42 and 45
kg ha-1 in monoculture and 21 and 22.5 kg ha-1 (50% of monoculture rate) in mixture at College
Rd and Jolly Rd, respectively, while rye was sown at 94 and 125 kg ha-1 in monoculture and 47
and 62.5 kg ha-1 in mixture. “Variety not stated” (VNS) vetch seed from Oregon was used at
both sites, while VNS rye from Minnesota and ‘Wheeler’ rye were used at the College Rd and
Jolly Rd sites, respectively (Albert Lea Seed House, Albert Lea, MN). At College Rd, plot sizes
were 6.7 x 8.5 m in 2009-2010 and 6.1 x 7.6 m in 2010-2011, while at Jolly Rd, plots were 3.8 x
18.3 m in 2008-2009 and 3.0 x 12.2 m in 2009-2010.

11

!
!
The dates of key field operations and data collection are summarized in Table 1.
Previous sorghum sudangrass summer cover crops were flail mowed and incorporated at least 2
wk prior to winter cover crop seeding in all trials. Incorporation was accomplished with a
rototiller at College Rd and with a moldboard plow and rotary spader at Jolly Rd in 2008 and
2009, respectively, followed by secondary tillage with a disk. The College Rd experiment was
managed organically with no additional soil amendments applied. The Jolly Rd experiment was
managed conventionally, and fall fertilizer applications included 224 kg ha-1 19-19-19 (N-P-K)
in 2008 and 45 kg ha-1 urea plus 67 kg ha-1 potash in 2009, in addition to 336 kg ha-1 elemental S
applied in both years to lower soil pH. All amendments were broadcast and incorporated using a
harrow. At College Rd, winter cover crop treatments were sown on September 1 in both 2010
and 2011. Rye and vetch seeds were broadcast by hand using a grid system to ensure uniformity,
and then incorporated to a depth of roughly 5 cm using a field cultivator. Cover crops at Jolly
Rd were drilled on September 2, 2008 using a Moore Unidrill no-till seeder (County Antrim,
Northern Ireland). In 2009, rye was drilled using a John Deere 750 no-till grain drill (Deere and
Company, Moline, IL) on September 4, while vetch was seeded using a Jang push seeder (JP-3,
Chungbuk, South Korea). All vetch seed was inoculated with N-DURE Rhizobium
leguminosarum inoculant prior to seeding (INTX Microbials LLC, Kentland, IN) at a rate of
approximately 10 g inoculant kg-1 seed.
Cover crop densities and aboveground biomass were sampled in the spring from four 25
x 50 cm (0.125 m2) quadrats in each plot at College Rd on May 10, 2010 and May 14, 2011, and
from two 0.5 m2 quadrats per plot at Jolly Rd on May 29, 2008 and May 27, 2009. At College
Rd, biomass and densities of weed populations were sampled from those same quadrats at the
time of cover crop sampling, and weeds sampled from C, V, RV, and R treatments were

12

!
!
subsequently sorted by species in the laboratory. At Jolly Rd, weed populations were not
sampled in 2009, and weed biomass was not collected in 2010. However, densities of weed
populations were sampled on April 2, 2010 using two 0.5 m2 quadrats in each plot. In addition,
individuals of field pennycress were transplanted into Jolly Rd cover crop plots in order to
evaluate the effects of cover crop treatments on the growth of this problematic mustard family
weed while also controlling for variability in field population densities and possible confounding
effects of cover crops on field pennycress emergence. A total of 12 (in 2008) and 5 (in 2009)
field pennycress individuals that emerged following pre-planting tillage (growing outside but
adjacent to the experimental area) were transplanted into two locations in each plot on September
29, 2008 and October 16, 2009. Field pennycress rosettes were 2.5 to 5 cm in diameter at the
time of transplanting, and were transplanted at a spacing of 30 cm within rows. Following
overwintering and spring growth periods, the number of surviving transplants was recorded and
their biomass sampled on May 13, 2009 and May 7, 2010. Average survival of transplants was
82 and 99 percent in 2009 and 2010, respectively, and cover crop treatment did not significantly
affect transplant survival in either year (data not shown). For both trials, all cover crop and weed
biomass samples were dried to constant weight at 38 °C.
Weeds were identified to species, and the data were then grouped into the following
categories for analysis: mustard family weeds, other broadleaves, grasses, and total weeds. The
fixed effect of rye-vetch cover crop treatment on all variables was evaluated using the Proc
MIXED procedure in SAS (Version 9.2, SAS Institute, Cary, NC) with block (replicate) included
as a random effect in the model. Data were natural-log or square-root transformed as necessary
to meet ANOVA assumptions of normality and homogeneity of variances. Data were analyzed
separately by trial location and year due to differences in management and the composition of

13

!
!
weed communities among the experimental fields. Where the global F test was significant (p <
0.05), treatment means were separated using Fisher’s protected LSD.

RESULTS AND DISCUSSION
Figure 1 presents aboveground biomass produced by rye and vetch in cover crop
monoculture and mixture treatments in the College Rd and Jolly Rd experiments. Average vetch
biomass in monoculture varied between 356 and 563 g m-2 across sites and years, while rye
biomass ranged from 330 to 587 g m-2, both typical ranges for these cover crop species in the
Great Lakes region (Clark, 2007; Sarrantonio, 1994). Seeding dates were similar across trials
and years, with cover crops at Jolly Rd seeded 3 to 4 days later in September than at College Rd
(Table 1). However, cover crops were sampled on average 2 weeks later in May at Jolly Rd than
at College Rd (Table 1), resulting in a longer period of cover crop growth for the Jolly Rd
experiment. Total rye-vetch mixture biomass tended to be dominated by rye at Jolly Rd (60 and
69 percent rye by dry weight in 2009 and 2010, respectively), while the proportions of rye to
vetch biomass in mixtures at College Rd were closer to 1:1 (46 and 55 percent rye in 2010 and
2011, respectively). Higher rye seeding rates, higher soil N fertility as a result of fall
fertilization, and slightly later fall planting dates at Jolly Rd could all favor rye over vetch in
mixture (Clark et al., 2007; Jannink et al., 1997; Jensen, 1996; Shipley et al., 1992).
Total aboveground biomass production among the cover crop treatments did not differ
significantly at College Rd in 2010 or 2011 (Figure 1A). In contrast, total biomass production in
the rye-vetch mixture was significantly greater than either monoculture at Jolly Rd in 2009, and
rye and rye-vetch mixture treatments produced similar biomass in 2010, both significantly
greater than the vetch monoculture (Figure 1B). The large amounts of total biomass produced by

14

!
!
the rye-vetch mixtures relative to the monoculture treatments at Jolly Rd were driven largely by
the rye component; although seeded at half the monoculture rate, rye biomass in mixture was 94
and 74 percent of that produced in monoculture in 2009 and 2010, respectively.
Weed community composition in experimental fields varied with location and year.
Table 2 lists dominant weed species at College Rd and Jolly Rd in 2010 and 2011. Collectively,
mustard family species made up 52 to 55 percent of the total weed density in control treatments
at College Rd in 2010 and 2011, respectively, but accounted for only 22 to 29 percent of the total
weed biomass, reflecting their relatively small size in the spring compared to other winter annual
broadleaf and grass species present in the fields. Common chickweed was an abundant broadleaf
weed in all years at both sites. Other broadleaf species present at College Rd included henbit in
2010 and corn chamomile (Anthemis arvensis L.) in 2011, while at Jolly Rd, broadleaf weeds
other than chickweed were not identified to the species level. No grass weeds were abundant at
College Rd in 2011, and annual bluegrass (Poa annua L.) was the only grass species abundant in
2010; however, annual bluegrass distribution across the field tended to be more variable than
other weed species at College Rd, likely reducing the power of statistical inference regarding
cover crop effects on annual bluegrass. At Jolly Rd, grasses comprised nearly 40 percent of the
total weed community by density, but data was not collected to the species level. Total weed
pressure in experimental fields, based on control treatments, was lowest at Jolly Rd in 2010 (139
plants m-2), followed by College Rd in 2010 (431 plants m-2), and highest at College Rd in 2011
(1120 plants m-2).
In both years at College Rd, all winter cover crop treatments reduced weed biomass (dry
weight, g m-2) compared to the no cover crop control, although the reduction was not significant
for annual bluegrass (Table 3A). In 2010, suppression of total weed biomass ranged from 91

15

!
!
percent in vetch monoculture to 95 and 98 percent in the rye-vetch mixture and rye monoculture,
respectively. Rye provided significantly greater suppression of weed biomass than vetch
monoculture for mustard species, but not for other broadleaf species (mostly common
chickweed). In 2011, total weed biomass suppression was 71 percent in vetch monoculture,
compared to 94 and 95 percent in the rye-vetch mixture and rye monoculture, respectively. Rye
suppressed both mustard and other broadleaf species significantly more than vetch. In both
years, the rye-vetch mixture provided suppression of weed biomass equivalent to that of the rye
monoculture.
Our results are generally consistent with previous studies showing greater weed
suppressive ability of cereal species compared to legumes (Brainard et al., 2011; Ofori and Stern,
1987), which may be a result of more effective resource competition or allelopathy (Barnes and
Putnam, 1986). Suppression by rye-containing cover crops was also less variable between 2010
and 2011 compared to suppression by vetch in monoculture, suggesting that the suppressive
ability of rye may be more robust in the face of year-to-year variability in environmental
conditions than vetch. We speculate that cooler spring temperatures in 2011 (data not shown)
may have limited vetch growth relative to both weeds and rye that year, resulting in less effective
weed suppression by vetch, and perhaps a greater proportion of rye in the rye-vetch mixture
(Figure 1A). The relative proportion of the component species in cereal-legume mixtures is
likely a key determinant of mixture performance. In a study of rye-pea cover crop mixtures,
Akemo et al. (2000) observed that weed biomass decreased with increasing proportion of rye in
the mixture, despite decreasing total cover crop biomass. Therefore, the weed suppressiveness of
cereal-legume mixtures is likely closely related to the relative proportion of the cereal species.

16

!
!
Although total weed biomass in the field is a relevant measure of actual weed pressure, it
is also a function of both the density of weeds present and their average individual biomass
production. Separately evaluating effects on weed density and weed dry weight per plant can
provide greater insight into the potential mechanisms behind suppression by cover crops. For
example, Kumar et al. (2008) found that both the inhibitory effects of cover crops on weeds
(including corn chamomile and shepherd’s-purse) and the mechanisms responsible for these
effects (nitrogen and fungal pathogens) differed for emergence and growth life stages.
At College Rd, all cover crops reduced the density of weeds (with the exception of annual
bluegrass) compared to the control in 2010 (Table 3B). Rye was significantly more suppressive
of mustard weed density than both the vetch monoculture and the rye-vetch mixture, but the
density of other broadleaf species was reduced equally by all cover crop treatments. In 2011,
mustard weed density was suppressed by all cover crop treatments, with the rye-vetch mixture
and rye monoculture providing equivalent control, both greater than the vetch monoculture. The
density of other broadleaf species, however, was not significantly affected by any of the cover
crops.
Comparable results were obtained at Jolly Rd in 2010, where the rye-vetch mixture and
rye monoculture both significantly reduced mustard and grass species density compared to the
control and vetch monoculture, but did not have an effect on other broadleaf density (Table 4).
In contrast to the results from College Rd, however, the vetch monoculture did not reduce weed
densities compared to the control at Jolly Rd. This difference may in part be attributable to the
early timing of weed sampling at Jolly Rd (April 2, compared to mid-May at College Rd). Vetch
often doesn’t achieve more than 30 percent ground cover in the fall, and produces the majority of
its biomass during the warmest months of the spring (Shipley et al., 1992; Teasdale et al., 2004).

17

!
!
Cover crops that can close their canopy earlier in the season are often better at suppressing weeds
(Brennan et al., 2011), and vetch biomass at the time weeds were sampled was likely close to
only half the amount observed in May. Rye, on the other hand, generally provides more
extensive soil cover in the fall (Boyd et al., 2009), in addition to its capacity to deplete soil
moisture and nutrients (particularly N) early in the season (Brainard et al., 2012c; Shipley et al.,
1992). The potential for rye to effectively interfere with weeds earlier than vetch may explain
why the rye monoculture and rye-vetch mixture (69 percent rye biomass) at Jolly Rd still
exhibited suppression of mustard and grass weed density at the earlier sampling date.
Furthermore, the suppression of grass species by rye-containing cover crops at Jolly Rd suggests
that the lack of significant cover crop effects for annual bluegrass at College Rd is not reflective
of grasses in general.
Interestingly, in all trials, the density of non-mustard broadleaf species (common
chickweed, corn chamomile, and henbit) was suppressed equally by all cover crop treatments,
while that of mustards was suppressed more by rye than by vetch (Tables 3B and 4). In a broad
sense, cover crops reduce weed density through a combination of inhibiting seed germination or
emergence, and promoting post-emergence mortality through interference or environmental
effects. Although we can only speculate, the greater suppression of mustards by rye compared to
vetch may have been due in part to mustard sensitivity to nitrogen depletion or shade. Many
members of the mustard family are highly sensitive to soil N levels in both their germination
(Kumar et al., 2008) and growth (Blackshaw et al., 2003) responses, so reductions in soil N due
to rye may have weakened these species and contributed to their mortality. Conversely, several
of the non-mustard species present in our trials, including common chickweed and corn
chamomile, are known to be relatively shade tolerant (Turkington et al., 1980) and less sensitive

18

!
!
to nitrogen depletion (Blackshaw et al., 2003; Kumar et al., 2008), and thus perhaps better able
to withstand competition from rye and vetch cover crops than mustards. In a study comparing
the effects of twelve weed species on kale and sugar beets under different levels of nitrogen,
Welbank (1963) found that the competitive effect of chickweed—one of the primary nonmustard broadleaf species in our trials—was relatively unaffected by nitrogen level. The greater
inhibitory effects of rye relative to vetch are likely to be most pronounced for weed species that
germinate and grow primarily in the fall when rye growth typically exceeds that of hairy vetch.
Conversely, for facultative winter annuals that have significant flushes of emergence in the
spring—potentially including common chickweed, henbit and corn chamomile, but also
shepherd’s purse and field pennycress (Cici and Van Acker, 2009; Kay, 1971)—fewer
differences in density suppression due to cover crops might be expected since growth of rye and
vetch is more similar at that time.
Cover crop treatments, particularly those containing rye, were more consistently
suppressive of weed biomass production (as measured by dry weight per plant) than of weed
density. All cover crop treatments significantly reduced the dry weight per plant of other
broadleaf species compared to the control in trials at College Rd, with rye monoculture
consistently providing the highest level of suppression (Table 3C). The rye-vetch mixture
equaled rye in 2011, but not in 2010. Mustard dry weight per plant was not significantly lower
in vetch monoculture than in the control in 2011, but otherwise, all cover crop treatments
suppressed the growth of mustard species as well. Rye monoculture and the rye-vetch mixture
provided equivalent suppression of mustards in both years at College Rd, a trend that was also
present for annual bluegrass in 2010, though not significant. Transplanted field pennycress
responded similarly to cover crop treatments at Jolly Rd in 2009 and 2010 (Figure 2). Both the

19

!
!
rye monoculture and the rye-vetch mixture provided over 95 percent suppression of field
pennycress dry weight per plant in each year, compared to less than 70 percent suppression by
vetch monoculture. Cereals are generally better competitors than legumes (Ofori and Stern,
1987), and as our results support, this quality tends to be manifest more in the suppression of
weed growth than weed density (Boyd et al., 2009).
This research supports that winter cover crops composed of rye and/or vetch can
significantly suppress winter annual weeds. Reductions in the biomass of winter annuals present
in fields may decrease the size of overwintering populations of important pests and diseases,
which could in turn lower disease pressure during the following season. However, hairy vetch is
also known to be a reproductive host for several species of plant parasitic nematodes, so growing
vetch before susceptible crops should be avoided (Aarssen et al., 1986; Rich et al., 2009; Timper
et al., 2006). Despite suppressing weed biomass production across taxonomic groups, the cover
crops failed to consistently reduce the density of non-mustard broadleaf weeds, including
common chickweed, henbit, and corn chamomile. Particularly in organic reduced tillage
systems, dense populations of these low-growing species could become a problem in subsequent
crops if they are able to survive mechanical kill of the cover crops. The persistence of these
weeds highlights the importance of integrating winter cover crops with other weed management
strategies to provide more complete control and avoid increases in problematic weeds over time
(Liebman and Gallandt, 1997).
Overall, cereal rye was the most effective weed suppressor. Therefore, where winter
annual weed control is a primary objective, rye would likely be the most effective and
inexpensive cover crop option. However, our results demonstrate that rye-vetch mixtures can
match the level of suppression achieved by rye monoculture, in addition to providing a potential

20

!
!
source of fixed nitrogen—a benefit particularly relevant for organic production. The weed
suppressiveness of the mixtures is likely tied to relative species composition, and mixtures
containing less than 50 percent rye biomass may sacrifice winter annual weed control.
Additional research is needed to relate rye-vetch seeding rates to resulting stand characteristics,
and to investigate how species proportions in rye-vetch mixtures influence the provision of other
agroecosystem services.

21

!
!

APPENDIX

22

!
!
Table 1.1. Dates of key field operations and data collection at College Rd and Jolly Rd.
College Rd
Jolly Rd
Activity
2009-10
2010-11
2008-09
2009-10
Cover crops seeded
1 Sept
1 Sept
4 Sept
5 Sept
THLAR† transplanted
29 Sept
16 Oct
Ambient weeds sampled
10 May
14 May
2 Apr
THLAR biomass sampled
13 May
7 May
Cover crop biomass sampled
10 May
14 May
29 May
27 May
† Field pennycress, Thlaspi arvense L.

23

!
!
Table 1.2. Dominant weed species present at College Rd and Jolly Rd in 2010 and 2011, including percent composition
based on density and biomass calculated from control treatments.

Species
Mustard family species
Field pepperweed - Lepidium campestre (L.) R. Br.
Mouse-ear cress - Arabidopsis thaliana (L.) Heynh.
Shepherd's purse - Capsella bursa-pastoris (L.) Medik.
Spring whitlowgrass - Draba verna L.
Field pennycress - Thlaspi arvense L.
Hoary alyssum - Berteroa incana (L.) DC.
Other broadleaf species
Common chickweed - Stellaria media (L.) Vill.
Henbit - Lamium amplexicaule L.
Corn chamomile - Anthemis arvensis L.
Other†
Grass species
Annual bluegrass - Poa annua L.

% Density
College Rd
2010
2011
52

55

12
8
20
11

29
21
3
2

Jolly Rd
2010
%
30

7

% Biomass
College Rd
2010
2011
22

29

6
5
10
2

20
5
4
<1

20
3
30

45

30

61

71

22
7

20

12

51
9

18

25

53
18

19
19

39‡

18
18

† With the exception of Stellaria media, "Other broadleaf species" data were not collected to the species level at Jolly Rd.
‡ Data for "Grass species" were not collected to the species level at Jolly Rd.

24

!
!
Table 1.3. Effect of cover crop treatment on (A) total shoot dry weight, (B) density, and (C) shoot dry weight per plant of dominant
winter annual weeds present in the College Rd experiment in 2010 and 2011.
Cover Crop

Mustards
2010
2011

Other broadleaves
2010
2011

25.9
1.5
0.6
0.1

a‡
b
bc
c

45.5
19.2
2.6
1.9

a
b
c
c

72.3
3.3
3.2
1.4

a
b
b
b

110.8
27.0
6.6
6.7

a
b
c
c

223
38
37
6

a
b
b
c

616
275
138
105

a
b
c
c

128
14
15
17

a
b
b
b

504
226
185
259

a
a
a
a

2011

21.1
5.5
2.6
0.9

a
a
a
a

119.3
10.3
6.3
2.4

a
a
a
a

431
108
123
48

a
a
a
a

279.3
102.0
85.0
46.5

a
b
bc
c

156.3
46.2
9.2
8.6

a
b
c
c

81
56
72
25

a
b
b
b

1120
500
323
364

a
b
b
b

a
b
bc
c

137.5
96.8
26.8
22.2

a
a
b
b

mg plant -1

C. Dry weight per plant
None
Vetch
Rye-Vetch
Rye

2010

plants m -2

B. Density
None
Vetch
Rye-Vetch
Rye

Total

g m -2

A. Dry weight
None
Vetch
Rye-Vetch
Rye

Grasses†
2010
2011

118.1
37.0
14.1
18.1

a
b
c
c

75.8
75.1
16.7
15.4

a
a
b
b

589.5
225.5
217.1
67.9

a
b
b
c

235.1
128.7
34.9
24.7

a
b
c
c

195.8
129.2
71.9
34.0

† Annual bluegrass (Poa annua L.) was the only abundant grass species present in 2010.
‡ For a given measurement, means within a column followed by the same letter are not significantly different (α=0.05).

25

!
!
Table 1.4. Effect of cover crop treatment on density of dominant winter
annual weeds present in the Jolly Rd experiment in 2010†.
Other
Mustards
Grasses
Total
broadleaves
Cover Crop
plants m -2
None
Vetch
Rye-Vetch
Rye

42
42
11
11

a‡
a
b
b

42
47
50
50

a
a
a
a

55
70
5
11

a
a
b
b

139
159
66
72

ab
a
c
bc

† Density data do not include transplanted field pennycress (Figure 1.2).
‡ Means within a column followed by the same letter are not significantly
different (α=0.05).

26

!
!

./001"234"5678/1"98":;$<"

+!!"
*!!"
)!!"

A. College Road
a

(!!"

a
a

'!!"

A

&!!"

A

A

-,"
2011

-"

%!!"
$!!"
#!!"
!"

./001"234"5678/1"98":;$<"

,"

-,"
2010

-"

+!!"

B. Jolly Road

*!!"

b

,"

B

)!!"
(!!"
'!!"

B

a

a

A

&!!"
%!!"
$!!"
#!!"
!"
,"

-,"
2009

-"

,"

-,"
2010

-"

Figure 1.1. Mean (±SE) shoot dry weight of vetch (V), rye-vetch mixture (RV), and rye (R)
winter cover crops in the spring at (A) College Rd in 2010 and 2011, and (B) Jolly Rd in 2009
and 2010. For RV treatments, the height of the stacked bar represents total cover crop biomass
(rye+vetch), while depicted standard errors correspond to the mean biomass of each component
species in the mixture. Within a given site and year, total cover crop biomass is not significantly
different for treatments labeled with the same letter (α=0.05).

27

+,-"./0123"4/,"45673"891"45673:$;"

!
!
'!!!"
&#!!"

a

&!!!"
%#!!"
%!!!"

A

$#!!"

b

$!!!"

B

#!!"
!"
("

)"

c C

c C

*)"

*"

Figure 1.2. Mean (±SE) dry weight per plant of transplanted field pennycress (Thlaspi arvense
L.) in control (C), vetch (V), rye-vetch mixture (RV), and rye (R) treatments at Jolly Rd in 2009
and 2010. Within a given year, means labeled with the same letter are not significantly different
(α=0.05).

28

!
!

REFERENCES

29

!
!
REFERENCES

Aarssen, L., V. Ivan, and K. Jensen. 1986. The biology of Canadian weeds: 76. Vicia
angustifolia L., V. cracca L., V. sativa L., V. tetrasperma (L.) Schreb. and V. villosa
Roth. Can. J. Plant. Sci. 66:711–737.
Akemo, M., E. Regnier, and M. Bennett. 2000. Weed suppression in spring-sown rye (Secale
cereale)-pea (Pisum sativum) cover crop mixes. Weed Technol. 14:545–549.
Barnes, J. P. and A. R. Putnam. 1986. Evidence for allelopathy by residues and aqueous extracts
of rye (Secale cereal). Weed Sci. 34:384-390.
Blackshaw, R. E., R. N. Brandt, H. H. Janzen, T. Entz, C. A. Grant and D. A. Derksen. 2003.
Differential response of weed species to added nitrogen. Weed Sci. 51:532-539.
Boyd, N. S., E. B. Brennan, R. F. Smith, and R. Yokota. 2009. Effect of seeding rate and
planting arrangement on rye cover crop and weed growth. Agron. J. 101:47–51.
Brainard, D. C., R. R. Bellinder, and V. Kumar. 2011. Grass–legume mixtures and soil fertility
affect cover crop performance and weed seed production. Weed Technol. 25:473–479.
Brainard, D. C., B. Henshaw, and S. Snapp. 2012a. Hairy vetch varieties and bi-cultures
influence cover crop services in strip-tilled sweet corn. Agron. J. 104:629-638.
Brainard, D. C., E. Peachey, E. Haramoto, J. Luna and A. Rangarajan. 2012b. Weed ecology and
management under strip-tillage: Implications for Northern U.S. vegetable cropping
systems. Weed Technol. (under review)
Brainard, D. C., J. Bakker, N. Myers and D. C. Noyes. 2012c. Rye living-mulch effects on soil
moisture and weeds in asparagus. HortScience 47: 58-63.
Brennan, E. B., N. S. Boyd, R. F. Smith, and P. Foster. 2011. Comparison of rye and legume–rye
cover crop mixtures for vegetable production in California. Agron J. 103:449–463.
Brennan, E. B. and R. F. Smith. 2005. Winter cover crop growth and weed suppression on the
central coast of California. Weed Technol. 19:1017–1024.
Carrera, L. M., A. A. Abdul-Baki, and J. R. Teasdale. 2004. Cover crop management and weed
suppression in no-tillage sweet corn production. HortScience 39:1262–1266.
Chen, M., A. M. Shelton, P. Wang, C. A. Hoepting, W. C. Kain, and D. C. Brainard. 2009.
Occurrence of the new invasive insect Contarinia nasturtii (Diptera: Cecidomyiidae) on
cruciferous weeds. J. Econ. Entomol. 102:115–120.
Cici, S.Z.H. and R. C. Van Acker. 2009. A review of the recruitment biology of winter annual
weeds in Canada. Can. J. Plant Sci. 89:575–589.

30

!
!
Clark, A. J., ed. 2007. Managing Cover Crops Profitably. 3rd ed. Beltsville, MD: Sustainable
Agriculture Network.
Clark, A. J., J. J. Meisinger, A. M. Decker, and F. R. Mulford. 2007. Effects of a grass-selective
herbicide in a vetch-rye cover crop system on nitrogen management. Agron. J. 99:36–42.
[CTIC] Conservation Technology Information Center. 2008. National Crop Residue
Management Survey. Available at http://www.ctic.purdue.edu/CRM/. Accessed March
10, 2012.
Creech, J. E., W. G. Johnson, J. Faghihi, and V. R. Ferris. 2007. Survey of Indiana producers and
crop advisors: A perspective on winter annual weeds and soybean cyst nematode
(Heterodera glycines). Weed Technol. 21:532–536.
Creech, J. E., A. Westphal, V. R. Ferris, J. Faghihi, T. J. Vyn, J. B. Santini, and W. G. Johnson.
2008. Influence of winter annual weed management and crop rotation on soybean cyst
nematode (Heterodera glycines) and winter annual weeds. Weed Sci. 56:103–111.
Duffus, J. E. 1971. Role of weeds in the incidence of virus diseases. Annu. Rev. Phytopathol.
9:319–340.
Groves, R. L., J. F. Walgenbach, J. W. Moyer, and G. G. Kennedy. 2001. Overwintering of
Frankliniella fusca (Thysanoptera: Thripidae) on winter annual weeds infected with
tomato spotted wilt virus and patterns of virus movement between susceptible weed
hosts. Phytopathology 91:891–899.
Hayhoe, K., C. P. Wake, T. G. Huntington, L. Luo, M. D. Schwartz, J. Sheffield, E. Wood, B.
Anderson, J. Bradbury, A. DeGaetano, T. J. Troy, and D. Wolfe. 2006. Past and future
changes in climate and hydrological indicators in the US Northeast. Clim. Dynam.
28:381–407.
Jannink, J. L., L. C. Merrick, M. Liebman, E. A. Dyck, and S. Corson. 1997. Management and
winter hardiness of hairy vetch in Maine. Orono, ME: Maine Agricultural and Forest
Exp. Stat. Tech. Bulletin no. 167. 35 p.
Jensen, E. S. 1996. Grain yield, symbiotic N2 fixation and interspecific competition for inorganic
N in pea-barley intercrops. Plant Soil 182:25–38.
Jordan, N. and C. Vatovec. 2004. Agroecological benefits from weeds. Pages 137-158 in Inderjit,
ed. Weed Biology and Management. Boston: Kluwer Academic Publishers.
Kay, Q.O.N. 1971. Anthemis cotula L. J. Ecol. 59:623–636.
Kumar, V., D. C. Brainard, and R. R. Bellinder. 2008. Suppression of Powell amaranth
(Amaranthus powellii), shepherd’s-purse (Capsella bursa-pastoris), and corn chamomile
(Anthemis arvensis) by buckwheat residues: Role of nitrogen and fungal pathogens.
Weed Sci. 56: 271-280.

31

!
!
Liebman, M. and E. A. Dyck. 1993. Crop rotation and intercropping strategies for weed
management. Ecol. Appl. 3:92–122.
Liebman, M. and E. R. Gallandt. 1997. Many little hammers: Ecological approaches for
management of crop-weed interactions. Pages 291-346 in L. E. Jackson, ed. Ecology in
Agriculture. San Diego, CA: Academic Press.
Liebman, M., C. L. Mohler, and C. P. Staver. 2001. Ecological Management of Agricultural
Weeds. 1st ed. Cambridge, UK: Cambridge University Press. 532 p.
McCracken, D. V., M. S. Smith, J. H. Grove, C. T. Mackown, and R. L. Blevins. 1994. Nitrate
leaching as influenced by cover cropping and nitrogen source. Soil Sci. Soc. Am. J.
58:1476–1483.
Mennan, H., M. Ngouajio, D. Isik, and E. Kaya. 2009. Effects of alternative winter cover
cropping systems on weed suppression in organically grown tomato (Solanum
lycopersicum). Phytoparasitica 37:385–396.
Mirsky, S. B., E. R. Gallandt, D. A. Mortensen, W. S. Curran, and D. I. Shumway. 2010.
Reducing the germinable weed seedbank with soil disturbance and cover crops. Weed
Res. 50:341–352.
Mohler, C. L. and M. Liebman. 1987. Weed productivity and composition in sole crops and
intercrops of barley and field pea. J. Appl. Ecol. 24:685-699.
Norris, R. F. and M. Kogan. 2005. Ecology of interactions between weeds and arthropods. Annu.
Rev. Entomol. 50:479–503.
Ofori, F. and W. R. Stern. 1987. Cereal–legume intercropping systems. Adv. Agron. 41:41-90.
Ogg, A. G., R. H. Stephens, and D. R. Gealy. 1993. Growth analysis of mayweed chamomile
(Anthemis cotula) interference in peas (Pisum sativum). Weed Sci. 41:394-402.
Pimentel, D., R. Zuniga, and D. Morrison. 2005. Update on the environmental and economic
costs associated with alien-invasive species in the United States. Ecol. Econ. 52:273–288.
Poggio, S. L. 2005. Structure of weed communities occurring in monoculture and intercropping
of field pea and barley. Agr. Ecosyst. Environ. 109:48–58.
Ranells, N. N. and M. G. Wagger. 1996. Nitrogen release from grass and legume cover crop
monocultures and bicultures. Agron. J. 88:777–782.
Rich, J. R., J. A. Brito, R. Kaur, and J. A. Ferrell. 2009. Weed species as hosts of Meloidogyne:
A review. Nematropica 39:157–185.
Sainju, U. M., W. F. Whitehead, and B. P. Singh. 2005. Biculture legume-cereal cover crops for
enhanced biomass yield and carbon and nitrogen. Agron. J. 97:1403–1412.

32

!
!
Sarrantonio, M. 1994. Northeast Cover Crop Handbook. Emmaus, PA: Rodale Institute. 118 p.
Schaad, N. W. and J. C. Dianese. 1981. Cruciferous weeds as sources of inoculum of
Xanthomonas campestris in black rot of crucifers. Phytopathology 71:1215-1220.
Shipley, P. R., J. J. Meisinger, and A. M. Decker. 1992. Conserving residual corn fertilizer
nitrogen with winter cover crops. Agron. J. 84:869–876.
Teasdale, J. R. 1996. Contribution of cover crops to weed management in sustainable
agricultural systems. J. Prod. Agric. 9:475–479.
Teasdale, J. R., T. E. Devine, J. A. Mosjidis, R. R. Bellinder, and C. E. Beste. 2004. Growth and
development of hairy vetch cultivars in the Northeastern United States as influenced by
planting and harvesting date. Agron. J. 96:1266–1271.
Teasdale, J. R. and C. L. Mohler. 1993. Light transmittance, soil temperature, and soil moisture
under residue of hairy vetch and rye. Agron. J. 85:673–673.
Timper, P., R. F. Davis, and P. G. Tillman. 2006. Reproduction of Meloidogyne incognita on
winter cover crops used in cotton production. J. Nematol. 38:83–89.
Turkington, R., C. K. Norman, and G. D. Franko. 1980. The biology of Canadian weeds: 42.
Stellaria media (L.) Vill. Can. J. Plant. Sci. 60:981–992.
Wagger, M. G., M. L. Cabrera, and N. N. Ranells. 1998. Nitrogen and carbon cycling in relation
to cover crop residue quality. J. Soil Water Conserv. 53:214-218.
Welbank, P. J. 1963. A comparison of competitive effects of some common weed species. Ann.
Appl. Biol. 51: 107-125.
Wisler, G. C. and R. F. Norris. 2005. Interactions between weeds and cultivated plants as related
to management of plant pathogens. Weed Sci. 53:914–917.
Young, B. G. 2006. Changes in herbicide use patterns and production practices resulting from
glyphosate-resistant crops. Weed Technol. 20:301–307.

33

!
!

CHAPTER II

RYE-VETCH MIXTURE PROPORTION TRADEOFFS: COVER CROP PRODUCTIVITY,
NITROGEN ACCUMULATION, AND WEED SUPPRESSION

34

!
!
Rye-vetch mixture proportion tradeoffs: Cover crop productivity, nitrogen accumulation,
and weed suppression

Zachary D. Hayden*, Mathieu Ngouajio, and Daniel C. Brainard

Department of Horticulture, Michigan State University, Plant and Soil Sciences Building, 1066
Bogue St., East Lansing, MI 48824.
*Corresponding author (haydenza@msu.edu).

This research was supported by funding from the Agriculture and Food Research Initiative
(AFRI) Competitive Grant No. 2012-67011-19716 from the U.S. Department of Agriculture
National Institute of Food and Agriculture (USDA-NIFA), from the USDA-NIFA Sustainable
Agriculture Research and Education (SARE) program (project no. GNC09-108), from MSU
AgBioResearch Project GREEEN (project no. GR09-068), and from The Ceres Trust Organic
Research Initiative. Additional support came from the C. S. Mott Group for Sustainable Food
Systems at Michigan State University. The authors would like to thank Drey Clark, Dr. Ajay
Nair, and Damen Kurzer for assistance in the field, Aaron Yoder for comments on an early
version of this manuscript, and Dr. Alexandra Kravchenko for guidance on statistical analysis.

Manuscript published in AGRONOMY JOURNAL, 106:904-914. 2014.

35

!
!
ABSTRACT
Cereal-legume cover crop mixtures have the potential to combine the unique strengths of
the component species while taking advantage of interspecific synergies. However, the relative
proportion of each species in mixture is likely to influence species interactions and entail
important tradeoffs in cover crop performance. The objective of this study was to evaluate how
the relative proportions of cereal rye (Secale cereale L.) and hairy vetch (Vicia villosa Roth)
sown in mixtures influenced cover crop biomass production, winter annual weed suppression,
vetch winter survival, and vetch N-fixation as measured by the 15N natural abundance method.
Following a replacement series design, treatments consisted of a gradient of seven rye-vetch
mixture proportions ranging from 100% vetch to 100% rye. Density and biomass composition in
the mixtures were highly correlated with rye and vetch seeding rates, with little evidence of
substantial interspecific interference. Total shoot biomass in all mixtures was equal to or greater
than that of either monoculture, but no differences were detected in vetch winter survival or the
efficiency of N fixation. Changing the proportions of rye and vetch in the mixtures resulted in
tradeoffs related to nitrogen and weed management goals. Increasing vetch in mixtures led to
greater fixed N accumulation, but also increased seed costs and reduced winter annual weed
suppression. A greater understanding of how rye-vetch mixture proportions influence cover crop
performance can support more-informed decision-making regarding cover crop selection and
mixture seeding rates.

INTRODUCTION
The grass-legume association has long been of interest to researchers working in both
natural and managed ecosystems. Mixtures of plants with complementary functional traits can

36

!
!
benefit from more efficient capture of light, water, and nutrients compared with monocultures,
and when interspecific interference is low, grass-legume mixtures can also benefit from potential
facilitative interactions (Vandermeer, 1992). In particular, reductions in soil inorganic N by the
grass can increase legume nodulation and N fixation (Izaurralde et al., 1992; Streeter, 1988),
while the presence of the legume can correspondingly increase N availability to the grass,
through biomass turnover or possible direct N transfer (Eaglesham et al., 1981; Ledgard and
Steele, 1992; van Kesse et al., 1985). Light interception by vining legumes can also be enhanced
by the opportunity to climb into the upright canopy of a companion grass or cereal (Keating and
Carberry, 1993). By reducing frost heaving or buffering temperature extremes at the soil surface
through reduced air movement or increased snow cover retention, the presence of a hardier grass
species may also improve legume winter survival in northern climates (Jannink et al., 1997;
Smith, 1975).
Cereal rye and the legume hairy vetch are widely studied as winter annual cover crop
species, in part because their cold hardiness makes them suitable for production across a broad
geographic range, including the northern United States (Clark, 2007). Rye is a rapid-growing,
N-responsive grass that has demonstrated substantial capacity to provide erosion control,
conservation of residual soil N, and weed suppression—both as a living cover crop and as a thick
surface mulch in reduced or no-tillage production systems (Clark, 2007; Ditsch et al., 1993;
Peachey et al., 2004). These strengths, in addition to inexpensive and widely available seed,
have made rye one of the most commonly grown cover crops in the United States. However,
while rye residues can build soil organic matter, they are not a significant source of available N
for subsequent cash crops, and depending on the stage at which rye is terminated, residue
incorporation can result in net N immobilization (Sainju et al., 2000; Clark et al., 2007a). In

37

!
!
contrast, hairy vetch is a legume known for its ability to accumulate considerable amounts of N
(up to 190 kg ha-1, even in northern temperate climates), much of which is rapidly available to
cash crops in the first season after termination (Ranells and Wagger, 1996; Teasdale, 2004).
However, vetch seed is relatively expensive, and its slower growth and readily decomposable
residues generally make it less effective than rye at soil conservation and weed suppression, both
before and after termination (Mohler and Teasdale, 1993; McCracken et al., 1994; Mennan et al.,
2009). Vetch can also become a problematic weed in winter annual and perennial cropping
systems due to hard seed (Aarssen et al., 1986), and in reduced tillage systems as a result of
incomplete kill and regrowth (Creamer and Dabney, 2002; Mischler et al., 2010).
Rye-vetch mixtures are often proposed as a way to combine the strengths of the two
species while moderating their individual weaknesses, in addition to taking advantage of
potential synergies that arise from the grass-legume association. In practice, the characteristics
of cereal-legume cover crop mixtures tend to reflect the relative proportions of each species
present, with the qualities of a given species becoming more prominent with greater proportion
in the mixture. Accordingly, many studies find that qualities such as weed suppression, total dry
matter production, and N availability in cereal-legume mixtures tend to be intermediate to the
corresponding monocultures (e.g., Ranells and Wagger, 1997c; Akemo et al., 2000b; HauggaardNielsen et al., 2003; Clark et al., 2007b; Benincasa et al., 2010; Dordas and Lithourgidis, 2011).
However, studies have also demonstrated that mixtures of vetch with a cereal can accumulate
more total dry matter and N than either species in monoculture (Clark et al., 1994; Sainju et al.,
2005), exhibit more efficient N accumulation by the components (Tosti et al., 2010), mineralize
N at rates approaching that of a vetch monoculture (Ranells and Wagger, 1996), and provide
weed suppression equivalent to that of a cereal monoculture (Teasdale and Abdul-Baki, 1998;

38

!
!
Hayden et al., 2012). In addition, because legumes are generally sown at reduced rates in
mixtures, overall seed costs for cereal-legume mixtures tend to be lower than for monoculture
legumes.
The performance of cereal-legume cover crop mixtures relative to monocultures depends
heavily on both environmental conditions and management decisions that influence the
competitive balance between species and final stand characteristics. Cereals, with their extensive
root systems and higher relative growth rates, tend to be stronger competitors for below-ground
resources than legumes (Mariotti et al., 2009). Therefore, when resources other than N,
including moisture, are limiting, cereals are likely to suppress legumes in mixtures (Ofori and
Stern, 1987). However, N limitation often shifts the competitive balance to benefit legumes in
mixtures, while high N fertility generally favors non-legumes (Jensen, 1996), including rye in
rye-vetch mixtures (Shipley et al., 1992; Clark et al., 2007a). In addition, rye can germinate and
grow at cooler temperatures than vetch (Nuttonson, 1958; Teasdale et al., 2004), which
contributes to its ability to establish later in the fall and accumulate biomass faster in the spring
(Shipley et al., 1992). As a result, colder weather, later planting dates, and earlier cover crop
termination are all likely to shift the balance toward rye over vetch in mixtures.
Considering the importance of relative stand composition in influencing the properties of
cereal-legume mixtures, seeding rates are a logical tool for managing mixture performance. The
fundamental importance of component species density is recognized and often thoroughly
investigated in studies of plant competition (Firbank and Watkinson, 1985) and agronomic
intercropping (Willey, 1979b). However, few experiments systematically evaluate the influence
of sown species proportions on the characteristics of cover crop mixtures (Akemo et al., 2000b;
Karpenstein-Machan and Stuelpnagel, 2000; Tosti et al., 2010), and most draw conclusions from

39

!
!
only a single mixture, often where the species are sown at 50% of their recommended
monoculture seeding rates. When multiple seeding rates of cereal-legume cover crops have been
evaluated side-by-side, the differences in stand characteristics among the mixtures are often
enough to significantly influence the provision of agroecosystem services by the cover crops,
such as weed suppression, soil N fertility, and effects on subsequent crop yields (Clark et al.,
1994; Akemo et al., 2000a, 2000b; Tosti et al., 2012).
Replacement series experiments, in which treatments consist of a monoculture of each
species and a gradient of species mixtures, provide one approach for investigating species
proportions in cover crop mixtures (Jolliffe, 2000). Replacement series are widely used in the
study of plant competition, and are commonly applied within agricultural contexts to evaluate
crop-weed interactions and intercrop productivity. While total plant density is held constant in a
traditional replacement series, a replacement series is proportional when the relevant
monoculture seeding rates differ for the component species, and total density therefore varies
across the mixtures. The limitations to inference from such designs have been thoroughly
reviewed (Snaydon, 1991; Jolliffe, 2000), but over-interpretation remains common in the
literature. In particular, the densities of component species are confounded with each other and
with total plant density in proportional designs. Therefore, without data on density-dependent
yield responses for each species (Firbank and Watson, 1985), the design is not suited for drawing
definitive conclusions regarding relative competitive abilities or the causal mechanisms
contributing to species performance in mixtures. Despite these caveats, proportional
replacement series are still effective in providing applied insights that can inform improvements
in cover crop mixture management.
A greater understanding of how species proportions influence mixture stands is an

40

!
!
important step toward maximizing potential benefits from rye-vetch cover crop mixtures to better
serve specific goals within cropping systems. Therefore, the objective of this study was to
evaluate how the relative proportions of rye and vetch sown in a winter annual cover crop
mixture influenced cover crop stand characteristics and performance with respect to
establishment, biomass productivity, winter annual weed suppression, and N accumulation,
fixation, and costs relative to rye and vetch monocultures.

MATERIALS AND METHODS
Site Description
The study was conducted at the Michigan State University (MSU) Horticulture Teaching
and Research Center in Holt, MI (42°40” N, 84°28” W) over the course of two seasons,
alternating between adjacent fields. Both fields were on level terrain and had a Spinks loamy
sand soil (sandy, mixed, mesic Lamellic Hapludalf). In each field, a summer cover crop of
sorghum sudangrass (Sorghum bicolor x S. bicolor var. Sudanense) was flail mowed and
incorporated using a rototiller at least 2 wk prior to sowing of winter cover crops in the fall.
Initial soil chemical characteristics for the fields in 2009 and 2010 were similar, and included on
average pH 6.6; CEC 7.1 cmol kg-1; and P (Bray P1 extract), K, and Mg levels of 133, 145, and
47 mg kg-1, respectively. The fields were managed according to National Organic Program
guidelines (Agricultural Marketing Service, 2013), and had been in organic transition with a
cropping history of warm-season vegetables under conventional tillage since 2008. No fertilizers
or soil amendments were applied in either year of this study.
Treatment and Experimental Design
Following a proportional replacement series design (Snaydon, 1991; Jolliffe, 2000), winter

41

!
!
cover crop treatments included rye and vetch sown in monocultures at rates of 94 and 42 kg ha-1,
respectively, and a gradient of species mixtures containing the following sown proportions of
rye:vetch relative to their monoculture seeding rates— 83:17, 67:33, 50:50, 33:67, 17:83. Rye
and vetch seeding rates and seed cost estimates for the cover crop treatments are listed in Table
1. A no-cover-crop control was also included for comparison. Experimental plots were 6.7 by
8.5 m in 2009-2010 and 6.1 by 7.6 m in 2010-2011, arranged in a randomized complete block
design with four replications.
Field Management and Data Collection
The dates of key field activities and data collection are summarized in Table 2. Rye and
vetch cover crops were broadcast sown by hand using a grid system that divided plots into
quarters to help ensure uniformity, and then incorporated to a depth of about 5 cm using a field
cultivator (Perfecta II, Unverferth Manufacturing Co. Inc., Kalida, OH). Variety not stated
(VNS) vetch seed grown in Oregon and VNS rye seed grown in Minnesota were used in both
years (Albert Lea Seed House, Albert Lea, MN). Vetch seed was inoculated with N-DURE
Rhizobium leguminosarum inoculant (INTX Microbials LLC, Kentland, IN) at a rate of
approximately 10 g inoculant kg-1 seed.
Rye and vetch plant densities were assessed in the fall from four 25- by 50-cm (0.125 m-2)
quadrats established in each plot. Spring cover crop densities, shoot biomass, and total weed
biomass were sampled later from the same quadrats, shortly before mowing and incorporation.
At the time of spring sampling in both years, vetch was pre-bloom and rye was between earemergence and anthesis. Potential differences in vetch winter survival across treatments were
evaluated by calculating the percentage change in vetch population density between fall and
spring sampling dates.

42

!
!
Shortly before sampling cover crop biomass in the spring, photosynthetically active
radiation (PAR) was measured both above the canopy and at the soil surface at four locations in
each plot at least 1 m from a plot edge. Measurements were taken midday under clear skies with
a 70-cm long quantum flux sensor (MQ-301, Apogee Instruments Inc., Logan, UT), and used to
calculate the percentage of PAR penetrating the cover crop canopies.
Rye, vetch, and total weed biomass samples were dried to a constant weight at 38°C prior
to taking dry weights. Cover crop dry weights per plant were calculated using rye and vetch
spring densities. Cover crop shoot biomass samples were then ground to pass through a 1 mm
screen, and subsamples were submitted to Midwest Laboratories Inc. (Omaha, NE) for analysis
of percent total C and N using a LECO TruSpec elemental analyzer (LECO Corp., St. Joseph,
MI). Subsamples were also submitted to the University of California-Davis Stable Isotope
Facility (SIF) for analysis of δ15N using a PDZ Europa ANCA-GSL elemental analyzer
interfaced to a PDZ Europa 20-20 continuous flow isotope ratio mass spectrometer (Sercon Ltd.,
Cheshire, UK).
Corrections for Soil Contamination on Vetch Shoots
Near-saturated soil moisture conditions at the time of cover crop sampling in 2011 resulted
in considerable amounts of soil adhering to vetch shoots that could not be completely removed
from samples prior to drying and analysis. Therefore, vetch dry weight and percent N data were
corrected for the presence of soil contamination following equations found in Hunt et al. (1999).
Briefly, the fraction by weight of vetch tissue (Fv) in a contaminated sample (vetch tissue plus
contaminating soil) was estimated using the following equation:
[1]

Fv = (A0 - As) / (A0 - Av)!

where As is the ash fraction of the contaminated sample, Av is the ash fraction of uncontaminated

43

!
!
vetch tissue, and A0 is the ash fraction of the contaminating soil. The ash fraction of
uncontaminated vetch tissue (Av) was estimated as 0.2, based on vetch samples taken from 2010
and 2011 that were not visually contaminated with soil and did not exhibit outlier δ15N values.
The ash fraction of the contaminating soil (Ao) was estimated from soil sampled shortly before
cover crop sampling in 2011, and sieved to remove large particles that were less likely to have
adhered to vetch tissue. Ash fractions were determined for the soil and for all vetch tissue
samples using a muffle furnace (2 h drying period at 105°C, followed by 4 h at 500°C).
The corrected vetch tissue dry weight (Mv) was then derived from the dry weight of the
contaminated sample (Ms) using Eq. 2:
[2]

Mv = Ms x Fv
Finally, the concentration of N in vetch tissue (Nv) was estimated from the sample N

concentration (Ns) and the concentration of N in the contaminating soil (N0) by:
[3]

Nv = (Ns – N0 x (1 - Fv)) / Fv
For consistency, the correction was applied to vetch samples from both years of the study;

however, it did not significantly affect the 2010 data due to the lack of soil contamination in that
year. No correction was necessary for rye data in either year because the upright nature of the
shoots minimized the risk of soil contamination during sampling.
Land Equivalent Ratios
Partial and total land equivalent ratios (LER) were calculated for each mixture on the basis
of rye and vetch dry weight (de Wit and Van der Bergh, 1965; Willey, 1979a). Partial LERs
(relative yields) for rye (Lr) and vetch (Lv) were calculated as the ratio of the dry weight of each
species in mixture to its dry weight in monoculture. The total LER for mixtures was then
calculated as the sum of the partial LERs. Relative yields greater than relative sown proportions

44

!
!
indicate greater biomass productivity in mixtures per unit rye or vetch seed sown, while total
LER > 1.0 demonstrates greater overall biomass productivity (dry weight per seed sown) for the
mixture relative to the monocultures.
Vetch N Fixation Estimates
Vetch N fixation was estimated using the 15N natural abundance method, which relies on
the slight enrichment of 15N generally present in soil N relative to atmospheric N2 (Shearer and
Kohl, 1986). Since biological N fixation exhibits little isotopic fractionation, legumes tend to
have a lower proportion of 15N in their tissues than plants that derive their N entirely from the
soil (Hogberg, 1997). Isotopic composition is expressed relative to atmospheric N2 using δ15N
values in parts per thousand (%0), with higher δ15N indicating greater 15N enrichment.
The percentage of vetch N derived from the atmosphere (%Ndfa) was calculated from Eq.
4 (Rochester and Peoples, 2005),
[4]

%Ndfa = (δ15Nref – δ15Nleg)/( δ15Nref – B) × 100

where δ15Nref is the shoot δ15N value for rye in monoculture (used as the non-N-fixing reference
plant), δ15Nleg is the shoot δ15N value for vetch, and B is the estimated δ15N of vetch when
grown entirely dependent on atmospheric sources of N. Following Hansen et al. (2002) and
Hansen and Vinther (2001), the lowest δ15Nleg value in each study year was used to approximate
B, resulting in conservative estimates of vetch %Ndfa. Rye monoculture δ15Nref values were
2.52 and 1.49 %0 in 2010 and 2011, respectively.
Vetch shoots contaminated with soil (ash content >0.2) in 2011 exhibited unreasonably
high δ15N values (1.10 – 5.58 %0) that could not be explained by contributions from soil 15N
alone. We speculate that microbial N transformations (such as denitrification) may have
occurred preferentially on contaminated vetch shoots prior to sampling in the field or during

45

!
!
sample processing before drying, potentially resulting in significant isotope fractionation
(Mariotti et al., 1981; Shearer and Kohl, 1986). Therefore, contaminated vetch tissue was
excluded from the analysis. As a result, 2011 mean %Ndfa estimates are reported only for the
three cover crop treatments where at least 3 field replicates of uncontaminated vetch tissue (ash
content < 0.2) were available.
The amount of vetch shoot N derived from N fixation (SNF, in kg ha-1) was estimated
using the following formulas:
[5]

ShootN = (shootdryweight)x(Nv)

[6]

SNF= (shootN)x(%Ndfa)

where Nv is the contamination-corrected concentration of N in the vetch shoots. Estimates of
SNF for 2011 were calculated assuming a consistent %Ndfa across the cover crop treatments
equal to the average %Ndfa obtained from the uncontaminated samples in that year.
Cost of Nitrogen Fixed
In order to estimate the relative costs incurred by growers for the additional fixed N
contributed to the system by vetch, the cost of N fixed (CNF) was estimated for each vetchcontaining cover crop treatment following Brainard et al. (2012), using the equation:
[7]

CNF= [(Pv x Qv) + (Pr x Qr) - (Pr x 94)] / SNF

where Pv and Pr are the prices ($ kg-1) of conventional vetch and rye seed, respectively; Qv and
Qr are the seeding rates (kg ha-1) of vetch and rye, respectively; and SNF is the vetch shoot N
fixed (kg ha-1) estimated from Eq. [6]. CNF is an estimate of the price paid for the N fixed by
vetch as influenced by legume performance and cover crop seed costs. The calculation controls
for costs associated with cover crop establishment by assuming that it applies to growers who
already seed rye at 94 kg ha-1, and therefore just have to substitute legume seed in their planting

46

!
!
equipment. Since rye is a common cover crop in many agronomic and vegetable cropping
systems, CNF provides a reasonable estimate of both the additional costs a grower may incur to
gain fixed N from vetch in their system, and the relative effects of rye-vetch mixture proportion
on the estimated price of additional fixed N. Because some fixed N also accumulates in root
tissue, calculations of CNF based on Eq. [7] likely overestimate the cost of total fixed N in vetch
biomass.
Statistical Analysis
The fixed effects of year, cover crop mixture proportion, and their interaction were
analyzed using mixed model ANOVA with the MIXED procedure of SAS (Version 9.2, SAS
Institute, Cary, NC). Block (replication) was included as a random factor in all models.
Assumptions of normality and equality of variances were evaluated, and unequal variance
models were used when necessary. Unless otherwise noted, effects where judged significant
when P < 0.05. Where the effect of cover crop mixture proportion was significant, the nature of
the response in each year was investigated using linear or polynomial regression with the REG
procedure of SAS. In the case of PAR, the effects of year and the interaction between year and
cover crop were not significant, so data were pooled over years and fitted with a single
regression equation. The significance and nature of the relationship between rye shoot N
concentration and cover crop mixture proportion were evaluated in the context of analysis of
covariance (ANCOVA) using the MIXED procedure of SAS, since the response was linear and
had a common slope in both years. Individual partial and total LER estimates for cover crop
mixtures were compared to their respective critical values using t-tests. Finally, due to missing
data in 2011, vetch %Ndfa data were analyzed separately by year.

47

!
!
RESULTS AND DISCUSSION
Weather Conditions
Table 3 presents total growing degree day (GDD) base 4 °C accumulation and rainfall at
the experiment station during the periods of cover crop growth. Heat accumulation and total
rainfall were similar in the fall of both study years, but the spring of 2011 saw 113 fewer GDD
and 125 more mm of rainfall than spring 2010. Previous research has demonstrated good
correlation between GDD base 4 °C accumulation and winter cover crop biomass in the spring
(Teasdale et al., 2004). While wetter soil conditions in 2011 contributed to difficulties with
vetch biomass sampling that season (see Materials and Methods), rainfall was sufficient in both
years to make water deficit during cover crop growth unlikely, suggesting that compared to heat
accumulation, soil moisture was probably a minor contributor to variability in cover crop growth
across study years.
Cover Crop Density and Vetch Winter Survival
Rye and vetch spring population densities across treatments were similar in the two years
of the study, with rye and vetch monocultures averaging 146 and 271 plants m-2, respectively
(Figure 1). The densities of both species in mixture were highly correlated with their relative
sown proportions, and the relationships were predominantly linear. This suggests that interspecific interactions had little effect on the germination and establishment of rye and vetch under
the conditions of our field study, and that seeding rates can be good predictors of final population
composition in rye-vetch mixtures.
The influence of mixture proportions on vetch winter survival, in particular, was
investigated by observing the percent change in vetch population density between counts taken
in the fall and spring from the same quadrats. There were no significant differences in vetch

48

!
!
population change among the treatments in either year (Table 4); however, contrary to
expectation, in most cases our estimates of vetch density actually increased between the fall and
spring, resulting in apparent overwinter survival estimates greater than 100%.
This has been observed in previous studies (Brainard et al., 2012), and may in part be
attributed to delayed germination from vetch hard seed (Aarssen et al., 1986). In addition,
individual vetch plants are increasingly difficult to distinguish from each other in-situ as the
shoots sprawl and intertwine. Therefore, we speculate that the in-situ counts taken in the fall
may have underestimated vetch density compared with the spring counts, in which quadrats were
destructively sampled and individual plants could more easily be identified by their roots. Later
fall sampling (and more advanced vetch growth) in 2009 (Table 3) may also have contributed to
greater under-counting that year, suggesting the larger population change estimates in 2009
relative to 2010 (P = 0.006) could be an artifact.
Assuming delayed vetch germination and potential sampling biases were consistent across
treatments, we found no evidence that vetch winter survival was improved by mixing with rye.
In contrast, Jannink et al. (1997) found that the presence of rye reduced overwinter mortality of
vetch at two locations in Maine, and Brainard et al. (2012) observed a similar response in one out
of two years in Michigan. While increased legume winter survival is an often-stated benefit of
cereal-legume cover crop mixtures, its significance will likely vary depending on both winter
weather conditions and the establishment (and hardiness) of the legume. The presence of rye
may have a greater influence on vetch winter survival, for example, when the vetch is seeded late
in the fall or when winter temperatures are particularly harsh.
Cover Crop Biomass Production and Land Equivalent Ratios
Vetch shoot biomass in monoculture was 551 and 305 g m-2 in 2010 and 2011,

49

!
!
respectively, while rye monocultures produced 414 and 330 g m-2 in those two years (Figure 2).
Cooler spring temperatures in 2011 (Table 3) likely contributed to the overall lower biomass
production observed that year, particularly for vetch (rye, P = 0.03; vetch, P = 0.004). As with
population density, relative seeding rates were strongly correlated with rye and vetch shoot dry
weight in mixtures, demonstrating that seeding rates can be a good predictor of biomass
composition as well. Over the course of the study, the total shoot biomass produced in mixtures
was generally equal to or greater than that produced in either monoculture. This was reflected in
the total LER values for the mixtures, which ranged from 1.13 to 1.42, and in all cases were
either equal to or significantly greater than one (Table 5). Furthermore, relative yields of both
rye (Lr) and vetch (Lv) met, and in some cases significantly exceeded, their relative sown
proportions in all mixtures, indicating that both species contributed to the equivalent or greater
efficiency of biomass production (dry weight per seed sown) observed in the mixtures relative to
the monocultures. In other words, just as much or more biomass was produced per dollar spent
on rye and vetch seed in mixtures as in monocultures.
Inference regarding the nature of interactions between species is limited from proportional
replacement series designs because total plant density is not constant across the treatments
(Jolliffe, 2000). In the case of this study, total density increased with higher proportions of rye in
the mixture (Figure 1). Without data on the yield response of each species to changes in density,
the balance between inter- and intra-specific interactions in mixtures cannot be determined with
confidence (Firbank and Watkinson, 1985; Jolliffe, 2000). However, given that neither rye nor
vetch productivity (dry weight per seed sown) was suppressed in any of the mixtures,
interspecific interference between the two species appears to have been minimal under the
conditions of this experiment. Furthermore, plants generally exhibit an inverse relationship

50

!
!
between density and biomass production per individual (Avci and Akar, 2006; Boyd et al.,
2009). Therefore, increased rye productivity in mixtures could simply be due primarily to the
decrease in total plant density associated with decreasing proportions of rye in the mixtures.
Conversely, vetch productivity was not reduced at the higher total plant densities associated with
higher rye proportions (in opposition to the expected density-dependent response), suggesting
that vetch may have benefitted from either weaker interspecific competition or possible
facilitation from rye.
Decades of intercropping research have demonstrated that cereals are usually the dominant
component in cereal-legume mixtures, suppressing the growth of the legume and contributing a
greater proportion of biomass to the total mixture yield (Ofori and Stern, 1987). This
competitive imbalance may be more severe for low-growing legumes than for those with
climbing growth habits (Davis et al., 1984; Fukai and Trenbath, 1993), but the performance of
vetch in mixture with cereals still varies across studies. Kurdali et al. (1996) and Tosti et al.
(2010) both saw reduced vetch dry matter productivity in mixture with barley (Hordeum vulgare
L.), while Dhima et al. (2007) also observed suppression of vetch by barley and oat (Avena
sativa L.), but not by wheat (Triticum aestivum L.) or triticale (xTriticosecale Wittmack). As
with our results, Ranells and Wagger (1996) observed limited suppression of vetch in mixture
with rye; however, Brainard et al. (2012) observed one out of two seasons where the productivity
of three vetch varieties was reduced on average by rye, which yielded over 400 g m-2 in the
50:50 mixtures. By comparison, rye yields were relatively low in our study. The sandy soils,
prior growth and incorporation of sorghum-sudangrass, and absence of N fertilization likely
contributed to low soil N conditions that favored vetch growth relative to rye, reducing the
potential for competitive inhibition in the mixtures.

51

!
!
An important qualification to note is that measures of yield advantage or efficiency depend
not only on the performance of the species in mixture, but also on the characteristics of their
respective monocultures. An implicit assumption in intercrop or mixture evaluation is that the
monocultures chosen are the optimal ones for sole-cropping of each species (Willey, 1979b). In
this study, monoculture seeding rates were chosen with the goal of maximizing cover crop
biomass production, especially for vetch. However, due to the high cost of seed, it is common to
see vetch sown at 28 kg ha-1 in monoculture, particularly in agronomic cropping systems (e.g.,
Clark et al., 1994; Sainju et al., 2005; Parr et al., 2011). Furthermore, rye monocultures are often
sown at rates of 125 kg ha-1 or more, which may better maximize biomass production,
particularly under conditions where soil N is not limiting (Clark, 2007; Boyd et al., 2009).
Outcomes in replacement-type mixtures and conclusions regarding mixture performance may
vary when different monoculture seeding rates are used (Willey, 1979b).
Winter Annual Weed Biomass and PAR
The percentage of PAR penetrating the cover crop canopies in early May ranged from 6 to
47% for vetch and rye monocultures, respectively, averaged over the two years of the study. In
the mixtures, PAR penetration increased with decreasing proportion of vetch and increasing
proportion of rye (Figure 3). All cover crop treatments significantly suppressed winter annual
weed biomass relative to control treatments in 2010 and 2011, and suppression increased with
increasing rye in the cover crops (Figure 4). Total weed biomass in the control treatments was
144 and 162 g m-2 in 2010 and 2011, respectively, and dominant species included common
chickweed [Stellaria media (L.) Vill.], annual bluegrass (Poa annua L.) and shepherd’s purse
[Capsella bursa-pastoris (L.) Medik.] in 2010, and corn chamomile (Anthemis arvensis L.), field
pepperweed [Lepidium campestre (L.) R. Br.], and common chickweed in 2011. While

52

!
!
suppression by high rye treatments was comparable between the two years of the study,
variability in weed suppression between 2010 and 2011 increased with greater proportion of
vetch sown. Although generally less of a management priority than summer annuals, winter
annual weeds can interfere directly with crops in reduced tillage systems and serve as alternative
hosts for important pests and diseases (Hayden et al., 2012).
Our results are consistent with previous studies demonstrating that weed suppression in
cereal-legume mixtures is generally driven by the presence and relative proportion of the more
competitive cereal component (Liebman and Dyck, 1993; Akemo et al., 2000b; Hayden et al.,
2012). The increasingly lower weed suppression in 2011 relative to 2010 with greater vetch in
the cover crops was likely a result of the cooler spring temperatures and lower biomass
production that year, and also suggests that the weed suppressiveness of vetch is less robust in
the face of year-to-year environmental variability than rye. While both rye and vetch are known
to exhibit allelopathic properties (Barnes and Putnam, 1986; White et al., 1989), shading is likely
a more dominant mechanism of competition for vetch than it is for rye, which suppresses
neighboring plants more predominately through competition for water and soil nutrients
(Teasdale and Daughtry, 1993; Mariotti et al., 2009). This dominance of below-ground
interference in the weed suppressiveness of rye is supported by the fact that weed biomass
decreased with increasing rye in the cover crops, despite a corresponding increase in the amount
of PAR reaching the soil surface in the spring. However, it is important to note that canopy
architecture changes substantially during rye growth, and PAR readings taken in the spring may
not reflect the level of competition for light exerted by rye on weeds the previous fall.
Cover Crop Nitrogen Content
Total shoot N content across the cover crop stands in 2010 and 2011 was largely driven by

53

!
!
rye and vetch biomass production. As expected, vetch biomass contributed the majority of N to
rye-vetch mixtures, and total N content generally increased with increasing proportion of vetch
sown (Figure 5). In 2010, total N content ranged from 3.8 g m-2 in the rye monoculture to 17.5 g
m-2 in the vetch monoculture, while in 2011, total N content ranged from 4.2 g m-2 in rye
monoculture to 11.6 g m-2 in the 83:17 vetch:rye mixture.
In both study years, a modest but significant increase in rye tissue N concentration was
observed in mixtures with increasing proportions of vetch (Figure 6). Combined with the
equivalent or greater rye biomass productivity observed in mixtures relative to the rye
monoculture (Table 5), this trend provides added support that the productivity of N accumulation
by rye in mixtures was greater as well. Similarly, Tosti et al. (2010, 2012) saw increases in the
tissue N concentration of barley with increasing vetch sown in replacement-style mixtures. In
both cases, the mechanism behind the increase cannot be clearly established. Improved cereal N
economy in cereal-legume mixtures is consistent with a facilitative increase in N availability due
to the legume. However, in proportional replacement series designs, greater legume proportions
in mixtures are also accompanied by a corresponding reduction in cereal density. Because rye is
a stronger competitor for soil N resources than vetch, reduced rye intra-specific competition for
soil N could also account for the increase in rye tissue N concentration across the gradient of ryevetch mixtures. The overall higher rye N concentrations in 2011 compared with 2010 (P <
0.001) likely reflect a more immature rye growth stage at sampling due to the lower GDD
accumulation during the 2010-2011 season.
Vetch Nitrogen Fixation
Estimates of vetch shoot %Ndfa were relatively high in both years, ranging from 81 to
88%. Although %Ndfa was marginally lower for vetch in monoculture than for vetch in mixture

54

!
!
with rye on average in 2010 (linear contrast, P = 0.065), estimates of N fixation did not differ
significantly across the rye-vetch mixtures. In 2011, soil contamination compromised N fixation
estimates from cover crop treatments containing greater than 50% vetch; however, for the three
mixtures where reliable data were available, vetch %Ndfa averaged 82% and did not differ
among the mixture rates.
Previous studies have observed increases in vetch N fixation efficiency in mixture with
cereals (Kurdali et al., 1996; Brainard et al., 2012). We expected that higher proportions of rye
in mixture might lead to greater inter-specific competition for available soil N and promote
greater N fixation by vetch. However, the relatively sandy soils, low organic matter content, and
absence of N fertilization at our study site may have reduced the potential for observing such an
effect. The presence of rye and its relative proportion in mixture may be more likely to influence
vetch N fixation efficiency under conditions where soil N availability is high. On the other hand,
excessive soil N may also shift the competitive balance in favor of the cereal, resulting in
inhibition of legume growth and potentially N fixation capacity. While Brainard et al. (2011)
observed increased nodulation of soybean when grown in combination with Japanese millet
[Echinochloa frumentacea (Roxb.) Link], the effect was not observed when composted manure
was added to the system.
The amount of shoot N fixed (SNF) was driven predominately by vetch biomass
productivity. Across the gradient of rye-vetch cover crops, SNF varied from zero in rye
monoculture to 142 and 85 kg N ha-1 in the 2010 and 2011 vetch monocultures, respectively
(Table 6). However, the total cost for conventional cover crop seed increased from $37 to $176
ha-1 over that same range of treatments (Table 1). As a result, the estimated cost of the N fixed
by vetch (CNF) ranged from $0.79 to $1.90 kg-1 N over the two years of the study, and despite a

55

!
!
numerical trend toward decreasing costs in mixture with greater rye, CNF was not affected
significantly by rye-vetch mixture rate (Table 6). A similar range of CNF estimates was
established by Brainard et al. (2012) for three vetch varieties in monoculture and in mixture with
rye. For conventional growers in particular, acknowledging the value of fixed N as a potential
fertilizer replacement may help justify the substantially higher seed costs of cover crops that
contain greater proportions of vetch. Comparisons of CNF with the cost of synthetic N fertilizers
are confounded by the incomplete first-season availability of N from organic residues and by the
potential value of additional services that are provided by cover crops beyond N fertility. But as
a reference, the average cost of urea fertilizer over the past five years was fairly comparable to
CNF from vetch in this study, ranging from $1.10 to $1.35 kg-1 N (USDA-Economic Research
Service, 2013). Furthermore, estimates of CNF in this study did not include fixed N
accumulated in root tissue. Since vetch roots can account for 10% or more of total tissue N
(Rogers and Sturkie, 1939), our figures likely overestimate CNF. Nevertheless, while seed costs
vary with time, source, and whether conventional or organic seed is used, our results suggest that
under the conditions of this study, the relative cost for the N fixed by vetch was not significantly
influenced by mixing with rye at any proportion.

CONCLUSIONS
The objective of this research was to evaluate the effects of rye-vetch species proportions
on cover crop mixture stand characteristics and performance relative to rye and vetch
monocultures. Seeding rates were good predictors of rye and vetch stand density and biomass
composition in mixtures, with little evidence of substantial interference between the two species
under the conditions of the study. While outcomes may vary depending on environment and

56

!
!
management, we also saw little evidence of synergistic benefits in the mixtures, such as
improved vetch winter survival, increased N fixation, or reduced costs per unit of N fixed.
However, changing the proportions of rye and vetch sown resulted in tradeoffs among some of
the agroecosystem services provided by the living cover crops—for example, greater fixed N
accumulation but higher seed costs and reduced winter annual weed suppression with increasing
vetch. A greater understanding of the tradeoffs among cover crop services, as well as costs, in
rye-vetch mixtures will support more-informed decision-making regarding cover crop selection
and mixture seeding rates.

57

!
!

APPENDIX

58

!
!
Table 2.1. Rye (R) and vetch (V) seeding rates and estimated seed costs across cover crop
mixture proportions.
Cover crop treatment
100 V : 0 R
83 V : 17 R
67 V : 33 R
50 V : 50 R
33 V : 67 R
17 V : 83 R
0 V : 100 R

Seeding rate
Vetch
kg ha-1
42
35
28
21
14
7
0

Rye

Vetch

0
16
31
47
63
78
94

176
146
118
88
58
30
0

Cost of seed†
Rye
$ ha-1
0
6
12
18
25
30
37

Total
176
152
130
106
83
60
37

† Seed costs were calculated using average prices paid for conventional vetch and rye seed at
the time of this experiment ($4.19 and $0.39 kg -1, respectively). Average prices paid for the
organic seed sown in the experiment were $4.63 and $0.48 kg-1 for vetch and rye, respectively.
Actual seed costs will vary with time, seed source, and whether conventional or organic seed is
used.

59

!
!
Table 2.2. Dates of key field activities and data collection.
Activity
Cover crop seeding
Cover crop fall density sampling
PAR† light penetration readings
Spring cover crop and weed biomass and density sampling
† Photosynthetically active radiation

60

2009-2010

2010-2011

1 Sept.
19-21 Oct.
6 May
10 May

1 Sept.
27-29 Sept.
11 May
14 May

!
!
Table 2.3. Monthly growing degree day (GDD) and rainfall totals in
Holt, MI during cover crop growth, 2009-2011.
Month

GDD†
2009-10 2010-11

Rainfall
2009-10
2010-11
mm
24
92
92
34
24
50

September‡
October
November

393
173
99

364
228
81

March
April
May§

81
233
91

28
125
139

16
74
52

66
135
65

Sept. to Nov.‡
Mar. to May§

665
405

673
292

141
141

176
266

† Calculated according to Baskerville-Emin method (Baskerville and
Emin, 1969) using base 4°C.
‡ Cumulative GDD beginning at cover crop seeding (see Table 2.2).
Data for Dec. through Feb. not presented due to negligible GDD
accumulation and unreliability of measurements of frozen
precipitation.
§ Cumulative GDD ending at cover crop biomass sampling (Table
2.2).

61

!
!
Table 2.4. Vetch apparent overwinter survival
(population change) across cover crop mixture
proportions.
Overwinter survival
Cover crop treatment
2010
2011
%
100 V : 0 R
137†
94
83 V : 17 R
119
98
67 V : 33 R
127
106
50 V : 50 R
157
105
33 V : 67 R
123
91
17 V : 83 R
126
95
F‡
ns§
ns
† See text for possible explanations for apparent
overwinter survival estimates > 100%.
‡ Significance of fixed effect of cover crop mixture
proportion within years.
§ ns, nonsignificant at the 0.05 probability level.

62

!
!
Table 2.5. Partial land equivalent ratios for vetch (Lv) and rye (Lr) and total land
equivalent ratios (LER) in rye-vetch cover crop mixtures.
2010
2011
Cover crop treatment
Lv
Lr
LER
Lv
Lr
83 V : 17 R
0.76
0.51*†
1.27
0.98
0.39*
67 V : 33 R
0.72
0.52
1.25
0.91*
0.51*
50 V : 50 R
0.54
0.66
1.20
0.53
0.62
33 V : 67 R
0.37
0.90
1.27
0.55‡
0.82‡
17 V : 83 R
0.22
1.00
1.23
0.31
0.82

LER
1.38‡
1.42*
1.15
1.37‡
1.13

* Significant at the 0.05 probability level.
† Indications of significance refer to whether individual Lv and Lr means are
significantly different from vetch and rye sown proportions in mixture, respectively,
and whether LER means are significantly different from 1.0 (based on t-tests).
‡ Significant at the 0.10 probability level.

63

!
!
Table 2.6. Vetch shoot N concentration, percentage of N derived from the atmosphere (Ndfa),
estimated shoot N fixed (SNF) and cost of N fixed (CNF) across cover crop mixture proportions.
N concentration
Ndfa†
SNF‡
CNF§
Cover crop
treatment
2010
2011
2010
2011¶
2010
2011
2010
2011
-1
-1
%
%
kg ha
$ kg
100 V : 0 R
3.17
3.37
80.9
-142.4
84.6
1.05
1.68
83 V : 17 R
3.18
3.30
87.9
-110.8
80.1
1.13
1.47
67 V : 33 R
3.13
3.31
86.5
-102.3
74.9
0.92
1.29
50 V : 50 R
3.12
3.36
88.4
84.0#
74.9
43.6
1.15
1.90
33 V : 67 R
3.28
3.37
87.5
82.8#
56.9
46.0
0.84
1.16
17 V : 83 R
3.18
3.36
84.5
80.1
31.5
25.5
0.79
1.11
F††
ns‡‡
ns
ns
ns
***
**
ns
ns
** Significant at the 0.01 probability level.
*** Significant at the 0.001 probability level.
† Calculated from Eq. [4].
‡ Calculated from Eq. [6].
§ Calculated from Eq. [7].
¶ Means are reported for treatments only where at least 3 replicates of uncontaminated vetch
tissue were available in 2011 (see Materials and Methods).
# Means calculated from three replicates (n=3)
†† Significance of fixed effect of cover crop mixture proportion within years.
‡‡ ns, nonsignificant at the 0.05 probability level.

64

!
!

Figure 2.1. Spring density of vetch (gray), rye (white), and vetch+rye (black) across cover crop
mixture proportions in 2010 and 2011. 2010 responses of vetch: y=1.34x, r2=0.81***; rye:
y=2.44x, r2=0.87***; and vetch+rye: y=1.14x+136.23, r2=0.73***. 2011 responses of vetch:
y=1.56x, r2=0.84***; rye: y=0.006x2+0.46x, r2=0.80***; and vetch+rye: y=1.12x+182.26,
r2=0.47***. Means (+/- SE). Single, double, and triple asterisks following r2 values denote
regression model significance at the 0.05, 0.01, and 0.001 levels, respectively.

65

!
!

Figure 2.2. Spring shoot biomass dry weights of vetch (gray), rye (white), and vetch+rye
(black) across cover crop mixture proportions in 2010 and 2011. 2010 responses of vetch:
y=5.38x, r2=0.71***; rye: y=-0.036x2+7.67x, r2=0.50***; and vetch+rye: y=-1.46x+607.32,
r2=0.16*. 2011 responses of vetch: y=-0.020x2+5.07x, r2=0.69***; rye: y=-0.025x2+5.62x,
r2=0.76***; and vetch+rye: y=-0.038x2+3.62x+332.52, r2=0.23*. Means (+/- SE). Single,
double, and triple asterisks following r2 values denote regression model significance at the 0.05,
0.01, and 0.001 levels, respectively. Asterisks in the figure indicate where total biomass
(vetch+rye) in a mixture was significantly greater than in either monoculture in that year at the
0.05 probability level.

66

!
!

Figure 2.3. Percentage of photosynthetically active radiation (PAR) penetrating the cover crop
stand canopy across mixture proportions. Means are for 2010 and 2011 data combined (n=8),
presented +/- SE. The triple asterisks following the r2 value in the figure denote regression
model significance at the 0.001 level.

67

!
!

Figure 2.4. Winter annual weed biomass (dry weight) across cover crop mixture proportions in
2010 and 2011, expressed as mean percentage of the no-cover-crop control treatment (+/- SE).
2010 response: y=0.089x+4.18, r2=0.18*. 2011 response: y=0.35x+2.27, r2=0.40***. Single,
double, and triple asterisks following r2 values denote regression model significance at the 0.05,
0.01, and 0.001 levels, respectively.

68

!
!

Figure 2.5. Shoot N accumulated in vetch (gray), rye (white), and vetch+rye (black) across
cover crop mixture proportions in 2010 and 2011. 2010 responses of vetch: y=0.92x,
r2=0.66***; rye: y=-0.00048x2+0.086x, r2=0.36***; and vetch+rye: y=-0.13x+17.83,
r2=0.67***. 2011 responses of vetch: y=-0.00066x2+0.17x, r2=0.66***; rye: y=0.00044x2+0.084x, r2=0.61***; and vetch+rye: y=-0.00097x2+0.032x+10.69, r2=0.58***.
Means (+/- SE). Single, double, and triple asterisks following r2 values denote regression model
significance at the 0.05, 0.01, and 0.001 levels, respectively.

69

!
!

Figure 2.6. Rye shoot N concentration across rye-vetch cover crop mixture proportions in 2010
and 2011. The slope of the response to rye-vetch proportion was equal in both years and
significant overall (analysis of covariance, P < 0.001). 2010 equation: y=0.0018x+0.96, r2=0.15.
2011 equation: y=0.0018x+1.31, r2=0.13. Means (+/- SE).

70

!
!

REFERENCES

71

!
!
REFERENCES

Aarssen, L.W., V. Ivan, and K.I.N. Jensen. 1986. The biology of Canadian weeds: 76. Vicia
angustifolia L., V. cracca L., V. sativa L., V. tetrasperma (L.) Schreb. and V. villosa
Roth. Can. J. Plant. Sci. 66:711–737.
Agricultural Marketing Service. 2013. National Organic Program. 7 CFR Section 205.
http://www.ecfr.gov (accessed 25 Sep. 2013)
Akemo, M.C., M.A. Bennett, and E.E. Regnier. 2000a. Tomato growth in spring-sown cover
crops. HortScience 35:843–848.
Akemo, M.C., E.E. Regnier, and M.A. Bennett. 2000b. Weed suppression in spring-sown rye
(Secale cereale)-pea (Pisum sativum) cover crop mixes. Weed Technol. 14:545–549.
Avci, M., and T. Akar. 2006. Ecological production of dryland hairy vetch by mechanical
control. Agron. Sustain. Dev. 26:29–34.
Barnes, J.P., and A.R. Putnam. 1986. Evidence for allelopathy by residues and aqueous extracts
of rye (Secale cereale). Weed Sci. 34:384-390.
Baskerville, G.L., and P. Emin. 1969. Rapid estimation of heat accumulation from maximum and
minimum temperatures. Ecology 50:514–517.
Benincasa, P., G. Tosti, F. Tei, and M. Guiducci. 2010. Actual N availability from winter catch
crops used for green manuring in maize cultivation. J. Sustain. Agric. 34:705–723.
Brainard, D., B. Henshaw, and S. Snapp. 2012. Hairy vetch varieties and bi-cultures influence
cover crop services in strip-tilled sweet corn. Agron. J 104:629–638.
Brainard, D.C., V. Kumar and R.R. Bellinder. 2011. Grass-legume mixtures and soil fertility
affect cover crop performance and weed seed production. Weed Technol. 25:473-479.
Boyd, N., E. Brennan, R. Smith, and R. Yokota. 2009. Effect of seeding rate and planting
arrangement on rye cover crop and weed growth. Agron. J. 101:47–51.
Clark, A. J., ed. 2007. Managing cover crops profitably. 3rd ed. Sustainable Agriculture
Network, Beltsville, MD.
Clark, A.J., A.M. Decker, and J.J. Meisinger. 1994. Seeding rate and kill date effects on hairy
vetch cereal rye cover crop mixtures for corn production. Agron. J. 86:1065–1070.
Clark, A.J., J.J. Meisinger, A.M. Decker, and F.R. Mulford. 2007a. Effects of a grass-selective
herbicide in a vetch-rye cover crop system on nitrogen management. Agron. J. 99:36–42.
Clark, A.J., J.J. Meisinger, A.M. Decker, and F.R. Mulford. 2007b. Effects of a grass-selective

72

!
!
herbicide in a vetch-rye cover crop system on corn grain yield and soil moisture. Agron.
J. 99:43–48.
Creamer, N.G., and S.M. Dabney. 2002. Killing cover crops mechanically: Review of recent
literature and assessment of new research results. Am. J. Alternative Agric. 17:32–40.
Davis, J.H.C., L. van Beuningen, M.V. Ortiz, and C. Pino. 1984. Effect of growth habit of beans
on tolerance to competition from maize when intercropped. Crop Sci. 24:751–755.
Dhima, K.V., A.S. Lithourgidis, I.B. Vasilakoglou, and C.A. Dordas. 2007. Competition indices
of common vetch and cereal intercrops in two seeding ratio. Field Crops Research 100:
249–256.
Ditsch, D.C., M.M. Alley, K.R. Kelley, and Y.Z. Lei. 1993. Effectiveness of winter rye for
accumulating residual fertilizer n following corn. J. Soil Water Conserv. 48:125–132.
Dordas, C.A., and A.S. Lithourgidis. 2011. Growth, yield and nitrogen performance of faba bean
intercrops with oat and triticale at varying seeding ratios. Grass Forage Sci. 66:569–577.
Economic Research Service. 2013. Fertilizer Use and Price. Table 7-Average U.S. farm prices of
selected fertilizers, 1960-2013.
http://www.ers.usda.gov/datafiles/Fertilizer_Use_and_Price/Fertilizer_Prices/table7.xls
(accessed 20 Aug. 2013)
Eaglesham, A.R.J., A. Ayanaba, V.R. Rao, and D.L. Eskew. 1981. Improving the nitrogen
nutrition of maize by intercropping with cowpea. Soil Biol. Biochem. 13:169–171.
Firbank, L.G., and A.R. Watkinson. 1985. On the analysis of competition within two-species
mixtures of plants. J. Appl. Ecol. 22:503–517.
Fukai, S., and B.R. Trenbath. 1993. Processes determining intercrop productivity and yields of
component crops. Field Crops Res. 34:247–271.
Hansen, E.M., H. Høgh-Jensen, and J. Djurhuus. 2002. Biological nitrogen fixation in a grazed
perennial grass/clover ley and correlation with herbage and soil variables. Eur. J. Agron.
16:309–320.
Hansen, J.P., and F.P. Vinther. 2001. Spatial variability of symbiotic N2 fixation in grass-white
clover pastures estimated by the 15N isotope dilution method and the natural 15N
abundance method. Plant Soil 230:257–266.
Hauggaard-Nielsen, H., P. Ambus, and E.S. Jensen. 2003. The comparison of nitrogen use and
leaching in sole cropped versus intercropped pea and barley. Nutr. Cycl. Agroecosys.
65:289–300.
Hayden, Z.D., D.C. Brainard, B. Henshaw, and M. Ngouajio. 2012. Winter annual weed
suppression in rye–vetch cover crop mixtures. Weed Technol. 26:818–825.

73

!
!
Hogberg, P. 1997. Tansley review No 95 - N-15 natural abundance in soil-plant systems. New
Phytol. 137:179–203.
Hunt, H.W., D.E. Reuss, and E.T. Elliott. 1999. Correcting estimates of root chemical
composition for soil contamination. Ecology 80:702–707.
Izaurralde, R.C., W.B. McGill, and N.G. Juma. 1992. Nitrogen fixation efficiency, interspecies N
transfer, and root growth in barley-field pea intercrop on a Black Chernozemic soil. Biol.
Fertil. Soils 13:11–16.
Jannink, J.L., L.C. Merrick, M. Liebman, E.A. Dyck, and S. Corson. 1997. Management and
winter hardiness of hairy vetch in Maine. Maine Agricultural and Forest Exp. Stat. Tech.
Bulletin no. 167., Orono.
Jensen, E.S. 1996. Grain yield, symbiotic N2 fixation and interspecific competition for inorganic
N in pea-barley intercrops. Plant Soil 182:25–38.
Jolliffe, P.A. 2000. The replacement series. J. Ecol. 88:371–385.
Karpenstein-Machan, M., and R. Stuelpnagel. 2000. Biomass yield and nitrogen fixation of
legumes monocropped and intercropped with rye and rotation effects on a subsequent
maize crop. Plant Soil 218:215–232.
Keating, B.A., and P.S. Carberry. 1993. Resource capture and use in intercropping: solar
radiation. Field Crops Res. 34:273–301.
Van Kessel, C., P.W. Singleton, and H.J. Hoben. 1985. Enhanced n-transfer from a soybean to
maize by vesicular arbuscular mycorrhizal (VAM) fungi. Plant Physiol. 79:562–563.
Kurdali, F., N. Sharabi, and A. Arslan. 1996. Rainfed vetch-barley mixed cropping in the Syrian
semi-arid conditions. Plant Soil 183:137–148.
Ledgard, S.F., and K.W. Steele. 1992. Biological nitrogen-fixation in mixed legume grass
pastures. Plant Soil 141:137–153.
Liebman, M., and E. Dyck. 1993. Crop rotation and intercropping strategies for weed
management. Ecol. Applic. 3:92–122.
Mariotti, A., J.C. Germon, P. Hubert, P. Kaiser, R. Letolle, A. Tardieux, and P. Tardieux. 1981.
Experimental determination of nitrogen kinetic isotope fractionation: Some principles;
illustration for the denitrification and nitrification processes. Plant Soil 62:413–430.
Mariotti, M., A. Masoni, L. Ercoli, and I. Arduini. 2009. Above- and below-ground competition
between barley, wheat, lupin and vetch in a cereal and legume intercropping system.
Grass Forage Sci. 64:401–412.
McCracken, D.V., M.S. Smith, J.H. Grove, C.T. Mackown, and R.L. Blevins. 1994. Nitrate
leaching as influenced by cover cropping and nitrogen source. Soil Sci. Soc. Am. J.

74

!
!
58:1476–1483.
Mennan, H., M. Ngouajio, D. Isık, and E. Kaya. 2009. Effects of alternative winter cover
cropping systems on weed suppression in organically grown tomato (Solanum
lycopersicum). Phytoparasitica 37:385–396.
Mischler, R., S.W. Duiker, W.S. Curran, and D. Wilson. 2010. Hairy vetch management for notill organic corn production. Agron. J. 102:355-362.
Mohler, C.L., and J.R. Teasdale. 1993. Response of weed emergence to rate of Vicia villosa Roth
and Secale cereale L. residue. Weed Res. 33:487–499.
Nuttonson, M.Y. 1958. Rye-climate relationships and the use of phenology in ascertaining the
thermal and photo-thermal requirements of rye. American Institute of Crop Ecology,
Washington, DC.
Ofori, F., and W.R. Stern. 1987. Cereal–legume intercropping systems. Adv. Agron. 41:41–90.
Parr, M., J.M. Grossman, S.C. Reberg-Horton, C. Brinton, and C. Crozier. 2011. Nitrogen
delivery from legume cover crops in no-till organic corn production. Agron. J. 103:1578–
1590.
Peachey, R.E., R.D. William, and C. Mallory-Smith. 2004. Effect of no-till or conventional
planting and cover crops residues on weed emergence in vegetable row crop. Weed
Technol. 18:1023–1030.
Ranells, N.N., and M.G. Wagger. 1996. Nitrogen release from grass and legume cover crop
monocultures and bicultures. Agron. J. 88:777–782.
Ranells, N.N., and M.G. Wagger. 1997. Winter annual grass-legume bicultures for efficient
nitrogen management in no-till corn. Agric. Ecosyst. Environ. 65:23–32.
Rochester, I., and M. Peoples. 2005. Growing vetches (Vicia villosa Roth) in irrigated cotton
systems: inputs of fixed N, N fertiliser savings and cotton productivity. Plant Soil
271:251–264.
Rogers, T.H., and D.G. Sturkie. 1939. Effect of fertilizers and method of their application on
nodulation, growth and nitrogen content of hairy vetch. Agron. J. 31:141–148.
Sainju, U.M., B.P. Singh, and W.F. Whitehead. 2000. Cover crops and nitrogen fertilization
effects on soil carbon and nitrogen and tomato yield. Can. J. Soil Sci. 80:523–532.
Sainju, U.M., W.F. Whitehead, and B.P. Singh. 2005. Biculture legume-cereal cover crops for
enhanced biomass yield and carbon and nitrogen. Agron. J. 97:1403–1412.
Shearer, G., and D.H. Kohl. 1986. N2-fixation in field settings: Estimations based on natural 15N
abundance. Funct. Plant Biol. 13:699–756.

75

!
!
Shipley, P.R., J.J. Meisinger, and A.M. Decker. 1992. Conserving residual corn fertilizer
nitrogen with winter cover crops. Agron. J. 84:869–876.
Smith, D. 1975. Forage management in the North. 4th ed. Kendall/Hunt Publ. Co., Dubuque, IA.
Snaydon, R.W. 1991. Replacement or additive designs for competition studies? J. Appl. Ecol.
28:930–946.
Streeter, J. 1988. Inhibition of legume nodule formation and N2 fixation by nitrate. Crit. Rev.
Plant Sci. 7:1–23.
Teasdale, J.R., and A.A. Abdul-Baki. 1998. Comparison of mixtures vs. monocultures of cover
crops for fresh-market tomato production with and without herbicide. HortScience
33:1163–1166.
Teasdale, J.R., and C.S.T. Daughtry. 1993. Weed suppression by live and desiccated hairy vetch
(Vicia villosa). Weed Sci. 41:207–212.
Teasdale, J.R., T.E. Devine, J.A. Mosjidis, R.R. Bellinder, and C.E. Beste. 2004. Growth and
development of hairy vetch cultivars in the northeastern United States as influenced by
planting and harvesting date. Agron. J. 96:1266–1271.
Tosti, G., P. Benincasa, M. Farneselli, R. Pace, F. Tei, M. Guiducci, and K. Thorup-Kristensen.
2012. Green manuring effect of pure and mixed barley – hairy vetch winter cover crops
on maize and processing tomato N nutrition. Eur. J. Agron. 43:136–146.
Tosti, G., P. Benincasa, and M. Guiducci. 2010. Competition and facilitation in hairy vetchbarley intercrops. Ital. J. Agron. 5:239–248.
Vandermeer, J.H. 1992. The ecology of intercropping. Cambridge Univ. Press, Cambridge.
White, R.H, A.D. Worsham, and U. Blum. 1989. Allelopathic potential of legume debris and
aqueous extracts. Weed Sci. 37:674–679.
Willey, R.W. 1979a. Intercropping - its importance and research needs. 1. Competition and yield
advantages. Field Crop Abstr. 32:1–10.
Willey, R.W. 1979b. Intercropping - its importance and research needs. 2. Agronomy and
research needs. Field Crop Abstr. 32:73–85.
de Wit, C.T., and J.P. Van Den Bergh. 1965. Competition between herbage plants. Neth. J.
Agric. Sci. 13:212–221.

76

!
!

CHAPTER III

RYE-VETCH COVER CROP RESIDUE QUALITY AND PLASTIC MULCH AFFECT SOIL
NITROGEN DYNAMICS, MICROBIAL COMMUNITIES, AND YIELDS IN VEGETABLE
PRODUCTION SYSTEMS

77

!
!
Rye-vetch cover crop residue quality and plastic mulch affect soil nitrogen dynamics,
microbial communities, and yields in vegetable production systems

Zachary D. Hayden*, Mathieu Ngouajio, and Daniel C. Brainard

Department of Horticulture, Michigan State University, Plant and Soil Sciences Building, 1066
Bogue St., East Lansing, MI 48824.
*Corresponding author (haydenza@msu.edu).

Funding for this research came from the Agriculture and Food Research Initiative (AFRI)
Competitive Grant no. 2012-67011-19716 from the USDA National Institute of Food and
Agriculture (NIFA), from the USDA-NIFA Sustainable Agriculture Research and Education
(SARE) program (Project no. GNC09-108), from MSU AgBioResearch Project GREEEN
(Project no. GR09-068), and from The Ceres Trust Organic Research Initiative. Additional
support came from the C. S. Mott Group for Sustainable Food Systems and The Graduate School
at Michigan State University. We would like to thank Drey Clark, Dr. Ajay Nair, Damen Kurzer,
Ben Henshaw, Erin Haramoto, and Carolyn Lowry for assistance in the field.

78

!
!
ABSTRACT
Cover crops can be an important source of nitrogen (N) for organic vegetable production,
but N availability depends heavily on both the characteristics of the cover crop residues and crop
management practices following cover crop termination. A study was conducted over two
seasons (2009-2010 and 2010-2011) in Michigan to investigate 1) how the proportions of cereal
rye (Secale cereale L.) and hairy vetch (Vicia villosa Roth) sown in a winter cover crop mixture
influence cover crop residue quantity and quality in the spring, and 2) how rye-vetch residue
characteristics and the use of black polyethylene mulch (PM) interact to affect soil N dynamics,
microbial communities, and yields of bell pepper (Capsicum annuum L.) and slicing cucumber
(Cucumis sativus L.) the following summer. Cover crop main plot treatments consisted of a
gradient of 7 sown proportions of rye:vetch, from 100% rye (94 kg/ha) to 100% vetch (42 kg/ha),
in addition to a no cover crop control. Subplots consisted of pepper and cucumber grown either
with or without PM. In both years, cover crop stands with higher sown proportions of vetch
generally contained more total N and had lower total residue C:N. Marketable vegetable yields,
soil inorganic N levels, and pore water nitrate concentrations below the root zone during the
summers were positively correlated with the total N content of cover crop mixtures. Yields and
soil N levels were generally higher when PM was used, but the differences between PM and bare
ground treatments were much less pronounced in 2011 than in 2010. Differences in the
magnitude and timing of summer precipitation events likely contributed to the discrepancy
between the two years, with potentially greater N leached from bare ground treatments in 2010
than in 2011. Lower microbial biomass was observed under PM compared to bare ground 3 wk
after plastic application, but cover crop and mulching treatments did not explain the majority of
variation observed in patterns of carbon substrate utilization across soil microbial communities

79

!
!
from Biolog-EcoplateTM data. Our results demonstrate that PM can be an important tool for
preserving N fertility benefits derived from high N cover crop residues, particularly on sandy
soils. While higher proportions of vetch in the cover crop mixtures and the use of PM generally
resulted in higher marketable vegetable yields, system performance must ultimately be based on
a balanced consideration of all the services provided, their costs, and their relationship to primary
production goals.

INTRODUCTION
The potential benefits of cereal rye (Secale cereale L.) and the legume hairy vetch (Vicia
villosa Roth) as winter annual cover crop species have been well documented in both agronomic
and vegetable cropping systems. Rye seed is inexpensive, and it possesses superior capabilities
in suppressing weeds, conserving residual soil N, and controlling erosion (Shipley et al., 1992;
Ditsch et al., 1993; Clark, 2007; Hayden et al., 2012). However, with increasing maturity, rye
residues can cause net N immobilization and yield loss in subsequent crops (Clark et al., 2007).
In contrast, the defining strengths of vetch are its considerable capacity for fixed N accumulation
and the rapid mineralization of that residue N following cover crop termination (Ranells and
Wagger, 1996; Teasdale et al., 2004). While outcomes vary depending on management and
environment, rye-vetch mixtures can combine these strengths while benefitting from potential
grass-legume functional complementarities (Ofori and Stern, 1987; Clark et al., 1994; Teasdale
and Abdul-Baki, 1998; Sainju et al., 2005). However, in most cases, cereal-legume cover crop
characteristics reflect relative species composition (Akemo et al., 2000; Karpenstein-Machan and
Stuelpnagel, 2000; Tosti et al., 2010). In Hayden et al. (2014), we demonstrated that relative
seeding rates can accurately govern the final proportions of rye and vetch in cover crop mixture

80

!
!
stands, and that changing the relative proportions of the species resulted in tradeoffs between
total fixed N accumulation, seed costs, and winter annual weed suppression. Understanding how
differences in total residue characteristics across the mixture proportions influence the soil-crop
system following residue incorporation is also essential in evaluating overall mixture
performance.
The availability of N from cover crop residues is governed by complex interactions
between soil biological activity, residue characteristics, and environment. Temperature,
moisture, soil texture, whether residues are incorporated or left as a surface mulch, and the
timing and method of cover crop termination are all factors that can influence residue
mineralization (Creamer and Dabney, 2002; Reberg-Horton et al., 2012). In addition, losses of
N through ammonia volatilization, denitrification, weed uptake, and perhaps most notably,
leaching, can reduce N availability for subsequent crops. Net mineralization from high N
legume residues typically occurs rapidly and out of sync with crop demand, resulting in excess
soil inorganic N early in the season that is susceptible to loss (Palm et al., 2001). In addition to
reducing the potential for N tie-up from cereal residues, the moderation of total residue quality in
cereal-legume mixtures may alter patterns of net N mineralization in ways that can reduce risks
for early-season N losses and improve synchrony with crop demand relative to a monoculture
legume (Kuo and Sainju, 1998; Clark et al., 2007; Crews and Peoples, 2005). Few studies have
evaluated the effects of cereal-legume mixture proportions on N dynamics and crop yields in a
field setting (Tosti et al., 2012), and results may vary in intensive vegetable production systems
for crops with differing durations and nutrient requirements, like bell pepper (Capsicum annuum
L.) and slicing cucumber (Cucumis sativus L.).
The use of black polyethylene (plastic) mulch is an industry standard for commercial

81

!
!
organic and conventional production of many warm-season vegetable crops in cooler climates.
Depending on color and chemical composition, plastic mulches can distinctly alter the soil and
surface microclimate in the crop row, while providing a barrier to weed emergence, precipitation,
soil splash, and the spread of soil borne diseases (Kasirajan and Ngouajio, 2012). Under dripirrigated black plastic, higher soil temperatures, more consistent soil moisture conditions, and
protection from leaching precipitation can have substantial effects on N mineralization and loss
from organic residues (Clarkson, 1960; Tarara, 2000; Schonbeck and Evanylo, 1998). However,
concerns over the economic and environmental costs associated with plastic mulch disposal, and
challenges with the performance and price of biodegradable alternatives, have discouraged the
use of plastic mulch in some agricultural sectors (Kyrikou and Briassoulis, 2007). Evaluating the
effects of cover crop residue quality on N dynamics and vegetable production both with and
without plastic mulch can provide insights into the influence of mulching on rye-vetch mixture
performance while producing results relevant to both production systems.
The size, activity, and functional characteristics of soil microbial communities are
important determinants of many agroecosystem processes, particularly residue decomposition
and nutrient cycling (Paul, 2007). Although innate soil characteristics and climate often have
greater effects, agricultural management practices—such as tillage, compost additions, and cover
cropping—can result in changes to microbial communities in both the short- and long-term
(Insam, 1990; Bossio et al., 1998; Nair and Ngouajio, 2012). Plastic films, sometimes in
combination with organic additions, are employed in practices such as soil solarization
(Katan,1981) and anaerobic disinfestation (Noriaki, 2008) to create extreme temperatures and/or
anoxic conditions for a short period with the goal of reducing populations of soil-borne pests and
pathogens. Under these conditions, plastic mulches reduce total microbial biomass and can

82

!
!
influence community composition (Scopa and Dumontet, 2007; Noriaki, 2008). However, few
studies (Moreno and Moreno, 2008; Carrera et al., 2007) have evaluated the influence on
microbial community function of the more modest but continuous increases in soil temperature,
and changes to aeration and moisture conditions (Khan et al., 2000), that result from standard
black plastic mulching during crop production, particularly in combination with residues of
varying quality.
The objectives of this research were 1) to investigate how the proportions of cereal rye and
hairy vetch sown in a mixture influence cover crop residue quantity and quality in the spring, and
2) how the characteristics of incorporated rye-vetch residues and the use of black plastic mulch
interact to affect vegetable yields and fruit quality, soil inorganic N dynamics, and soil microbial
communities in bell pepper (Capsicum annuum L.) and slicing cucumber (Cucumis sativus)
production systems.

MATERIALS AND METHODS
Site Description
The study was conducted over two seasons (2009-2010 and 2010-2011), alternating
between adjacent fields on a level Spinks loamy sand soil (sandy, mixed, mesic Lamellic
Hapludalf) at the Michigan State University Horticulture Teaching and Research Center in Holt,
Michigan (42°40” N, 84°28” W). A summer cover crop of sorghum sudangrass (Sorghum
bicolor x S. bicolor var. Sudanense) was flail mowed and incorporated into each field with a
rototiller at least 2 wk prior to seeding rye and vetch in the fall. The fields had been in organic
transition with a history of warm-season vegetable production under conventional tillage since
2008. Initial soil chemical characteristics for the fields averaged pH 6.6; CEC 7.1 cmol kg-1; 903

83

!
!
ppm Ca; and P (Bray-P1), K, and Mg levels of 133, 145, and 47 mg kg-1, respectively.
Production practices during the experiment complied with National Organic Program guidelines
(Agricultural Marketing Service, 2013). No fertilizers or soil amendments were applied in either
year.
Treatment and Experimental Design
The treatment design was a 3-way factorial, with the factors including cover crop mixture,
cash crop grown, and plastic mulch use. The experiment was arranged as a split-split plot RCBD
with 4 replications, where the whole plot factor was rye-vetch winter cover crop treatment.
Levels followed a proportional replacement series design (Jolliffe, 2000; Hayden et al., 2014),
and included a gradient of the following sown proportions of rye:vetch (relative to their
monoculture seeding rates): 100:0, 83:17, 67:33, 50:50, 33:67, 17:83, 0:100, in addition to a no
cover crop control (0:0). The rye and vetch monocultures (100:0 and 0:100) were sown at 94
and 42 kg ha-1, respectively. Following cover crop kill and incorporation in the spring, bell
pepper and slicing cucumber were grown during the summer (as subplots) in each main plot,
both on black plastic mulch and bare ground (sub-subplots).
Field Management
Table 1 summarizes the dates of key field activities and data collection during the study.
Rye and vetch seed was broadcast sown by hand and incorporated to a depth of about 5 cm using
a field cultivator. Whole plots were 6.7 by 8.5 m in 2009-2010 and 6.1 by 7.6 m in 2010-2011.
Organic-certified variety not stated (VNS) vetch seed grown in Oregon and VNS rye seed grown
in Minnesota were used in both years (Albert Lea Seed House, Albert Lea, MN). Vetch seed
was inoculated with N-DURE Rhizobium leguminosarum inoculant (INTX Microbials LLC,
Kentland, IN) at a rate of approximately 10 g inoculant kg-1 seed.

84

!
!
Cover crop stands were flail mowed and incorporated using a rototiller to a depth of about
15 cm in the spring, taking care to drive tractors slow enough to minimize residue movement. In
2010, residue incorporation was delayed by 1 wk following mowing due to heavy rainfall and
subsequent wet soil conditions.
Approximately 2 wk after residue incorporation, fields were rototilled again prior to bed
preparation. For plastic beds, black polyethylene mulch (1.25 mil thickness; Mid South
Extrusion Inc., Monroe, LA) was laid on 75 cm wide raised beds (15 cm height) with a single
drip irrigation line (30 cm emitter spacing) using a combined plastic mulch layer and bed shaper.
Bare ground beds were similarly drip irrigated, but in keeping with general grower practices for
un-mulched systems on sandy soils, beds were not raised. Plastic and bare ground beds within
subplots were spaced on 1.22 m centers, while adjacent beds in separate main or subplots were
on 1.83 m centers. All beds were at least 91 cm inside any main plot (cover crop treatment)
border.
Bell pepper (var. ‘Paladin’) and slicing cucumber (var. ‘Cobra’) were grown from organiccertified seed (when available) in a greenhouse over the course of 9 and 2 wk prior to field
transplanting, respectively. Pepper and cucumber transplants were grown in 98- and 72-cell plug
trays, respectively, using a media composed of 50% peat moss, 25% vermiculite, and 25% plantbased compost by volume. Fish emulsion was applied for supplemental fertility as needed
during transplant production. Transplants were hardened outdoors for 1 wk in a lath house prior
to being transplanted into the field.
Peppers were grown in staggered double rows (46 cm in-row plant spacing and 30 cm
inter-row spacing) with each treatment bed containing 20 data plants. Cucumbers were grown in
single rows (46 cm plant spacing) with each treatment bed containing 12 data plants. All beds

85

!
!
were drip irrigated simultaneously on a weekly basis and as needed, receiving approximately 2.5
cm per week during the summer. Crops were not fertilized, relying solely on N fertility derived
from soil organic matter and incorporated cover crop residues.
Bare ground beds were maintained weed-free during the summer through hoeing every 2-3
wk within and between rows, and transplant holes in plastic beds were hand-weeded at those
times as well. Preventative sprays of Bacillus thuringiensis (Dipel® DF, Valent BioSciences
Corp., Libertyville, IL) were applied to peppers to limit damage from cutworms (family
Noctuidae) and tomato hornworm (Manduca quinquemaculata), and pyrethrin (Pyganic®,
McLaughlin Gormley King Company, Minneapolis, MN) applications were made to cucumbers
to manage cucumber beetles (Diabrotica spp. and Acalymma spp.) prior to harvest.
Data Collection and Analysis
Cover Crop Residue Quantity and Quality
Rye and vetch shoot biomass was sampled in the spring from four 25- by 50-cm (0.125 m–
2)

quadrats in each main plot, dried to a constant weight at 38 ºC before measuring dry weights,

and then ground to pass through a 1-mm screen. Subsamples were submitted to Midwest
Laboratories (Omaha, NE) for analysis of total C and N concentrations using a Leco TruSpec
elemental analyzer (Leco Corp., St. Joseph, MI). Vetch tissue dry weight and nutrient
concentration data were corrected to account for soil adherence to vetch shoot samples in 2010,
the details of which are discussed in Hayden et al. (2014). More extensive analysis of the
influence of rye-vetch mixture proportions on cover crop stand establishment, biomass
productivity, N fixation, and winter annual weed suppression are also presented.
Vegetable Crop Yields and Fruit Quality
Bell peppers were picked five times at approximately 1 wk intervals in both years. Fruit

86

!
!
were counted and weighed after being categorized as either marketable (including U.S. Fancy,
U.S. No. 1, and U.S. No. 2 grades) or non-marketable (deformed or damaged by sunscald,
disease, decay, or insects) according to USDA standards (Agricultural Marketing Service, 2005).
Cucumbers were picked nine times at approximately 3 d intervals in both years, and were
similarly counted, weighed, and categorized. Marketable cucumber fruit included U.S. Fancy,
U.S. Extra No. 1, U.S. No. 1, U.S. No. 1 Small, U.S. No. 1 Large, and U.S. No. 2 grades, while
non-marketable fruit were deformed or damaged by cuts, scarring, disease, decay, or insects
(Agricultural Marketing Service, 1958). Yields from individual harvests were summed to obtain
total pepper and cucumber yields for each season. In the event that one or more data plants in a
bed died at random prior to harvest but more than 2 wk after transplanting (6% of pepper beds
lost a maximum of one plant and 27% of cucumber beds lost a maximum of up to two plants),
total yield values were adjusted to a per 20- or 12- plant basis for pepper and cucumber,
respectively. The average weight of marketable fruit was calculated by dividing total marketable
fruit weights by total marketable fruit counts, and the percentage non-marketable yield by weight
was obtained by dividing the total weight of non-marketable fruit by the combined weight of
total marketable and non-marketable fruit. Pepper leaf absorbance readings were also taken at
the time of flowering using a SPAD-502 Chlorophyll Meter (SPAD; Minolta, Osaka, Japan) as
an estimate of chlorophyll content and plant N status. Readings were taken consistently on
mature lower leaves that were axilary to the first branching of the plant stem.
Soil Inorganic Nitrogen Dynamics
Composite soil samples were collected from each whole plot prior to cover crop seeding in
the fall and 1 wk after cover crop incorporation in the spring. Beginning 1 wk after vegetable
bed preparation and plastic mulch application, composite soil samples (12 cores per bed) were

87

!
!
taken from pepper beds every 2 wk during the growing season through mid-August. Cucumber
beds were not sampled in the interest of time and cost of analysis. All soil samples were taken to
a depth of 15 cm using a soil probe and stored at 4 ºC until analyzed. Soil was collected from
beneath plastic mulch by cutting small slits prior to inserting the probe, resulting in only
superficial damage to the mulch in terms of soil coverage, even by the end of the season. Postcover-crop-incorporation soil samples were analyzed for inorganic N content. Soils were dried
at 38 ºC for 36 h, 10 g dry soil were extracted in 50 ml 1M KCl, and extracts were analyzed for
NO3- and NH4+ concentrations using a Lachat injection flow autoanalyzer (Lachat QuickChem,
Hach Company, Loveland, CO).
Suction lysimeters were also installed in pepper beds as a way to estimate relative
differences in nitrate leaching potential during the summer among plastic mulch and bare ground
management following select cover crop treatments (100 V: 0 R, 50 V: 50 R, 0 V: 100 R, and
0:0). Each lysimeter consisted of a porous ceramic cup (4.8 cm outside diameter and 5.8 cm
height) at the base of a 91 cm long PVC tube sealed at the top with a one-hole rubber stopper
(Model 1900, SoilMoisture Equipment Corp., Santa Barabara, CA). Lysimeters were installed
within the crop row roughly 15 cm from the drip line and with the porous cup at a depth of 0.6 m
from the bed surface. Installation involved auguring a hole to the desired depth, inserting the
lysimeter into a slurry of silica flour to surround the porous cup and improve hydraulic contact
with the surrounding soil, adding bentonite clay pellets to form a seal around the PVC tube
above the ceramic cup to limit preferential flow down the tube into the sampling area, and
backfilling the remaining height around the lysimeter with reserved soil. Pore water samples
were collected weekly during the summer beginning 1 wk after plastic mulch application. To
collect the samples, a 35 cb vacuum was applied to each lysimeter using a vacuum hand pump

88

!
!
generally 1 d following an irrigation event. The total volume of accumulated pore water was
then removed and measured 18 h after vacuum application, and subsamples were frozen at -20
ºC until analysis for NO3- concentration on a Lachat injection flow autoanalyzer.
Soil Microbial Communities
Subsamples of soil collected 3 wk following bed preparation (21 and 20 June in 2010 and
2011, respectively) from plastic mulch and bare ground pepper beds following select cover crop
treatments (100 V: 0 R, 50 V: 50 R, 0 V: 100 R, and 0:0) were sieved (4 mm) and used for
microbial analyses. Soil microbial biomass carbon (MBC) was estimated by the chloroform
fumigation-incubation (CFI) method (Jenkinson and Paulson, 1976; Paul et al., 1999). For each
treatment replicate, 20 g of soil were weighed into six 50-ml beakers, three of which were
fumigated with CHCl3 (stabilized with non-polar hydrocarbons) for 24 h, and three of which
served as non-fumigated controls. Soils in beakers were then re-inoculated with approximately 1
g of soil from their respective treatments and adjusted to 55% water holding capacity. Each
beaker was incubated at 23 ºC for 10 d in a 1 L air-tight mason jar with a rubber septum in the
lid. Headspace samples where extracted with a 1 ml syringe following incubation and analyzed
for CO2 concentration on an infrared gas analyzer (Qubit Model S-151, Qubit Systems Inc.,
Kingston, ON, Canada). Following Horwath et al. (1996), MBC was calculated as:
[1] MBC = 1.73CF - 0.56CC
where CF and CC are the evolved CO2-carbon from fumigated and non-fumigated control
beakers per g dry soil, respectively (average of three lab replicates). Soil respiration was taken
as CC, and metabolic quotient (qCO2) was calculated by dividing MBC by soil respiration.
Differences in the functional diversity of soil microbial communities were evaluated based
on patterns of substrate utilization by populations cultured on Biolog EcoPlatesTM (Biolog Inc.,

89

!
!
Hayward, CA). Each 96-well plate contains 31 distinct carbon sources and a control well (water)
replicated three times. Following a procedure adapted from Garland and Mills (1991), 10 g
field-moist soil from each treatment field replicate sample were shaken in 90 ml of a sterilized
saline solution (0.85% NaCl, w/v) for 60 min, pre-incubated for 18 h at 23 ºC, and then brought
to a 10-3 final dilution before adding 150 µl aliquots to each of the 96 wells (1 plate per treatment
field replicate). Carbon source utilization is indicated by color change (clear to purple) due to
the reduction of tetrazolium dye in the wells by respiratory CO2. Plates were incubated at 23 C
for 7 d, during which color development in each well was measured at 24 h intervals as
absorbance at 590 nm using a microplate reader (Model 680, Bio-Rad Laboratories, Hercules,
CA). The overall degree of substrate utilization was expressed as average well color
development (AWCD), calculated as the mean difference between carbon source absorbances
(Ri) and the absorbance reading for the control well (C) within plate replicates (Garland and
Mills, 1991):
[2]

AWCD =

∑R − C
i

31

Patterns of substrate utilization were investigated after first dividing the control-corrected
absorbances (Ri - C) by AWCD within plate replicates to reduce potential bias due to differences
in inoculum density among samples (Garland and Mills, 1991). Carbon source responses were
then averaged based on categorization of the substrates into 6 chemical classes (Preston-Mafham
et al., 2002) prior to principal component analysis.
Statistical Analysis
For cover crop residue variables, the fixed effects of year, cover crop mixture proportion,
and their interaction were analyzed using mixed model ANOVA with the MIXED procedure of
SAS (Version 9.2, SAS Institute, Cary, NC), where block (nested in year) was included as a
90

!
!
random factor. Where the effects of cover crop mixture proportion on total residue N content
and C:N were significant, the nature of each response within years was investigated using
polynomial or linear regression with the REG procedure of SAS. For vegetable yield and quality
variables and all univariately-analyzed microbial community variables, the fixed effects of year,
cover crop treatment, plastic mulch use, and all interactions were analyzed as a split-plot design
using mixed model ANOVA with the MIXED procedure of SAS. Block(year) and
block(year)*cover crop treatment were included as random factors in the model. Where
mulch*cover crop and higher order interactions were not significant (α = 0.05), only main effects
and marginal means of mulch use and cover crop treatment within years are presented (Tables 3,
4, and 5). Total soil inorganic N and pore water NO3- concentrations during the summer were
analyzed using repeated measures mixed model ANOVA with the MIXED procedure of SAS.
Model structure was similar to above, but with years analyzed separately due to model
complexity, and sample date included as the repeated factor. Due to the even temporal spacing
of the measurements, a first-order autoregressive variance-covariance structure was used. For all
of the above analyses, residuals were inspected to evaluate consistency with ANOVA
assumptions of normality and equality of variances, and data transformations (natural log or
square root) and unequal variance models were employed when necessary. Effects were judged
significant when P < 0.05.
Principal components analysis (PCA) was conducted on BiologTM data using the
correlation matrix of the six substrate class absorbance variables. Data for each year were
analyzed separately, and plate replicates were considered subsamples and averaged prior to
analysis. The first two principal component (PC) axes accounted for a majority of the total
variance in both years, and PC scores labeled by treatments were plotted on these axes to

91

!
!
evaluate patterns of substrate utilization. Correlations (loadings) between the PC axes and the
original substrate class variables were plotted as vectors, with correlations > |0.50| considered
significant.

RESULTS
Weather Conditions
Daily air temperature and rainfall data from April through September during the two years
of the experiment are presented in Figure 1. Spring temperatures during cover crop growth were
cooler in 2011 than in 2010. However, during the period of vegetable production, average
temperatures were similar in the two years (21.7 and 21.5 ºC in 2010 and 2011, respectively),
with 1284 and 1232 growing degree days (GDD) base 10 ºC accumulating between pepper
transplanting and last harvest in the two seasons. While cumulative rainfall during that period
was higher in 2011 than in 2010 (204 and 286 mm in 2010 and 2011, respectively), 56% of the
summer 2011 precipitation fell over a period of just 3 d near the start of pepper harvests. In
contrast, during the first month following plastic application, fields received 106 mm of rainfall
in 2010, compared with only 44 mm in 2011.
Rye-vetch Mixture Residues
Rye and vetch shoot dry weights were highly correlated with the proportion of each species
sown in mixture (Table 2). Biomass production was highest for each species when grown in
monoculture, with vetch yielding 551 and 305 g m-2 in 2010 and 2011, respectively, and rye
yielding 415 and 330 g m-2 (Table 2). In 2010, total shoot biomass (vetch+rye) in mixtures were
equivalent or intermediate to yields in the two monocultures, while in 2011, three mixtures (83
V: 17R, 67V: 33R, and 33V: 17R) produced significantly greater total biomass than either

92

!
!
monoculture (Table 2).
The shifts in cereal-legume composition across mixtures influenced the quality of the
combined cover crop residues. Total residue N content generally increased with greater
proportions of vetch in mixture (Figure 2). The increase was linear in 2010, when vetch
monoculture residues contained 175 kg N ha-1. In 2011, increases in total N leveled off for cover
crops containing greater than 50% vetch, due in part to reduced vetch biomass productivity and
greater rye tissue N concentrations that year (Hayden et al., 2014). Correspondingly, total
residue C:N increased with greater proportions of rye in mixture in both years, but still remained
below 25:1 for all cover crop treatments except the rye monoculture.
Vegetable Yields and Fruit Quality
Marketable yields (weight) of both bell pepper and slicing cucumber were generally higher
following cover crop mixtures that contained greater proportions of vetch (Figure 3). Yield
increases were predominately linear in 2011, while in 2010, the relative gains diminished or
ceased for mixtures containing more than 33 or 50% vetch. With the exception of treatments
containing greater than 83% rye, cover crops generally provided a significant yield benefit
relative to the no-cover-crop control. However, for vegetables grown on bare ground in 2010,
yield benefits even from high vetch mixtures were comparably modest and not statistically
significant. Overall, yields were higher for peppers and cucumbers grown on plastic mulch than
on bare ground, but the magnitude of the yield benefits differed between the two study years. In
2010, plastic mulch yields were 155% and 152% higher than bare ground on average (P<0.001)
for pepper and cucumber, respectively, but only 77% (P<0.001) and 9% (P = 0.04) higher in
2011.
Differences in the number of marketable pepper (Table 3) and cucumber (Table 4) fruit

93

!
!
produced largely mirrored the trends observed with marketable weights. The use of plastic
mulch and cover crops containing greater proportions of vetch were associated with modest
increases in average pepper and cucumber fruit weights in some cases (Tables 3 and 4). While
cover crop treatment had a significant effect in 2011 on the percentage of pepper yields that were
non-marketable, it had no significant influence on cucumber non-marketable percentage in either
year (Tables 3 and 4). Furthermore, plastic mulch use reduced pepper non-marketable
percentage from 9.2 to 4.7% in 2010 and from 19.6 to 15.3% in 2011 (Table 3). In contrast,
cucumber non-marketable percentages actually increased when grown on plastic, from 15.6 to
20.9% in 2010 and from 19.1 to 22.5% in 2011 (Table 4).
Soil Inorganic Nitrogen
Average soil inorganic N availability during the summer in pepper beds was highly
correlated with both marketable yields (Figure 4a) and SPAD chlorophyll content readings
(Figure 4b) over the course of the study. Cover crops containing greater proportions of vetch
generally resulted in higher soil inorganic N concentrations, with levels for rye monocultures
remaining less than or equivalent to the no-cover-crop control (Figure 5). Soil N levels increased
during the first several weeks after residue incorporation and then fell as pepper uptake increased
in late June. No significant differences in the timing of N release could be detected among the
varying cover crop residues (Figure 5). Plastic mulch was overall associated with higher soil N
levels, but the effect varied considerably between years. In 2010, the average inorganic N
concentration during the summer under bare ground management was 56% less than under
plastic mulch (Figure 5 a, b; P<0.001). In contrast, average bare ground soil N was only 19%
less than plastic mulch in 2011 (P<0.001), with no significant differences between the two
practices on 3 of the 6 sampling dates following plastic application (June 6, July 6, and July 18;

94

!
!
Figure 5 c, d), compared with no dates in 2010 (Figure 5 a, b).
Lysimeter Pore Water Nitrate Concentrations
Similar to concentrations of total inorganic N in the top 15 cm of soil, the concentration of
nitrate in pore water extracted from 0.6 m depth with lysimeters was generally higher following
cover crop residues containing greater proportions of vetch (Figure 6). Nitrate concentrations
following rye monocultures were less than or equivalent to no-cover-crop controls. Under
plastic mulch, the changes in pore water nitrate over the course of the summer displayed a
pattern comparable to N availability in the overlying soil—increasing early in the season and
then decreasing with greater pepper uptake later (Figure 6 a, b). Under bare ground, pore water
nitrate concentrations showed influence from precipitation events, particularly in 2011 (Figure 6
b, d). As with soil N levels, plastic mulch had higher pore water nitrate concentrations at depth
than bare ground. The effect varied between years as well, but in a manner opposite to that
observed with the upper layer soil N. In 2010, average bare ground pore water nitrate
concentrations were 10% less than plastic mulch (Figure 6 a, b; P<0.001), while in 2011, nitrate
concentrations were 24% lower under bare ground—56% lower if sampling dates following the
3 d, 160 mm rainfall event in late July are excluded (Figure 6 c, d; P<0.001).
Microbial Communities
The incorporation of rye-vetch cover crops resulted in equivalent or greater MBC, soil
respiration, and metabolic quotient than the no-cover-crop control on the dates microbial
communities were sampled in 2010 and 2011 (Table 5). Where differences among the cover
crop treatments were significant, values tended to be higher following the rye monoculture or the
rye-vetch mixture than following the vetch monoculture. In both years, MBC and soil
respiration were higher for soils under bare ground management than under plastic mulch (Table

95

!
!
5). However, the metabolic quotient was not influenced by the mulching practice.
Principal component analyses of data from Biolog EcoplatesTM revealed variation in
microbial community substrate utilization across soils sampled during the experiment (Figure 7).
The percentage of total variance captured on PC 1 was 39 and 38% in 2010 and 2011,
respectively, and PC 2 represented 21 and 24% in the two years. In both years, PC 1 was
positively correlated with utilization of substrates classified as polymers and miscellaneous,
while negatively correlated with carboxylic acids. The predominant loadings for PC 2 were
carbohydrates (r=-0.67) and amines/amides (r=0.73) in 2010, but included carbohydrates
(r=0.89) and amino acids (r=-0.79) in 2011 (Figure 7 b, d). Despite variation across samples,
considerable overlap in principal component scores was observed among cover crop treatments
and between plastic mulch and bare ground management, with no clear distinctions in microbial
community substrate utilization resolved in either year. However, microbial communities in nocover-crop control soils exhibited greater variability than cover crop treatments along PC 2 in
2010 (Figure 7a) and along PC 1 in 2011 (Figure 7c).

DISCUSSION
Rye-Vetch Residue Effects on N Dynamics and Vegetable Yields
The availability of N from cover crop biomass depends in large part on the characteristics
of the residues. While the prevalence of specific classes of compounds (i.e., cellulose,
hemicellulose, lignins, polyphenols) can have important impacts on N mineralization (Crews and
Peoples, 2005), first-season N availability is broadly related to total residue N and C content.
Laboratory incubation studies suggest that the percentage of total N in incorporated vetch and
vetch-cereal residues that becomes available as inorganic N in the first season is likely at

96

!
!
maximum 50-60%, and decreases with increasing total residue C:N in mixtures (Vigil and
Kissel, 1991; Honeycutt et al., 1993; Garrett, 2009), with residues that have total C:N > 25
potentially resulting in short term net N immobilization (Allison, 1966). Our results generally
conformed to these principles, with rye-vetch residues that contained more total N and had lower
C:N leading to higher early-season soil inorganic N levels, and inorganic N following rye
monocultures remaining less than or equivalent to the background levels found in the no-covercrop control. Not surprisingly, increasing inorganic N levels in the plow layer also translated
into higher pore water nitrate concentrations below the root zone (Figure 6). Similarly,
Rosecrance et al. (2000) and Tosti et al. (2014) found higher N leaching from vetch residues than
from rye or barley alone or in mixture with vetch, and Campiglia et al. (2010) observed a linear
relationship between total N content of vetch and vetch-oat residues and the amount leached
under the cover crops when used as mulches during pepper production.
Mineralization from high N cover crop residues occurs rapidly following incorporation,
typically resulting in an excess of soil inorganic N vulnerable to leaching in the weeks prior to
peak warm-season crop demand (Ranells and Wagger, 1996). It is often proposed that the
combination of contrasting residue qualities in cereal-legume mixtures may alter patterns of N
release to mitigate this asynchrony (Crews and Peoples, 2005). Without information on pepper
N uptake during the growing season, data on soil inorganic N concentrations during the summer
within the pepper rows alone are insufficient for resolving differences in temporal patterns in net
N mineralization following the various rye-vetch mixtures. Nevertheless, rapid mineralization
leading to excess early-season N was evident following all cover crop treatments in this study
that exhibited net N mineralization (Figure 5). Research suggests that combinations of residues
of contrasting quality most often exhibit patterns of net N mineralization that are more or less the

97

!
!
weighted average of the patterns exhibited by the high and low quality residues individually,
rather than resulting in release that is notably more synchronous with crop demand (Palm et al.,
2001). Over the course of a 30 wk laboratory incubation, Kuo and Sainju (1998) tracked net N
mineralization from mulched rye-vetch residue combinations along a gradient from 100% vetch
to 100% rye. While early season mineralization rates varied within a narrow range between the
two extremes, the cumulative amount of N mineralized followed the total N content in the
mixtures. In replacement-style mixtures (where a proportion of the high N component is
replaced with an equivalent proportion of the low N component), the resulting reductions in total
N and lower percent availability in the first season will likely have much greater relevance for
warm-season crop N nutrition than the magnitude of any potential changes in patterns of net N
mineralization. However, in the event that species interactions or management result in a ryevetch mixture accumulating significantly greater total N than a vetch monoculture (Sainju et al.,
2005), the relative reduction in net N mineralization could have the benefit of increasing soil N
retention without limiting potential N fertility (Lawson et al., 2013).
The correlations between pepper marketable yields, SPAD readings, and average soil
inorganic N support that N fertility was a key driver of treatment effects on vegetable yields
(Figure 4). In 2011, both pepper and cucumber exhibited predominately linear increases in
marketable yield following cover crops containing greater proportions of vetch. Despite cover
crop residues containing more total N than in 2011, relative yield gains in 2010 diminished with
increasing vetch proportion. A number of possibilities could account for this difference.
Nitrogen requirements may have been largely met by levels following the 50:50 rye:vetch
mixture within the context of the 2010 field, or another limiting factor not measured may have
prevented yield gains with increasing N availability that year. Alternatively, differences in cover

98

!
!
crop management between the timing of termination and plastic application in the two years
could have led to greater N loss from high vetch residues in 2010. Cover crops were mowed and
incorporated 2 wk before plastic application in 2011, while in 2010, cover crops were mowed 3
wk before plastic application and the mowed residues were left on the soil surface during the first
wk after termination because heavy rainfall prevented tillage. Preferential N leaching,
denitrification (Rosecrance et al., 2000), or ammonia volatilization (Quemada and Cabrera,
1995; Crews and Peoples, 2004) from higher vetch cover crop residues during this extended time
could have reduced the relative total season N availability from these treatments. However, the
close association between vetch proportion and soil inorganic N levels still observed early in the
2010 season (Figure 5a) would appear to contradict this explanation.
Plastic Mulch Effects
The higher soil inorganic N concentrations observed under plastic, as well as the close
correspondence between relative yield benefits and differences in N levels between plastic and
bare ground treatments during the two study years, suggest that N fertility was an important
factor behind the yield increases observed with plastic mulch. In regions with cooler summers,
black polyethylene mulch has the potential to promote greater N mineralization by warming the
soil, and when paired with drip irrigation, by maintaining more consistent soil moisture
conditions (Tarara, 2000; Agehara and Warncke, 2005). The process of soil redistribution in
forming raised beds in a plastic mulch system may also serve to concentrate incorporated
residues from a wider area into the crop root zone, providing proportionally greater cover-cropderived N fertility localized beneath the crop relative to a flat bed. In addition, plastic mulch
protects inorganic N in the crop root zone from leaching during precipitation events, which can
be considerable on sandy soils like those present at the experimental site (Clarkson, 1960).

99

!
!
Differences in the amount and timing of precipitation in 2010 and 2011 may account for the
differences in N availability and yields between plastic mulched and bare ground pepper beds in
the two years. In 2010, twice as much precipitation fell in the first month after plastic
application than in 2011, likely leading to relatively greater leaching from bare ground beds that
year. Indeed, a substantial drop in soil N concentrations occurred only in bare ground beds
following 56 mm of rain in the 6 d after plastic application in 2010, and levels remained lower
than plastic mulch for the rest of the summer. Correspondingly, pore water N concentrations
below bare ground beds were relatively higher in 2010 than in 2011.
Interestingly, pore water nitrate concentrations at depth were generally equivalent or higher
under plastic mulch than bare ground management. This indicates that there was still
considerable potential for leaching losses from plastic beds under our drip irrigation regime. In
this study, irrigation was applied uniformly across the fields to meet the requirements of crops
grown on bare ground, but a lower rate for plastic mulch beds may have been able to satisfy crop
demands while reducing the depth of irrigation (Quemada et al., 2013). The potential for
leaching due to precipitation events is clearly greater in the absence of plastic mulch, as evinced
by nitrate concentrations under bare ground following heavy rains in 2011 (Figure 6d).
However, data from suction lysimeters provide an incomplete picture of total leaching losses.
Even with accurate estimates of drainage through the soil profile, suction lysimeters generally
sample resident pore water solute concentrations rather than flux concentrations (Brandi-Dohrn
et al., 1996), with the potential to miss high concentrations of N leached below the sampling
depth prior to pore water extraction. As a result, suction lysimeters tend to underestimate
quantities of leached N compared with other methods, including drainage lysimeters (Zotarelli et
al., 2007). This bias should be more pronounced for soils exposed to heavy rain events that

100

!
!
infiltrate below the root zone, suggesting our data may underestimate the N leaching potential of
bare ground relative to plastic mulch management.
Even when N is not limiting, the temperature and moisture microclimate effects of black
plastic mulch can promote more rapid growth and development of warm-season vegetable crops
leading to higher and earlier yields (Kasirajan and Ngouajio, 2012), although reaching
temperatures above crop-specific thresholds can be detrimental (Tarara, 2000). Plastic mulches
can also influence yield quality by promoting cleaner fruit and reducing the incidence of
soilborne disease, in addition to quality benefits related to water and fertility (Kasirajan and
Ngouajio, 2012). The lower percentage of non-marketable pepper fruit with plastic mulch and
with greater proportions of vetch in cover crop mixtures suggest that quality improvement in
pepper was related to N availability. However, the relationship may have been indirect—solar
injury (sunscald) was the predominant pepper fruit defect during the study, and shading from
fuller plant canopies under higher N fertility may have reduced incidence (Roberts and
Anderson, 1994). The increase in non-marketable percentage for cucumbers grown on plastic in
both years was surprising, particularly for a crop whose fruit rest on the soil surface. The lack of
cover crop effect on cucumber quality suggests that differences were not driven by N fertility.
The predominant cucumber defects were deformed or misshapen fruit and excessive scarring
from cucumber beetle damage. Surface temperature highs may be elevated by up to 20 ºC on
black plastic mulch relative to bare ground (Tarara, 2000), and we speculate that these higher
temperatures may have contributed to more incomplete pollination in cucumbers, either by
directly harming pollen growth or survival or by discouraging pollinators, leading to higher
incidence of deformed fruit (Maestro and Alvarez, 1988; Swiader and Ware, 2002). The less
extreme temperature increases typical >2.5 cm above the surface of black plastic may also have

101

!
!
promoted greater reproduction and/or activity of cucumber beetles (Radin and Drummond, 1994;
Tarara, 2000).
Responses of Soil Microbial Communities
Soil microbes are typically carbon limited, and the addition of organic residues generally
prompts an increase in microbial activity and growth proportional to the quantity of labile C
added, which slows as the materials are decomposed over time (Jackson, 2000; Paul, 2007). The
tendency towards higher MBC and soil respiration following rye or rye-vetch cover crops,
particularly in 2011, is most likely related to larger total C additions. The metabolic quotient is a
measure of the amount of C respired per unit microbial biomass, or the efficiency of microbial C
utilization. Higher metabolic quotients are often considered an indication of stress, suggesting
that more carbon is being allocated to energy and maintenance functions (and thus respired)
rather than contributing to growth (Anderson and Domsch, 1993), although this interpretation
has been questioned (Wardle and Ghani, 1995). Larger metabolic quotients following rye or ryevetch mixtures than vetch alone may reflect a greater degree of N limitation for microbes relative
to the quantity of carbon added (Wardle and Ghani, 1995). Nair and Ngouajio (2012) saw a
similar response to increasing carbon additions with incorporation of rye and vetch cover crops
and compost.
Overall, plastic mulch exhibited higher MBC than bare ground, which likely accounted for
the correspondingly higher soil respirations also observed (Table 5). Moreno and Moreno (2008)
also found reduced indicators of microbial biomass following black polyethylene mulch, which
they suggested may be a result of inhibition due to higher temperatures. Alternatively, one might
expect that moderate soil warming under plastic would lead to increased microbial activity and
population growth (Nicolardot et al., 1994), and the lack of differences in metabolic quotient

102

!
!
between plastic and bare ground treatments in our study do not support that microbes were under
particular stress under plastic. Li et al. (2004) documented higher MBC with plastic mulching
relative to bare ground, although differences varied depending on the time after plastic
application in one year. Another possible explanation is that conditions under plastic mulching
may alter the temporal dynamics of microbial growth, such that comparisons of MBC with bare
ground may vary depending on the date of sampling.
We hypothesized that incorporation of cover crop residues of differing quality and the
fertility and microclimate effects related to plastic mulch use would result in shifts in the
functional diversity of microorganisms present. While variation in substrate utilization by the
communities on Biolog EcoplatesTM was evident, distinct treatment effects were not resolved by
our analysis (Figure 7). By contrast, previous studies have documented changes in microbial
community composition in the first season as a result of cover cropping and plastic mulching
(Carrera et al., 2007; Wu et al., 2009; Nair and Ngouajio, 2012). However, results from Buyer et
al. (2010) comparing in part a vetch residue mulch with black and white plastic mulches during
tomato production suggest that organic inputs from cover cropping are more important than
management induced variations in temperature and moisture for controlling microbial
community structure. Utilizing fatty acid and DNA marker techniques, distinctions in
community composition have been identified specifically between soils following rye and vetch
cover cropping as well (Buyer et al., 2010; Maul et al., 2014). The greater variability we
observed for substrate utilization in no-cover-crop treatments in both years was an interesting
finding. Those treatments also tended to have the lowest MBC, so one possible explanation for
the greater variability is that there was a higher probability of losing rare species of functional
significance during dilution of the no-cover-crop relative to the cover crop treatments (Preston-

103

!
!
Mafham et al., 2002).
The BiologTM assay is a culture-based technique, and has been demonstrated to select for
certain components of the microbial community over others—particularly favoring rapidlygrowing (r-strategist) species (Ros et al., 2008). In addition, due to the great diversity of
microbes in the soil, communities can exhibit considerable functional redundancy (Strickland et
al., 2009), which may further limit the ability of substrate utilization approaches to resolve
differences in microbial communities even if structural changes are present, particularly in the
short term. For example, Carrera et al., (2007) found relatively weaker cover cropping effects
with substrate utilization data than with analysis of phospholipid fatty acids. These results and
our data support that potential changes in microbial community composition resulting from rye
and vetch cover crop incorporation or mulching with black plastic film in the first season (short
term) may have limited significance for actual agroecosystem functioning.
Management Implications
Rye-vetch cover crops containing greater proportions of vetch can be expected to result in
greater soil N availability following incorporation, and when N is limiting, to lead to higher
yields of bell pepper and slicing cucumber. Along the gradient of rye-vetch proportions
evaluated in this study, changes in total N content appeared to have a larger influence on
vegetable yields than did potential changes in patterns of N mineralization. However, Hayden et
al. (2014) demonstrated that greater proportions of vetch also resulted in increased seed costs and
tradeoffs with other agroecosystem services, such as winter annual weed suppression. The
optimal rye-vetch mixture proportion will ultimately depend on cover crop objectives, crop
management practices, and site-specific characteristics of the field.
Regardless of rye-vetch proportion, net N that mineralized from mixtures accumulated in

104

!
!
the soil prior to peak vegetable demand, leaving it vulnerable to leaching from early-season
precipitation, particularly on the sandy soils used in this study. Our results demonstrate that
plastic mulch can be an important tool for preserving N fertility benefits from high N cover crop
residues, contributing substantially to the yield increases observed relative to bare ground
management. Where mulch film is used following cover crop incorporation, it should be applied
as soon as residue conditions allow to maximize N preservation. Although the benefits of plastic
mulch must be weighed against the costs associated with its use and disposal, the importance of
protecting soils against N leaching will likely grow over time due to predicted increases in
extreme precipitation events in many regions (Hayhoe et al., 2006).

105

!
!

APPENDIX

106

!
!
Table 3.1. Dates of key field activities and data collection.
Activity
2009-2010
Cover crop seeding
1 September
Cover crop biomass sampling
10 May
Cover crop termination (by flail mower)
10 May
Cover crop incorporation (by rototiller)
17 May
Plastic mulch application
1 June
Bell pepper transplanting
3 June
Slicing cucumber transplanting
3 June
Suction lysimeter installation
8 June
Bell pepper SPAD† readings
12-13 July
Slicing cucumber harvests (9)
12 July - 2 August
Bell pepper harvests (5)
27 July - 31 August
† Minolta SPAD-502 Chlorophyll Meter

107

2010-2011
1 September
14 May
17 May
17 May
31 May
2 June
6 June
7 June
15 July
18 July - 5 August
27 July - 29 August

!
!
Table 3.2. Shoot residue dry weight of vetch, rye, and vetch+rye (total) across cover
crop mixture proportions in 2010 and 2011.
Cover Crop
Treatment
2010 Season
100 V : 0 R
83 V : 17 R
67 V : 33 R
50 V : 50 R
33 V : 67 R
17 V : 83 R
0 V : 100 R
Prob > F‡
2011 Season
100 V : 0 R
83 V : 17 R
67 V : 33 R
50 V : 50 R
33 V : 67 R
17 V : 83 R
0 V : 100 R
Prob > F‡

Vetch

Rye
g m-2

Total
.

550.9 (64.9)†
400.5 (67.5)
379.5 (27.2)
270.6 (68.2)
199.2 (22.6)
117.8 (16.1)
--

-217.7 (65.0)
216.2 (29.1)
235.4 (46.2)
349.4 (14.9)
387 (26.2)
414.5 (58.8)

550.9 (64.9)
618.2 (51.3)
595.7 (23.0)
506.0 (113.2)
548.6 (32.9)
504.8 (22.9)
414.5 (58.8)

< 0.001

0.002

0.30

305.0 (22.7)
294.4 (16.8)
274.9 (27.5)
156.3 (32.0)
165.8 (34.6)
90.8 (20.6)
--

-128.3 (12.5)
167.0 (11.5)
200.7 (12.3)
266.3 (32.4)
267.8 (13.0)
330.2 (31.0)

305.0 (22.7)
422.7 (20.3)
441.9 (34.4)
357.0 (35.5)
432.1 (53.8)
358.5 (20.8)
330.2 (31.0)

< 0.001

< 0.001

0.01

† Treatment mean (standard error)
‡ Significance of fixed effect of cover crop mixture proportion within years (P value).

108

!
!
Table 3.3. Bell pepper total number of marketable fruit harvested, average weight of marketable
fruit, and percentage non-marketable yield by weight when grown either on black plastic mulch
or bare ground (no plastic) following 7 rye-vetch cover crop mixture proportions or a no cover
crop control in 2010 and 2011.
Treatment
Number of
Average Weight
Percentage NonMarketable Fruit
of Marketable
Marketable Weight
Fruit
2010
2011
2010
2011
2010
2011
-1
number 20-plants
g
%
Mulch Main Effect†
No plastic
28 (3)‡
54 (4)
184 (2) 185 (2)
9.2 (1.1) 19.6 (1.7)
Plastic Mulch
68 (4)
73 (4)
190 (2) 190 (2)
4.7 (0.5) 15.3 (1.1)
Prob > F§
Cover Crop Main Effect†
Control
100 V : 0 R
83 V : 17 R
67 V : 33 R
50 V : 50 R
33 V : 67 R
17 V : 83 R
0 V : 100 R
Prob > F§

< 0.001

< 0.001

< 0.001 < 0.001

0.006

0.009

38 (8)
57 (12)
62 (9)
52 (10)
50 (10)
56 (11)
44 (8)
25 (7)

39 (6)
86 (8)
75 (10)
83 (6)
55 (6)
65 (8)
51 (6)
51 (5)

182 (2)
191 (3)
196 (3)
186 (2)
186 (4)
186 (3)
189 (3)
181 (4)

173 (4)
197 (4)
197 (3)
196 (3)
186 (4)
189 (6)
184 (1)
178 (4)

8.3 (1.5)
6.0 (1.7)
4.4 (0.7)
6.8 (2.2)
4.3 (1.0)
5.6 (1.2)
7.2 (1.7)
13.1 (3.2)

22.2 (2.7)
12.4 (1.4)
17.1 (4.6)
13.0 (1.0)
15.8 (2.0)
18.3 (4.0)
20.9 (1.8)
19.8 (2.7)

0.03

< 0.001

0.17

< 0.001

0.21

< 0.001

† Main effects and marginal means are presented because mulch*cover crop and higher order
interactions were not significant at the 0.05 probability level.
‡ Marginal mean (standard error)
§ Significance of fixed main effects within years (P value).

109

!
!
Table 3.4. Slicing cucumber total number of marketable fruit harvested, average weight of
marketable fruit, and percentage non-marketable yield by weight when grown either on black
plastic mulch or bare ground (no plastic) following 7 rye-vetch cover crop mixture proportions or
a no cover crop control in 2010 and 2011.
Treatment
Number of
Average Weight
Percentage NonMarketable Fruit
of Marketable
Marketable Weight
Fruit
2010
2011
2010
2011
2010
2011
-1
number 20-plants
g
%
Mulch Main Effect†
No plastic
29 (2)‡
71 (4)
241 (4) 283 (2)
15.6 (1.6) 19.1 (0.8)
Plastic Mulch
66 (4)
76 (4)
268 (2) 290 (2)
20.9 (1.4) 22.5 (0.7)
Prob > F§
Cover Crop Main Effect†
Control
100 V : 0 R
83 V : 17 R
67 V : 33 R
50 V : 50 R
33 V : 67 R
17 V : 83 R
0 V : 100 R
Prob > F§

< 0.001

0.10

< 0.001

0.07

0.01

< 0.001

32 (7)
59 (8)
60 (9)
55 (9)
57 (10)
49 (9)
41 (7)
26 (7)

50 (4)
97 (6)
90 (5)
92 (5)
66 (7)
80 (8)
60 (4)
55 (3)

251 (7)
256 (7)
253 (11)
265 (7)
254 (7)
253 (7)
254 (6)
252 (11)

283 (3)
284 (3)
292 (4)
291 (2)
281 (6)
292 (3)
286 (5)
283 (6)

22.3 (3.8)
17.5 (2.1)
15.5 (2.5)
15.6 (2.4)
16.7 (2.9)
16.4 (2.5)
16.8 (2.7)
25.0 (4.9)

21.2 (1.3)
21.5 (1.3)
19.5 (1.4)
22.4 (1.8)
19.0 (1.3)
19.8 (2.7)
20.8 (1.7)
22.0 (1.3)

< 0.001

< 0.001

0.23

0.27

0.33

0.86

† Main effects and marginal means are presented because mulch*cover crop and higher order
interactions were not significant at the 0.05 probability level.
‡ Marginal mean (standard error)
§ Significance of fixed main effects within years (P value).

110

!
!
Table 3.5. Microbial biomass carbon (MBC), soil respiration, and metabolic quotient in soils
sampled from pepper rows grown either on black plastic mulch or bare ground (no plastic)
following select rye-vetch cover crop mixture proportions or a no cover crop control. Soils were
sampled on 21 and 20 June in 2010 and 2011, respectively.
Treatment

MBC
2010
2011
ug CO2-C g-1 dry
soil

Soil Respiration
2010
2011
ug CO2-C g-1 dry
soil

Metabolic Quotient
2010
2011

Mulch Main
Effect†
No plastic
Plastic Mulch

105 (4)‡
95 (4)

118 (5)
105 (4)

67 (2)
63 (3)

76 (4)
69 (3)

0.65 (0.02)
0.66 (0.02)

0.65 (0.03)
0.66 (0.03)

Prob > F§

0.003

< 0.001

0.12

0.02

0.58

0.78

91 (6)
102 (5)
102 (6)
103 (9)

103 (7)
117 (5)
100 (7)
126 (6)

57 (4)
62 (4)
69 (4)
71 (5)

64 (5)
69 (4)
73 (3)
85 (4)

0.63 (0.02)
0.61 (0.03)
0.68 (0.03)
0.71 (0.03)

0.63 (0.04)
0.59 (0.03)
0.74 (0.04)
0.68 (0.03)

0.65

0.07

0.13

0.02

0.10

0.007

Cover Crop Main
Effect†
Control
100 V : 0 R
50 V : 50 R
0 V : 100 R
Prob > F§

† Main effects and marginal means are presented because mulch*cover crop and higher order
interactions were not significant at the 0.05 probability level.
‡ Marginal mean (standard error)
§ Significance of fixed main effects within years (P value).

111

!
!

Figure 3.1. Daily rainfall amounts (bars) and minimum, maximum, and average daily air
temperatures (lines) at the Michigan State University Horticulture Teaching and Research Center
(Holt, Michigan) from April through September in 2010 and 2011. The first and second dotted
lines denote the dates of cover crop termination and plastic mulch application, respectively, in
each year.

112

!
!

Figure 3.2. Total (vetch+rye) shoot residue N content (a) and C:N ratio (b) across cover crop
mixture proportions in 2010 (open circles) and 2011 (closed circles). The horizontal dotted line
references a C:N value of 25, a common estimate for the cutoff between expectations of net N
mineralization and immobilization in soils following residue incorporation. Means +/- SE. The
responses of total N content in 2010: y=-1.3x+178.3, r2=0.67, P<0.001; and 2011: y=0.0097x2+0.32x+106.9, r2=0.58, P<0.001. The responses of C:N in 2010: 0.0038x20.13x+13.73, r2=0.90, P<0.001; and 2011: 0.0019x2-0.01x+12.77, r2=0.85, P<0.001.

113

!
!

Figure 3.3. Total marketable yields of bell pepper (a, c) and slicing cucumber (b, d) when
grown either on black plastic mulch (black squares) or bare ground (white squares) following 7
rye-vetch cover crop mixture proportions or a no cover crop control (0:0) in 2010 (a, b) and 2011
(c, d). Means +/- SE.

114

!
!

Figure 3.4. Correlations between the average soil inorganic N concentration in bell pepper rows
during the summer and a) total bell pepper marketable yield and b) bell pepper SPAD meter leaf
absorbance readings (dimensionless units) for 2010 and 2011 data combined. Rye-vetch cover
crop mixture treatments are designated by grayscale gradient, while black plastic mulch and bare
ground treatments are differentiated by shape. Pearson correlation coefficients (R) and their
significance are indicated on the plots.

115

!
!

Figure 3.5. Total soil inorganic N concentrations during the summer in bell pepper rows grown
either on black plastic mulch (a, c) or bare ground (b, d) in 2010 (a, b) and 2011 (c, d) following
7 rye-vetch cover crop mixture proportions or a no cover crop control. Means +/- SE. For
clarity, error bars are only presented for select cover crop treatments (Control, 100 V: 0 R, 50 V:
50 R, and 0 V: 100 R).

116

!
!

Figure 3.6. NO3--N concentrations of pore water extracted during the summer from suction
lysimeters installed to a depth of 61 cm below bell pepper rows grown either on black plastic
mulch (a, c) or bare ground (b, d) in 2010 (a, b) and 2011 (c, d) following select rye-vetch cover
crop mixture proportions or a no-cover-crop control. Means +/- SE.

117

!
!

Figure 3.7. Principal component analysis of soil microbial substrate utilization data from
Biolog-EcoPlatesTM in 2010 (a, b) and 2011 (c, d). Component scores for black plastic mulch
(black) and bare ground (white) management following select rye-vetch cover crop mixture
treatments (shapes) are plotted for the first 2 principal components (a, c), as well as the
correlations (loadings) between the 2 principal components and the original 6 substrate class
variables (b, d). The percentages of the total variance accounted for by each principal
component in a given year are indicated in parentheses in axis titles.

118

!
!

REFERENCES

119

!
!
REFERENCES

Agehara, S., and D.D. Warncke. 2005. Soil moisture and temperature effects on nitrogen release
from organic nitrogen sources. Sci. Soc. Am. J. 69:1844-1855.
Agricultural Marketing Service. 2013. National Organic Program. 7 CFR Section 205.
http://www.ecfr.gov (accessed 25 Sep. 2013)
Agricultural Marketing Service. 1958. United States standards for grades of cucumbers.
http://www.ams.usda.gov/AMSv1.0/ams.fetchTemplateData.do?template=TemplateN&p
age=FreshMarketVegetableStandards (accessed 15 Aug. 2011).
Agricultural Marketing Service. 2005. United States standards for grades of sweet peppers.
http://www.ams.usda.gov/AMSv1.0/ams.fetchTemplateData.do?template=TemplateN&p
age=FreshMarketVegetableStandards (accessed 15 Aug. 2011).
Akemo, M.C., E.E. Regnier, and M.A. Bennett. 2000. Weed suppression in spring-sown rye
(Secale cereale)-pea (Pisum sativum) cover crop mixes. Weed Technol. 14:545–549.
Allison, F.E. 1966. The fate of nitrogen applied to soils. Adv. Agron. 18:219–258.
Anderson, T., and K.H. Domsch. 1993. The metabolic quotient for CO2 (qCO2) as a specific
activity parameter to assess the effects of environmental conditions, such as pH, on the
microbial biomass of forest soils. Soil Biol. Biochem. 25:393–395.
Bossio, D.A., K.M. Scow, N. Gunapala, and K.J. Graham. 1998. Determinants of soil microbial
communities: Effects of agricultural management, season, and soil type on phospholipid
fatty acid profiles. Microb. Ecol. 36:1–12.
BrandiDohrn, F., R. Dick, M. Hess, and J. Selker. 1996. Suction cup sampler bias in leaching
characterization of an undisturbed field soil. Water Resour. Res. 32:1173–1182.
Buyer, J.S., J.R. Teasdale, D.P. Roberts, I.A. Zasada, and J.E. Maul. 2010. Factors affecting soil
microbial community structure in tomato cropping systems. Soil Biol. Biochem. 42:831–
841.
Campiglia, E., R. Mancinelli, E. Radicetti, and S. Marinari. 2010. Legume cover crops and
mulches: effects on nitrate leaching and nitrogen input in a pepper crop (Capsicum
annuum L.). Nutr. Cycl. Agroecosys. 89:399–412.
Carrera, L.M., J.S. Buyer, B. Vinyard, A.A. Abdul-Baki, L.J. Sikora, and J.R. Teasdale. 2007.
Effects of cover crops, compost, and manure amendments on soil microbial community
structure in tomato production systems. Appl. Soil Ecol. 37:247–255.
Clark, A. J., ed. 2007. Managing cover crops profitably. 3rd ed. Sustainable Agriculture
Network, Beltsville, MD.

120

!
!
Clark, A.J., A.M. Decker, and J.J. Meisinger. 1994. Seeding rate and kill date effects on hairy
vetch cereal rye cover crop mixtures for corn production. Agron. J. 86:1065–1070.
Clark, A.J., J.J. Meisinger, A.M. Decker, and F.R. Mulford. 2007. Effects of a grass-selective
herbicide in a vetch-rye cover crop system on nitrogen management. Agron. J. 99:36–42.
Clarkson, V.A. 1960. Effect of black polyethylene mulch on soil and microclimate temperature
and nitrate level. Agron. J. 52:307–309.
Creamer, N.G., and S.M. Dabney. 2002. Killing cover crops mechanically: Review of recent
literature and assessment of new research results. Am. J. Alternative Agric. 17:32–40.
Crews, T.E., and M.B. Peoples. 2004. Legume versus fertilizer sources of nitrogen: ecological
tradeoffs and human needs. Agric. Ecosyst. Environ. 102:279–297.
Crews, T.E., and M.B. Peoples. 2005. Can the synchrony of nitrogen supply and crop demand be
improved in legume and fertilizer-based agroecosystems? A review. Nutr. Cycl. in
Agroecosys. 72:101–120.
Ditsch, D.C., M.M. Alley, K.R. Kelley, and Y.Z. Lei. 1993. Effectiveness of winter rye for
accumulating residual fertilizer n following corn. J. Soil Water Conserv. 48:125–132.
Garland, J.L., and A.L. Mills. 1991. Classification and characterization of heterotrophic
microbial communities on the basis of patterns of community-level sole-carbon-source
utilization. Appl. Environ. Microbiol. 57:2351–2359.
Garrett, A. 2009. Improving nitrogen management with cover crops in organic broccoli
production. M.S. thesis., Oregon State Univ., Corvallis, OR.
Hayden, Z.D., D.C. Brainard, B. Henshaw, and M. Ngouajio. 2012. Winter annual weed
suppression in rye–vetch cover crop mixtures. Weed Technol. 26:818–825.
Hayden, Z.D., M. Ngouajio, and D.C. Brainard. 2014. Rye-vetch mixture proportion tradeoffs:
Cover crop productivity, nitrogen accumulation, and weed suppression. Agron. J.
106:904–914.
Hayhoe, K., C.P. Wake, T.G. Huntington, L. Luo, M.D. Schwartz, J. Sheffield, E. Wood, B.
Anderson, J. Bradbury, A. DeGaetano, T.J. Troy, and D. Wolfe. 2006. Past and future
changes in climate and hydrological indicators in the US Northeast. Clim. Dynam.
28:381–407.
Honeycutt, C., L. Potaro, K. Avila, and W. Halteman. 1993. Residue quality, loading rate and
soil-temperature relations with hairy vetch (Vicia villosa Roth) residue carbon, nitrogen
and phosphorus mineralization. Biol. Agric. Hortic. 9:181–199.
Horwath, W., E. Paul, D. Harris, J. Norton, L. Jagger, and K. Horton. 1996. Defining a realistic
control for the chloroform fumigation-incubation method using microscopic counting and
C-14-substrates. Can. J. Soil Sci. 76:459–467.

121

!
!
Insam, H. 1990. Are the soil microbial biomass and basal respiration governed by the climatic
regime? Soil Biol. Biochem. 22:525–532.
Jackson, L.E. 2000. Fates and losses of nitrogen from a nitrogen-15-labeled cover crop in an
intensively managed vegetable system. Soil Sci. Soc. Am. J. 64:1404–1412.
Jenkinson, D.S., and D.S. Powlson. 1976. The effects of biocidal treatments on metabolism in
soil—V: A method for measuring soil biomass. Soil Biol. Biochem. 8:209–213.
Jolliffe, P.A. 2000. The replacement series. J. Ecol. 88:371–385.
Karpenstein-Machan, M., and R. Stuelpnagel. 2000. Biomass yield and nitrogen fixation of
legumes monocropped and intercropped with rye and rotation effects on a subsequent
maize crop. Plant Soil 218:215–232.
Kasirajan, S., and M. Ngouajio. 2012. Polyethylene and biodegradable mulches for agricultural
applications: A review. Agron. Sustain. Dev. 32:501–529.
Katan, J. 1981. Solar heating (solarization) of soil for control of soilborne pests. Annu. Rev.
Phytopathol. 19:211–236.
Khan, A.R., D. Chandra, S. Quraishi, and R.K. Sinha. 2000. Soil aeration under different soil
surface conditions. J. Agron. Crop Sci. 185:105–112.
Kyrikou, I., and D. Briassoulis. 2007. Biodegradation of agricultural plastic films: a critical
review. J. Polym. Environ. 15:125–150.
Lawson, A., A.M. Fortuna, C. Cogger, A. Bary, and T. Stubbs. 2013. Nitrogen contribution of
rye-hairy vetch cover crop mixtures to organically grown sweet corn. Renew. Agric.
Food Syst. 28:59–69.
Li, F., Q. Song, P.K. Jjemba, and Y. Shi. 2004. Dynamics of soil microbial biomass C and soil
fertility in cropland mulched with plastic film in a semiarid agro-ecosystem. Soil Biol.
Biochem. 36:1893–1902.
Maestro, M.C., and J. Alvarez. 1988. The effects of temperature on pollination and pollen tube
growth in muskmelon (Cucumis melo L.). Sci. Hortic. 36:173–181.
Maul, J.E., J.S. Buyer, R.M. Lehman, S. Culman, C.B. Blackwood, D.P. Roberts, I.A. Zasada,
and J.R. Teasdale. 2014. Microbial community structure and abundance in the
rhizosphere and bulk soil of a tomato cropping system that includes cover crops. Appl.
Soil Ecol. 77:42–50.
Moreno, M., and A. Moreno. 2008. Effect of different biodegradable and polyethylene mulches
on soil properties and production in a tomato crop. Sci. Hortic. 116:256–263.
Nair, A., and M. Ngouajio. 2012. Soil microbial biomass, functional microbial diversity, and
nematode community structure as affected by cover crops and compost in an organic

122

!
!
vegetable production system. Appl. Soil Ecol. 58:45–55.
Nicolardot, B., G. Fauvet, and D. Cheneby. 1994. Carbon and nitrogen cycling through soil
microbial biomass at various temperatures. Soil Biol. Biochem. 26:253–261.
Noriaki. 2008. Biological soil disinfestation (BSD) of soilborne pathogens and its possible
mechanisms. Jpn Agric. Res. Q. 42:7–12.
Ofori, F., and W.R. Stern. 1987. Cereal–legume intercropping systems. Adv. Agron. 41:41–90.
Palm, C.A., K.E. Giller, P.L. Mafongoya, and M. Swift. 2001. Management of organic matter in
the tropics: translating theory into practice. Nutr. Cycl. Agroecosys. 61:63–75.
Paul, E.A., ed. 2007. Soil microbiology, ecology and biochemistry. 3rd ed. Academic Press,
Burlington, MA.
Paul, E.A., D. Harris, M.J. Klug, and R.W. Ruess. 1999. The determination of microbial
biomass. In: G.P. Robertson, D.C. Coleman, C.S. Bledsoe, P. Sollins, eds., Standard soil
methods for long-term ecological research. Oxford University Press, New York. p. 291–
317
Preston-Mafham, J., L. Boddy, and P.F. Randerson. 2002. Analysis of microbial community
functional diversity using sole-carbon-source utilisation profiles: A critique. FEMS
Microbiol. Ecol. 42:1–14.
Quemada, M., M. Baranski, M.N.J. Nobel-de Lange, A. Vallejo, and J.M. Cooper. 2013. Metaanalysis of strategies to control nitrate leaching in irrigated agricultural systems and their
effects on crop yield. Agric. Ecosyst. Environ. 174:1–10.
Quemada, M., and M. Cabrera. 1995. Carbon and nitrogen mineralized from leaves and stems of
4 cover crops. Soil Sci. Soc. Am. J. 59:471–477.
Radin, A., and F. Drummond. 1994. Patterns of initial colonization of cucurbits, reproductive
activity, and dispersion of striped cucumber beetle, Acalymma-vittata (F.) (Coleoptera:
Chrysomelidae). J. Agric. Entomol. 11:115–123.
Ranells, N.N., and M.G. Wagger. 1996. Nitrogen release from grass and legume cover crop
monocultures and bicultures. Agron. J. 88:777–782.
Reberg-Horton, S.C., J.M. Grossman, T.S. Kornecki, A.D. Meijer, A.J. Price, G.T. Place, and
T.M. Webster. 2012. Utilizing cover crop mulches to reduce tillage in organic systems in
the southeastern USA. Renew. Agric. Food Syst. 27:41–48.
Roberts, B., and J. Anderson. 1994. Canopy shade and soil mulch affect yield and solar injury of
bell pepper. HortScience 29:258–260.
Ros, M., M. Goberna, J.A. Pascual, S. Klammer, and H. Insam. 2008. 16S rDNA analysis reveals
low microbial diversity in community level physiological profile assays. J. Microbiol.

123

!
!
Methods 72:221–226.
Rosecrance, R., G. McCarty, D. Shelton, and J. Teasdale. 2000. Denitrification and N
mineralization from hairy vetch (Vicia villosa Roth) and rye (Secale cereale L.) cover
crop monocultures and bicultures. Plant Soil 227:283–290.
Sainju, U.M., W.F. Whitehead, and B.P. Singh. 2005. Biculture legume-cereal cover crops for
enhanced biomass yield and carbon and nitrogen. Agron. J. 97:1403–1412.
Schonbeck, M.W., and G.K. Evanylo. 1998. Effects of mulches on soil properties and tomato
production II: Plant-available nitrogen, organic matter input, and tilth-related properties.
J. Sustain. Agric. 13:83–100.
Scopa, A., and S. Dumontet. 2007. Soil solarization: effects on soil microbiological parameters.
J. Plant Nutr. 30:537–547.
Shipley, P.R., J.J. Meisinger, and A.M. Decker. 1992. Conserving residual corn fertilizer
nitrogen with winter cover crops. Agron. J. 84:869–876.
Strickland, M.S., C. Lauber, N. Fierer, and M.A. Bradford. 2009. Testing the functional
significance of microbial community composition. Ecology 90:441–451.
Swiader, J.M., and G.W. Ware. 2002. Producing vegetable crops. 5th ed. Interstate Publishers,
Danville, IL.
Tarara, J.M. 2000. Microclimate modification with plastic mulch. HortScience 35:169–180.
Teasdale, J.R., and A.A. Abdul-Baki. 1998. Comparison of mixtures vs. monocultures of cover
crops for fresh-market tomato production with and without herbicide. HortScience
33:1163–1166.
Teasdale, J.R., T.E. Devine, J.A. Mosjidis, R.R. Bellinder, and C.E. Beste. 2004. Growth and
development of hairy vetch cultivars in the northeastern United States as influenced by
planting and harvesting date. Agron. J. 96:1266–1271.
Tosti, G., P. Benincasa, M. Farneselli, R. Pace, F. Tei, M. Guiducci, and K. Thorup-Kristensen.
2012. Green manuring effect of pure and mixed barley – hairy vetch winter cover crops
on maize and processing tomato N nutrition. Eur. J. Agron. 43:136–146.
Tosti, G., P. Benincasa, M. Farneselli, F. Tei, and M. Guiducci. 2014. Barley–hairy vetch
mixture as cover crop for green manuring and the mitigation of N leaching risk. Eur. J.
Agron. 54:34–39.
Tosti, G., P. Benincasa, and M. Guiducci. 2010. Competition and facilitation in hairy vetchbarley intercrops. Ital. J. Agron. 5:239–248.
Vigil, M.F., and D.E. Kissel. 1991. Equations for estimating the amount of nitrogen mineralized
from crop residues. Soil Sci. Soc. Am. J. 55:757–761.

124

!
!
Wardle, D.A., and A. Ghani. 1995. A critique of the microbial metabolic quotient (qCO2) as a
bioindicator of disturbance and ecosystem development. Soil Biol. Biochem. 27:1601–
1610.
Wu, M., L. Wu, L. Zhao, and M. Chen. 2009. Effects of continuous plastic film mulching on
paddy soil bacterial diversity. Acta Agric. Scand. 59:286–294.
Zotarelli, L., J.M. Scholberg, M.D. Dukes, and R. Muñoz-Carpena. 2007. Monitoring of nitrate
leaching in sandy soils. J. Environ. Qual. 36:953-962.!!

125

!
!

CHAPTER IV

PLANTING DATE AND STAGGERED SEEDING OF RYE-VETCH MIXTURES:
BIOMASS, NITROGEN, AND LEGUME WINTER SURVIVAL

126

!
!
Planting date and staggered seeding of rye-vetch mixtures: Biomass, nitrogen, and legume
winter survival

Zachary D. Hayden*, Mathieu Ngouajio, and Daniel C. Brainard

Department of Horticulture, Michigan State University, Plant and Soil Sciences Building, 1066
Bogue St., East Lansing, MI 48824.
*Corresponding author (haydenza@msu.edu).

The authors would like to thank Drey Clark, Dr. Ajay Nair, Damen Kurzer, Rebekah Struck, and
Ben Henshaw for their assistance in the field. Funding to support this study came from the
Agriculture and Food Research Initiative (AFRI) Competitive Grant No. 2012-67011-19716
from the U.S. Department of Agriculture National Institute of Food and Agriculture (USDANIFA), from the USDA-NIFA Sustainable Agriculture Research and Education (SARE) program
(project no. GNC09-108), from MSU AgBioResearch Project GREEEN (project no. GR09-068),
from The Ceres Trust Organic Research Initiative, and from the C. S. Mott Group for Sustainable
Food Systems and the Graduate School at Michigan State University.

127

!
!
ABSTRACT
Mixtures of cereal rye (Secale cereale L.) and the legume hairy vetch (Vicia villosa Roth)
are used to provide fixed N in balance with other important cover crop services, but late planting
and interference from rye can limit vetch productivity. This two-year study in Michigan
investigated how fall planting dates influence rye-vetch cover crop biomass quantity and quality
in the spring, and evaluated whether staggering (delaying) rye seeding could improve vetch
performance (biomass production, N2 fixation, and winter survival) in mixtures. Treatments
consisted of a two-way factorial of three vetch planting dates (late August, mid September, and
late September) and three lengths of rye seeding stagger (co-seeded, short stagger, and long
stagger). Later planting of co-seeded mixtures generally led to reduced total shoot biomass and
lower proportions of vetch, resulting in cover crop residues with less fixed N and a higher total
C:N. For earlier planting dates, delaying rye seeding until vetch emergence (short stagger)
increased vetch shoot biomass by 760 - 1,060 kg ha-1 (30-36 kg vetch N ha-1) relative to coseeding. Staggered seeding provided no benefit to vetch biomass at later planting dates, and
delaying rye seeding until the vetch 3-4 leaf stage (long stagger) resulted in significant
reductions in vetch winter survival compared with co-seeding. Additional research would be
needed to determine whether potential on-farm benefits of rye-vetch staggered seeding justify the
additional management complexity involved and possible tradeoffs with other cover crop
services.

INTRODUCTION
Winter annual cover crop mixtures composed of cereal rye and the legume hairy vetch
have been well-studied for their potential to accumulate significant fixed N (Ranells and Wagger,

128

!
!
1996) while providing more effective weed suppression (Teasdale and Abdul-Baki, 1998;
Hayden et al., 2012), erosion control (Clark, 2007), N scavenging (Shipley et al., 1992), and
greater biomass production (Clark et al., 1994; Sainju et al., 2005) at lower overall seed costs
(Hayden et al., 2014) than a vetch monoculture. Cereal-legume functional complementarity may
contribute to mixture advantages, including greater resource-use efficiency (Fukai and Trenbath,
1993), improved N availability for the cereal (Ledgard and Steele, 1992), and enhanced N
fixation (Streeter, 1988; Izaurralde et al., 1992), light interception (Keating and Carberry, 1993),
and winter survival (Jannink et al., 1997) for the legume. The moderation of total residue quality
in mixtures may also alter patterns of net N mineralization following cover crop termination,
alleviating problems with N tie-up associated with cereal cover crops while potentially
improving the synchrony of N release with crop demand compared to a monoculture legume
(Kuo and Sainju, 1998; Clark et al., 2007). However, interference between the components can
limit mixture benefits—most often resulting in reductions in legume performance (Ofori and
Stern, 1987).
Both management decisions and environmental conditions can have considerable influence
on species interactions and outcomes in cereal-legume mixtures. As winter annuals, rye and
vetch are sown in the late summer or early fall, and after overwintering, put on the majority of
their growth the following spring. The popularity of rye and vetch as cover crops can be
attributed in part to their cold hardiness, which makes production possible across a broad
geographic range, including the Northeast United States and the Great Lakes region (Clark,
2007). Fall planting dates influence the developmental stage reached before the onset of winter
conditions, with later planting reducing spring biomass production (Teasdale et al., 2004; Farsad
et al., 2011) and increasing the potential for overwinter mortality, particularly for vetch

129

!
!
(Brandsaeter et al., 2008). Rye can germinate, grow, and survive at lower temperatures than
vetch (Nuttonson, 1957; Teasdale et al., 2004), contributing to its ability to establish later in the
fall and accumulate biomass faster in the spring (Shipley et al., 1992). Later planting dates and
colder winter conditions are therefore likely to favor rye growth over that of vetch in mixtures.
However, the presence of rye may also improve vetch winter survival by reducing frost heaving
or buffering low temperature extremes through greater snow cover retention or reduced air
movement at the soil surface (Smith, 1975; Jannink et al., 1997). Despite a body of research on
these two species individually, little information exists on how planting dates influence rye and
vetch in mixture. Since inclement weather and the harvest of previous cash crops often delay
planting past recommended dates, such information is important for helping farmers understand
what they can expect from rye-vetch mixtures, particularly in terms of biomass production and N
fixation.
Owing in part to extensive root systems and high relative growth rates, cereals tend to be
stronger competitors than legumes (Mariotti et al., 2009). Suppression of the legume component
is a common challenge in cereal-legume mixtures, particularly under conditions of excess soil N,
moisture stress, or non-N nutrient limitation (Ofori and Stern, 1987; Jensen, 1996). Previous
studies have investigated seeding rates (e.g. Clark et al., 1994), termination dates (e.g. Clark et
al., 1997), and planting arrangements (e.g. Brennan et al., 2009) in rye-vetch and other cereallegume mixtures as means to manage species interactions for improved cover crop performance.
Relay intercropping, in which the seeding of mixture components is staggered temporally,
has also been used in cereal-legume cash crop and cash crop-cover crop combinations as a
strategy for moderating competition to improve yields of one or both components or to achieve
other system benefits (Francis et al., 1982; Ofori and Stern, 1987; Amosse et al., 2013). While

130

!
!
the concept is rarely considered for cover crop mixtures, staggered seeding of rye and vetch
could foster advantages of agronomic significance over the standard co-seeded practice. In
particular, delayed seeding of rye after vetch could reduce interference from rye during both fall
establishment and spring growth periods, with the potential to improve vetch establishment,
productivity, and N fixation in mixtures. Alternatively, less rye growth in the fall could
negatively impact vetch winter survival (Smith, 1975), and reductions in the uptake of
rhizosphere N by stagger-sown rye could limit possible improvements to vetch N fixation
efficiency compared with a co-seeded mixture (Streeter, 1988). Staggered seeding also entails
additional management complexity, including the labor, fuel, and soil compaction costs
associated with an added tractor pass. However, if the practice resulted in significant
improvements in the total amount of N fixed per vetch seed sown in mixtures while also
maintaining services more characteristic of rye (such as weed suppression, N scavenging, and
erosion control), staggered seeding of rye-vetch mixtures may provide a net benefit.
The objectives of this study were 1) to investigate the influence of fall planting date on
spring biomass quantity and quality in a co-seeded rye-vetch cover crop mixture, and 2) to
evaluate the interactive effects of staggered (delayed) rye seeding and planting date on cover
crop mixture outcomes, with particular emphasis on possible improvements in vetch biomass
production, N fixation, and winter survival.

MATERIALS AND METHODS
The study site was located at the Michigan State University (MSU) Horticulture Teaching
and Research Center in Holt, MI (42°40” N, 84°28” W) on a level Spinks loamy sand soil
(sandy, mixed, mesic Lamellic Hapludalf). Prior to the start of the experiment, the field had

131

!
!
been fallow for two summers with fall and spring rototilling to control weeds. Initial soil
chemical characteristics in fall 2009 included pH 6.4; CEC 5.3 cmol kg-1; 602 ppm Ca; and P
(Bray-P1), K, and Mg levels of 131, 85, and 101 mg kg-1, respectively. No fertilizers or soil
amendments were applied during the study.
The treatment design was a two-way factorial consisting of three vetch planting dates (late
August, mid September, and late September) and three lengths of rye seeding stagger (co-seeded,
short stagger, and long stagger), resulting in nine rye-vetch cover crop mixture treatments. Rye
and vetch were sown at the same time in co-seeded mixtures, whereas rye seeding was delayed
in staggered mixtures. Table 1 presents the dates of rye and vetch seeding for each mixture in
2009 and 2010. To ensure that rye was sown at a consistent vetch growth stage in each stagger
treatment, the delay in rye seeding was based on growing degree day (GDD) base 4°C
accumulation following vetch seeding. In short and long stagger treatments, rye was sown
approximately 120 and 230 GDD after vetch, respectively, corresponding to the timing of vetch
emergence and the appearance of the first runner at the base of vetch seedlings (3-4 leaf stage).
Rye-vetch mixture treatments were sown into permanent 1.5 x 1.5m experimental plots
arranged in a randomized complete block design with four replications. Adjacent plots were
separated by 0.6m wide paths. Rye and vetch seeds were planted by hand using custom-made
grids for each species to ensure uniform seed densities and spacings. Rye was sown at 266 seeds
m-2 and vetch at 82 seeds m-2 in all mixtures, approximating field seeding rates of 47 and 21 kg
ha-1 for rye and vetch, respectively. All seeds were planted to a depth of about 3 cm, and care
was taken not to disturb vetch seedlings during staggered rye seedings. In both years of the
study, variety not stated (VNS) vetch seed grown in Oregon and VNS rye seed grown in
Minnesota were used (Albert Lea Seed House, Albert Lea, MN). Vetch seed was inoculated

132

!
!
with N-DURE Rhizobium leguminosarum inoculant (INTX Microbials LLC, Kentland, IN) prior
to each seeding event at a rate of approximately 10 g inoculant kg-1 seed.
Rye and vetch densities and shoot biomass were sampled from two 25 x 50 cm quadrats
established in each plot, oriented such that they encompassed a consistent number of sown rye
and vetch seeds. Fall counts of rye and vetch seedlings within the quadrats were taken following
emergence of each species in a given treatment, and plant counts were taken again in the spring
from the same quadrats immediately prior to biomass harvest, on 14 and 17 May in 2010 and
2011, respectively. Potential differences in vetch winter survival were evaluated by calculating
the percentage change in vetch plant counts between fall and spring samplings. Rye and vetch
shoot biomass was harvested from quadrats by clipping at the soil surface, and dry weights were
obtained after drying to a consistent weight at 38 ºC. Cover crop biomass remaining in the plots
after sampling in 2010 was mowed with a line trimmer and incorporated into the soil using a
walk-behind rototiller. Weeds were managed during the summer with a line trimmer, and plots
were rototilled twice more at 2 wk and 1 d prior to sowing rye and vetch for the second year of
the study. Treatment locations were re-randomized.
Rye and vetch tissue samples were ground to pass through a 1mm screen, and submitted to
the University of California-Davis Stable Isotope Facility (SIF) for analysis of percentage total C
and N and δ15N using a PDZ Europa ANCA-GSL elemental analyzer interfaced to a PDZ
Europa 20-20 continuous flow isotope ratio mass spectrometer (Sercon Ltd., Cheshire, UK).
Cover crop shoot C and N content was calculated as the product of shoot dry weight and
percentage shoot C or N. Vetch N fixation was estimated using the 15N natural abundance
method, which relies on the slight enrichment of 15N generally present in soil N relative to
atmospheric N2 (Shearer and Kohl, 1986; Hogberg, 1997). Isotopic composition is expressed

133

!
!
relative to atmospheric N2 using δ15N values in parts per thousand (%0), with higher δ15N
indicating greater 15N enrichment. The percentage of vetch N derived from the atmosphere
(%Ndfa) was calculated from Eq. 1 (Rochester and Peoples, 2005),
[1]

%Ndfa = (δ15Nref – δ15Nleg)/( δ15Nref – B) × 100

where δ15Nref is the shoot δ15N value for rye in monoculture (used as the non-N-fixing reference
plant), δ15Nleg is the shoot δ15N value for vetch, and B is the estimated δ15N of vetch when
grown entirely dependent on atmospheric sources of N. Following Hansen et al. (2002) and
Hansen and Vinther (2001), the lowest δ15Nleg value in each study year was used to approximate
B, resulting in conservative estimates of vetch %Ndfa. Rye monoculture δ15Nref values were
2.08 and 1.44 %0 in 2010 and 2011, respectively.
Statistical Analysis and Data Presentation
In keeping with the factorial design of the experiment, the fixed effects of year, planting
date, stagger, and their interactions were analyzed through mixed model ANOVA using the
MIXED procedure of SAS, where block (replication) was included as a random factor. Unequal
variance models and/or square root transformations were used when necessary to satisfy
ANOVA assumptions of normality and equality of variances. Where slicing indicated
significant differences (P < 0.05) between levels of stagger within a given year and planting date,
means were separated using Fisher’s Protected LSD at the 0.05 probability level.
Differences in planting dates and weather conditions between the two years of the study
resulted in variable heat accumulation during cover crop growth in each treatment (Table 1),
with important implications for interpreting the results. Data for both years were therefore
combined in figures on the basis of total GDD base 4 ºC accumulation during the period of vetch
growth in mixtures, in order to visualize the effects of later fall planting and staggered seeding

134

!
!
within the context of seasonal temperature variability. With years combined, the nature of the
overall effect of heat accumulation (GDD, as influenced by planting date and weather) on cover
crop mixture outcomes within each level of stagger was investigated through linear or
polynomial regression using the REG procedure of SAS (Version 9.2, SAS Institute, Cary, NC).

RESULTS AND DISCUSSION
Weather and GDD Accumulation
Table 1 presents the influence of planting dates and lengths of rye stagger on the total GDD
accumulated during vetch and rye growth in the various mixture treatments during the 20092010 and 2010-2011 seasons. Within years, relative differences in heat accumulation among
treatments were driven by planting dates, while between years differences in heat accumulation
for a given treatment were influenced by both planting date and variable weather conditions.
Due to later planting dates and cooler temperatures, cover crops grown during the 2010-2011
season experienced 51 fewer GDD in the fall and 123 fewer GDD in the spring than those grown
during the 2009-2010 season. Furthermore, minimum winter temperatures were lower overall in
2010-2011 than in 2009-2010. The average and absolute minimum air temperatures,
respectively, during the coldest month (January) were -8.5 and -20.8 ºC in 2009-2010 and -11.2
and -22.7 ºC in 2010-2011. Final spring biomass production by rye and vetch cover crops is
highly correlated with GDD base 4 ºC accumulation during both the fall and spring (Teasdale et
al., 2004; Farsad et al., 2011). While colder winter conditions will tend to result in lower overall
heat accumulation, the magnitude and frequency of minimum temperature extremes, which are
not directly captured in the calculation of GDD, may be a more influential factor in cover crop
overwinter survival.

135

!
!
Effect of Planting Date on Co-seeded Mixture Biomass
Later fall planting dates and lower overall GDD accumulation were generally associated
with lower total shoot biomass of co-seeded rye-vetch mixtures by the time of cover crop harvest
in May, but the responses differed for the component species (Figure 1A). Vetch biomass
production decreased substantially in mixtures planted later in the fall, with reductions from
355.0 to 82.9 and 153.2 to 20.6 g m-2 between the late August and late September planting dates
in 2009-2010 and 2010-2011, respectively. In contrast, rye biomass production in later planted
mixtures increased or remained unchanged relative to earlier planting dates for all but the mid
September planting in 2010-2011. As a consequence, spring stands of mixtures planted later in
the fall contained increasingly greater proportions of rye over vetch biomass. The percentage of
total mixture biomass that was vetch decreased from 58 to 21% between the late August and late
September planting dates in 2009-2010, and from 42 to 12% between the same planting date
treatments in 2010-2011. This compositional change was accompanied by a corresponding
decrease in total N accumulated by the cover crops. While mixtures planted in late August
accumulated 16.8 and 8.7 g N m-2 in 2010 and 2011, respectively, delaying planting until late
September reduced total N by 55 and 57% in the two years (Figure 1B). Furthermore, the total
residue C:N ratio of the mixture stands exhibited a general increase with later planting and lower
GDD accumulation, ranging from 15.2 to 24.0 in 2010 and from 17.2 to 19.9 in 2011 (Figure
1C).
These results are consistent with physiological and growth response differences between
rye and vetch. While reduced heat accumulation in the fall or spring will limit the biomass
production of both species (Guldan and Martin, 2003; Teasdale et al., 2004; Farsad et al., 2011),
rye productivity is generally more robust in the face of delayed planting or colder temperatures

136

!
!
than vetch (Shipley et al., 1992). The increases in rye biomass observed in later-planted
mixtures, particularly in 2010, were likely a result of reduced competition from the legume.
The reduced overall productivity and shifts in biomass composition that accompanied
delayed planting of rye-vetch mixtures have important implications for the green manure benefits
that can be expected from the cover crop following termination. Not only do potential fixed N
additions decline in tandem with reduced vetch productivity, but the increasing biomass C:N of
later planted mixtures may negatively impact the first season availability of any accumulated
residue N as well (Vigil and Kissel, 1991; Kuo and Sainju, 1998). Total C:N was below 25 for
all mixtures in this study, suggesting that the residues would not have caused extended net N
immobilization if incorporated at the time of sampling (Allison, 1966). However, the quality of
the mixed residues at the time of harvest will be influenced by the maturity of rye and vetch in
addition to the relative species composition of the mixture, and may vary considerably depending
on when the cover crop is terminated (Clark et al., 1997; Clark et al., 2007).
Given the relatively high cost of vetch seed, farmers should consider whether seeding a
mixture is worth the investment in the legume after a certain point in the fall, when compared
with sowing a less expensive cereal like rye alone (Hayden et al., 2014). Seeding a rye-vetch
mixture, even later than recommended, does maintain the possibility of significant fixed N
additions in the event that winter and/or spring temperatures are warmer than average. Farmers
may also be able to compensate by altering mixture management to promote greater vetch
biomass production— such as by increasing the proportion of vetch sown in the mixture (Hayden
et al., 2014), or if possible, by delaying cover crop termination to allow more time for cover crop
growth in the spring (Teasdale et al., 2004; Clark et al., 2007). The potential for delayed rye
seeding to improve vetch biomass production in mixtures was also a consideration in designing

137

!
!
this experiment.
Staggered Seeding Effects
Staggered seeding of rye resulted in greater vetch shoot biomass production than coseeding in mixtures planted in late August and mid September in 2009 and in late August in
2010 (Figure 2A). Delaying rye seeding on those planting dates until vetch emergence (short
stagger) increased vetch biomass by between 30 - 62% (76 - 106 g m-2) relative to the co-seeded
mixtures. The effect on vetch biomass of further delaying rye seeding until the vetch 3-4 leaf
stage (long stagger) did not differ significantly from the shorter stagger.
Vetch productivity and stand establishment may have benefitted from reduced interference
during seedling establishment, though biomass production more likely increased as a result of
lower competitive inhibition from the delayed rye component during fall and especially spring
growth periods. Total GDD accumulation during cover crop growth appeared to regulate the
interactions between vetch planting date and the effects of staggered seeding over the two
seasons of the study.
No significant benefits to vetch biomass production were observed with staggered seeding
in mixtures where vetch experienced fewer than 850 GDD before harvest, which included the
late September planting in both years and the mid September planting in 2010 (Figure 2A).
Below that threshold, vetch productivity may have been more limited by heat accumulation than
by interference from rye, or was perhaps tempered by greater stand loss from overwinter
mortality in the staggered mixtures (Figure 3C). In most cases, rye biomass declined in mixtures
with increasing length of stagger due to the shorter period available for growth (Figure 2B). The
magnitude of the stagger-driven reductions in rye biomass generally increased with later planting
dates and more limited GDD accumulation, as well as greater overwinter stand loss (data not

138

!
!
shown). As a result, where vetch accumulated greater than 950 GDD, stagger-sown mixtures
yielded equivalent total biomass to co-seeded mixtures, while in the case of lower GDD
accumulation, staggering seeding led to significant reductions in the total biomass produced
(Figure 2C).
The shoot N content of the cover crop mixtures was driven predominately by rye and vetch
biomass yields. For the earlier planting date treatments where >950 GDD accumulated during
vetch growth, staggered seeding resulted in the equivalent of an additional 30-36 kg N ha-1 in
vetch shoots relative to co-seeding due to an increase in vetch biomass (Table 2). While the
average effect of delayed rye seeding on vetch N did not differ between the two lengths of
stagger, the change in the total N content of mixtures generally decreased from short to long
staggers due to greater reductions in rye biomass (Table 2). However, by increasing vetch and/or
decreasing rye biomass, the compositional shifts in mixture biomass fostered by longer delays in
rye seeding typically resulted in lower total residue C:N at the time of sampling (Table 2).
Predicting the dynamics of inorganic N availability from mixed organic materials can be
complex, but both N content and residue quality will influence net N mineralization in the first
season following cover crop incorporation (Kuo and Sainju, 1998).
Neither planting date nor staggered seeding significantly influenced vetch N fixation
efficiency (%Ndfa) under the conditions of this study (Table 2). The estimated percentage of
vetch N derived from fixation was 88% in 2009-10, and 77% in 2010-11 but did not vary by
treatment within years. Depending on genetic and environmental contexts, the presence of a
grass species in mixture may alternately enhance legume N fixation by depleting soil inorganic N
(Jensen, 1996) or impair it as a result of competitive inhibition (Brainard et al., 2011). Previous
studies have documented increases in vetch %Ndfa when vetch is grown in mixture with a cereal

139

!
!
species rather than in monoculture (Kurdali et al., 1996; Brainard et al., 2012), but the impact of
relatively small compositional changes in rye-vetch mixtures (such as those resulting from
staggered seeding) on vetch N fixation are probably minor or difficult to detect, especially under
conditions of low soil N availability. Working on land adjacent to this experiment with similar
sandy and unfertilized soils, Hayden et al. (2014) observed no significant differences in vetch
%Ndfa across a gradient of different rye and vetch proportions in mixture.
Vetch Winter Survival
In co-seeded mixtures, vetch winter survival was consistent across the planting dates and
study years, with 90% of vetch seedlings surviving through the winter on average (Figure 3C).
Staggered seeding did not influence winter survival in mixtures where >950 GDD accumulated
during vetch growth; however, for later planting dates where mixtures experienced more limited
GDD accumulation, delaying rye seeding resulted in significant reductions in vetch winter
survival. For the late September seeding in 2010-11, delaying rye seeding until the vetch 3-4
leaf stage (long stagger) resulted in winter survival of 46% compared to 90% with co-seeding
(Figure 3C).
Hairy vetch is the hardiest of winter annual legume cover crops, but its cold tolerance
varies depending on cultivar and developmental stage (Wilke et al., 2008). Although freezing
resistance declines after flower initiation (Brandsaeter et al., 2002), vetch stands generally
experience greater winter mortality when planted later in the fall (Brandsaeter et al., 2008;
Guldan and Martin, 2003), suggesting that in the vegetative phase younger plants or seedlings
are more cold sensitive. For example, during a season in New York when the average and
absolute minimum temperatures in January were -13 and -25 ºC, respectively, Teasdale et al.
(2004) observed 85% winter survival for hairy vetch sown on 25 August versus 39% survival

140

!
!
when sown 14 September.
The likely mechanisms by which rye growth may improve legume winter survival in
mixtures include reducing frost heaving of the soil, slowing boundary layer air movement, and
increasing snow cover retention—all of which lower effective minimum temperature extremes or
otherwise limit legume exposure (Smith, 1975). Jannink et al. (1997) observed reduced vetch
overwinter mortality in mixture with rye at two locations in Maine, and Brainard et al. (2012)
found a similar response for three vetch cultivars in one out of two years in Michigan. In
contrast, Hayden et al. (2014) found no evidence that the presence of rye influenced vetch winter
survival in mixtures that were sown on 1 September over two years in Michigan. Our staggered
seeding results demonstrate that the presence of rye can provide a clear benefit to vetch winter
survival, particularly when vetch is planted late or is otherwise exposed to cold temperatures at a
less freeze-resistant developmental stage. Delaying rye seeding at later planting dates reduced
rye biomass production in the fall (Figure 2B), such that it was likely no longer effective at
moderating soil surface conditions and/or microclimate during the winter months. However,
from a practical standpoint, it is worth noting that the conditions under which rye is likely to
have the greatest influence on vetch winter survival, including exceptionally cold winter
temperatures or late fall planting, are also conditions that tend to greatly limit vetch biomass
production at mid-May termination dates, irrespective of stand density.

CONCLUSIONS
Rye-vetch cover crop stands will vary considerably depending on when the mixture is
planted. Later planting and less GDD accumulation during cover crop growth generally led to
reduced total shoot biomass and a greater proportion of rye biomass in mixtures. As a

141

!
!
consequence, later planted co-seeded mixtures accumulated less fixed and total N, and produced
residues with a higher combined C:N.
A novel finding from this study was that for earlier planted rye-vetch mixtures where >950
GDD accumulated during vetch growth, staggering (delaying) rye seeding until vetch emergence
increased vetch shoot biomass by 760 - 1,060 kg ha-1 and vetch N content by 30 - 36 kg ha-1
compared with co-seeding. For later plantings, staggered seeding provided no benefit to vetch
biomass production, only reductions in rye and total biomass. In addition, staggering rye seeding
until the vetch 3-4 leaf stage at these later planting dates resulted in lower vetch winter survival
compared with co-seeded mixtures.
Although staggered seeding resulted in increased fixed N accumulation in some mixtures
planted earlier in the season, this study did not evaluate possible tradeoffs in the provision of
other important services by the cover crop. For example, weed suppression has been shown to
decrease with reductions in rye biomass in rye-legume mixtures (Akemo et al., 2000; Hayden et
al., 2014). Furthermore, any benefits of the practice must ultimately be weighed against the
costs associated with additional management complexity, including the time, labor, and fuel of
an added tractor pass for rye seeding. In this regard, it is important to note that the seeding
methodology used in this study did not reflect conditions that would result from the field-scale
application of staggered seeding, such as the effects of additional traffic on young vetch
seedlings or the establishment of broadcast rye without incorporation.
Several alternative approaches to realizing the benefits of staggered seeding seen in this
study may be worth further exploration. Seed coatings have been used to delay germination of
cash crops to allow for simultaneous seeding in relay intercropping systems, but coating seeds is
itself expensive and the resulting germination kinetics can be challenging to manage (Stendahl,

142

!
!
2005). Alternatively, selection for more rapid germinating vetch varieties, or use of seedpriming techniques, may be useful for increasing the rate of germination and emergence of vetch
relative to rye. Seed priming is commonly used to speed germination (Parera and Cantliffe,
1994), although its efficacy for vetch has not been studied, and the costs associated with priming
can be substantial. Additional research would be required to determine whether potential onfarm benefits of rye-vetch staggered seeding would justify additional management costs or
possible tradeoffs in cover crop services.

143

!
!

APPENDIX

144

!
!
Table 4.1. Seeding schedule and accumulated growing degree days (GDD) during rye and vetch growth in cover crop mixture
treatments.

Treatment
Late Aug. vetch seeding
Co-seeded mixture
Short rye stagger‡
Long rye stagger§
Mid Sept. vetch seeding
Co-seeded mixture
Short rye stagger
Long rye stagger
Late Sept. vetch seeding
Co-seeded mixture
Short rye stagger
Long rye stagger

2009-2010
Date sown
Accumulated GDD†
Vetch
Rye
Vetch
Rye

2010-2011
Date sown
Accumulated GDD†
Vetch
Rye
Vetch
Rye

8/25/09
8/25/09
8/25/09

8/25/09
9/3/09
9/11/09

1189
1189
1189

1189
1078
961

8/31/10
8/31/10
8/31/10

8/31/10
9/8/10
9/17/10

1020
1020
1020

1020
898
800

9/9/09
9/9/09
9/9/09

9/9/09
9/18/09
9/24/09

993
993
993

993
864
780

9/14/10
9/14/10
9/14/10

9/14/10
9/23/10
10/7/10

832
832
832

832
720
599

9/22/09
9/22/09
9/22/09

9/22/09
10/2/09
10/27/09

815
815
815

815
702
571

9/28/10
9/28/10
9/28/10

9/28/10
10/12/10
10/29/10

665
665
665

665
540
417

† Total growing degree day (GDD) base 4°C accumulation between date of vetch or rye seeding and cover crop sampling on 14 or 17
May in 2010 and 2011, respectively.
‡ In short stagger mixtures, rye seeding was delayed 120 GDD following vetch seeding, approximately at vetch seedling emergence.
§ In long stagger mixtures, rye seeding was delayed 230 GDD following vetch seeding, approximately at the vetch 3-4 leaf stage.

145

!
!
Table 4.2. Percentage vetch shoot N derived from the atmosphere (Ndfa), vetch and rye shoot N content, and total (vetch+rye) shoot N content and C:N ratio in
co-seeded, short stagger, and long stagger mixtures sown in late August, Mid September, and Late September in 2009 and 2010.
Vetch Ndfa
Vetch shoot N content
Rye shoot N content
Total shoot N content
Total shoot C:N
2009-10
2010-11
2009-10
2010-11
2009-10
2010-11
2009-10
2010-11
2009-10
2010-11
%
g m -2
Late Aug. vetch seeding
Co-seeded mixture
92 (0.4) ns†‡ 74 (6) ns 12.7 (1.3) a§ 5.9 (0.6) a 4.1 (0.2) ns 2.9 (0.1) ab 16.8 (1.2) a 8.7 (0.6) a
15.2 (1.0) ns 17.2 (0.3) a
Short rye stagger
74 (12)
78 (4)
16.3 (2.0) b 9.2 (0.9) b 4.3 (0.5)
3.7 (0.5) a 20.6 (2.4) b 12.9 (1.2) b 14.2 (0.4)
15.8 (0.6) ab
Long rye stagger
67 (17)
76 (4)
13.1 (0.8) a 9.5 (1.9) b 3.2 (0.3)
2.2 (0.2) b 16.3 (1.0) a 11.7 (2.0) ab 14.9 (0.3)
14.5 (0.7) b
Mid Sept. vetch seeding
Co-seeded mixture
94 (2) ns
86 (8) ns 6.1 (0.9) a
4.2 (0.7) ns 4.1 (0.4) a 4.9 (0.9) a 10.2 (1.2) a 9.0 (1.4) ns 20.8 (0.7) a 19.9 (2.1) a
Short rye stagger
93 (2)
79 (3)
9.1 (1.2) ab 3.0 (0.7)
3.7 (0.3) a 3.6 (0.6) b 12.8 (1.5) ab 6.6 (0.5)
17.0 (0.5) b 18.0 (0.9) a
Long rye stagger
93 (3)
72 (6)
12.2 (2.2) b 4.1 (1.6)
2.7 (0.3) b 1.6 (0.3) c 14.9 (2.1) b 5.7 (1.6)
14.3 (0.6) c 13.2 (0.6) b
Late Sept. vetch seeding
Co-seeded mixture
95 (2) a
74 (5) ns 3.0 (0.4) ns 0.8 (0.3) ns 4.6 (1.2) a 2.8 (0.1) a 7.6 (0.9) a 3.7 (0.3) ns 24.0 (1.0) a 18.4 (2.2) a
Short rye stagger
97 (1) a
77 (5)
3.5 (0.6)
1.1 (0.3)
3.1 (0.2) b 1.7 (0.2) b 6.6 (0.4) a 2.8 (0.4)
20.6 (1.2) a 16.4 (1.3) a
Long rye stagger
91 (3) b
79 (2)
1.8 (0.5)
1.1 (0.8)
1.3 (0.2) c 0.6 (0.04) c 3.1 (0.6) b 1.7 (0.8)
16.5 (0.9) b 10.1 (0.5) b
† Mean (SE), n = 4.
‡ ns, not significant at the 0.05 probability level.
§ Within a given year and planting date, means followed by the same letter indicate levels of stagger are not significantly different at the 0.05 probability level.

146

!
!

Figure 4.1. Vetch (black), rye (white), and total (vetch+rye, gray) shoot dry biomass (A), total
residue N content (B), and total residue carbon:nitrogen ratio (C) in co-seeded mixtures sown in
late August, Mid September, and Late September in 2009 (circles) and 2010 (triangles). Data
from the two years are plotted based on total growing degree days (GDD) base 4 ºC accumulated
during vetch growth on a reversed scale, with planting date treatments indicated above the xaxis. Means +/- SE. Responses to GDD accumulation (years combined) of vetch shoot biomass:
y=0.0008x2-0.95x+298.53, r2= 0.82***; rye shoot biomass: y=-0.0016x2+2.99x-1122.92, r2=
0.26*; total shoot biomass: y=0.69x-216.93, r2= 0.56***; total N content: y=0.02x-10.32 , r2=
0.71***; and total C:N: y=-0.0001x2+0.10x-21.08, r2= 0.36**. Single, double, and triple
asterisks following r2 values denote regression model significance at the 0.05, 0.01, and 0.001
levels, respectively.

147

!
!

Figure 4.2. Vetch (A), rye (B), and total (vetch+rye, C) shoot dry biomass in short rye stagger
(gray) and long rye stagger (white) mixtures relative to co-seeded mixtures (black) sown in late
August, Mid September, and Late September in 2009 (circles) and 2010 (triangles). Data from
the two years are plotted based on total growing degree days (GDD) base 4 ºC accumulated
during vetch growth on a reversed scale, with planting date treatments indicated above the xaxis. Actual GDD accumulation for rye is less than that for vetch in staggered mixtures (Table
1). Within a given year and planting date treatment, levels of stagger are not significantly
different when means (+/- SE) are followed by the same letter (α = 0.05). Responses to GDD
accumulation (years combined) of vetch co-seeded: y=0.0008x2-0.95x+298.53, r2= 0.82***;
vetch short stagger: y=0.0011x2-1.22x+342.41, r2= 0.89***; vetch long stagger: y=0.79x531.85, r2= 0.71***; rye co-seeded: y=-0.0016x2+2.99x-1122.92, r2= 0.26*; rye short stagger:
y=-0.0012x2+2.51x-1048.44, r2= 0.62***; rye long stagger: y=0.38x-232.79, r2= 0.81***; total
co-seeded: y=0.69x-216.93, r2= 0.56***; total short stagger: y=1.13x-630.00, r2= 0.85***; total
148

!
!
Figure 4.2 (cont’d)
long stagger: y=1.17x-764.63, r2= 0.80***. Single, double, and triple asterisks following r2
values denote regression model significance at the 0.05, 0.01, and 0.001 levels, respectively.

149

!
!

Figure 4.3. Percentage vetch winter survival in co-seeded (black), short rye stagger (gray), and
long rye stagger (white) mixtures sown in late August, Mid September, and Late September in
2009 (circles) and 2010 (triangles). Data from the two years are plotted based on total growing
degree days (GDD) base 4 ºC accumulated during vetch growth on a reversed scale, with
planting date treatments indicated above the x-axis. Within a given year and planting date
treatment, levels of stagger are not significantly different when means (+/- SE) are followed by
the same letter (α=0.05). Responses to GDD accumulation (years combined) of co-seeded: y=0.0023x+92.10, r2=0.0011; short stagger: y=-0.00021x2+0.41x-107.20, r2=0.29*; and long
stagger: y=-0.00034x2+0.71x-277.40 , r2=0.77***. Single, double, and triple asterisks following
r2 values denote regression model significance at the 0.05, 0.01, and 0.001 levels, respectively.

150

!
!

REFERENCES

151

!
!
REFERENCES

Akemo, M.M., E.E. Regnier, and M.A. Bennett. 2000b. Weed suppression in spring-sown rye
(Secale cereale)-pea (Pisum sativum) cover crop mixes. Weed Technol. 14:545–549.
Allison, F.E. 1966. The fate of nitrogen applied to soils. Adv. Agron. 18:219–258.
Amossé, C., M.-H. Jeuffroy, and C. David. 2013. Relay intercropping of legume cover crops in
organic winter wheat: Effects on performance and resource availability. Field Crops Res
145: 78–87.
Brainard, D., B. Henshaw, and S. Snapp. 2012. Hairy vetch varieties and bi-cultures influence
cover crop services in strip-tilled sweet corn. Agron. J 104:629–638.
Brainard, D.C., V. Kumar and R.R. Bellinder. 2011. Grass-legume mixtures and soil fertility
affect cover crop performance and weed seed production. Weed Technol. 25:473-479.
Brandsaeter, L.O., H. Heggen, H. Riley, E. Stubhaug, and T.M. Henriksen. 2008. Winter
survival, biomass accumulation and N mineralization of winter annual and biennial
legumes sown at various times of year in northern temperate regions. Eur. J. Agron.
28:437–448.
Brandsaeter, L.O., A. Olsmo, A.M. Tronsmo, and H. Fykse. 2002. Freezing resistance of winter
annual and biennial legumes at different developmental stages. Crop Sci. 42:437–443.
Brennan, E., N. Boyd, R. Smith, and P. Foster. 2009. Seeding rate and planting arrangement
effects on growth and weed suppression of a legume-oat cover crop for organic vegetable
systems. Agron. J. 101:979–988.
Clark, A. J., ed. 2007. Managing cover crops profitably. 3rd ed. Sustainable Agriculture
Network, Beltsville, MD.
Clark, A.J., A.M. Decker, and J.J. Meisinger. 1994. Seeding rate and kill date effects on hairy
vetch cereal rye cover crop mixtures for corn production. Agron. J. 86:1065–1070.
Clark, A.J., A.M. Decker, J.J. Meisinger, and M.S. McIntosh. 1997. Kill date of vetch, rye, and a
vetch-rye mixture: I. Cover crop and corn nitrogen. Agron. J. 89:427–434.
Clark, A.J., J.J. Meisinger, A.M. Decker, and F.R. Mulford. 2007. Effects of a grass-selective
herbicide in a vetch-rye cover crop system on nitrogen management. Agron. J. 99:36–42.
Farsad, A., T.O. Randhir, S.J. Herbert, and M. Hashemi. 2011. Spatial modeling of critical
planting date for winter rye cover crop to enhance nutrient recovery. Agron. J. 103:1252–
1257.
Francis, C.A., M. Prager, and G. Tejada. 1982. Effects of relative planting dates in bean

152

!
!
(Phaseolus vulgaris L.) and maize (Zea mays L.) intercropping patterns. Field Crops Res.
5:45–54.
Fukai, S., and B.R. Trenbath. 1993. Processes determining intercrop productivity and yields of
component crops. Field Crops Res. 34:247–271.
Guldan, S.J., and C.A. Martin. 2003. Hairy vetch biomass yield as affected by fall planting date
in the irrigated steppe of the southern Rocky Mountains. J. Sustain. Agric. 22:17–23.
Hansen, E.M., H. Høgh-Jensen, and J. Djurhuus. 2002. Biological nitrogen fixation in a grazed
perennial grass/clover ley and correlation with herbage and soil variables. Eur. J. Agron.
16:309–320.
Hansen, J.P., and F.P. Vinther. 2001. Spatial variability of symbiotic N2 fixation in grass-white
clover pastures estimated by the 15N isotope dilution method and the natural 15N
abundance method. Plant Soil 230:257–266.
Hayden, Z.D., D.C. Brainard, B. Henshaw, and M. Ngouajio. 2012. Winter annual weed
suppression in rye–vetch cover crop mixtures. Weed Technol. 26:818–825.
Hayden, Z.D., M. Ngouajio, and D.C. Brainard. 2014. Rye-vetch mixture proportion tradeoffs:
Cover crop productivity, nitrogen accumulation, and weed suppression. Agron. J.
106:904-914.
Hogberg, P. 1997. Tansley review No 95 - N-15 natural abundance in soil-plant systems. New
Phytol. 137:179–203.
Izaurralde, R.C., W.B. McGill, and N.G. Juma. 1992. Nitrogen fixation efficiency, interspecies N
transfer, and root growth in barley-field pea intercrop on a Black Chernozemic soil. Biol.
Fertil. Soils 13:11–16.
Jannink, J.L., L.C. Merrick, M. Liebman, E.A. Dyck, and S. Corson. 1997. Management and
winter hardiness of hairy vetch in Maine. Maine Agricultural and Forest Exp. Stat. Tech.
Bulletin no. 167., Orono.
Jensen, E.S. 1996. Grain yield, symbiotic N2 fixation and interspecific competition for inorganic
N in pea-barley intercrops. Plant Soil 182:25–38.
Keating, B.A., and P.S. Carberry. 1993. Resource capture and use in intercropping: solar
radiation. Field Crops Res. 34:273–301.
Kuo, S., and U.M. Sainju. 1998. Nitrogen mineralization and availability of mixed leguminous
and non-leguminous cover crop residues in soil. Biol. Fertil. Soils 26:346–353.
Kurdali, F., N. Sharabi, and A. Arslan. 1996. Rainfed vetch-barley mixed cropping in the Syrian
semi-arid conditions. Plant Soil 183:137–148.
Ledgard, S.F., and K.W. Steele. 1992. Biological nitrogen-fixation in mixed legume grass

153

!
!
pastures. Plant Soil 141:137–153.
Mariotti, M., A. Masoni, L. Ercoli, and I. Arduini. 2009. Above- and below-ground competition
between barley, wheat, lupin and vetch in a cereal and legume intercropping system.
Grass Forage Sci. 64:401–412.
Nuttonson, M.Y. 1958. Rye-climate relationships and the use of phenology in ascertaining the
thermal and photo-thermal requirements of rye. American Institute of Crop Ecology,
Washington, DC.
Ofori, F., and W.R. Stern. 1987. Cereal–legume intercropping systems. Adv. Agron. 41:41–90.
C.A. Parera, D.J. Cantliffe. 1994. Presowing seed priming. Hortic. Rev. 16:109–141.
Ranells, N.N., and M.G. Wagger. 1996. Nitrogen release from grass and legume cover crop
monocultures and bicultures. Agron. J. 88:777–782.
Rochester, I., and M. Peoples. 2005. Growing vetches (Vicia villosa Roth) in irrigated cotton
systems: inputs of fixed N, N fertiliser savings and cotton productivity. Plant Soil
271:251–264.
Sainju, U.M., W.F. Whitehead, and B.P. Singh. 2005. Biculture legume-cereal cover crops for
enhanced biomass yield and carbon and nitrogen. Agron. J. 97:1403–1412.
Shearer, G., and D.H. Kohl. 1986. N2-fixation in field settings: Estimations based on natural 15N
abundance. Funct. Plant Biol. 13:699–756.
Shipley, P.R., J.J. Meisinger, and A.M. Decker. 1992. Conserving residual corn fertilizer
nitrogen with winter cover crops. Agron. J. 84:869–876.
Smith, D. 1975. Forage management in the North. 4th ed. Kendall/Hunt Publ. Co., Dubuque, IA.
Stendahl, F. 2005. Seed coating for delayed germination: A tool for relay cropping of annual
crops. Licentiate thesis, Swedish Univ. of Agricultural Sciences, Uppsala.
Streeter, J. 1988. Inhibition of legume nodule formation and N2 fixation by nitrate. Crit. Rev.
Plant Sci. 7:1–23.
Teasdale, J.R., and A.A. Abdul-Baki. 1998. Comparison of mixtures vs. monocultures of cover
crops for fresh-market tomato production with and without herbicide. HortScience
33:1163–1166.
Teasdale, J.R., T.E. Devine, J.A. Mosjidis, R.R. Bellinder, and C.E. Beste. 2004. Growth and
development of hairy vetch cultivars in the northeastern United States as influenced by
planting and harvesting date. Agron. J. 96:1266–1271.
Vigil, M.F., and D.E. Kissel. 1991. Equations for estimating the amount of nitrogen mineralized
from crop residues. Soil Sci. Soc. Am. J. 55:757–761.

154

!
!
Wilke, B.J., and S.S. Snapp. 2008. Winter cover crops for local ecosystems: linking plant traits
and ecosystem function. J. Sci. Food Agric. 88:551–557.

155

!
!

CHAPTER V

CONCLUSIONS

156

!
!
CONCLUSIONS

Cover crop mixtures composed of cereal rye and the legume hairy vetch have the
potential to combine the unique strengths of the component species while taking advantage of
interspecific synergies. However, outcomes depend on both environmental and management
factors that influence species interactions during cover crop growth, the final composition of
mixture stands, and the decomposition of residues following cover crop termination. The
unifying objective of this work was to document the performance of rye-vetch mixtures relative
to rye and vetch monocultures while leveraging our understanding of ecological principles to
develop and evaluate strategies for improving mixture management within vegetable cropping
systems.
In evaluating how the proportion of rye and vetch sown in a mixture influenced species
interactions and cover crop performance, seeding rates were found to be good predictors of rye
and vetch stand density and biomass composition in the spring. While outcomes may vary
depending on environment and management, we observed little evidence of substantial
interference between the two species in mixture, but also saw no significant evidence of
facilitative benefits, such as improved vetch winter survival, increased N fixation efficiency, or
reduced costs per unit of N fixed. However, changing the proportions of rye and vetch sown
resulted in important tradeoffs among some of the agroecosystem services provided by the living
cover crops—for example, greater fixed N accumulation, but higher seed costs and reduced
winter annual weed suppression, with increasing vetch. While greater rye content was associated
with greater suppression of winter annual weed density and biomass, differences in the relative
effects of rye and vetch on the density of specific components of the weed community were

157

!
!
documented.
Following cover crop termination, N content was a dominant factor in the effects of
incorporated rye-vetch residues and the use of black plastic mulch on subsequent vegetable
production. Cover crops containing greater proportions of vetch generally resulted in higher soil
N availability following incorporation, and under conditions where N was limiting, led to higher
yields of bell pepper and slicing cucumber. However, greater N availability in the plough layer
also translated into higher pore water NO3- concentrations below the crop root zone.
Along the gradient of rye-vetch proportions evaluated in this experiment, changes in total
cover crop N content appeared to have a larger influence on vegetable yields than did any
potential changes in temporal patterns of N mineralization. Regardless of rye-vetch proportion,
net N mineralized from mixtures accumulated in the soil prior to peak vegetable demand, leaving
it vulnerable to leaching from early-season precipitation, particularly on the sandy soils present
at the experimental site. Although the benefits of plastic mulch must be weighed against the
costs associated with its use and disposal, our results demonstrate that plastic mulch can be an
important tool for preserving N fertility benefits from high N cover crop residues, contributing
substantially to the yield increases observed with plastic mulch relative to bare ground
management. Although lower microbial biomass was also observed under plastic mulch relative
to bare ground 3 wk after plastic application, changes to the microbial communities resulting
from cover crop and plastic mulch treatments would likely have limited practical significance for
agroecosystem functioning in the short term.
Results from a separate experiment demonstrated that fall planting dates and staggered
(delayed) seeding of rye after vetch can have important impacts on biomass composition of ryevetch mixture stands in the spring. Later planting and less GDD accumulation during cover crop

158

!
!
growth generally led to reduced total shoot biomass and a greater proportion of rye biomass in
mixtures. As a consequence, delayed planting of rye-vetch mixtures resulted in spring stands
containing less fixed and total N, and residues having a higher combined C:N. Staggering rye
seeding just until vetch emergence in the fall was successful in increasing vetch shoot biomass
production by 760 - 1,060 kg ha-1 and vetch N content by 30-36 kg ha-1 compared with co-seeded
mixtures for earlier planting dates, but benefits were not observed when mixtures were planted
later in the fall. Staggering rye seeding until the vetch 3-4 leaf stage at these later planting dates
resulted in higher vetch winter morality compared with co-seeded mixtures, demonstrating a
clear benefit to vetch winter survival due to the presence of rye, particularly when vetch is
planted late or is otherwise exposed to cold temperatures at a less freeze-resistant developmental
stage. Additional research would be needed to determine whether potential on-farm benefits of
rye-vetch staggered seeding justify the additional management complexity involved and possible
tradeoffs with other cover crop services.
Future improvements in the economic and environmental sustainability of agriculture will
depend heavily on our ability to translate understanding of complex ecological interactions into
applied management solutions. Further optimizing the performance of cover crops within
agricultural production systems will rest broadly on our understanding and application of the
biotic and abiotic controls on plant competition and decomposition of organic materials. The
results of this research have the potential to support more-informed decision making regarding
rye and vetch cover crop selection, mixture seeding rates, and management practices both before
and after cover crop termination, which can take into account a farmer’s principal objectives for
the cover crop, as well as the contexts of the cropping system and environment.

159