“nun-.auv vu-

. ‘ 3.. E.
5.... .3.ku igﬁmﬂu

’
. . n. {8.52

an L
, mmﬁﬁmt , a?
\o. haw-u.- .. 5..) “A”. :, .WNXWWM‘WMS y.» av
2. . , ‘ . .71

gnu—ﬂ-

Irhaa ) f
it .u

D. .

..
i
.33. . 2r“! .4...
- $.93 sum
’

in.
In. .

.2 !
.5103...
4.1.3:)

11“.?"

n
5.3."

s
.

 

 

 

Tm ~"s
"’ | LIBRARY
1 card . iviicmgun State
I Umversuty

This is to certify that the
dissertation entitled

EVALUTATION OF SWITCHGRASS AND BIG BLUESTEM FOR
USE IN COOL«SEASON GRAZING SYSTEMS TO IMPROVE
SEASONAL FORAGE YIELD AND LIVESTOCK GAINS

presented by
Daniel John Hudson

has been accepted towards fulﬁllment
of the requirements for the

Doctoral degree in Crop and Soil Sciences

 

 

W»?

Major Professor's Silnature

ZT/xz/af/

Date

MSU is an affirmative-action, equal-opportunity employer

- -. ---.----.-.—._l—.—.-¢—.- -.------.--t.

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K IPro;/Acc&Pres/ClRC/DateDue Indd

EVALUTATION OF SWITCHGRASS AND BIG BLUESTEM FOR USE IN COOL-
SEASON GRAZING SYSTEMS TO IMPROVE SEASONAL FORAGE YIELD AND
LIVESTOCK GAINS
By

Daniel John Hudson

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Department of Crop and Soil Sciences

2008

ABSTRACT

USING SWITCHGRASS AND BIG BLUESTEM IN COOL-SEASON GRAZING
SYSTEMS TO IMPROVE SEASONAL FORAGE YIELD AND LIVESTOCK GAINS

By

Daniel John Hudson

Michigan livestock producers are faced with a choice that arises annually: tO
move their livestock from dormant summer pastures to locations where they are fed
stored forage, or to allow their livestock to continue grazing, risking damage to the land.
Keeping livestock on high-quality pasture is in the economic interest of the farmer, the
physical interest of the animal, and the land itself. In this study we compared a typical
Michigan grazing system with two other grazing systems, each of which integrate one
native warm season grass species into the typical grazing system to provide quality mid-
summer forage. The grazing experiment was conducted at the W.K. Kellogg Biological
Station in Hickory Comers, MI. A completely randomized design with four replications
was used. Treatments for this trial include: 1) traditional Michigan cool season
grass/legume pasture; 2) a treatment two-thirds of which is the same as ‘1)' and one-third
of which is composed of switchgrass (Panicum virgatum); 3) a treatment two-thirds of
which the same as to ‘ l)‘ and one-third which is composed of big bluestem (Andropogon
gerardii). A variable stocking rate was be used and pastures were rotationally grazed.
Grazing began in late-April and ended by October 15 Of each year, depending upon
conditions. Pasture performance is described in terms of forage dry matter offered

throughout the grazing season, botanical composition, forage quality, number and

distribution of animal unit grazing days, average daily gain Of animals on pasture, and
total animal weight gain/hectare. The animals used in this study were Holstein steers,
weighing approximately 240 kg at the beginning of the grazing season. Results reveal
that although average daily gains do not differ among treatments,total animal weight
gain per hectare is signiﬁcantly different among treatments. Inclusion of big bluestem or
switchgrass in a grazing system may reduce the risk Of extreme declines in available
pasture dry matter, but management constraints of these grasses reduce total grazable
acreage in the Spring and early fall. Economic comparisons that include variable costs
Show that integrating switchgrass or big bluestem into traditional cool-season grass and
legume pastures in the proportions used in this experiment will result in an economic
disadvantage for southwest Lower Michigan livestock producers. However, these
differences in economic performance could be offset by moderate financial incentives
from organizations that wish to encourage livestock producers to enhance wildlife habitat
by including switchgrass and big bluestem in their grazing systems. There were no
Signiﬁcant differences in the relative abundance of grassland birds among the treatments.
The size and design of the grazing experiment was not suitable for rigorous research on

the effect of native warm season grasses in grazing systems on grassland bird species.

Acknowledgments
I thank God who has given me strength and encouragement through Jesus Christ
throughout my graduate program. May He be glorified. I would also like to thank my
wife (Julia), children, parents, and extended family for their support and encouragement
and toleration of my preoccupied mind and frequent extended hours. Further, I would
hketothank:

0 Dr. Richard Leep, my major professor, for his guidance and support throughout
my graduate program.

0 My committee members, Dr. Donald Penner, Dr. James Kells, and Dr. Henry
Campa for their advice and guidance.

0 The Forage Crew, Tim Dietz, Matt Leep, Matt Smith, Stephanie Little, James
DeYoung, Sheri Weisbeck, Debi Warnock, Tim Boring, Nasser Al-Ghumaiz,
Mildred Lyon, and Joe Brooks for their major roles in data collection and project
management.

0 The staff of the WK. Kellogg Dairy, including Jim Bronson, Rob Ashley, Scott
Neudeck, Rick Lawrence, Larry Langshaw, Shannon Kimbrue, and Sandra
Welker, for their assistance in supporting and managing animals and equipment
associated with my research.

0 MSU Extension for their patience as I completed this project.

0 Dwight Fischer of the USDA-ARS for taking time to consult with me in planning
this project.

0 Ernst Conservation Seed Company for donation of switchgrass and big bluestem

seed.

iv

Ampac Seed Company, and Dave Robison for donating seed and providing
discounted seed.

Craig Newland for donation of seed and making land available for an on-farm
demonstration.

VetLife, LLC, for donation Of Encore estradiol steer implants

Elanco/Merial, for donation Of Eprinex pour-on parasite and insect control

product.

TABLE OF CONTENTS

Page
LIST OF TABLES .............................................................................. viii
LIST OF FIGURES ............................................................................... xi
INTRODUCTION AND BACKGROUND INFORMATION ............................... 1
Literature reVIew .......................... 2
Objectives and Hypotheses ........................................................................ 5
Literature Cited .................................................................................... 7

CHAPTER 1
SEASONAL FORAGE DYNAMCIS OF COOL-SEASON PASTURES INTEGRATED
WITH SWITCHGRASS AND BIG BLUESTEM

Abstract ............................................................................................. 10
Introduction ......................................................................................... 1 1
Literature Review .................................................................................. 12
Materials and Methods ............................................................................ 16
Results and Discussion ........................................................................... 35

‘ Conclusions ......................................................................................... 40
Literature Cited .................................................................................... 64
CHAPTER 2

INTEGRATED WARM- AND COOL- SEASON GRASS AND LEGUME PASTURES:
STEER AND PASTURE PERFORMANCE

Abstract ............................................................................................. 68
Introduction ......................................................................................... 70
Literature Review .................................................................................. 71
Materials and Methods ............................................................................ 75
Results and Discussion ............................................................................ 85
Conclusions ......................................................................................... 94
Literature Cited .................................................................................... l 18
CHAPTER 3

ECONOMIC PERFORMANCE OF COOL-SEASON PASTURES INTEGRATED
WITH SWITCHGRASS AND BIG BLUESTEM

Abstract ............................................................................................. 121
Introduction ......................................................................................... l 22
Materials and Methods ............................................................................ 126
Results and Discussion ........................................................................... 133
Literature Cited .................................................................................... 146

vi

CHAPTER 4

RELATIVE ABUNDANCE OF GRASSLAND BIRDS IN COOL-SEASON
PASTURES AND GRAZING SYSTEMS INTEGRATED WITH SWITCHGRASS
AND BIG BLUESTEM

Abstract .................................................................... p ......................... 149
Introduction ......................................................................................... 1 50
Literature Review .................................................................................. 155
Materials and Methods ............................................................................ 170
Results and Discussion ........................................................................... 177
Conclusions ......................................................................................... 179
Literature Cited .................................................................................... 183
APPENDICES

Appendix I .......................................................................................... 189
8iterature Cited .................................................................................... 194
Appendix II ......................................................................................... 195
Literature Cited .................................................................................... 209
Appendix III.’. ....................................................................................... 210
Literature Cited .................................................................................... 212

vii

LIST OF TABLES

Page
CHAPTER 1
Table 1.1. Dates and rates of urea application ............................................... 42
Table 1.2. Grazing initiation and termination dates by year ............................... 43
Table 1.3. Transformations used to normalize studentized residuals of forage
quantity and quality parameters ................................................................ 44
Table 1.4. DOY and corresponding dates .................................................... 45
Table 1.5. Crude protein percentage of forage dry matter offered over time,
modeled with yield data from 2003-2005 ..................................................... 49
Table 1.6. Acid detergent ﬁber content Of forage dry matter Offered over time,
modeled with yield data from 2003-2005 ..................................................... 51
Table 1.7. Neutral detergent fiber content of forage dry matter offered over time,
modeled with yield data from 2003-2005 ..................................................... 53
CHAPTER 2
Table 2.1. Weighing events by year ............................................................ 97
Table 2.2a. Example of animal unit grazing day calculation (continued in
Table 2.2b) ........................................................................................ 98
Table 2.2b. Example of animal unit grazing day calculation (continued) ................ 99

Table 2. 3. Guaranteed analysis of Land O’ Lakes Pro Phos 8 granular mineral

supplement ........................................................................................ 1 00
Table 2.4. Beginning and ending dates of grazing intervals used to calculate
time-series ADG for the 2003 through 2005 grazing seasons. ............................ 101
Table 2.5. Fixed effects and interactions used in the full model Of gain ha",
animal unit grazing days (AUGD), and average daily gain (ADG) ...................... 102
Table 2.6. Gain ha'] by treatment, year, and alfalfa abundance ............................ 103

viii

Table 2.7. Average daily gain of Steers by treatment and year at time intervals

throughout the 2003-2005 grazing seasons ................................................... 104
Table 2.8. Average daily gain of steers by treatment and year at intervals

throughout the 2003-2005 grazing seasons ................................................... 108
Table 2.9. Animal unit grazing days accumulated by treatment and year at

intervals throughout the 2003-2005 grazing seasons ........................................ 1 10
Table 2.10. Treatment across year comparison Of AUGD per hectare over time. .....I 14
CHAPTER 3

Table 3.1. Weighing events by year ............................................................ 137
Table 3.2 Average daily gain Of steers by treatment and year at time intervals
throughout the 2003-2005 grazing seasons .................................................. 138
Table 3.3. Animal unit grazing days accumulated by treatment and year

at intervals throughout the 2003-2005 grazing seasons ..................................... 139
Table 3.4. Animal gain haJ by treatment ..................................................... 140

Table 3.5. The cost Of three years of cool-season grass and legume pasture
maintenance ...................................................................................... l 4 I

Table 3.6. The cost of establishment of big bluestem pasture and three years of
maintenance ...................................................................................... 1 42

Table 3.7. The cost of establishment of switchgrass pasture and three years of
maintenance. ..................................................................................... 1 43

Table 3.8. Summary Of maintenance and establishment costs Of treatments ............ 144

Table 3.9. Summary of gross income, expenses, and net income of treatments

by year ............................................................................................. 145
CHAPTER 4

Table 4.1. Birds Observed within each treatment during Observation periods. . 181
Table 4.2. Average number of birds per transect as affected by treatment .............. I82
APPENDICES

ix

Table A. 1. Latin and common names of forage and palatable non-forage Species
found in the CS portions of all treatments throughout the experiment .................... 199

Table A2. Latin and common names of undesirable plant species found
in the CS portions of experimental pastures .................................................. 200

Table A3. Contribution of major pasture plant species to botanical
composition of the cool-season portion Of the CS-BBS treatment ........................ 205

Table A4. Contribution of major pasture plant Species to botanical
composition of the CS-only treatment ........................................................ 206

Table A5. Contribution of major pasture plant species to botanical
composition of the cool season portion of the CS-SG treatment .......................... 207

LIST OF FIGURES

CHAPTER 1

Page
Figure 1.1. Forage dry matter Offered over time during the 2003 grazing season
for the CS-BBS and CS-Only treatments ..................................................... 46

Figure 1.2. Forage dry matter offered over time during the 2004 grazing season. 47
Figure 1.3. Forage dry matter Offered over time during the 2005 grazing season ...... 48

Figure 1.4. Crude protein content Of forage dry matter offered over time,
modeled with yield data from 2003-2005 ...................................................... 50

Figure 1.5. Acid detergent ﬁber content of forage dry matter offered over
time, modeled with yield data from 2003-2005 .............................................. 52

Figure 1.6. Neutral detergent ﬁber content of forage dry matter Offered over
time, modeled with yield data from 2003-2005 ............................................. 54

Figure 1.7. Crude protein content of forage dry matter offered over time ,
from the CS-BBS and CS-Only treatments, modeled with yield data from 2003 ...... 55

Figure 1.8. Crude protein content of forage dry matter Offered over time
from all treatments, modeled with yield data from 2004 ................................... 56

Figure 1.9. Crude protein content of forage dry matter offered over time
from all treatments, modeled with yield data from 2005 ................................... 57

Figure 1.10. Acid detergent ﬁber content of forage dry matter offered over time
from the CS-BBS and CS-Only treatments, modeled with yield data from 2003 ...... 58

Figure 1.11. Acid detergent ﬁber content of forage dry matter offered over time
from all treatments, modeled with yield data from 2004 ................................... 59

Figure 1.12. Acid detergent ﬁber content of forage dry matter offered over time
from all treatments, modeled with yield data from 2005 ................................... 60

Figure 1.13. Neutral detergent ﬁber content of forage dry matter offered over time
from the CS-BBS and CS-Only treatments, modeled with yield data from 2003 ...... 61

xi

Figure 1.14. Neutral detergent ﬁber content of forage dry matter offered over time
from all treatments, modeled with yield data from 2004 ................................... 62

Figure 1.15. Neutral detergent ﬁber content Of forage dry matter Offered over time

from all treatments, modeled with yield data from 2005 ................................... 63
CHAPTER 2

Figure 2.1. Average daily gain of steers by treatment and year at intervals

throughout the 2003 grazing seasons .......................................................... 105
Figure 2.2. Average daily gain of steers by treatment and year at intervals

throughout the 2004 grazing seasons .......................................................... 106
Figure 2.3. Average daily gain of steers by treatment and year at intervals

throughout the 2005 grazing seasons .......................................................... 107
Figure 2.4. Average daily gain Of steers between 2003 and 2005 ......................... 109
Figure 2.5. Number of AUGD ha"l accumulated by each treatment in 2003 ............ I 1 1
Figure 2.6. Number of AUGD ha'] accumulated by treatment in 2004. 112

Figure 2.7. Number of AUGD per hectare accumulated by each treatment in 2005.1 13

Figure 2.8. Mean number of AUGD per hectare accumulated by the CS-BBS

treatment from 2003 through 2005 ............................................................ 1 15
Figure 2.9. Mean number of AUGD per hectare accumulated by the CS-Only

treatment from 2003 through 2005 ............................................................ 116
Figure 2.10. Mean number of AUGD ha'I accumulated by the CS-SG treatment

from 2003 through 2005 ........................................................................ l 17
APPENDIX 1

Figure A. 1. Positioning Of treatment replications within the experimental area ....... 189
Figure A.2. Pasture layout with respect to alfalfa abundance ............................ 190

Figure A.3. Precipitation by month for Hickory Corners, M1 (KBS-LTER, 2005). .. 191

Figure A.4. Mean air temperature by month for Hickory Comers, MI
(KBS-LTER, 2005) .............................................................................. 192

xii

Figure A.5. Maximum air temperature by month for Hickory Comers,
MI (KBS-LTER, 2005) ......................................................................... 193

APPENDIX II

Figure 8.1. Forage grass percentage of total CS pasture dry matter from
2003-2005. ....................................................................................... 201

Figure 8.2. Forage legume percentage of total CS pasture dry matter from
2003-2005. ....................................................................................... 202

Figure 8.3. Palatable weed percentage Of total CS pasture dry matter from
2003-2005. ....................................................................................... 203

Figure 3.4. Undesirable weed percentage of total CS pasture dry matter from
2003-2005. ....................................................................................... 204

Figure 8.5. Crop percentage of big bluestem and switchgrass pastures from the
CS-BBS and CS-SG treatments, respectively ................................................ 208

xiii

INTRODUCTION AND BACKGROUND INFORMATION

During the mid-summer months of most years, Michigan producers who keep
livestock on pasture are faced with a choice: to remove their livestock from dormant
summer pastures to locations where they are fed stored forage, or to allow their livestock
to continue grazing. The choice to leave livestock on pasture during periods of
inadequate pasture productivity can result in overgrazing. Overgrazing often reduces
pasture productivity (Coleman, 2007) resulting in weed invasion (Hazell, 1967), and
therefore may require renovation or replanting. In southwest Lower Michigan, spring
and fall cool season grass-legume growth is excellent due to mild temperatures and
adequate rainfall, but drought stress during the summer months often causes these grasses
and legumes to become dormant (Laude, 1953). Several state universities in the North
Central Region promote the use of native perennial warm season grasses, such as
switchgrass and big bluestem as summer forage in some circumstances (Anderson, 2000;
Bamhart, I994; Bartholomew). These grasses have the potential to provide large
volumes of high quality mid-summer forage when cool season grasses and legumes are
less productive (Balasko, 2003). While both species are indigenous to Michigan, no
grazing research has been conducted to determine if using these species in a pasture
system is a viable option. Beyond their capacity to produce large quantities of biomass,
switchgrass and big bluestem grow well in marginal soils and can provide excellent cover
for nesting birds (George et al., 1979; Tober, 1992) and for other types Of wildlife and
can reduce the loss of sediment, nitrogen, and phosphorus in areas prone to erosion

(Blanco-Canqui, 2004). Switchgrass (Ma, 2000) and other native tallgrass prairie species

are believed to sequester large amounts of carbon in the soil, and grasslands in general
present great aesthetic appeal (Keeney, 2007), particularly in the summer and fall.

The Michigan Hay and Grazing Council has described alternative forage research
as one of its top priorities (Lindquist, 2002). Both the Michigan Department of Natural
Resources (Sargent, 1999) and the USDA Natural Resource Conservation Service
(USDA-NRCS, 1996) have encouraged Michigan graziers to include native warm season
grasses in their grazing systems. However, graziers may be reluctant to use these grasses
without knowledge of their performance in Michigan. This project was designed to
demonstrate how livestock perform when these grasses are integrated into the typical
grazing systems in Southern Michigan.

Literature review

Switchgrass and big bluestem are both erect—growing native perennial warm
season grasses. They have C4 metabolism, so peak biomass production occurs at a leaf
temperature of 37°C, in contrast to cool season grasses which peak between 20 and 25°C
and dramatically decline at temperatures above 27°C (Nelson, 1995). This physiological
characteristic makes warm season grasses most productive in mid-summer. Big
bluestem, switchgrass, and indiangrass have been used extensively in US. government
programs such as the Soil Bank and the Conservation Reserve Program since the 1930’s,
resulting in several million hectares being replanted with either mixtures or monocultures
of these grasses (Moser, 1995). The extensive use of these grasses in the Conservation
Reserve Program demonstrates their contribution to grassland ecosystems and their role
in providing wildlife habitat (USDA-NRC S, 1999; USDA-NRCS, 2007). The ecological

importance of these grasses is widely recognized (Harvey, 2000; Henning, 1993; USDA-

NRCS, 1999) and they can be planted along railroad right-of-ways and roadsides, near
waterways, for wildlife cover (Moser, 1995), for erosion control (Blanco-Canqui, 2004;
Rankins, 2001), and can function well as vegetative conservation buffers to prevent soil
pesticide loss via surface water (Rankins, 2001). Both big blueStem and switchgrass
tolerate a soil pH of 4.5 and 4.9, respectively (Duke, 1978) and thus can be planted in
areas that are not typically used for crop production. Switchgrass and big bluestem both
have extensive root systems that penetrate the soil to depths Of more than two meters
(Weaver, 1954) and Show great potential for carbon sequestration. Beyond having vast
potential for wildlife habitat, erosion control, and carbon sequestration, these grasses are
recommended by many research and extension institutions for summer grazing. Mitchell
(1996) suggests that summer grazing is more efﬁcient if separate pastures of warm and
cool season grasses are maintained; placing grazing animals in the warm season pastures
during mid-summer, and then returning them to cool season pastures after cool season
pasture recovery. Henning (1993) points out that a combination of cool season and warm
season grass pastures can provide a more constant supply of high-quality feed through the
summer than either cool or warm season grass pastures can provide alone. Bamhart
(1994) indicates that pasture efﬁciency may be increased by converting one-fourth to
one-third of the cool season grass pasture acreage to a warm season grass pasture to be
grazed at different times during the grazing season. This allows the cool season grasses
to be grazed in the Spring and early summer, occasionally throughout the summer, and in
the fall, while utilizing the warm season grass pasture more intensively during periods of

low cool season grass productivity. Also, this strategy allows greater rest periods for cool

season pastures, which will increase their vigor and productivity for late summer through
early fall grazing.

Recommendations for inclusion Of warm season grasses in grazing systems are
supported by numerous studies in many states. Kreuger and Curtis (1979) conducted a
study in South Dakota comparing switchgrass, big bluestem, indiangrass (Sorghastrum
nutans (L.) Nash), and sideoats grama (Bouteloua curtipendula) for mid—summer grazing
and concluded that switchgrass and big bluestem are useful species for beef production in
July and August. George (1996) demonstrated that grazing either big bluestem or

switchgrass pastures in mid-summer can result in impressive steer weight gains (1.42 and
-l . . . . .
1.1 1 kg day , respectively) and that rotational grazmg management 15 superior to

continuous grazing management for both switchgrass and big bluestem. In an Iowa
study, (Moore et al., 2004) concluded that, in some circumstances, sequentially grazing
cool- and warm-season grasses may be a desirable alternative to remaining on cool-
season pastUreS throughout the summer. This management strategy is an option because
warm-season grasses are highly productive during the mid-summer when cool-season
grasses are typically less productive. Much of this improved seasonal productivity is due
to the resource use-efﬁciency of warm season grasses for water (Ehleringer, 1993; Moser,

2004; Wedin, 2004) phosphorus (Morris, 1982), and nitrogen (Brown, 1985). Brown
(1985) observed that the C4 grasses in his experiment synthesized twice as much dry

matter per unit of nitrogen taken up from a soil with low levels of N. While the
productivity of warm-season grasses is high, the forage quality is low and tends to

decline rapidly, which can result in depressed livestock performance (Moore et al., 2004).

Bamhart (1994) presents data that shows dry matter yield increasing with
increasing nitrogen inputs, but the relationship is not linear; approximately 75 kg ha'l of
nitrogen is optimal, considering biomass production and crude protein responses.
Switchgrass matures earlier in the summer than big bluestem and the forage quality of big
bluestem does not decline as rapidly as it does with switchgrass (Moser, 1995). Both
switchgrass and big bluestem pastures must be managed differently than cool season
grass pastures. Neither species tolerates close grazing and both require rest periods of 21
to 45 days between grazing events, depending on environmental conditions (Anderson,
2000; Henning, 1993; Mitchell, 1996).

Objectives and Hypotheses
0 Objective 1: To compare the productivity of a traditional Michigan cool season
grass-legume grazing system with the productivity Of two model systems that
include warm season grasses, in the ﬁrst three years after planting the warm
season grasses.

o Hypothesis la: The grazing system models that are integrated with warm
season grasses will result in higher livestock weight gains (kg/ha) than
typical Michigan grazing system model which includes only cool season
grasses and legumes.

o Hypothesis lb: An integrated grazing system that includes big bluestem
will result in higher livestock weight gains (kg/ha) than an integrated
grazing system that includes switchgrass.

0 Objective 2: To compare the proﬁtability of the above systems over three years of

experimentation.

o Hypothesis Za: When considering all inputs and outputs, the initial cost of
establishing the warm season grass-integrated grazing systems will be
offset by their higher productivity by the end of the third year.

0 Hypothesis 2b: Big bluestem-integrated grazing systems will not be
Signiﬁcantly more proﬁtable than switchgrass-integrated grazing systems
by the third year of experimentation.

Objective 3: To describe songbird species diversity in the respective pasture
systems, describing apparent relationships between species Spatial distribution
and type of pasture system.

0 Hypothesis 3: Pasture systems that include switchgrass and big bluestem
will have a higher songbird species diversity, nesting success, and

evidence greater use by raptors feeding on mice and other rodents

LITERATURE CITED

Anderson, B. 2000. Grazing management on warm season grasses Missouri Forage and
Grassland Council Annual Meeting. Universiy of Missouri, Lake Ozark, MO.

Balasko, J.A., and C. J. Nelson. 2003. Grasses for northern areas, p. 125-147, In R. F.
Barnes, Nelson, J. C., Moore, K. J., and M. Collins, ed. Forages: an introduction
to grassland agriculture, Vol. I.

Bamhart, SK. 1994. Warm-season grasses for hay and pasture - Pm-569, pp. 4, In I. S.
U. Exension, (ed.). Iowa St. U. Ext.

Bartholomew, H.M., Sulc, R.M., Hendershot, R., and J. Cline. Perennial warm season
grasses for Ohio AGF-022-95. Ohio St. U. Ext. Dept. of Hort. and Crop Sci.,
Columbus.

Blanco-Can‘qui, H., Gantzer, C. J., Anderson, S. H., Alberts, E. E., and A. L. Thompson.
2004. Grass barrier and vegetative ﬁlter strip effectiveness in reducing runoff,

sediment, nitrogen, and phosphorus loss. Soil Science Society of America:l670-
1678.

Brown, RH. 1985. Growth of C3 and C4 grasses under low N levels. Crop Science
25:954-957.

Coleman, S.W., Sollenberger, LE. 2007. Plant-Herbivore Interactions, p. 123-136, In R.
F. Barnes, Nelson, J. C., Moore, K. J., and M. Collins, ed. Forages: the science of
grassland agriculture, Vol. 2, 6th ed.

Duke, J.A. 1978. The quest for tolerant germplasm, p. 1-61, In M. Stelly, ed. Crop
tolerance to suboptimal land conditions. American Society of Agronomy, Crop
Science Society of America, Soil Science Society of America, Madison.

Ehleringer, J.R., and Monson, R. K. 1993. Evolutionary and ecological aspects of
photosynthetic pathway variation. Annual Review of Ecology and Systematics
24:41 1-439.

George, RR, A.L. Farris, C.C. Schwartz, D.D. Humburg, and J.C. Coffey. 1979. Native
prairie grass pastures as nest cover for upland birds. Wildlife Society Bulletin 7:4-
9.

Harvey, R.G. 2000. Establishing prairie plants for CRP or wildlife habitat with
herbicides. University of Wisconsin Extension.

Hazell, DB. 1967. Effect of Grazing Intensity on Plant Composition, Vigor, and
Production. Journal of Range Management 20:249-252.

Henning, J.C. 1993. G4673: big bluestem, indiangrass, and switchgrass [Online].
Available by University of Missouri Extension
http://muextension.missouri.edu/explore/agguides/crops/g04673.htm (veriﬁed
03/22/06).

Keeney, D.R., Sanderson, M. A. 2007. Forages and the environment, p. 167-176, In R. F.
Barnes, Nelson, J. C., Moore, K. J., and M. Collins, ed. Forages: the science of
grassland agriculture, Vol. 2.

Krueger, CR, and DC. Curtis. 1979. Evaluation Of Big Bluestem, Indiangrass, Sideoats

Grama, and Switchgrass Pastures with Yearling Steers. Agronomy Journal
71:480-482.

Laude, HM. 1953. The nature Of summer dormancy in perennial grasses. Botanical
Gazettez284-292.

Lindquist, J. 2002. 2002 Michigan Hay and Grazing Council Industry Priorities.
Ma, 2., Wood, W., and D. I. Bransby. 2000. Soil management impacts on soil carbon
sequestration by switchgrass. Biomass and Bioenergy 18:469-477.

Mitchell, R., Moser, L., Anderson, 3., and S. Waller. 1996. Switchgrass and big bluestem
for grazing and hay, G94-1198-A [Online] expired (veriﬁed 01/29/2003).

Moore, K.J., T.A. White, R.L. Hintz, P.K. Patrick, and EC. Brummer. 2004. Forages and
pasture management - Sequential grazing of cool- and warm-season pastures.
Agronomy Journal 96:] 103-1111.

Mon'is, R.J., Fox, R. H., and Jung, G. A. 1982. Growth, P upake, and quality of warm
and cool—season grasses on low available P soil. Agronomy Journal 74:125-129.

Moser, LE, and K. P. Vogel. 1995. Switchgrass, big bluestem, and indiangrass, In R. F.
Barnes, Darrell, A.M., and CJ. Nelson, ed. Forages: An introduction to grassland
agriculture, Vol. 1, Fifth ed. Iowa State UP, Ames.

Moser, L.E., Burson, B. L., and Sollenberger, L. E. 2004. Warm-season (C4) grass
overview, p. 1-14, In L. E. Moser, Burson, B. L., and Sollenberger, L. E., ed.
Agronomy NO. 45: Warm-season (C4) grasses. American Society of Agronomy:

Crop Science Society of America: Soil Science Society of America, Madison,
Wis.

Nelson, C]. 1995. Photosynthesis and carbon metabolism, p. 31-43, In D. R.F., A.M.,
and CJ. Nelson, ed. Forages: an introduction to grassland agriculture, Vol. 1, Fifth
ed. Iowa State UP, Ames.

Rankins, A.J., Shaw, D. R., and Boyette, M. 2001. Perennial grass ﬁlter strips for
reducing herbicide losses in runoff. Weed Science 49:647-651.

Sargent, M.S., Carter, K. S., eds. 1999. Grassland management, p. 297 Managing
Michigan wildlife: 3 landowners guide. Michigan United Conservation Clubs,
East Lansing.

Tober, DA, and A. D. Chamrad. 1992. Warm-season grasses in the northern great
plains. Rangelands 14:227-230.

USDA-NRCS. 1996. Establishing warm-season grasses, pp. 3, In USDA-NRCS, (ed.)
Hay and pasture management. USDA-NRCS, Columbia, Missouri.

USDA-NRCS. 1999. Job sheet for native warm season grass establishment as wildlife
habitat (645), South Carolina.

USDA—NRCS. 2007. Georgia Job sheet (645): Native grasses for wildlife habitat, pp. 3.

Weaver, J.E. 1954. North American Prairie Johnsen Publishing Company, Chicago.

Wedin, DA. 2004. C4 grasses: resource use, ecology, and global change, p. 15-50, In L.
E. Moser, Burson, B. L., and Sollenberger, L. E., ed. Agronomy No. 45: Warm-

season (C4) grasses. American Society of Agronomy, Crop Science Society of
America, Soil Science Society of America.

Chapter 1

SEASONAL FORAGE DYNAMCIS OF COOL-SEASON PASTURES INTEGRATED
WITH SWITCHGRASS AND BIG BLUESTEM

ABSTRACT
High temperatures and lack Of precipitation Often cause the productivity and quality of
cool-season pastures in Michigan to severely decline for an extended period during the
summer. This study was conducted to determine whether integration of switchgrass or
big bluestem into cool-season grazing systems would mitigate this period Of low pasture
productivity and quality. Cool-season grass and legume pastures (CS-Only) were
compared to integrated big bluestem and cool-season grass and legume pastures (C S-
BBS) and integrated switchgrass and cool-season grass and legume pastures (CS-SG).
The seasonal dynamics of forage dry matter offered, crude protein, neutral detergent
ﬁber, and acid detergent ﬁber were used to compare the dynamics of the pasture
productivity and quality in each of three pasture systems. Dry matter offered in the C S-
Only treatment was higher earlier in the grazing season and the peak in dry matter-
offered occurred between 20 and 30 days earlier in the C S-Only treatment than in the
other treatments. The seasonal trend for crude protein was generally higher and more
constant for the CS-Only treatment than for the CS-BBS and CS-SG treatments, which
were very similar to each other. The seasonal trend for acid detergent ﬁber and neutral
detergent ﬁber (NDF) was generally lower and more constant for the CS-Only treatment
than for the CS-BBS and CS-SG treatments, which were very similar to each other.
These ﬁndings suggest that, compared to the CS-Only treatment, the CS-SG or the C S-
BBS treatments probably will not improve livestock gain or improve the distribution Of

animal grazing days throughout the summer.

10

Introduction

Switchgrass and big bluestem have been promoted as forage for livestock
(Anderson, 2000; Bamhart, 1994; Bartholomew, 1995) and are also known to provide
valuable wildlife habitat (Delisle and Savidge, 1997; Sample and Mossman, 1997). Thus
far, little research has been conducted in the Great Lakes region to describe livestock
performance on grazing systems that include big bluestem or switch grass. Because
livestock performance is closely related to forage quality and abundance, it is important
to measure these parameters in addition to the response of the grazing animal. In order to
explain animal response, it is helpful to describe pasture system dynamics in terms of
forage dry matter yield, and forage quality parameters including crude protein, acid
detergent ﬁber, and neutral detergent ﬁber.

One of the primary objectives of the study was to compare the productivity of a
traditional Michigan cool season grass-legume grazing system with the productivity of
two model systems that include warm season grasses, in the ﬁrst three years after
planting the warm season grasses. The associated hypotheses were:

1. The grazing system models that are integrated with warm season grasses will result in
higher livestock weight gains (kg ha'l) than a typical Michigan grazing system model

which includes only cool season grasses and legumes.

2. An integrated grazing system that includes big bluestem will result in higher livestock
weight gains (kg ha'l) than an integrated grazing system that includes switchgrass.

These hypotheses were evaluated in Chapter 3 of this dissertation. The subject of this

chapter is pasture quality and the dynamics of forage dry matter Offered to the grazing

ll

animals, both of which will help to understand why the above hypotheses were or were
not validated.

LITERATURE REVIEW
Measurement of pasture forage yield

While there are numerous parameters that can be measured in pasture research,
the high cost of conducting these investigations dictates that only those most directly
related to the objectives be measured. Burns (1989) recommends that herbage mass,
green leaf mass, forage quality, herbage density, and botanical composition data be
collected for all grazing experiments. Depending on the objectives of the experiment,
measurements such as grazing time, bite Size, and leaf area index may also be taken.

Accurate appraisal of the quantity of forage present is a difﬁcult procedure
because of the high degree of natural variation across most pastures. The most accurate
means of estimating forage yield is direct sampling (i.e. clipping a quadrat of known
area), but the number of samples required to obtain a reliable mean and the amount of
labor required to perform this work is usually prohibitive, and indirect methods are
necessary (Wilm, 1944).

There are several methods which can be used to estimate forage yield including
capacitance meters, rising plate meters, visual estimates (O’Donovan, 2002), and grazing
rulers (Sanderson et al., 2001). The electronic capacitance meter uses the difference of
the dielectric constant between the air and the herbage to describe the quantity of herbage
present (Sanderson et al., 2001). Although they can be fabricated in many ways, a basic
type of rising plate meter can be made by taking a wood, metal, or plastic plate of known

area and mass and cutting a hole in the middle; the hole should be large enough for a 2.5

12

cm dowel to easily pass through. The plate, which rests at the bottom of the dowel, is
connected by the corners to the top of the dowel by string. The dowel is then marked
along the axis, from the bottom upward, with length graduations. The rising plate meter
is used by holding the top end of the dowel and setting the plate on the pasture forage
surface (Scrivner et al., 1986). The forage supports the plate at a certain height above the
ground, while the dowel passes through to the ground. The graduation closest to the
intersection of the plate and the dowel is the height measurement recorded. This
measurement describes the total standing crop (Scrivner et al., 1986) and is calibrated to
actual pasture dry matter per unit area by means of double—sampling (Wilm, 1944).
(Earle and Mcgowan, 1979) indicate that more readings are necessary to estimate herbage
biomass in pastures where there is a high degree of yield variability. While the rising
plate meter is commonly accepted as a way to determine pre-grazing herbage, Stockdale
and Kelly (Stockdale and Kelly, 1984) suggest that it is not appropriate for estimating
heavily trampled post—grazing residue.

When using the rising plate meter to determine total herbage dry matter weight, a
wide range of dry matter values should be chosen for double-sampling in the pasture
being evaluated (Lucas, 1990). Linear regressions are used to relate rising-plate meter
readings to dry matter yield by either pooling all data from the grazing season (Michell,
1982; Scrivner et al., 1986) or by pooling data from different parts of the grazing season
(Michell and Large, 1983). If conditions (e.g. lodged forage) indicate that the rising-plate
meter method is not appropriate (Stockdale, 1984), then direct sampling is an appropriate

means to estimate pasture dry matter yield (Stockdale and Kelly, 1984).

13

Opinions about the usefulness of these techniques for indirectly measuring forage
dry matter in pastures vary widely. Scrivner et al. (1986) concluded that “the rising plate
meter was a quick and effective way Of assessing total forage growth and utilization.”
(Gabriels, 1993) concluded that “at the present it is impossible to make an accurate
prediction of the dry matter of a certain site on the basis of non-destructive
measurements.” However, the reason that there is so much research on indirect pasture
measurements is that most pasture researchers recognize that measurement of pasture
productivity is necessary and that direct measurement is not practical or ﬁnancially
feasible.

Measurement of pasture forage botanical composition and forage quality

The purpose of planting a pasture with a diverse mixture of species is to increase
productivity, improve the quality of the forage, and to provide more uniformity in forage
availability throughout the grazing season (Henson, 1941). Thus, monitoring botanical
composition is an important aspect of pasture studies. Changes in botanical composition
can reflect the impact of management practices (VanKeuren, 1957) as well as animal
preference and hence, should be monitored periodically through the grazing season.
Weed dynamics in particular agricultural systems are influenced by the patterns of
disturbance of plants and soil. Grazing systems range from continuous to intensive-
rotational in nature and speciﬁc weed species can either be favored or selected against.
For example, Canada thistle is favored by continuously grazing pastures, while high-
intensity, low-frequency grazing reduces the density of Canada thistle (De Bruijn, 2006).
Alternately, excessive grazing under particular conditions can cause other weed Species

to proliferate (Harker, 2000).

14

One method of determining botanical composition is hand-separation of the
forage into the component Species and calculating the relative contribution of each
species to the dry matter weight. This method is labor intensive, expensive, and time
consuming and there have been numerous attempts to develop indirect methods of
determining botanical composition (VanKeuren, 1957). VanKeuren and Algren
compared (1957) the inclined point quadrat method (Amy, 1942; Mannetje, 1963) of
estimating botanical composition with the vertical point quadrat method, visual
estimation, and hand-separations. They concluded that the inclined point quadrat method
had the highest correlation and the least variation Of the indirect methods. This method
uses quadrats with ten needles ﬁxed within them. These quadrats are placed randomly
multiple times in the area to be assessed. The species of plant through which each needle
passes every time the quadrat is placed is recorded, and the botanical composition of the
pasture is derived using the resulting data.

Mannetje (1963) describes another indirect method of estimating botanical
composition called the dry-weight-rank method. This procedure is carried out by taking
50 to 80 visual ratings from within quadrats placed within the pasture that is being
described. Visual estimation is used to rank which species is ﬁrst (1), second (2), and
third (3) in dry matter production within each quadrat. The total number of times that a
species receives any of the rankings (1, 2, or 3) is calculated as well as the total number
of times that species is given a speciﬁc rating of l, 2, or 3. Relative proportions of ‘1’,
‘2’, and ‘3’ ratings are calculated and multiplied by 70.2, 21.1, and 8.7, respectively. The

sum of these three numbers represents the percent composition of that Species in the

15

pasture at that time. Mannetje concluded that the dry-weight-rank method provides an
accurate estimation of botanical composition by weight in a wide variety of pasture types.

Forage quality including crude protein (crude protein), neutral detergent ﬁber
(neutral detergent ﬁber), and acid detergent ﬁber (acid detergent ﬁber) are frequently
appear in forage quality reports (Collins, 2003) and are frequently used in studies where
livestock performance is of concern (Karsli et al., 1999; Wen et al., 2002). There are
standard wet-chemistry techniques that are used for each component, but they are Often
used in conjunction with near infra-red reﬂectance spectroscopy (N IRS) (Hart, 1989).

MATERIALS AND METHODS

This experiment was conducted at W. K. Kellogg Biological Station in Hickory
Corners, M1, on Kalamazoo series (ﬁne-loamy, mixed, mesic Typic Hapludalfs) soils.
Four replications of three treatments were assigned to the experimental area using a
completely randomized design. The experimental layout is described in Appendix I
Figure 1.1.
The three treatments included:

1. A cool-season grass-legume pasture (CS-Only) representative of pastures found
throughout the Lower Peninsula of Michigan. These pastures were comprised
primarily of perennial ryegrass (Lolium perenne), quackgrass (A gropyron repens),
alfalfa (Medicago sativa), white clover (T rifolium repens), red clover (Trifolium
pratense), orchardgrass (Dactylis g/omerata), and tall fescue (F estuca
arundinacea).

2. An integrated big bluestem and cool-season grass-legume pasture. One-third of

the system is a monoculture of big bluestem that was planted in 2002, and two-

16

thirds is composed of the cool-season grasses and legumes listed above (Bamhart,
1994; Undersander, 2002).

3. An integrated switchgrass and cool—season grass-legume pasture. One-third of the
system is a monoculture of switchgrass that was planted in 2002, and two—thirds is
composed of the cool-season grasses and legumes listed above (Bamhart, 1994;
Undersander, 2002).

All pastures had an area of 1.6 ha with the exception of two which, because of the
shape of the experimental area and existing permanent fencing, were somewhat smaller
or larger. The cool-season grass and legume (henceforth, ‘cool-season’) portions Of all
treatments predated this experiment. The major cool season Species included: perennial
ryegrass, Kentucky bluegrass, white clover, quackgrass, and alfalfa. In the planning
stages of this experiment, differences in species abundance between replications were
considered negligible and blocking was not considered. After the experiment began, it
became apparent that the variability in ‘alfalfa abundance’ was signiﬁcant and should be
considered a blocking effect. All pastures which were planted with alfalfa in 1994 were
considered were included in ‘alfalfa abundant’ block. The pasture layout with respect to
experimental treatments and alfalfa abundance is described in Appendix I Figure A2.
2002 Warm-season Pasture Pre-planting Preparation

On April 26 all areas which were to be planted with switchgrass or big bluestem

. -l . . . .
were treated wrth glyphosate at a rate1 of 1.12 kg ha . At this time, eXIstIng cool-season

grasses were approximately 20-25 cm tall. On May 22 and 23 we attempted to burn the

 

‘ All pesticide rates are described in terms of kg ha"of active ingredient applied.

17

residue of the herbicide-killed pasture, but the low volume of biomass prevented a
complete bum.
2002 Planting

The big bluestem and switchgrass monoculture portions of the CS—BBS and CS-
SG treatments were planted between May 28 and May 30. The seeding equipment was a

Truax rangeland drill with a 1.5 m planting width. Niagara big bluestem was planted

with a target rate of 10 kg pure live seed (PLS) ha". F orestburg switchgrass was planted
in the CS-SG pastures at a rate of 10 kg PLS ha-l.

2002 Weed and Insect Control

On June 7 the area that had been planted with switchgrass or big bluestem was

sprayed with a herbicide combination that delivered 0.07 kg ha.1 imazethapyr and 0.91

kg ha-1 glyphosate. At the time Of herbicide application, less than 5% of crop seedlings

had emerged. On June 12, scouting revealed that many of the emerging switchgrass
seedlings were being damaged by insects; some were being consumed to ground level
while others were fenestrated. Further investigation suggested that the fenestration was
caused by Stewart’s bacterial wilt (Munkvold, 2002), which can be transmitted by com
flea beetles (Chaetocnema pulicaria M.) as well as other species of ﬂea beetles (N CSU,
no date). Other cutting and chewing damage was attributed to thrips (order

T hysanoptera). On June 20, big bluestem and switchgrass pastures were sprayed with

1.12 kg ha”1 of carbaryl. On August 3 clopyralid was Sprayed on all big bluestem and

. ' -l .
swrtchgrass pastures at a rate of 0.28 kg ha , In order to control broadleaf weeds.

18

2003 Switchgrass Pasture Reseeding and Subsequent Insect and Weed Control

The switchgrass pastures that had been planted in 2002 were assessed in the
spring of 2003. At that time, it was evident that ﬂea beetle and Stewart’s bacterial wilt
damage from summer of 2002 had damaged the switchgrass stand enough to cause stand

failure and replanting was deemed necessary. Switchgrass pastures were replanted with
Forestburg switchgrass on May 19, 2003 at a rate of 14.3 kg ha]. Because the ground
was relatively free Of debris and living plants, a conventional three meter drill was used
for replanting. The big bluestem pastures planted in 2002 had not been damaged by

Stewart’s bacterial wilt and did not need re-planting.

On June 10, 2003, when the switchgrass seedlings in the re-seeded pastures had
reached a height of approximately 2.5 cm, carbaryl insecticide was applied at a rateI of
1.68 kg ha], in order to control the flea beetle and thrip populations. The switchgrass

portion of the CS-SG treatment was not grazed in 2003.
On July 17, 2003, a combination of herbicide and insecticide mixture was used to

control weeds and insects:

o Carbaryl (flea beetles/thrips): 1.68 kg ha-1
. 2,4-D amine (broadleaf weeds): 1.06 kg ha.1

0 Imazethapyr (broadleaf and grass weeds): 0.07 kg ha.1

2002-2003 C ool-season Pasture Improvement
Some of the pasture systems had been planted with alfalfa in 1994 while others had not.
On April 18, 2003, white clover (KOpu II, Ampac Seed Company) was drilled into all

cool-season pastures in order to make the legume composition more uniform among

19

treatments and replications. The John Deere 750 drill was set to plant 1.63 kg of pure
live seed ha.I into the living sod of all cool season pastures of all treatments.

On August 22, 24, and 25 of 2003, orchardgrass, tall fescue, and white clover
were planted into the dormant (but living) sod of all cool-season portions of all

treatments. A John Deere 750 nO-till drill was used for planting:

o Orchardgrass (Tekapo, Ampac Seed Company): 8.4 kg PLS ha-1
0 Endophyte free tall fescue (Bronson, Ampac Seed Company): 6.0 kg PLS ha.1

0 White clover (Kopu II, Ampac Seed Company): 3.9 kg PLS ha-l

2004-2005 weed control in switchgrass and big bluestem pastures
2004: Broadleaf weeds were controlled in 2004 by spraying the switchgrass and big
bluestem portions of the C S-SG and CS-BBS treatments on April 14-15 with a

combination Of:

o 0.84 kg ha.1 of2,4-D ester

0 0.73 kg ha.1 Of dicamba

2005: Due to competition from quackgrass (Elytrigia repens (L.)), the following
herbicide combination was applied to all big bluestem and switchgrass pastures on April

6 in order to control broadleaf and grass weeds:

. 0.54 kg ha" of dicamba
. 0.81 kg ha" of2,4-D ester

0 1.08 kg ha'1 of glyphosate

20

At the time of application, the quackgrass was approximately 10 cm tall and very few
big bluestem or switchgrass crown buds were showing. The inclusion of glyphosate in
the herbicide mixture virtually eliminated the quackgrass while leaving big bluestem and
switchgrass unharmed.

Control of pokeweed and bull thistle

Due to their noxious and aggressive characteristics, pokeweed (Phytolacca
americana L.) and bull thistle ( C ircisum vulgare) were controlled manually as part of
daily pasture observation.

Pasture Fertility Management

The soils on which‘the experiment was conducted are Kalamazoo series (ﬁne-
loamy, mixed, mesic Typic Hapludalfs). Soil testing was conducted by taking 20 20-cm
soil cores per pasture in 2003-2004. In the case Of the CS-BBS and CS-SG treatments,
twenty soil cores were collected from both cool season and warm—season grass (i.e.
switchgrass and big bluestem) portions and tested separately by the MSU .Soil and Plant
Nutrient Laboratory. Soil tests from 2003 indicated that the soil pH of pastures ranged
between 5.5 and 6.2, indicating a need for lime application. On April 7, 2003, lime was
applied according to soil test recommendations.

Pasture nitrogen fertility was managed by applying urea twice per grazing season
to the cool-season portions of the treatments and once to the switchgrass and big
bluestem monocultures as described in Table 1.1.

Pastures were managed as uniformly as possible, but each was managed
separately and optimally (Bransby, 1989). When grazing was initiated each spring, cattle

on all treatments were given access to the entire cool-season portion of their respective

21

pastures, with the exception of the fourth replication of the CS-Only treatment, where a
concurrent experiment was being carried out which required that livestock be excluded
from a portion of the pasture at certain points of time. As the spring ﬂush of cool-season
pasture growth intensiﬁed, pastures were “staged” (Barker, 1999): cattle were limited to
one half, and later, one quarter of the cool-season portion of their treatments at which
point rotational grazing began and continued until the end of the season. During the ﬁnal
grazing event of the season, the cattle were again given access to the entire cool-season
portion of their respective pastures. In 2003, the big bluestem portion of the CS-BBS
treatments was split into three parts each of which was grazed in rotation (the switchgrass
portion of the CS-SG treatment was not grazed in 2003). In 2003, residue Of refused big
bluestem forage was mowed after grazing as needed in order to make the pasture height
more uniform at subsequent grazing events. However, it was noted that the small size of
the subdivisions within the big bluestem pasture resulted in excessive trampling of forage
which caused a high proportion of plants to re-grow from the crown rather from
intemodes, thereby increasing the amount of time required for regrowth. Further, the
tractor tire tracks left from mowing caused the crushed plants to regrow from the crown;
this regrowth was lush and tended to be grazed preferentially by the livestock. Thus, in
2004 and 2005 the big bluestem and switchgrass portions of the CS-BBS and CS-SG
treatments were not subdivided or mowed and each portion was left undivided and grazed
as one of the rotations.

Decisions about timing and duration of rotations were dictated by current and
anticipated pasture forage availability rather than using strict dates or interval lengths.

Variables such as soil moisture, physiological and re-growth stages of forage species,

22

typical seasonal weather pattems, and weather forecasts were used to make rotation
decisions. After the Spring flush, the grazing intervals were usually seven days for the
CS-BBS and CS-SG treatments; C S-Only treatments were usually rotated every ten days
and consisted of four subdivisions.

Mid—season grazing pressure was adjusted using the put-and-take method2 of
pasture stocking (Mott, 1960). ‘Tester steers’ were left on the same replication of the
same treatment for the entire grazing season. Put-and-take steers were added to or
removed from a particular replication to increase or reduce the grazing pressure in order
to:

o optimize utilization of pasture forage
o prevent the accumulation low-qualityI forage
0 meet but not exceed the current or anticipated forage dry matter production

The protocol for this experiment required equivalent treatment for each
replication to the extent possible, while managing each replication optimally, relative to
grazing intensity, rotation frequency, and grazing initiation/termination dates. Decisions
about timing and duration of rotations were not managed using strict dates or interval
lengths but rather were dictated by current and anticipated pasture forage availability.
Variables such as soil moisture, physiological and re-growth stages of forage species,
typical seasonal weather patterns, and weather forecasts were considered in deciding
when to rotate a group of steers from one pasture location to another and whether to
reduce grazing intensity by removing put-and-take steers from particular pastures.

Within an individual pasture, animals were rotated when one or more of the following

 

2 See Technical Note 1 in Appendix III

23

was true: 1) paddocks forward in the rotations were about to become over-mature; 2)
animals or pasture might be harmed by low forage levels in the current rotation; 3) rapid
pasture growth indicated that Shorter intervals were needed in the immediate future to
prevent accumulation of low-quality forage which would likely be refused.

The anticipated date for termination of grazing was October 1, but pasture
conditions including plant stress, drought, and forage dry matter availability were the
primary considerations for termination of grazing for the season (Bransby, 1989; Leep,
2003). Dates of grazing and termination dates for each year are listed in Table 1.2.

All forage samples collected in this study were placed in paper bags and dried for
a minimum of 48 hours at 43°C3, and then weighed. When grazing was initiated each
year, dry matter availability was determined by harvesting plant tissue from within six
0.25 m2 quadrats. Each time the rotation sequence brought steers to a cool-season portion
of their respective treatment, plants within the quadrats were clipped at a height of
approximately 5 cm. When steers were moved into the big bluestem or sWitchgrass
portions of treatments, plants within quadrats were harvested at approximately 13 cm.
When rotations began, dry matter availability was determined using a rising-plate meter
(Gabriels, 1993; Harmoney et al., 1997; Michell, 1982) unless conditions (lodging of
forage, rain) were not appropriate, in which case quadrat clippings were used instead.
When the rising-plate meter was used to measure forage availability, 100 rising-plate
meter readings4 were taken from the paddock into which the animals were about to be
moved. Three of the rising-plate meter measurements from each pasture were double-

sampled at the time of rotation in order to calibrate the rising-plate meter (Harmoney et

 

3 See Technical Note 2 in Appendix 3
4 See Technical Note 3 in Appendix 3

24

al., 1997). Double-sampling was performed by measuring the forage with the RPM and
then overlaying the rising—plate meter with a 0.209 m2 quadrat, removing the rising-plate
meter, and harvesting the forage within the quadrat as described above. A range of
quadrat rising-plate meter values were chosen for double-sampling at the time of rotation
in order to represent the range of values present in the pastures being assessed (Lucas,
1990). Linear regression was used to relate rising-plate meter readings to dry matter
yield (Michell, 1982; Scrivner et al., 1986; Stockdale and Kelly, 1984). When a rising-
plate meter needed to be replaced and the replacement rising-plate meter had a different
weight, a new regression was performed for the replacement rising—plate meter.
Regressions for cool-season, big bluestem, and switchgrass pastures were performed
separately. Following weighing, the sub-samples were combined and ground through a 1
mm screen in preparation for laboratory analysis.

For cool-season pastures, regressions for Spring pasture production were based on
rising-plate meter data collected between May 4 and June 23 of all years. Data collected
from June 24 through September 10 of all years was used for the second regression. For
big bluestem and switchgrass monocultures, regressions for spring pasture production
were based on rising-plate meter data collected up to and including June 20 of 2003 and
2004. Data collected from June 21 onward was used for the second regression.
Regression coefﬁcients changed and correlation coefﬁcients were improved when the
separate regressions were performed for spring and summer pastures. In many cases,
logarithmic transformation of rising-plate meter data was necessary (Gabriels, 1993).

If conditions in the CS-portion of treatments were not appropriate for using the

rising-plate meter (Stockdale, 1984), then six 0.209 m2 quadrats were clipped from each

25

paddock before rotation (Stockdale and Kelly, 1984). In 2003 and 2004, yield for the big
bluestem and switchgrass monocultures yield was determined using the rising-plate meter

when conditions (precipitation, wind) permitted; in 2005, quadrats were used exclusively.

When quadrats were used to estimate forage yield, forage from within 0.209 m2 quadrats

was harvested at a height of approximately 13 cm to Obtain an estimate of forage dry
matter present. In 2003, four quadrats were used to determine yield and quality. In 2004
and 2005 dry matter yield in big bluestem and switchgrass monocultures was estimated
for to each grazing event by taking the average weight of dried forage from six quadrat
clippings.
Botanical Composition

Botanical composition was determined differently for the big bluestem
monocultures than for the cool-season portions of the treatments. The botanical
composition of the warm-season grass portions, was determined by clipping four quadrats
to grazing height (approximately 13 cm) and manually separated into ‘weed’ and ‘crop’
components. Each component from each quadrat was placed into a separate bag, dried
for 48 hours at 43°C, and weighed. Warm season grass botanical composition is
described by the percentage of the total sample dry matter made up of the crop (i.e., big
bluestem or SG) planted to that pasture. This data was collected from each paddock at
the time of each rotation (Burns, 1989).

The botanical composition of the cool-season portions Of the systems were
assessed two times in 2003 and 2004 and three times in 2005 by using the dry—weight
rank method (DWRM) (Mannetje, 1963). The DWRM was performed by making visual

ratings of the forage contained in a 0.209 m2 quadrat that was non-selectively placed on

26

the ground 80 or more times in each pasture while following a zig—zag (Z-shaped) pattern
or making multiple linear transects. Workers visually estimated and ranked the forage
species which contributed most to the total plant dry matter contained within the quadrat
as ‘1’; the second most prevalent plant species was given a rank of ‘2’; and the third most
prevalent as ‘3’. If only one or two plant species was contained by the quadrat, the plant
species present did not receive more than one ranking each. If the worker was unable to
distinguish which of two or three species was most prevalent, they were both given the
same two or three numbers. For example, if perennial ryegrass was the most prevalent
species, but the worker was unable to tell whether orchardgrass or white clover was the
next most prevalent species, both the orchardgrass and the white clover would receive a
ranking of ‘2 and 3’. The total number of times that a species receives any of the
rankings (1, 2, or 3) was calculated as well as the total number of times that species was
given a speciﬁc rating of l, 2, or 3. Relative proportions of ‘1,’ ‘2,’ and ‘3’ ratings were
calculated for each species and then multiplied by 70.2, 21.1, and 8.7, respectively. The
sum of these three numbers represents the percent composition of that species in the
pasture (Mannetje, 1963).

In all years, botanical composition of big bluestem and switchgrass monocultures
was based on hand-separations of the harvested forage from four quadrats. Forage
quality of the harvested big bluestem and switchgrass subsamples was determined using
the same procedures used to determine the forage quality of samples collected from the
cool-season portions of the treatments. Because the rising-plate meter is not useful for

determining post-grazing residue (Stockdale, 1984).

27

After drying, forage dry matter was ground ﬁrst using a Wiley Mill with a 2 mm
screen and then through a 1 mm screen with either a UDY Cyclone Mill (Udy Mill
Corporation, Fort Collins, CO) or a Christy-Turner Lab Mill (Christy-Turner Ltd.,
Ipswich Suffolk, UK); subsamples from the same date, paddock, and rotation were
combined for analysis. Forage quality parameters including crude protein, neutral
detergent ﬁber, and acid detergent ﬁber were determined by using wet-chemistry
techniques. In-house modiﬁcations5 of the Modiﬁed Van Soest method (Stern, 1991)
were used to determine neutral detergent ﬁber and acid detergent ﬁber concentrations.
The Hach modiﬁed Kjeldahl procedure (Hach, 1985; Watkins, 1987) was used to
determine total N concentration in the forage.

Pasture and animal performance were also described in this experiment. The
animals used in this study were Holstein steers that weighed approximately 240 kg at the
beginning of the grazing season. The put-and-take method of pasture stocking (Mott,
1960) was used (modiﬁcations to the protocol are described in Appendix 11). Data
calculated from animal measurements includes: average daily gain, gain per hectare,
animal unit grazing days per hectare (see Chapter 3 of this dissertation), and net income
per hectare (see Chapter 4 of this dissertation).
Statistical Analysis
PROC MIXED of SAS (SAS Institute, 2000-2003) was used to conduct analysis of
variance tests for forage quality and quantity parameters. Transformations necessary to
normalize the distribution Of studentized residuals are given in Table 1.3.

Because rotation decisions were made according to the current and anticipated

forage dry-matter availability rather than by calendar date, the forage quality and quantity

 

5 See Technical Note 4 in Appendix 3

28

dynamics are modeled according to day of year (DOY) rather than particular date
intervals. Dynamics of forage quality and quantity are modeled using all dry matter,
crude protein, acid detergent ﬁber, and neutral detergent ﬁber data points from the
experiment together with the signiﬁcant ﬁxed and blocking effects for each parameter to
describe the behavior or those parameters at ten day intervals between DOY 1 17 (April
28) and DOY 257 (September 15).

The statistical model for dry matter Offered includes treatment, year, day Of year,
and WSG (i.e. whether either big bluestem or switchgrass was included within the
system) nested within treatment, as ﬁxed effects; these effects were also tested for
interactions. Date was used as a repeated measurement, and the SP(POW) covariance

structure for unequally spaced measurements was used to model the covariance between
measurements (e. g., DOY A and DOYB).

The ﬁIll model for forage dry matter offered is:

Yijk = II + Ai + Bj + Ck + Czk + DI + (AB)ij +(AC)iIt + (BC)th + (CD)ItI + (3C2)th + eijkl
where:

. . .th .th th
yijkj Is the dry matter value observed In I treatment for the j year at the k day of year

at on the 1th pasture type (WSG or non-WSG)

and:

u is the overall mean

A; is a ﬁxed effect, the ith treatment (CS-BBS, CS-Only, CS-SG)
Bj is a ﬁxed effect, the 1"“ year (2003, 2004, 2005)

29

ck is a ﬁxed effect, the It’h DOY

Czk is the square of the kth day of year

D1 is a blocking effect, the 1th pasture type sampled (WSG or non-WSG)
(AB)ij is the interaction Of treatment and year

(AC)ik is the interaction Of treatment and DOY

(BC)J-k is the interaction Of year and DOY
(CD)kl is the interaction Of DOY and WSG(trt)

(BC )jk IS the Interaction used tO describe the quadratIc relationship Of DOY and
year
eijkl is the error term

The statistical model for the crude protein of forage offered includes treatment,
alfalfa abundance, year, DOY, and WSG nested within treatment as ﬁxed effects; these
effects were also tested for interactions. Date was used as a repeated measurement, and

the SP(POW) covariance structure for unequally spaced measurements was used to

model the covariance between measurements (e. g., DOY A and DOYB).

The full model for the crude protein Of the forage dry matter offered is:
2
Yijk = 11 + AI + Bj + Ck + C k + DI + Em + (AC)ik +(BC)th +(CD)ItI + eijkl

where:

30

. . . .th .th rh
Yijkl Is the crude protein value Observed In I treatment for the j year at the k day of

th th
year at on the l pasture type (warm-season grass or non-warm-season grass) at the m

alfalfa abundance

and:

(AC)iIt
(BC)th
(CD)ItI

eijklm

is the overall mean

is a ﬁxed effect, the ith treatment (CS-BBS, CS-Only, CS-SG)

is a ﬁxed effect, the f“ year (2003, 2004, 2005)

is a ﬁxed effect, the It’h DOY

is a blocking effect, the 1th pasture type (WSG or non-WSG)

is a blocking effect, the 1th alfalfa abundance (abundant, not abundant)
is the interaction of treatment and DOY

is the interaction Of year and DOY

is the interaction of DOY and WSG(trt)

is the error term

The statistical model for the acid detergent ﬁber of forage Offered includes

treatment, year, day of year, and WSG nested within treatment as ﬁxed effects; these

effects were also tested for interactions. Date was used as a repeated measurement, and

the SP(POW) covariance structure for unequally spaced measurements was used to

model the covariance between measurements (e.g., DOY A and DOYB).

31

The full model for the acid detergent ﬁber of the forage dry matter Offered is:

2 2
Yijk = 11 + Ai + Bj + Ck + C k + DI + (AC)ik + (BC)th + (CD)ItI + (BC )jk + eithI

where:

. . . .th .th th
Yijkl Is the acrd detergent ﬁber value Observed In I treatment for the j year at the k

day Of year at on the 1th pasture type (WSG or non-WSG)

and:

(AC)iIt
(BC)th
(CD)ItI

(3C2)th

eijkl

is the overall mean

is a ﬁxed effect, the ith treatment (C S-BBS, CS-Only, C S-SG)

is a ﬁxed effect, the jth year (2003, 2004, 2005)

is a ﬁxed effect, the kth DOY

is the square of the kth day of year

is a blocking effect, the 1th pasture type sampled (WSG or non-WSG)
is the interaction of treatment and DOY

is the interaction Of year and DOY

is the interaction of DOY and WSG(trt)

is the interaction used to describe the quadratic relationship of DOY and

year

is the error term

32

The statistical model for the neutral detergent ﬁber of forage Offered includes
treatment, year, day Of year, and WSG nested within treatment as ﬁxed effects; these
effects were also tested for interactions. Date was used as a repeated measurement, and

the SP(POW) covariance structure for unequally spaced data'was used to model the
covariance between measurements (e.g., DOY A and DOYB).

The full model for the neutral detergent ﬁber of the forage dry matter offered is:

Yijk = 11 + Ai + Bi + Cit + Czk + DI + Em + (AC)iIt + (BC)th + (CD)itI + (3C2)th + 6ide
where:

. . .th .th th
Yijkl IS the neutral detergent ﬁber value Observed In I treatment for the j year at the k

day of year at on the 1th pasture type (WSG or non-WSG) ) at the mth alfalfa abundance

and:
u is the overall mean
A; is a ﬁxed effect, the ith treatment (CS-BBS, CS-Only, CS—SG)
Bj is a ﬁxed effect, the jth year (2003, 2004, 2005)
ck is a ﬁxed effect, the It’h DOY
,2 . . th
C k 15 the square Of the k day Of year
D1 is a blocking effect, the 1th pasture type sampled (WSG or non-WSG)
Em is a blocking effect, the 1th alfalfa abundance (abundant, not abundant)
(AC)ik is the interaction of treatment and DOY

(BC)jk is the interaction Of year and DOY

33

(B.E)ik is the interaction of year and alfalfa abundance
(CD)kl is the interaction Of DOY and WSG(trt)

(BC2)J-k is the interaction used to describe the quadratic relationship Of year and DOY

Second Analysis

Because the 2002 switchgrass seeding failed to be established by the summer of
2003, one could argue that failing to examine the treatment by year interaction could lead
to erroneous conclusions about treatment performance. To address this concern, the acid
detergent ﬁber, neutral detergent ﬁber, and crude protein data was analyzed a second
time with the treatment by year interaction included in the model, even though it was
known that the interactions were not signiﬁcant. The treatment by year interaction was

already known to be statistically signiﬁcant for dry matter and it is included in the

original full model (above). This resulted in the term (AB),-j, representing the interaction

of treatment and year, being inserted into the model for crude protein, acid detergent
ﬁber, and neutral detergent ﬁber; this term was already part of the full model of dry
matter. The Tukey method was used to reduce Type I error. The best covariance
structure for acid detergent ﬁber when the treatment*year interaction was included was
spherical exponential.
Weather

Southwest Lower Michigan is a dry area of the state and the agronomic
consequences of this are accentuated by the relatively course textured Kalamazoo series
soils. Tabular data for precipitation, average temperatures, and average high ‘

temperatures are presented in Appendix 1, Figures A.3-A.5.

34

April and May 2003 were typical relative to precipitation and mean air
temperature; while June and July precipitation was lower than average, causing a
pronounced decline in forage production. Average monthly air temperatures throughout
the 2004 grazing season were very close to the 17-year average, while precipitation was
higher than normal, particularly in May, which had 25 cm of rainfall. This, coupled with
other well distributed precipitation throughout the 2004 grazing season led to relatively
uniform forage growth throughout the grazing season.

March, April, May, August and September of 2005 were extremely dry while
June and July both received more precipitation than normal. The March-May 2005
precipitation patterns seem to have prevented the normal spring ﬂush of pasture growth,
while the unusually high precipitation in June and July caused the “spring ﬂush” of
forage growth to occur in July-August. June through September 2005 was also
characterized by higher than normal average temperatures. The potential dormancy-
inducing effect of these above average temperatures on pastures was likely offset by the
unusually high precipitation in June and July as well as the lack of precipitation in
March-May.

RESULTS AND DISCUSSION
Forage dry matter and yield quality: ﬁrst analysis
Dry matter offered over time

Forage dry matter and quality were modeled using day of year (DOY) as the
independent variable. Table 1.4 relates DOY to dates. In this model, DOY 117 is the
starting date for grazing, and the data is plotted over time using ten day intervals which

end on DOY 257. The interaction between treatment and year existed for dry matter-

35

offered is described in Figures 1.1-1.3. These ﬁgures demonstrate the two results of
setting aside one-third of the pasture acreage to big bluestem and switchgrass. First, it is
intuitive that the dry matter offered in the CS-Only treatment was higher earlier in the
grazing season because the entire pasture was available for grazing by early-May whereas
the big bluestem and switchgrass portions of the CS-BBS and CS-SG treatments,
respectively, were not grazed until early- to mid-June (Bamhart, 1994). Second, as
expected, the peak in dry matter-offered occurred between 20 and 30 days earlier in the
CS-Only treatment than in the other treatments (Bartholomew, 1995)

It is important to note that in 2004 and 2005, refused big bluestem and
switchgrass forage from the CS-BBS and CS-SG treatments was not mowed, which
resulted in the rapid accumulation big bluestem and switchgrass residue, much of which
reached the reproductive stage. By mid-summer the big bluestem and switchgrass
pastures had accumulated a large volume of low quality forage (Coleman, 2004), much of
this residue remained at the time that grazing of the big bluestem and switchgrass
pastures was terminated in 2004 and 2005. Thus, while the overall amount of forage
offered to the livestock at certain times was higher for the CS-SG and CS-BBS
treatments, much of it was selectively grazed and did not correlate with total pasture
system productivity (i.e., animal gain per hectare).

Forage quality modeled over time

The effect of treatment also was signiﬁcant for crude protein, acid detergent ﬁber,
and neutral detergent ﬁber, but there was no interaction between treatment and year
(Tables 1.5-1.7, Figures 1.4-1.6). Acid detergent ﬁber and neutral detergent ﬁber both

had a quadratic relationship with time, while crude protein had a linear relationship.

36

When data from all years of the experiment was used to model the changes in forage
percent crude protein over time the slope of the line is ﬂatter (i.e., less negative) for the
CS-Only treatment than it is for the other treatments (see Figure 1.4 and Table 1.5). The
crude protein levels of the CS-BBS and CS-SG treatments did not differ signiﬁcantly
from each other throughout the grazing season, probably because the cool-season
components of each system contributed to the overall model and diluted any differences
that existed. During the beginning of the grazing season, forage percent crude protein is
actually higher in the CS-BBS and CS-SG systems, most likely because the cool-season
portions Of those pastures were rotated slightly more frequently during the early part of
the grazing season because one-third of the entire pasture area had been converted to
either big bluestem or switchgrass pasture.

AS the grazing season continued, the livestock did not graze the big bluestem or
switchgrass pastures completely or uniformity due to the sheer volume produced within a
relatively short time and because of the rapid decline in forage quality (Coleman, 2004)
as the plants became more physiologically mature. The crude protein data, modeled with
data from 2003-2005 shows that forage in the CS—Only treatment maintained a crude
protein level between 15 and 20% while the both other treatments were near or below
10% crude protein by the end of the season. These results correlate with the ﬁndings of
Moore et a1. (2004) who concluded that, under most circumstances it is more
advantageous to leave livestock on cool season pastures at a lower stocking rate than to
move the whole larger group of animals to native grasses, primarily because of the low

and rapidly declining forage quality of mid-summer warm-season grasses.

37

As with crude protein, forage acid detergent ﬁber and neutral detergent ﬁber
percentages remained more constant in the CS-Only treatment than in the CS-BBS or CS-
SG treatments (Moore et al., 2004) (see Figures 1.5-1.6 and Tables 1.6-1.7). Although
big bluestem is generally known to maintain higher forage quality during late summer
(Moser, 1995), neither the acid detergent ﬁber nor neutral detergent ﬁber levels of the
CS-BBS and CS-SG treatments differed from each other, probably because cool-season
forages were also part of the grazing system. Acid detergent ﬁber levels for the CS-Only
treatment peaked at approximately 30%, remaining between 25% and 30% from DOY
147 through DOY 25 7, signiﬁcantly lower than the levels described for the CS-BBS and
CS-SG treatments at each interval except from DOY 167-187 (see Table 1.4 to correlate '
DOY to date). The levels of acid detergent ﬁber found in the CS-BBS and CS-SG
treatments between DOY exceeded 30% acid detergent ﬁber from DOY 167 onward,
peaking around DOY 237. Neutral detergent ﬁber for the CS-Only treatment peaked at
nearly 54%, remaining between 38% and 51% from DOY 117 through DOY 257,
Signiﬁcantly lower than the levels described for both the CS-BBS and CS-SG treatments
from DOY 157 onward.

Forage dry matter and yield quality: second analysis

Analysis of the treatment by year interactions for ADF and NDF revealed
analogous results to the effect of treatment alone: the CS-Only treatment generally had a
ﬂatter curve and lower levels of ADP and NDF than the CS-BBS and CS-SG treatments
(Moore etal., 2004), which were very similar to each other (see Figures 1.10—1.15).
Analysis of the same interaction for crude protein revealed results analogous to the effect

of treatment alone: the slope of the line representing CP over time is generally higher

38

(Moore et al., 2004) and always more positive for the CS-Only treatment than for either
of the other treatments (see Figures 1.7-1.9). The effect of environment on overall levels
of CP and grazing management are obvious in 2004 when there was record rainfall
(Appendix 1, Figure A3) and explosive pasture growth. The CS-Only treatment model
for 2004 actually shows increasing crude protein of forage offered from the beginning
through the end of the season due to the time it took for the livestock to consume the
rapidly accumulating forage.

While the interactions of treatment and year for ADF, NDF, and CP were not
signiﬁcant, considering the forage quality and yield data in each year may give direction
to future investigation. Dry matter offered was higher in the late summer for the C S-
BBS and CS-SG systems, but the forage quality was uniformly poor. Any of the
treatments offer opportunities to manage forage quality and production by harvesting
excess forage rather than stockpiling, but in the CS—BBS and CS-SG treatments, this
practice would seem to undermine the objective of having forage available during the
summer slump: if big bluestem or switchgrass were harvested on about July 1, when the
forage is very high quality, there would be very little forage available in the event of an
August drought.

Botanical composition”

Fifty nine plant species were found in pastures throughout the experiment. Pasture
botanical composition is described in terms of species and plant category over time. The
four categories used to describe pasture plants include: forage grass, forage legume,

palatable weed, and undesirable weed. In this context, ‘palatable’ means that when a

 

6 The botanical composition trends over time and among treatments are discussed more exhaustively in
Appendix II

39

given plant species is in a vegetative stage the plant is not selectively refused by grazing
animals (e.g., dandelion). “Undesirable” pasture plants are either clearly selectively
reﬁised plants (e. g., bull thistle) or those that are generally known to be eaten rarely by
cattle (e.g., corn Speedwell).
Botanical Composition by Experimental Treatment: Pasture Plant Categories

Palatable pasture weed dry matter ﬂuctuated between 3% and 12% throughout the
experiment; trends were similar among treatments. Treatment differences of percent
pasture dry matter made up of undesirable pasture weeds were greatest in 2003 and
tended to converge over time, although seasonal ﬂuctuations were evident.

In all treatments of this study three trends emerged from botanical composition
data relative to plant type:
0 Forage grass dry matter percent remained at similar levels among treatments over

time.
0 Grass and legume dry matter percent remained relatively stable over time.
o Undesirable weeds declined from 2003 to 2005.
o Palatable weeds were stable or increased Slightly over the course of the experiment.
CONCLUSIONS

As expected, the inclusion Of big bluestem and switchgrass pastures as
components of grazing systems did not improve forage quality. The effect of treatment
on the distribution of forage dry matter offered varied by year; inclusion of either
switchgrass or big bluestem as a component of the grazing season signiﬁcantly Shifted
the distribution of forage dry matter offered to the later part of the grazing season. While

each livestock production Operation is somewhat unique and judgments about the

40

distribution of treatment dry—matter distribution will be somewhat subjective, many
producers would not accept the loss of one third of the early-spring and early-fall pasture
due to the management constraints of big bluestem and switchgrass pastures. The CS-
BBS and CS-SG systems tended to have later and higher peaks of forage productivity and
were less likely to have overly mature cool-season pastures early in the spring.

Based on this research, any perceived advantages to using switchgrass rather than big
bluestem in an integrated cool-season/warm-season pasture grazing system would not be
related to forage quality or productivity. These ﬁndings suggest that, compared to the
CS-Only treatment, the CS-SG or the CS-BBS treatments probably will not improve

livestock gain or improve the distribution Of animal grazing days throughout the summer.

41

Table 1.1. Dates and rates of urea application.

 

 

pate kg ha'1 actual N cs (all) BBS and 36
May 8, 2003 67 x x
June 18, 2003 66 x
May 27, 2004 73 x x
August 2, 2004 57 x
May 11, 2005 63 x x
July 13, 2005 70 x

 

42

Table 1.2. Grazing initiation and termination dates by year.

 

 

Year Grazing Initiated Grazing Terminated
2003 29-Apr 9-Sep
2004 27-Apr 8-Oct
2005 21-Apr 8-Sep

 

43

Table 1.3. Transformations used to normalize studentized residuals Of forage quantity and
quality parameters.

 

 

Parameter Effect Transformation

DM (Mg/ha) sqrt (abs (abs (1 — DM) - 0.3))

CP (%) treatment and treatment*year sqrt ((CP X 0.05))

ADF (%) treatment sqrt (abs (abs (31-ADF) -2 ) )

ADF (%) treatment*year sqrt ( abs ( abs ( 31 - ADF ) - 2.4 ) )
NDF (%) treatment sqrt ( abs ( abs ( 52 - NDF) - 3 ))
NDF (%) treatment*year sqrt ( abs ( abs ( 55 - NDF ) - 3))

 

44

Table 1.4. DOY and corresponding dates.

 

 

DOY Date

1 17 27-Apr
127 7-May
137 17-May
147 27-May
157 6-Jun
167 16-Jun
177 26-Jun
187 6-Jul
197 16-Jul
207 26-Jul
217 5—Aug
227 15—Aug
237 25-Aug
247 4-Sep
257 14-Sep

 

45

Figure 1.1. Forage dry matter offered over time during the 2003 grazing season for the
CS-BBS and CS—Only treatments.

3
2.5
2
_._ CS—BBS
. *7 , 9.5-9“!

DM Offered (Mg per Ila)
I— III

.°
01

 

AAAA
{\ééw“

AAAAA
A‘PPPP

’\’\
‘5

AAA A
S‘OQSI‘ b

x
Day of Year

 

46

 

Figure 1.2. Forage dry matter offered over time during the 2004 grazing season.

2004 DM Production

DM Offered (Mg per ha)

 

AAAAAAAAAAAAAAA
s‘OOx‘xbxb0S‘50A99'9‘Pw‘AP

Day of Year

 

47

 

Figure 1.3. Forage dry matter offered over time during the 2005 grazing season.

 

 

2.5
.22
I-
0
9-
Bl
ELS
E
g]
E
:11.5
0
’\’\'\’\’\’\’\ AAAAAAA
eoeeeeee‘eeeaeee
Dayonear

48

 

Table 1.5. Crude protein percentage of forage dry matter Offered over time, modeled
with yield data from 2003-2005.

 

 

CS-BBS CS-Only CS-SG
DOY % CP SE % CP SE % CP SE
117 21.5 at 0.8 19.2 b 0.6 22.0 a 0.9
127 20.7 a 0.7 19.0 b 0.6 21.0 a 0.8
137 19.9 a 0.6 18.8 a 0.5 20.0 a 0.7
147 19.0 a 0.6 18.6 a 0.4 19.0 a 0.6
157 18.2 a 0.5 18.3 a 0.4 18.1 a 0.5
167 17.3 a 0.4 18.1 a 0.4 17.1 a 0.5
177 16.5 b 0.4 17.9 a 0.4 16.1 b 0.4
187 15.7 b 0.4 17.7 a 0.4 15.1 b 0.4
197 14.8 b 0.4 17.5 a 0.4 14.1 b 0.4
207 14.0 b 0.4 17.3 a 0.4 13.1 b 0.4
217 13.1 b 0.5 17.0 a 0.5 12.1 b 0.5
227 12.3 b 0.6 16.8 a 0.5 11.1 b 0.6
237 11.4 b 0.7 16.6 a 0.6 10.1 b 0.7
247 10.6 b 0.7 16.4 a 0.7 9.2 b 0.8
257 9.8 b 0.8 16.2 a 0.7 8.2 b 0.9

 

'1 Means within a row with different letters are signiﬁcantly different at P < 0.05

49

Figure 1.4. Crude protein content of forage dry matter offered over time, modeled with
yield data from 2003-2005.

 

+ CS-BBS
—I-— CS-Only ‘
CS-SG

 

 

AAAAAAAAAAAAAAA
00¢¢¢4°0¢9AP¢0¢~P~5

Day of Year

 

 

50

Table 1.6. Acid detergent ﬁber content of forage dry matter Offered over time, modeled
with yield data from 2003-2005.

 

 

CS-BBS CS-Only CS-SG
DOY ADF SE ADF SE ADF SE
117 19.5 b 0.9 19.6 a 0.6 19.3 b 1.0
127 22.1 b 0.7 21.7 a 0.5 21.9 b 0.8
137 24.6 b 0.6 23.5 a 0.4 24.3 b 0.7
147 26.7 b 0.5 25.0 a 0.4 26.5 b 0.6
157 28.7 b 0.4 26.4 a 0.3 28.4 b 0.5
167 30.4 a 0.4 27.5 a 0.3 30.1 a 0.5
177 32.0 a 0.3 28.4 a 0.3 31.6 a 0.4
187 33.2 a 0.3 29.1 a 0.3 32.9 a 0.4
197 34.3 b 0.3 29.5 a 0.3 33.9 b 0.4
207 35.1 b 0.4 29.7 a 0.4 34.7 b 0.4
217 35.7 b 0.4 29.7 a 0.4 35.3 b 0.4
227 36.1 b 0.5 29.5 a 0.5 35.7 b 0.5
237 36.3 b 0.6 29.1 a 0.5 35.8 b 0.6
247 36.2 b 0.7 28.4 a 0.7 35.7 b 0.8
257 35.9 b 0.9 27.5 a 0.8 35.4 b 0.9

 

‘1' Means within a row with different letters are signiﬁcantly different at P < 0.05

51

Figure 1.5. Acid detergent ﬁber content of forage dry matter offered over time, modeled
with yield data from 2003-2005.

40

1 —°— CS-BBS
CS-Only ‘
+ CS-SG

 

AAAAAAAAAAAA
S‘OSPS‘SPSP'OSPQ'BA}

r91

AAA
City"

Day of Year

 

 

52

Table 1.7. Neutral detergent ﬁber content of forage dry matter offered over time,
modeled with yield data from 2003-2005.

 

 

CS-BBS CS-Only CS-SG
DOY NDF SE NDF SE NDF SE
117 43.9 a 1.8 38.7 a 1.3 42.0 a 1.9
127 47.2 a 1.5 41.5 a 1.1 45.6 a 1.6
137 50.2 a 1.3 44.0 a 0.9 48.9 a 1.4
147 53.0 b 1.1 46.2 a 0.8 51.9 ab 1.2
157 55.5 b 0.9 48.2 a 0.8 54.6 b 1.1
167 57.6 b 0.8 49.8 a 0.8 57.1 b 0.9
177 59.5 b 0.8 51.2 a 0.7 59.2 b 0.9
187 61.2 b 0.8 52.2 a 0.7 61.1 b 0.8
197 62.5 b 0.8 53.0 a 0.8 62.7 b 0.8
207 63.5 b 0.8 53.5 a 0.8 64.0 b 0.8
217 64.3 b 0.9 53.8 a 0.9 65.0 b 0.9
227 64.8 b 1.1 53.7 a 1.0 65.8 b 1.1
237 65.0 b 1.3 53.4 a 1.2 66.2 b 1.3
247 64.9 b 1.5 52.7 a 1.5 66.4 b 1.6
257 64.5 b 1.9 51.8 a 1.8 66.2 b 2.0

 

1 Means within a row with different letters are signiﬁcantly different at P < 0.05

53

Figure 1.6. Neutral detergent ﬁber content of forage dry matter offered over time,

modeled with yield data from 2003-2005.

NDF (%)

 

80
70
60 ,

Ublll
96°

’3;

A"

, ”(kg
[Jar

A
{b

Day of Year

.e’é‘

“why-..” . aw- thew-«~41.

 

L," __,L
, —.— CS-BBS

; +cs-OnIy
; ,,,C,Srs,(i_._

 

 

54

Figure 1.7. Crude protein content of forage dry matter offered over time from the CS-
BBS and CS-Only treatments, modeled with yield data from 2003.

30 ..

* *+* asses
“.FT‘.C§'9_“IY,

 

 

s“$e“e“~<‘4§‘49

Day of Year

 

 

 

55

Figure 1.8. Crude protein content of forage dry matter Offered over time from all
treatments, modeled with yield data from 2004.

 

 

 

 

 

 

25
20
Q 15 » + CS-BBSM
E: —-—- CS-Only
U 10 CS-SG
5 _. a A v”. E - LL
0 t T
’\ ’\ ’\ ’\ ’\ ’\ ’\ ’\
x‘ {’5 \‘9 x“ x9 A} '9’ “\3’
Day of Year

56

Figure 1.9. Crude protein content of forage dry matter offered over time from all
treatments, modeled with yield data from 2005.

 

AAAAAAAAAAAA
OOSRSPSPSSOSPG'P'Q'I“

AA
«1'4"

80:)

Day of Year

 

57

Figure 1.10. Acid detergent ﬁber content of forage dry matter offered over time from the
CS-BBS and CS-Only treatments, modeled with yield data from 2003.

50
45
40
35
3° ;::—’cs:§ss
tseeuy

ADF (%)
N
u-

 

AAAAAAA AAAAAAA
s‘x‘x'5x‘x5xbx"$9~°¢99¢~5

 

 

Day of Year

58

Figure 1.11. Acid detergent ﬁber content of forage dry matter offered over time from all
treatments, modeled with yield data from 2004.

 

AAAAAAAA AAAAA
s‘O¢x"x"x‘°<\x“’~“\A§\~‘~”¢~"\P

 

 

Day of Year

59

Figure 1.12. Acid detergent ﬁber content of forage dry matter offered over time from all
treatments, modeled with yield data from 2005.

,_., , 777--..Lﬁ

_._ "cs-BBS
. —-— CS—Only
,CstSG,

 

’\’\'\'\’\’\’\’\ ‘\’\’\’\’\’\
eoeeeeeeé‘eeeeee

Day of Year

 

 

60

Figure 1.13. Neutral detergent ﬁber content of forage dry matter offered over time from
the CS-BBS and CS-Only treatments, modeled with yield data from 2003.

i-Qéic's-irris
L ‘7',“ 95:0}!!!

NDF(%)
assasssass

 

’\’\’\’\’\’\’\‘\’\’\'\’\’\’\’\
x‘0~5~’~9xb<\€°'3'\9w‘9'€5¢~5
Day of Year (DOY)

 

61

 

Figure 1.14. Neutral detergent ﬁber content of forage dry matter offered over time from
all treatments, modeled with yield data from 2004.

 

7o
..JA”""‘"‘”+"*M—\_ ﬁx ,
so
so
:3 40 +cséans .
E +CS-Only
z 30 cs-sG .

N
G

n-
G

 

AAAAAAAAAAAAAAA
S‘OOSFSESSOG’G'PA‘PP'PAP
Day of Year (DOY)

 

62

Figure 1.15. Neutral detergent ﬁber content of forage dry matter offered over time from
all treatments, modeled with yield data ﬁ'om 2005.

 

_._ cs-an
+ CS—Only .
CS-SG

 

 

AAAAAAAAA
OSPG~°A>A20¢AP

I!)
’9)
’J)
I?)
(r)
’6‘)

Day of Year (DOY)

 

 

63

LITERATURE CITED

Anderson, B. 2000. Grazing management on warm season grasses Missouri Forage and
Grassland Council. University of Missouri.

Amy, AC, and A. R. Schmid. 1942. A study of the inclined point quadrat method Of
botanical analysis of pasture mixtures. Journal of the American Society of
Agronomy 34:238-247.

Barker, J.M., Buskirk, D. D., Ritchie, H. D., Rust, S. R., Leep, H. R., and K. J. Barclay.
1999. Intensive grazing management of smooth bromegrass with or without
alfalfa or birdsfoot trefoil: heifer performance and sward characteristics. The
Professional Animal Scientist 15: 1 30-135.

Bamhart, SK. 1994. Warm-season grasses for hay and pasture - Pm-569. Iowa St. U.
Ext.

Bartholomew, H.M., Sulc, R.M., Hendershot, R., and J. Cline. 1995. AGF-
02295:Perennial warm season grasses for Ohio. Ohio St. U. Ext. Dept. of Hort.
and Crop Sci.

Bransby, DJ. 1989. Compromises in the design and conduct of grazing experiments, p.
136, In G. C. Marten, ed. Grazing research: design, methodology, and analysis.
CSSA Spec. Pub. No. 16. Crop Science Society of America and the American
Society of Agronomy, Madison.

Burns, J.C., Lippke, H., and D. S. Fisher. 1989. The relationship of herbage mass and
characteristics to animal responses in grazing experiments, p. 7-19, In G. C.
Marten, ed. Grazing research: design, methodology, and analysis; CSSA Spec.
Pub. No. 16. Crop Science Society of America, Madison.

Coleman, S.W., Moore, J. E., and Wilson, J. R. 2004. Quality and Utilization, p. 1171, In
L. E. Moser, Burson, B. L., and Sollenberger, L. E., ed. Warm Season (C4)
Grasses, Vol. 45. ASA-CSSA-SSA of America, Madison.

Collins, M., and J. 0. Fritz. 2003. Forage Quality, p. 363-390, In R. F. Barnes, Nelson, J.
C., Collins, M., and Moore, K. J., ed. Forages: an introduction to grassland
agriculture, Vol. 1. Iowa State University Press, Ames.

De Bruijn, S.L., and E. W. Bork. 2006. Biological control of Canada thistle in temperate

pastures using high density rotational cattle grazing. Biological Control 36:305-
315.

64

Delisle, J.M., and J .A. Savidge. 1997. Avian use and vegetation characteristics of

conservation reserve program ﬁelds. Journal of Wildlife Management 61 :318-
325.

Earle, D.F., and AA. Mcgowan. 1979. Evaluation and Calibration of an Automated
Rising Plate Meter for Estimating Dry-Matter Yield Of Pasture. Australian Journal
of Experimental Agriculture 19:337-343.

Gabriels, P.C.J., and J .V. Van Den Berg. 1993. Calibration of two techniques for
estimating herbage mass. Grass and Forage Science 48:329-335.

Hach, C.C., Brayton, S. V, and Kopelove, A. B. 1985. A powerful Kjeldahl nitrogen
method using peroxymonosulfuric acid. Journal of Agricultural Food Chemistry
33:1117-1123.

Harker, N.K., Baron, V. S., Chanasyk, D. S., Naeth, M. A., and F. G. Stevenson. 2000.
Grazing intensity effects on weed populations in annual and perennial pasture
systems. Weed Science 48:231-238.

Harmoney, K.R., K.J. Moore, J.R. George, BC. Brummer, and JR. Russell. 1997.
Determination of pasture biomass using four indirect methods. Agronomy Journal
89:665-672.

Hart, RH, and CS. Hoveland. 1989. Objectives of grazing trials, p. xi, 136 p., In G. C.
Marten, ed. Grazing research: design, methodology, and analysis - CSSA special
publication number 16. Crop Science Society Of America and the American
Society of Agronomy, Madison.

Henson, RR, and M. A. Hein. 1941. A botanical and yield study of pasture mixtures at
Beltsville, Maryland. Journal of the American Society of Agronomy 33:700-708.

Karsli, M.A., J .R. Russell, and MJ. Hersom. 1999. Evaluation of berseem clover in diets

of ruminants consuming corn crop residues. Journal Of Animal Science 7722873-
2882.

KBS-LTER. 2005. The Kellogg Biological Station Long-Term Ecological Research
Climate Database. Michigan State University Board of Trustees.

Leep, R. 2003. Forage Specialist and Assistant Professor of Crop and Soil Sciences,
Michigan State University.

Lucas, R.J., and K.F. Thompson. 1990. Pasture assessment for livestock managers, p. vii,

499 p., In R. H. M. Langer, ed. Pastures, their ecology and management. Oxford
University Press, Auckland; New York.

65

Mannetje, LL, and KP. Haydock. 1963. The dry-weight—rank method for the botanical
analysis of pasture. Journal of the British Grassland Society 18:268-275.

Michell, P. 1982. Value of a rising-plate meter for estimating herbage mass of grazed
perennial ryegrass-white clover swards. Grass and Forage Science 37:81-87.

Michell, P., and RV. Large. 1983. The estimation of herbage mass of perennial ryegrass
swards - a comparative-evaluation of a rising-plate meter and a single-probe
capacitance meter calibrated at and above ground-level. Grass and Forage Science
38:295-299.

Moore, K.J., T.A. White, R.L. Hintz, P.K. Patrick, and EC. Brummer. 2004. Forages and
pasture management - Sequential grazing of cool- and warm-season pastures.
Agronomy Journal 96:1103-1111.

Moser, LE, and K. P. Vogel. 1995. Switchgrass, big bluestem, and indiangrass, In R. F.
Barnes, Darrell, A.M., and C]. Nelson, ed. Forages: An introduction to grassland
agriculture, Vol. 1, Fifth ed. Iowa State UP, Ames.

Mott, GO. 1960. Grazing pressure and the measurement Of pasture production. Proc VIII
Int. Grassl. Cong.:606-6I 1.

Munkvold, G. 2002. Electronic communication regarding photograph of diseased
switchgrass seedling, pp. diagnosis, In D. J. Hudson, (ed). Iowa State University,
electronic correspondance.

NCSU. no date. Corn ﬂea beetle [Online]. Available by NCSU-IPM
http://ipm.ncsu.edu/AGZ7l/com sorghum/com ﬂea beetlehtml (veriﬁed
03/27/06).

O’Donovan, M., Dillon, P., Rath, M., and G. Stakelum. 2002. A comparison of four
methods of herbage mass estimation (abstract). Irish Journal of Agricultural and
Food Research 40.

Sample, D.W., and MJ. Mossman. 1997. Managing habitat for grassland birds: a guide
for Wisconsin Wisconsin Department of Natural Resources, Madison.

Sanderson, M.A., C.A. Rotz, S.W. Fultz, and EB. Rayburn. 2001. Estimating forage
mass with a commercial capacitance meter, rising plate meter, and pasture ruler.
Agronomy Journal 93:1281-1286.

SAS Institute, Inc. 2000-2003. SAS. Release 8 and 9. SAS Institute, Inc., Cary, NC.

Scrivner, J.H., D.M. Center, and MB. Jones. 1986. A rising plate meter for estimating
production and utilization. Journal of Range Management 39:475-477.

66

Stern, M.D., and MI. Endres. 1991. Laboratory manual: research techniques in ruminant
nutrition University of Minnesota, Department of Animal Science.

Stockdale, CR. 1984. Evaluation of techniques for estimating the yield of irrigated
pastures intensively grazed by dairy-cows.2. The rising plate meter. Australian
Journal of Experimental Agriculture 24:305-311. .

Stockdale, CR, and KB. Kelly. 1984. A comparison of a rising-plate meter and an
electronic capacitance meter for estimating the yield of pastures grazed by dairy-
cows. Grass and Forage Science 39:391-394.

Undersander, D. 2002. Conversation about using warm-season grasses as a component of
grazing systems.

VanKeuren, R.W., and H.L. Ahlgren. 1957. A statistical study of several methods used in

determining the botanical composition of a sward: I. A study of established pastures.
Agronomy Jounal 49:532-536.

Watkins, K.L., Veum, T. L., and Krause, G. F. 1987. Total nitrogen determination of
various sample types: a comparison of the Hach, Kjeltec, and Kjeldahl methods.
Journal of the Association of Analytical Chemists 70:410-412.

Wen, L., R.L. Kallenbach, J.E. Williams, GA. Roberts, P.R. Beuselinck, R.L. McGraw,
and HR. Benedict. 2002. Performance of steers grazing rhizomatous and

nonrhizomatous birdsfoot trefoil in pure stands and in tall fescue mixtures.
Journal of Animal Science 80: 1970-1976.

Wilm, H.G., Costello, DE, and GE. Klipple. 1944. Estimating forage yield by the

double sampling method. Journal of the American Society of Agronomy 36:194-
203.

67

Chapter 2
INTEGRATED WARM- AND COOL- SEASON GRASS AND LEGUME PASTURES:
STEER AND PASTURE PERFORMANCE
ABSTRACT
High temperatures and lack of precipitation often cause the productivity of cool-season
pastures in Michigan to decline during the summer. Our objective was to determine the
effect of integrating monocultures of switchgrass or big bluestem with cool-season
pastures on pasture and steer productivity. Cool-season grass and legume pastures (CS-
Only) were compared to integrated big bluestem and cool-season grass and legume
pastures (CS-BBS) and integrated switchgrass and cool-season grass and legume pastures
(CS-SG). Dynamics of pasture quality and productivity were compared by describing
total livestock weight gain per hectare, accumulation and distribution of animal-unit
grazing days, and average daily gain of Holstein steers. The CS-Only treatment resulted
in more total animal weight gain per hectare than either the CS-BBS Or the CS-SG
treatments, and the CS-BBS treatment resulted in more total animal weight gain per
hectare than the CS-SG treatment. The inclusion of big bluestem or switchgrass in a
grazing system did not result in the accumulation of more animal unit grazing days per
hectare in any year and only resulted in the accumulation of more animal unit grazing
days per hectare during two time intervals. Treatment had no effect on steer average
daily gain, likely because livestock tend to graze selectively in poor pastures and because
the they were only on the big bluestem and switchgrass portions of their respective
treatments for short time intervals. Although the inclusion of big bluestem or switchgrass

in the grazing systems did not reduce the average daily gain, it did not regularly improve

68

that parameter or the other measurements of pasture performance. Producer adoption of
the grazing systems similar to the CS-BBS and CS-SG treatments, will depend on: the
class of livestock that characterizes the livestock Operation; the ﬂexibility of the producer
to either sell cattle or feed stored forage during periods of time when cool-season pasture
productivity is inadequate; and the willingness of the producer to design, establish, and
manage a more complex pasture system while waiting several years or more before

recapturing their investment.

69

INTRODUCTION

Switchgrass and big bluestem have been promoted as forage for livestock
(Anderson, 2000; Bamhart, 1994; Bartholomew, 1995) and are also known to provide
valuable wildlife habitat (Delisle and Savidge, 1997; Sample and Mossman, 1997). Thus
far, little research has been conducted in the Great Lakes region to describe livestock on
grazing systems that include big bluestem or switchgrass. Because livestock performance
is closely related to forage quality and abundance, it is important to measure these
parameters in addition to the response of the grazing animal. In order to explain animal
response, it is helpful to describe pasture system dynamics in terms of forage dry matter
yield, and forage quality parameters including crude protein, acid detergent ﬁber, and
neutral detergent ﬁber.
One of the primary objectives of the experiment was to compare the productivity of a
traditional Michigan cool season grass and legume grazing system with the productivity
of two model systems that include warm season grasses, in the ﬁrst three years after
planting the warm season grasses. The associated hypotheses were:

1. The grazing system models that are integrated with warm season grasses will result in
higher total livestock weight gains per hectare (henceforth, ‘gain ha.l ’) than a typical

Michigan grazing system model which includes only cool season grasses and

legumes.
2. An integrated grazing system that includes big bluestem will result in higher gain ha.l

than an integrated grazing system that includes switchgrass.
These hypotheses were evaluated in Chapter 3 of this dissertation. The subject of this

chapter is pasture quality and the dynamics of forage dry matter offered to the grazing

70

animals, both of which will help to understand why the above hypotheses were or were
not validated.
LITERATURE REVIEW

Grazing experiment concepts: method of stocking

Grazing experiments are conducted either using: 1) ﬁxed stocking rates, where a
certain number of animals are kept on a pasture for the season, year, or the duration of the
experiment; or 2) variable stocking rates (put-and-take), where the grazing pressure is
adjusted as forage availability requires (Wheeler et al., 1973), but a certain minimum
number of livestock are left on the pasture to determine parameters such as average daily
gain (ADG). Bransby (1989) argues that grazing experiments should have multiple
levels of grazing pressure (i.e. high, medium, and low), but acknowledges that resources
frequently may not permit this. Because of this, the put-and-take method described by
Mott (1960) is often the method of choice for experiments that have limited resources
(Bransby, 1989). Mott explains that because pasture productivity within the same
experiment can vary widely, it is not reasonable to attempt to maintain identical stocking
rates among treatments or replications. Rather, it is more reasonable to manage each
pasture optimally, adjusting the grazing pressure so that no treatment or replication is
grazed more intensely than another. Wheeler et a1. (1973) points out that the put-and-take
method is frequently criticized for lacking objectivity, but contends that this objection
can be overcome by using a team of plant and animal scientists; using multiple grazing
pressures can also help overcome this objection (Mott, 1952), but this very expensive and

is rarely done (Wheeler et al., 1973), because of the expense involved.

71

The debate between those in favor Of ﬁxed stocking rate studies vs. variable
stocking rate studies has been intense, but Wheeler et al. (1973) provides a balance by
pointing out that ﬁxed stocking rates are used successfully in areas where: 1) forage does
not ﬂuctuate drastically between seasons; 2) excess forage can be left standing for later
use; or 3) economic animal production takes place before or during the time when the
available forage becomes insufﬁcient to maintain animal weight. The put-and-take
method is used where research is being performed on cultivated and improved pastures
rather than on native pastures (Mott, 1952); in Short-term instead of year-round studies; in
temperate-humid zones rather than rangeland; and in the US. and Europe rather than
Australia or New Zealand (Wheeler et al., 1973).

Bransby (1989) suggests that grazing trials can be terminated (i.e. concluded for
the year] by either terminating the grazing of all pastures simultaneously or at different
times. Termination of grazing of different pastures at different times should be based on
pasture descriptor goals or some predetermined per-animal production level.

Grazing experiment concepts: animal measurement and treatment

The experimental unit in grazing trials is the pasture unit, not the animals on the
pasture. Using the animals as the experimental units artiﬁcially increases the degrees of
freedom and increases the likelihood of declaring differences between treatments when
no difference exists. Animals in the experiments are the sub-units and the source of the
experimental error for the trial (Fisher, 1999). Much of the data used to describe
differences between systems is related to animal weight and weight gain: average daily
gain (ADG) per animal, animal unit days (AUD) per hectare, average seasonal gain per

animal, etc. Because grazing animals are large and the status of their digestive systems

72

can cause the weight of individual animals to ﬂuctuate drastically within and between
days, a method to accurately determine animal weights must be used. Baker (1942)
explains that daily variation in cattle weight occurs for two reasons. First, there are
environmental conditions that affect all animals in the same way. Second, there are
individual characteristics of animals that cannot be explained by environmental
phenomena. Because individual characteristics cannot be controlled, the goal of the
researcher is to minimize the variation caused by the environment. This is generally done
in one of two ways: ‘Shrinking’, or multiple consecutive weighing events. Shrinking is
usually accomplished by depriving all animals of food and water for 12-16 hours before
weighing (Stuedemann, 1989), and some managers prefer not to use this method. Koch
et al. (1958) demonstrated that using three-day average weight was an effective way to
reduce variation due to ﬁll. Baker (1942), Koch et al. (1958), and Patterson (1947)
contend that greater accuracy is achieved by weighing more animals per experiment than
weighing fewer animals multiple times. Patterson also concluded that weighing animals
three times in three days reduced the standard error for the weight, although not as
dramatically as one might expect. Koch et al. (1958) concluded that using an average of
three weights taken over the course of three days can help minimize the problems
presented by ﬂuctuations in rumen fullness and that two or three day averages are
recommended when comparing individual animals. Mott (1959) states that
environmental factors can cause weight to vary within a day, and that all treatments
should be weighed on the same day in the same relatively short period of time.

Animals must be weighed at the beginning and end of the experiment, and should

be weighed periodically throughout the experiment. Single-day monthly weighing is

73

recommended to monitor animal health (Buskirk, 2003), but these weights should not be
used to compare treatments because of the differences in rumen ﬁll that are due to factors
other than the treatment (such as freshness of the paddock in the rotation) (Mott, 1959).
When the put-and-take stocking method is used and adjustments in the stocking rate are
necessary, animals should be weighed each time they are added to or taken from a
pasture; those weights should be used to calculate animal unit grazing days (AUGD)
(Fisher, 2003).

Because the conditions (i.e., pasture height, density, and quality) among the
pasture treatments and replications may vary at the time of the trial termination, there
may be differences in rumen fullness among animals that could affect the seasonal gain
calculations and the standard error. Buskirk (2003) suggests that gut—ﬁll can be
standardized to a great degree if all animals are fed bay for seven days, beginning on the
day that the grazing trial is terminated, and then weighing them on the eighth and ninth
days. The mean of these two weights can be used to calculate the ﬁnal animal weight,
which can be used to calculate the seasonal gain for each animal.

Units such as “animal grazing days” are sometimes used. However, when the
sizes of experimental animals are variable, the amount of forage dry-matter consumed in
one day of grazing is also variable. Because of this, it is helpful to standardize the
animals mathematically. An ‘animal unit’ is deﬁned as, “one mature, non-lactating

bovine weighing 500 kg and fed at maintenance level, or the equivalent, expressed as
(weight)o'75, in other kinds or classes of animals” (Allen, 1991). This terminology allows

units such as animal unit days (AUD) on pasture, and animal unit grazing days per

hectare (AUGD ha”).

74

Use Of hormone implants is accepted industry practice in Michigan (Bartlett,
2003; Ritchie, 1998) and improve proﬁtability by increasing animal gains and improving
several carcass characteristics (Chester-Jones, 2002).

The objective of this experiment was to compare the productivity of a

conventional cool-season grazing system over three years with the productivity of two

C3/C4 integrated grazing systems that included either switchgrass or big bluestem. Our
hypotheses were 1) that the C 3/C 4 integrated grazing systems would have greater gain

ha" than typical Michigan grazing system model which includes only C3 pasture species,
and 2) that the integrated system that included big bluestem would result in greater gain
ha'l than an integrated grazing system that included switchgrass.

MATERIALS AND METHODS
This grazing experiment was conducted at W. K. Kellogg Biological Station in
Hickory Comers, M1, on Kalamazoo series (ﬁne-loamy, mixed, mesic Typic Hapludalfs)
soils. Four replications of three treatments were assigned to the experimental area using a
completely randomized design. The experimental layout is described in Appendix I
Figure A1
The three treatments included:

0 A cool-season grass-legume pasture representative of pastures found throughout
the Lower Peninsula of Michigan. These pastures were comprised primarily of
perennial ryegrass (Loliz/m perenne), quackgrass (A gropyron repens), alfalfa
(Medicago sativa), white clover (T rifolium repens), red clover (Trifolium
pratense), orchardgrass (Doctvlis glomerata), and tall fescue (F estuca

amndinacea).

75

0 An integrated switchgrass and cool-season grass-legume pasture. One-third of the
system was a monoculture of switchgrass that was planted in 2002, and two-thirds
was composed of the cool-season grasses and legumes listed above (Bamhart,
1994; Undersander, 2002). l

0 An integrated big bluestem and cool-season grass-legume pasture. One-third of
the system was a monoculture of big bluestem that was planted in 2002, and two-
thirds was composed of the cool-season grasses and legumes listed above
(Bamhart, 1994; Undersander, 2002).

Mid-season grazing pressure was adjusted using the put-and-take method’ of
pasture stocking (Mott, 1960). The animals used in this study wereHolstein steers that
weighed approximately 240 kg at the beginning of the grazing season. After the steers
were initially received at the experimental area, they were weighed and then ranked by
weight. Proceeding from the heaviest to the lightest, complete sets of randomized
numbers between one and twelve were applied to the ranked animals. The number that
each steer receives corresponded to the pasture to which they were assigned.

All animals were weighed twice before the experiment began in order to establish
a more accurate beginning weight (Koch et al., 1958). After the weighing events, the
steers were placed on pasture and were weighed once every four weeks (Table 2.1).
Weights recorded for each steer on consecutive dates within year were averaged to
establish beginning and ending weights for each animal with the exception of individual
replications of treatments that were terminated early due to poor pasture conditions.

Tester animals from treatment replications that had to be terminated early due to pasture

 

' See Technical Note 1 in Appendix III

76

conditions were weighted twice consecutively after being taken from their respective
pastures.

Weighing Of the steers every four weeks was primarily intended to monitor
animal health. Data from the 28-day weighing intervals was not used to describe
treatment differences in ADG. At the end of the season, the animals were removed from
the pasture and fed bay for seven days and weighed on the eighth and ninth days after
being removed from the pasture. The mean of these weighing events represented the

end-Of-season weight and was used to calculate the seasonal gain for each steer (Buskirk,

2003)
Gain ha'lwas calculated by multiplying the number of animal grazing days per '

hectare accumulated on each pasture throughout each grazing season by the ADG of the

testers on that pasture during that grazing season:
Total gain ha'l = (animal grazing days per hectare) X (average daily gain)
While gain ha"I is an important comparison to make among treatments, it is an

incomplete description of livestock performance over time. The put-and-take method of
pasture stocking was used: steers were added to and taken from the pasture depending on
existing and anticipated forage dry matter availability. This management strategy
allowed the number of AUGD accumulated by a pasture during a period of time to be
directly linked to the forage productivity of that pasture during that time period, while the
average ADG of a group of steers reﬂected the forage quality. Together, the ADG,
AUGD, and forage quality parameters allowed us to describe the dynamics of forage

quality and quantity over time.

77

Animals that remained on the same pasture throughout the entire grazing seasons
are referred to as ‘testers,’ while those that are used to adjust grazing pressure are referred
to as put-and-take animals (PT). The original protocol called for a minimum of four
testers for each treatment; in certain circumstances, however, it was necessary to reduce
the number to two or three, in order to continue collecting data without degrading the

pasture.

In the context of this study, one animal unit (AU) is standardized at 500 kg 0'75

(Allen, 1991). This standard was used to calculate the metabolic body weight of

individual animals at particular times as follows:

Number of AU = (body weight (kg))075 / 500 kgO'75

The following rules were used to calculate the number of AU represented by a steer
during a particular pasture during each time interval that they were on that pasture.

0 Each time a PT was added to or taken from a pasture in order to adjust grazing
intensity, it was weighed (Fisher, 2003).

o If a PT was put on a pasture and taken off the pasture between 28-day weighing
events, the ‘on’ and ‘off’ weights were averaged to calculate the average AU of
that animal during that time period.

0 If a PT was on pasture through two or more consecutive weighing events, the
weights at the beginning and end of each of the 28-day intervals were averaged to
calculate AU for that time interval.

0 If a PT was taken off pasture between weighing events, the ‘off’ weight and the
weight from the most recent 28-day weighing event day were used for the AU

calculation.

78

o If a PT was introduced to a pasture on a weigh day and taken off pasture before

the next weigh day, the ‘Off’ weight and ‘on’ weights were averaged to determine

AU.

Animal unit grazing days (AUGD) were used to standardize the units used to describe
grazing intensity in each system. Animal unit grazing days were calculated for each
animal by multiplying the number of days by the number of AU calculated for that
animal during that time interval. Using these procedures, total AUGD accumulated by
each treatment in a particular time period or over the whole year can be compared. Tables
2.2a and 2.2b contain an example of calculating the AUGD of a hypothetical case of a
steer being put on a pasture on the day of a 28-day weighing event, weighed 28 days later
(and put back on pasture), and taken off pasture after 17 more days, for a total of 45
grazing days (GD).

Time-series average daily gains (ADG) were calculated based on change in body
weight between two weighing events (generally 56 days) in order to reduce error related
to large variations in rumen fullness. The exception to this was the calculation of the
ADG for the entire grazing season, in which case the total number of days on pasture was
used. Table 2.4 contains the dates used to calculate time-series average daily gain.
Statistical analysis for all parameters was performed using PROC MIXED in SAS (SAS
Institute, 2000-2003).

All pastures had an area of 1.6 ha with the exception of two which, because of the
shape of the experimental area and existing permanent fencing, differed were somewhat
larger or smaller. The cool-season portions of all treatments predated this experiment and

were composed primarily of perennial ryegrass, Kentucky bluegrass, white clover,

79

quackgrass, and alfalfa. In the planning stages of this experiment, differences in Species
abundance between replications were considered negligible and blocking was not
considered. After the experiment began, it became apparent that the variability in ‘alfalfa
abundance’ was Signiﬁcant and should be considered a blOcking effect. All pastures
which were planted with alfalfa in 1994 were considered were included in ‘alfalfa
abundant’ block. The pasture layout with respect to experimental treatments and alfalfa
abundance is described in Appendix I, Figure A2.

A11 forage samples collected in this study were placed in paper bags and dried for
a minimum of 48 hours at 43°C2, and then weighed. When grazing was initiated each
year, dry matter availability was determined by harvesting plant tissue from within six
0.25 m2 quadrats. Each time the rotation sequence brought steers to a cool-season portion
of their respective treatment, plants within the quadrats were clipped at a height of
approximately 5 cm. When steers were moved into the big bluestem or switchgrass
portions of treatments, plants within quadrats were harvested at apprOximately 13 cm.
When rotations began, dry matter availability was determined using a rising-plate meter
(Gabriels, I993; Harmoney et al., 1997; Michell, 1982) unless conditions (lodging of
forage, rain) were not appropriate, in which case quadrat clippings were used instead. In
2003 and 2004, yield for the big bluestem and switchgrass monocultures yield was
determined using the rising-plate meter when conditions (precipitation, wind) permitted;
in 2005, quadrats were used exclusively.

Forage quality parameters including crude protein, neutral detergent ﬁber, and

acid detergent ﬁber were determined by using wet-chemistry techniques. In-house

 

2 See Technical Note 2 in Appendix III

80

modiﬁcations3 of the Modiﬁed Van Soest method (Stern, 1991) were used to determine
neutral detergent ﬁber and acid detergent ﬁber concentrations. The Hach modiﬁed
Kjeldahl procedure (Hach, I985; Watkins, 1987) was used to determine total N
concentration in the forage. For a full description of this data, see Chapter 2 of this
dissertation. In addition to forage quality and forage dry matter presented, data was
collected to describe pasture botanical composition over the course of the experiment (see
Appendix 11). Cost and value assumptions were also applied to pasture establishment,
maintenance, and animal performance in order to compare the net income per hectare of
the treatments (see Chapter 4 of this dissertation).

Herd Health

tit
One estradiol implant (EncoreJ, VetLife) was placed between the skin and

cartilage of the middle one-third of one ear of each steer used in this experiment. At the
time the implants were installed, an ear tag identifying each animal was also placed in
one ear. Insecticidal ear tags were also used to reduce stress from ﬂies and to reduce the
incidence Of pinkeye spread by face ﬂies. Animal health was monitored by all workers
and problems were addressed by the W. K. Kellogg Dairy herdsman, who consulted with
veterinarians as needed. Flies and parasites were controlled by applying a pour-on
iverrnectin (see Table 2.1 for dates of application). While on pasture, the steers were
given free access to a mineral supplement described in Table 2.3. During part of the
2004 grazing season, the steers were over-consuming the supplement to the degree that a

decision was made to add increasing amounts of granular sodium chloride to the Pro Phos

. 8® (Land O’ Lakes) supplement until the steers reduced their intake to recommended

 

3 See Technical Note 4 in Appendix III

81

rates. When it became apparent that the addition of sodium chloride was not reducing

consumption of the supplement, the practice of adding more sodium chloride to the

mineral supplement was ended and Pro Phos 8® again offered free choice.

Statistical analysis procedures
Gain ha", A UGD, and ADG
For analysis of gain ha'l, AUGD, and ADG for each year and for the whole experiment,

‘proc mixed’ (SAS V8) was used to conduct an ANOVA test. The ﬁxed effects used in

the analysis of each parameter are detailed in Table 2.5.

.For gain ha", the statistical model is:

Yijk=u+Ai+Bj+Ck+€ijk

where:
. . -l . .th .th th

Yijk IS the gain ha Observed In 1 year for the j treatment at the k alfalfa
abundance

u is the overall mean

A, is a ﬁxed effect, the ith year (2003, 2004, 2005)

E,- is a ﬁxed effect, the 1"“ treatment (CS-BBS, CS—Only, CS-SG)

Ck is a blocking effect, the km alfalfa abundance (abundant, not abundant)

eijk is the error term

Average daily gain (ADG) of steers over time
To describe of ADG data over time, ‘proc mixed’ (SAS V8) was used to conduct a

repeated measures ANOVA test as described in Table 2.5. Fixed effects included

82

treatment, time period, year, and alfalfa abundance; these effects were also tested for

interactions.

Time period (e.g., Time 1-2, Time 2-3) was used as a repeated

measurement, and a heterogeneous compound symmetry covariance structure was used.

Yijk = 11 + Ai + Bj + Ck + DI+ (ABC)ijk1+ eijkl

where:

Yijkl

(ABCD)ijkl

eithI

. . .th .th th .
IS the ADG Observed InI year for the j treatment at thek time

is the overall mean

is a ﬁxed effect, the ith year (2003, 2004, 2005)

is a ﬁxed effect, the f“ treatment (CS-BBS, CS-Only, CS-SG)

is a ﬁxed effect, the It’h time (Times 1, 2, 3, 4, 5, 6)

is a blocking effect, the 1th alfalfa abundance (abundant, not abundant)

is the interaction among year, treatment, time, and alfalfa abundance, and

is the error term

To describe whole-year ADG data, the statistical model is:

Yijk = 11 + Ai + Bj + (AB)ij + 6i

where:

Yi

is the ADG Observed in ith year

is the overall mean

is a ﬁxed effect,'the ith year (2003, 2004, 2005)

is a ﬁxed effect, the jth treatment (CS-BBS, CS-Only, CS-SG)

83

(AB);; is the interaction among year and

e; is the error term

Animal unit grazing days per hectare over time

To describe the distribution of AUGD over time, ‘proc mixed’ (SAS V8) was used to
conduct a repeated measures ANOVA test as described in Table 2.5. Fixed effects
included treatment, year, and time period (e.g. Time 1, Time 2); these effects were also
tested for interactions. Time was used as a repeated measurement, and an autoregressive

covariance structure was used.

Yijk = 11 + Ai + B;- + Cit + (ABClith + eijit

where:
. -l . .th .th .

y;;k Is the number of AUGD ha Observed at in 1 treatment for the j year In the
kth time period

It is the overall mean

A; is a ﬁxed effect, the ith treatment (CS-BBS, CS-Only, CS-SG)

B; is a ﬁxed effect, the 1"“ year (2003, 2004, 2005)

Ck is a ﬁxed effect, the kth time period (Times I, 2, 3, 4, 5, 6)

(ABC); jk is the interaction among year, treatment, and time period

e;;k is the error term

84

To describe whole-year AUGD data, the statistical model is:

Yijk = 11 + Ai + Bj + Cit + (AB)ijk + 6ith

where:
. . .th .th th
y;;k IS the number Of AUGD observed In I year for the j treatment for the k
alfalfa abundance
u is the overall mean
A; is a ﬁxed effect, the 1th year (2003, 2004, 2005)
B; is a ﬁxed effect, the jth treatment (CS-BBS, CS-Only, CS-SG)
Ck ' is a blocking effect, the krh alfalfa abundance (abundant, not abundant)
(AB);; is the interaction between year and treatment
e;;k is the error term
RESULTS AND DISCUSSION
Gain per hectare

While ADG and total grazing days per season are important pieces of
infonnation, gain ha'I provides crucial information for economic comparisons of
treatments. Hypothesis Ia and 1b (from Chapter 1) are:

0 Hypothesis 1a: The grazing system models that are integrated with warm season
grasses will result in greater gain ha'l than a typical Michigan grazing system
model which includes only cool season grasses and legumes.

o Hypothesis lb: An integrated grazing system that includes big bluestem will result

in greater gain ha'l than an integrated grazing system that includes switchgrass.

85

Statistical analysis Of gain ha'l revealed that treatment, presence Of alfalfa, and ‘year’ all
were signiﬁcant variables, but there were no interactions (Table 2.6).

Gain ha.1 in CS-Only system was Signiﬁcantly higher than the CS-BBS and CS-

SG treatments. Also, the CS-BBS treatment yielded more gain ha'l than the CS-SG.

This result should be qualiﬁed by noting that one-third of the CS-SG pasture was
unavailable for grazing in 2003 due to the failure Of the switchgrass portion to establish.

If the switchgrass portion of pasture had not failed to establish by 2003, it is likely that
there would not be a signiﬁcant difference in gain ha'l between the CS-BBS and CS-SG
treatments.

The effect ofiyear on steer gain ha.l was signiﬁcant and likely due to precipitation
patterns (Appendix I, Figure A.3). May Of 2004 had record rainfall, and 2004 had the
highest gain ha"; 2005 had the next highest level of gain ha'l, as well as high levels of
precipitation in June and July; 2003 had the lowest level of gain ha" due partially to the
failure of switchgrass pastures (which were not grazed), and a generally dry late-summer.
Overall, 2004 yielded more than twice the gain ha'1 of 2003. Alfalfa abundance was
included in the statistical model for gain ha"I and average daily gain over time because it
was determined to be a'signiﬁcant variable and because the presence of alfalfa as a

dominant forage species is known to Signiﬁcantly increase animal weight gain ha-l

(Barker, 1999).

Hypothesis 1a must be rejected because the CS-Only treatment resulted in more

gain ha'l than either the CS-BBS or the CS-SG treatments. Hypothesis 1b cannot be

86

rejected because the CS-BBS treatment resulted in more gain ha'l than the CS-SG

treatment.
Average daily gain

Seasonal average daily gain was signiﬁcantly inﬂuenced by ‘year,’ but not by
‘treatment’ or alfalfa abundance (Figures 2.1-2.4). Differences in ADG among years are
related primarily to differences in temperature and amount and distribution of
precipitation in each year (Appendix 1, Figures A.3-A.5). High temperatures increase the
proportion of plant cell wall constituents, thereby reducing forage digestibility (Ford,
1979), which leads to lower livestock gains . Altemately excessive moisture can lead to
such rapid accumulation of physiologically mature forage plants. In general, cool-season
pasture grasses and alfalfa decrease in digestibility as the plants advance in physiological
maturity (Kam, 2006; Morrison, 1956) which can reduce average daily gain.

Analysis of time-series ADG data indicated a signiﬁcant interaction between year,
treatment, and alfalfa abundance (Table 2.7). The presence of abundant alfalfa was
sometimes beneﬁcial and other times detrimental, without an obvious pattern among or
within years, times, or treatments. Anecdotally, it appeared that alfalfa contributed to
forage quality and pasture performance during periods of heat and moderate drought
(Sleugh, 2000). On the occasion that the alfalfa component of a pasture was overly
mature and thus of lower forage quality than the other signiﬁcant pasture Species, gains
were likely reduced by the low forage quality of that alfalfa.

Animal unit grazing days
The distribution of grazing days within each treatment and each year are summarized in

Table 2.9 and Figures 2.5-2.7.

87

A signiﬁcant reduction in forage dry matter availability in the CS-Only system
relative to the other two treatments was expected during the warmer and drier weeks of
the summer. However, the data does not Show that the CS-BBS or the CS-SG systems
accumulated more AUGD than the CS-Only system during any time periods except for
Times 4 and 5 of 2005.

Animal unit grazing day trends

Because of the CS-SG treatment failed to establish by the summer of 2003 was
due to insect damage (and Stewart’s bacterial wilt) which, if anticipated, could have been
prevented, general comments comparing treatments will be made with the assumption
that failure to establish switchgrass is atypical, and performance of the CS-SG treatment
in 2004 and 2005 is typical.

During times I and 2 in each year, the AUGD accumulated by the CS—BBS and
CS-SG do not differ signiﬁcantly from each other, but the C S-Only treatment
accumulated signiﬁcantly more AUGD than either of the other treatments. This is
because during Times 1—2, entire pasture area of the CS-Only treatment was available to
graze, whereas the big bluestem and switchgrass portions of their respective treatments
were not generally ready to be grazed until mid-June. By Time 3 of 2004 and 2005, the
big bluestem and switchgrass portions of the CS-BBS and CS-SG treatments were ready
to graze, and the difference between the numbers of AUGD accumulated by treatment
diminish or disappear altogether until Time 5. In Time 5 of each year, the CS-BBS and
CS-SG treatments equaled or exceed the number Of AUGD accumulated by the CS-Only
treatment. By early September, grazing on the big bluestem and switchgrass portions of

their respective treatments was terminated for the season. Thus, the pasture area

88

available for grazing on those treatments was essentially reduced by 33% compared to
the CS-only treatment, until the end of the grazing season for the cool-season pasture. In
years when the grazing season for the cool-season pastures goes beyond early-September,
this can result in signiﬁcantly fewer AUGD being accumulated by the CS-BBS and CS-
SG treatments. However, this situation only occurred in the second year of this
experiment.
After a killing frost, it is possible to re-introduce livestock to dormant switchgrass
or big bluestem pastures, but this may not be optimal for three reasons:
1. Dormant warm season grasses have lower forage quality and mineral
supplementation may be necessary.
2. The possibility of damage to the plant crown by the hooves of grazing animals
3. A reduction in above-ground plant material would make burning a less effective
management option.
Yearly accumulation of A U GD by treatment
Because the livestock on all treatments were increasing in size throughout the
summer, the accumulated grazing days needed to be standardized for the size of the
animals on the pasture, which is what AUGD represents. A ﬂat (as opposed to volatile or
sharply declining) distribution of AUGD over the course of the grazing season would
indicate that the grazing system was successful in distributing forage availability across
the grazing system.
Animal weight gain data from this study does not correlate with the ﬁndings of
Moore et a1. (2004), who found that, except in limited circumstances, the inclusion of

switchgrass and/or big bluestem in an integrated cool- and warm-season grazing system

89

reduced animal weight gain. The primary weakness of the integrated system in the
environment of Southern Michigan is related to the total number AUGD per season and
the failure of the integrated systems to signiﬁcantly improve the distribution of AUGD
across the grazing season. 1

Due to the failure of the stand to establish in 2002, the switchgrass portion of the
CS-SG treatment had to be re-seeded and was in the establishment phase again in 2003.
In 2003, the switchgrass portion of the CS-SG treatment was not grazed but the total area
for each replication (including the ungrazed switchgrass portion) was used in the
calculation of AUGD for this treatment.

During Times 3, 4, and 5 of 2003, the CS-SG treatment accumulated significantly
fewer AUGDS than either of the other treatments due to the fact that the switchgrass
portion of the treatment had failed to establish in 2002 and had to be re—established in
2003. In 2003, the CS—Only treatment accumulated the most AUGD, followed by the
CS-BBS treatment, and then the CS-SG treatment. Figure 2.5 demonstrates that none of
the treatments yielded a ﬂat distribution of AUGD across the 2003 grazing season.

The weather throughout the 2004 grazing season was unusually cool and wet
(record precipitation in the month of May), favoring the cool-season grasses and legumes
in general (Appendix I, Figure A.3). In 2004, the CS-Only treatment accumulated more
AUGD than either of the other treatments in Times 1, 2, and 6. In the case of Times 1
and 2, the big bluestem and switchgrass had not yet accumulated enough top-growth to
allow grazing without potentially reducing stand vigor. By Time 6 of each year, the big
bluestem and switchgrass portions of the CS-BBS and CS-SG treatments was terminated

to allow root carbohydrate reserves to regenerate prior to a killing frost (Bamhart, 1994;

90

Wolf, 1996), which comes much earlier for warm-season grasses than for cool-season
grasses.

As in the previous two years, in 2005 the CS-BBS and CS-SG treatments
accumulated fewer AUGD in Times 1 and 2. Because the steers from all treatments were
removed their respective pastures by September 8, the CS-Only treatment did not
accumulate more AUGD than the CS-BBS and CS-SG treatments during the ﬁnal grazing
interval as had been the case in 2004. All steers were removed from all treatments by
September 8 because the soil was dry enough and the plants weakened to the point that
many whole pastures plants were being plucked from the soil as the steers attempted to
graze the remaining forage.

In the third year after establishment (2005) the CS-BBS and CS-SG distributions
of AUGD over time were almost identical. While the CS-Only treatment differed from
the other treatments in its seasonal distribution of AUGD, it was not substantially more or
less-‘ﬂat,’ although the AUGD tended to increase over time rather than decreasing. In
2005, precipitation was scarce in early spring and a series of rainfall events occurred in
mid-July (Appendix I, Figure A.3). This result of this rainfall event, coupled with a lack
of early-spring rain, was an unusual abundance of high-quality forage from late-July
through early-August.

The effect of ‘year’ on the forage distribution among treatments experiment can
be observed by comparing the total number of AUGD accumulated by each treatment
during each growing season. In each treatment-by-time comparison, the AUGD

accumulated in 2004 is greater than or equal to those accumulated in 2003 and 2005.

91

In all treatments, the distribution of AUGD throughout each year is affected by
levels of precipitation. Spring and summer of 2003 were characterized by generally low
levels of precipitation while 2004 had record high precipitation in May, and 2005 had
low levels of spring precipitation and unusually high levels of precipitation in mid-June
through early-July.

In 2003 the CS-Only treatment had a typical decline of AUGD over time. In
2004, which had unusually high rainfall and low temperatures, there was a decline in
AUGD over time from Time I through Time 4, but the difference between the maximum
and minimum number accumulated per time period was fewer than 20 AUGD, whereas
in 2003 the difference was 46 AUGD. In 2005, low levels of spring precipitation
combined with high levels of summer precipitation caused a ﬂatter distribution of AUGD
over time than in either 2003 or 2004.

The low number of AUGD accumulated in the CS-SG treatment in 2003 and the
downward trend over time is due to the failure of the switchgrass portion of the pasture to
establish between 2002 and 2003. The switchgrass portion of the CS-SG treatment
accounted for one third of the total pasture area, to establish.

Implications of A UGD data
Animal unit grazing day (AUGD) data is useful for determining whether big

bluestem and switchgrass are capable of delivering enough forage of acceptable quality
during the mid-summer to meet the needs of livestock when the C3 pasture productivity

has declined. The only time that either the CS-BBS or CS-SG treatments accumulated
signiﬁcantly more AUGD in the mid-summer grazing period than the CS-Only treatment

was in Times 4 and 5 of 2005. This is likely because the unusually dry Spring of 2005

92

prevented the usual ﬂush of spring pasture growth. When signiﬁcant quantities of

precipitation ﬁnally occurred in 2005 (late-June and early-July), the pastures produced
large volumes of high quality C3 forage. This was the case for all treatments, allowing

the livestock on the CS-BBS and CS-SG treatments to accumulate more AUGD from the
cool-season portions of their pastures than otherwise would be expected. In 2003 and
2004, however, there were no Signiﬁcant differences among the treatments in the number
of AUGD accumulated in Time 3 through Time 5, except those differences which were
due to the failure of the switchgrass portion of the CS-SG treatment to establish by 2003.
Many graziers have more acreage than their livestock can effectively graze in the
spring, and they often set aside about one third of the pasture to harvest one or two
cuttings of bay for sale or feeding at a later time. After the ﬁrst or second cutting has
been harvested, the pastures that had been set aside for hay production can begin to be
grazed. Some producers, however, would prefer to keep their livestock on pasture as
much as possible and would prefer to purchase hay of known quality rather than be
forced to produce hay — especially if they have limited time, equipment, or labor
resources. Many have suggested that warm-season grasses can help accomplish this

(Balasko, 2003; Bamhart, 1994; Peterson, 2007). However, in all years of this study, the
big bluestem and switchgrass were ready for grazing before the C3 pasture species

declined in productivity. In this situation, the Anderson (2000) recommends that mixed
or solid stands of switchgrass be grazed when they are ready, regardless of whether the

cool-season pastures are still productive. He advises that it is better is to harvest or
stockpile excess C3 forage and to graze the big bluestem and switchgrass while their

forage volume and quality are optimal. This was practiced in this experiment, but it adds

93

more complexity than most graziers are accustomed to in traditional grazing systems.
The producer’s willingness to set aside one third of his pasture acreage for hay
production or warm-season grass pasture will, in part, be related to their capacity
(equipment, storage, time) to harvest quality hay from a traditional grazing system and/or
their willingness to increase the complexity of their system in order to keep their
livestock on pasture.
CONCLUSIONS

Depending on the availability of a dry-lot and the price/availability of feed, a
livestock manager may or may not be willing to reduce stocking density as pasture forage
productivity declines as the summer progresses. An assumption in this experiment was
that including big bluestem and switchgrass monocultures species as part of the grazing
system would effectively reduce the number of acres available for grazing in the early
spring months, eliminating the need for the pasture manager to mow or harvest a portion

Of the pasture for hay due to excessive growth and/or declining forage quality associated
increasingly mature C3 forage species. Altemately, if the number of livestock on Spring

pasture is sufﬁcient to keep pace with spring pasture growth, that same number of

livestock will need more mid-summer forage than that pasture can prOduce because of

livestock growth and/or pasture decline. In this case, some of the livestock would need to

be sold or would need stored forage during the summer decline in pasture productivity.
Because the livestock on all treatments were increasing in size throughout the

summer, the accumulated grazing days needed to be standardized for the Size of the

. animals on the pasture, which is what AUGD represents. A ﬂat (as Opposed to volatile or

sharply declining) distribution Of AUGD over the course Of the grazing season would

94

indicate that the grazing system was successful in distributing forage availability across
the grazing system. Neither the CS-BBS nor CS-SG treatments resulted in consistently
ﬂatter distributions of AUGD over time (Tables 2.9 and 2.10; Figures 2.5-2.7 and 2.8-
2.10). A

Because pasture managers vary in their ability to harvest and otherwise control
excessive spring growth and/or move livestock to a dry lot to be fed stored forage during
periods of pasture shortage, the acceptable distribution of grazing days over the course of
the grazing season will vary among producers; thus, sweeping statements about the
relative ﬁtness of the CS-BBS, CS-Only, and the CS-SG systems for Southern Michigan

cannot be made. However, it is clear that Hypothesis la should be rejected, because
neither CS-BBS nor the CS-SG treatments yielded more gain ha"l than the CS-Only
treatment; in fact, the Opposite was true. Also, Hypothesis 1b should not be rejected,
because the CS-BBS treatment did yield more gain ha'l than the CS-SG treatment.

Although forage quality of big bluestem and switchgrass harvested from quadrats
is much lower than that from a cool-season pasture during the mid- to late-summer, it did
not have a signiﬁcant impact on ADG. This is likely due to the ability of livestock to
selectively graze in poor pastures as well as the short time intervals that the livestock
were on the big bluestem and switchgrass monocultures.

The practical implications of this research will vary by livestock production
operation, depending on several characteristics:

0 The class of livestock that characterizes the livestock operation.

95

The operational ﬂexibility of the producer to either sell cattle or feed stored forage
during periods of time when C3 pasture productivity is inadequate (Moore et al.,
2004).

The willingness of the producer to:

0 design, establish, and manage a more complex pasture system.

0 wait for several years before recapturing their investment.

96

Table 2.1. Weighing events by year)r
2003 2004 2005
28-A pr 1’ 26-A pr 20-A pr
29—Apr 27-Apr 2l-AprI
28-May1: 25-MayI 19-May
24 JuneI 22 June: 16 June};
22-July',t 20-JulyI l4-Ju1y
I9-Aug l4-Aug l l-Augi
l 7-Sep 22-Sep 8-Sep
l 8-Sep 23-Sep l 6-Sep

l 7-Sep

 

IWeights from italicized dates were used to establish beginning- and end- of season
weights.
ilndicates dates Of application of ivermectin pour-on

97

Table 2.2a. Example of animal unit grazingday calculation (continued in Table 2.2b)

 

 

Date Event/Action Weight AU Calculation Animal Units
21-Apr Put PT on pasture 273 kg (273 kg 0.75)/ 5000.75 0.55
l9-May Weigh day 309 kg (3090.75)/ 5000.75 0.62

5-Jun Remove PT 325 kg (3250.75)/ 5000.75 0.65

 

98

Table 2.2b. Example of animal unit grazing day calculation (continued).

 

 

Time AU Average AU Calculation for
Interval (GD) During Interval AUGD AUGD
21-Apr - l9-May 28 (0.55 + 0.62)/2 0.585 0.585 AU X 28 GD 16.4
ZO-May - 5-June 17 (0.62 + 0.65)/2 0.635 0.635 AU X 17 GD 10.8
Total AUGD 27.2

 

99

Table 2.3. Guaranteed analysis of Land O’Lakes Pro Phos 8 granular mineral

 

 

supplement.

Mineral/vitamin Content Unit
Calcium (Ca), Min 12.00%

Calcium (Ca), Max 14.00%

Phosphorus (P), Min 8.00%

Salt (NaCl), Min 16.00%

Salt (NaCl), Max 18.50%

Magnesium (Mg), Min 2.00%

Potassium (K), (Min) 0.10%

Zinc (Zn), Min 4,375 ppm
Manganese (Mn), Min 2,500 ppm
Copper (Cu), Min 1,300 ppm
Iodine (1), Min 130 ppm
Selenium (Se), Min 22 ppm
Vitamin A, Min 330,000 I.U./kg
Vitamin D, Min 26,000 I.U./kg
Vitamin E, Min 130

 

l.U./kg

100

Table 2.4. Beginning and ending dates of grazing intervals used to calculate time-series
ADG for the 2003 through 2005 gazing seasons.

 

 

Year 2003 2004 2005
Interval Beginning Ending Beginning Ending Beginning Ending
Time 1-2 29-Apr 24-June 27-Apr 22-June 21-Apr l6-June
Time 2-3 29-May 22-Ju1y 26-May 20 July 20-May l3-Ju1y
Time 3-4 25-June 19-Aug 23-June 17-Aug l7-June ll-Aug
Time 4-5 23-July 9-Sep 21-June I4-Sep 14-July 8-SejL

 

101

Table 2.5. Fixed effects and interactions used in the ﬁrll model of gain ha'], animal unit
grazing days (AUGD), and average daily gain (ADG).

 

 

Parameter Treatment Year AAT Treatment X Year
Gain per hectare x x x
AUGD x x x x
ADG x f

 

TAA: alfalfa abundance
I Occasional interaction with alfalfa abundance (also see Table 7).

102

Table 2.6. Gain ha'l by treatment, year, and alfalfa abundance.

 

 

Effect Gain per hectare SE
Treatment
CS-BBS 624.4 b 16.2
CS-Only 721.0 a 16.2
CS-SG 572.9 c 14.0
Year
2003 453.4 c 15.5
2004 907.1 a 15.5
2005 558.9 b 15.5

Alfalfa abundance
AA 669.4 a 12.7
ANA 610.1 b 12.7

Means within effect with different letters are signiﬁcantly different (p < 0.05).

103

Table 2.7. Average daily gain of steers by treatment and year at time intervals throughout
the 2003-2005 grazingseasons.

 

 

 

 

Time 1-2 Time 2-3 Time 3-4 Time 4-5

Year Treatment ADG SE ADG SE ADG SE ADG SE
kg day‘l

2003 CS-BBS 1.72 a 0.09 1.38 a 0.09 0.79 a* 0.06 0.98 a* 0.08
2003 CS-Only 1.81 a 0.07 1.50 a 0.08 0.69 a* 0.06 0.99 a* 0.07
2003 CS-SG 1.72 a 0.08 1.48 a 0.08 0.72 a 0.06 1.09 a* 0.09
2004 CS-BBS 1.72 a 0.05 1.32 a 0.06 1.11 a* 0.04 0.46 b* 0.05
2004 CS-Only 1.84 a 0.06 1.34 a 0.06 0.99 ab* 0.05 0.73 a* 0.06
2004 CS-SG 1.75 a 0.05 1.26 a 0.06 0.96 b* 0.04 0.57 b 0.05
2005 CS-BBS 1.81 a* 0.06 1.48 a* 0.06 0.79 a* 0.05 0.91 b“ 0.06
2005 CS-Only 1.52 b* 0.06 1.32 ab* 0.07 0.72 ab“ 0.05 1.10 a* 0.06
2005 CS-SG 1.44 b* 0.06 1.29 b* 0.06 0.65 b* 0.04 0.83 b* 0.05

 

Means within year with different letters are signiﬁcantly different (p < 0.05).

*Indicates an interaction with alfalfa abundance during that time period

104

Figure 2.1. Average daily gain of steers by treatment and year at intervals throughout the
2003 grazing seasons.

 

+CS-BBS I
—I- CS-Oniy “

.. CS-SG W: ;

 

Time 1-2 Time 2-3 Time 3-4 Time 4-5

Time Interval

 

105

Figure 2.2. Average daily gain of steers by treatment and year at intervals throughout the
2004 grazing seasons.

ADG (kg)

 

Time l-2 Time 2-3 Time 34 Time 4—5

Time Interval

 

106

Figure 2.3. Average daily gain of steers by treatment and year at intervals throughout the
2005 grazing seasons.

; _________.__._ _____.

 

 

 

 

 

 

 

 

 

 

 

 

 

2
1.5 .- .. '”
30 +CS—BBS
; l r ~-~~~~ +CS-On1y
3 " CS-SG-
0.5 r *— w ~ ~ ~ ~4~ — e ~— ~~
0 -
Time 1-2 Time 2-3 Time 3-4 Time 4—5
Time Interval

 

 

107

Table 2.8. Average daily gain of steers by treatment and year at intervals throughout the
2003-2005 grazing seasons.

 

 

Year Treatment ADG
2003 CS-BBS 1.30 a
2003 CS-Only 1.34 a
2003 CS-SG 1.36 a
2004 CS-BBS 1.45 a
2004 CS-Only 1.47 a
2004 C S-SG 1.42 a
2005 CS-BBS 1.29 a
2005 CS-Only 1.27 a
2005 CS-SG 1.12 b

 

Means within year with different letters are Signiﬁcantly different (SE = 0.04; p < 0.10).

108

Figure 2.4. Average daily gain of steers between 2003 and 2005.

Kilograms
P
as

P
A

 

2003 2004 2005

 

' Year
1 ,, L ,, ,
Means with different letters are signiﬁcantly different (p < 0.05). SE = 0.02.

109

 

Table 2.9. Animal unit grazing days accumulated by treatment and year at intervals
throughout the 2003-2005 grazig seasons

 

 

 

 

Year Treatment Time 1 Time 2 Time 3 Time 4 Time 5 Time 6 Season
AUGDma
2003 CS-BBS 65.1 b 42.2 b 54.0 a 44.13 34.8 a 0.0 a 240.2 b
2003 CS-Only 80.9 a 58.8 a 56.9 a 47.3 a . 34.6 a 2.5 a 281.0 a
2003 CS-SG 61.0 b 42.9 b 41.1 b 28.5 b 23.0 b 0.0 a 196.5 c
2004 CS-BBS 72.8 b 76.2 b 88.7 a 74.2 a 71.1 a 52.7 b 435.6 b
2004 CS-Only 95.5 a 97.9 a 88.9 a 78.6 a 77.5 a 87.8 a 526.1 a
2004 CS-SG 74.9 b 75.9 b 91.8 a 72.1 a 74.5 a 46.0 b 435.1 b
2005 CS-BBS 62.9 b 58.4 b 67.0 a 76.0 a 70.6 a 0.0 a 335.0 a
2005 CS-Only 81.1 a 74.4 a 64.8 a 61.2 b 58.8 b 0.0 a 340.3 a
2005 CS-SG 62.3 b 55.9 b 67.2 a 74.3 a 61.6 ab 0.0 a 321.3 a

 

Means within the same column and year with different letters are signiﬁcantly different

(p < 0.05). The standard error for CS—BBS = 3.7; CS-Only = 3.7; CS-SG = 3.2.

110

Figure 2.5. Number of AUGD ha~l accumulated by each treatment in 2003.

g , a. -
g tics-BBS
E lcs-Onry
:2 1395-59 ,,
<

 

3,:
iii
g
3
3.1
g;
3’5
Ti

1:44:2th :35 $315111 inlet

Time 1 Time 2 Time 3 Time 4 Time 5 Time 6

Interval

 

 

Means within the same time interval with different letters are signiﬁcantly different (p <
0.05).

111

Figure 2.6. Number of AUGD ha-l accumulated by treatment in 2004.

i—n—I
AGAWGN
GGGGO

AUGD per hectare
N
G

 

6

Time 1 Time 2 Time 3 Time 4 Time 5 Time 6

Interval

 

 

 

112

Figure 2.7. Number of AUGD per hectare accumulated by each treatment in 2005.

AUGD per hectare

 

Time 1 Time 2 Time 3 Time 4 Time 5

Interval

 

Means within the same time interval with different letters are signiﬁcantly different (p <
0.05). '

113

Table 2.10. Treatment across year comparison of AUGD per hectare over time

 

 

 

 

Treatment Year Time 1 Time 2 Time 3 Time 4 Time 5 Time 6 Season
AUGD/ha
CS-BBS 2003 65.1 a 42.2 c 54.0 c 44.1 b 34.8 b 0.0 b 240.2 c
CS-BBS 2004 72.8 a 76.2 a 88.7 a 74.2 a 71.1 a 52.7 a 435.6 a
CS-BBS 2005 62.9 a 58.4 b 67.0 b 76.0 a . 70.6 a 0.0 b 335.0 b
CS-Only 2003 80.9 b 58.8 c 56.9 b 47.3 c 34.6 c 2.5 b 281.0 c
CS-Only 2004 95.5 a 97.9 a 88.9 a 78.6 a 77.5 a 87.8 a 526.1 a
CS-Only 2005 81.1 b 74.4 b 64.8 b 61.2 b 58.8 b 0.0 b 340.3 b
CS-SG 2003 61.0 b 42.9 c 41.1 c 28.5 b 23.0 c 0.0 b 196.5 c
CS-SG 2004 74.9 a 75.9 a 91.8 a 72.1 a 74.5 a 46.0 a 435.1 a
CS-SG 2005 62.3 b 55.9 b 67.2 b 74.3 a 61.6 b 0.0 b 321.3 b

 

Means within the same column and treatment with different letters are signiﬁcantly
different (p < 0.05). The standard error for CS-BBS = 3.7; CS-Only = 3.7; CS-SG = 3.2.

114

Figure 2.8. Mean number of AUGD per hectare accumulated by the C S-BBS treatment
from 2003 through 2005.

 

AUGD per hectare

Time 1 Time 2 Time 3 Time 4 Time 5 Time 6

Interval

 

 

 

Means within the same time interval with different letters are signiﬁcantly different (p <
0.05).

115

Figure 2.9. Mean number of AUGD per hectare accumulated by the CS-Only treatment
from 2003 through 2005.

2

a

E [12003
is .2004
D-

c [12005
o L

D

<

 

Time 1 Time 2 Time 3 Time 4 Time 5 Time 6

Interval

 

 

 

Means within the same time interval with different letters are signiﬁcantly different (p <
0.05).

116

Figure 2.10. Mean number of AUGD ha'1 accumulated by the CS-SG treatment from
2003 through 2005.

120
100

56‘”
99:

AUGD per hectare
N
e

 

a

Time 1 Time 2 Time 3 Time 4 Time 5 Time 6

Interval

 

Means within the same time interval with different letters are signiﬁcantly different (p <
0.05).

117

LITERATURE CITED

Allen, V.G. 1991. Terminology for grazing lands and grazing animals Pocahontas Press,
Blacksburg, VA. -

Balasko, J.A., and C. J. Nelson. 2003. Grasses for northern areas, p. 125-147, In R. F.
Barnes, Nelson, J. C., Moore, K. J., and M. Collins, ed. Forages: an introduction
to grassland agriculture, Vol. 1.

Barker, J.M., Buskirk, D. D., Ritchie, H. D., Rust, S. R., Leep, H. R., and K. J. Barclay.
I999. Intensive grazing management of smooth bromegrass with or without
alfalfa or birdsfoot trefoil: heifer performance and sward characteristics. The
Professional Animal Scientist 15:130-135.

Bamhart, SK. 1994. Warm-season grasses for hay and pasture - Pm-569 [Online].
Available by Iowa St. U. Ext
http://www.extension.iastate.edu/PublicationS/PM569.pdf (veriﬁed 03/22/2006).

Bartlett, B. 2003. Michigan State University District Livestock Agent, Upper Peninsula
Michigan State University Extension, pp. Ben said that it is fairly normal for
stocker steers to be implanted on pasture. If they are not, it is more because the
operator did not get around to it (or something like that). In D. Hudson, (ed.).

Bransby, DJ. 1989. Compromises in the design and conduct of grazing experiments, p.
136, In G. C. Marten, ed. Grazing research: design, methodology, and analysis.
CSSA Spec. Pub. No. 16. Crop Science Society Of America and the American
Society of Agronomy, Madison.

Buskirk, D. 2003. Associate Professor of Animal Sciences, Michigan State University.

Chester-Jones, H. 2002. Implant strategies for dairy steers. University of Minnesota
Southern Research and Outreach Center, Waseca, MN.

Fisher, D.S. 1999. Deﬁning the experimental unit in grazing trials [Online]
lmﬂ/wwwasas.Org/ias/svmposia/proceedings/0905.pdf (veriﬁed March 16,
2006).

Fisher, D.S. 2003. USDA-ARS Rangeland Scientist, Watkinsville, GA.

Ford, C.W., Morrison, 1. M., Wilson, J. R. 1979. Temperature effects on lignin,

hemicellulose and cellulose in tropical and temperate grasses. Australian Journal
of Agricultural Research 30:621-633.

118

Gabriels, P.C.J., and J.V. Van Den Berg. 1993. Calibration of two techniques for
estimating herbage mass. Grass and Forage Science 48:329-335.

Harmoney, K.R., K.J. Moore, J.R. George, BC. Brummer, and JR. Russell. 1997.
Determination of pasture biomass using four indirect methods. Agronomy Journal
89:665-672. ‘

Kam, J.F., Berdahl, J. D., and Frank, A. B. 2006. Nutritive quality of four perennial
grasses as affected by species, cultivar, maturity, and plant tissue. Agronomy
Journal 98: 1400-1409.

Koch, R.M., E.W. Schleicher, and V.H. Arthaud. 1958. The Accuracy of Weights and
Gains of Beef Cattle. Journal of Animal Science 17:604-611.

Michell, P. 1982. Value of a rising-plate meter for estimating herbage mass of grazed
perennial ryegrass-white clover swards. Grass and Forage Science 37:81-87.

Moore, K.J., T.A. White, R.L. Hintz, P.K. Patrick, and EC. Brummer. 2004. Forages and
pasture management - Sequential grazing of cool- and warm-season pastures.
Agronomy Journal 96:1103-1111.

Morrison, F .B. 1956. Feeds and Feeding. 22nd ed. Morrison Publishing Company, Ithaca,
NY.

Mott, GO. 1959. Symposium on forage evaluation: IV. animal variation and
measurement of forage quality. Agronomy Jounal 51:223-226.

Mott, GO. 1960. Grazing pressure and the measurement of pasture production. Proc VIII
Int. Grassl. Cong.:606-611.

Mott, GO, and H.L. Lucas. 1952. The design, conduct, and interpretation of grazing
trials on cultivated and improved pastures. Proc VI Int. Grassl. Cong. 6: 1380-
1385.

Peterson, P.R., Rayburn, E. B., Cropper, J. B., and Belesky, D. P. 2007. Perennial wann-
season grasses, In E. B. Rayburn, ed. Forage utilization for pasture-based

livestock production. Natural Resource, Agriculture, and Engineering Service,
Ithaca, NY.

Ritchie, H., Rust, S., and R. Black. 1998. Looking ahead to Michigan’s beef cattle
industry in the year 2000: Special Report SR449201. Michgian State University

Agricultural Experiment Station.

SAS Institute, Inc. 2000-2003. SAS. Release 8 and 9. SAS Institute, Inc., Cary, NC.

119

Sleugh, B., Moore, K. J., George, J. R., and Brummer, E. C. Brummer. 2000. Binary
Legume-Grass Mixtures Improve Forage Yield, Quality, and Seasonal
Distribution. Agronomy Journal 92:24-29.

Stuedemann, J.A., and A. G. Matches. 1989. Measurement of animal response in grazing
research, p. 136, In G. C. Marten, ed. Grazing research: design, methodology, and
analysis. CSSA Spec. Pub. No. 16. Crop Science Society of America, American
Society of Agronomy, Madison.

Undersander, D. 2002. Conversation about using warm-season grasses as a component of
grazing systems.

Wheeler, J.L., J.C. Burns, R.D. Mochrie, and H.D. Gross. 1973. Choice of ﬁxed or
variable stocking rates in grazing experiments. Experimental Agriculture 9:289-
302.

Wolf, D.D., Fiske, D. A. 1996. Planting and managing switchgrass for forage, wildlife,
and conservation: Publication Number 418-016. Virginia Cooperative Extension,
Blacksburg, VA.

120

Chapter 3

ECONOMIC PERFORMANCE OF COOL-SEASON PASTURES INTEGRATED
WITH SWITCHGRASS AND BIG BLUESTEM

ABSTRACT
High temperatures and lack of precipitation Often cause the productivity and quality of
cool-season pastures in Michigan to severely decline for an extended period during the
summer. Our objective was to compare the proﬁtability of three pasture systems over
three years of experimentation. Cool-season grass and legume pastures (CS-Only) were
compared to integrated big bluestem and cool-season grass and legume pastures (CS-
BBS) and integrated switchgrass and cool-season grass and legume pastures (CS—SG).
Accumulated animal grazing-days per hectare and average daily gain per steer and
variable expenses related to treatment were used to calculate net income per hectare for
each treatment. The CS-Only, CS-BBS and CS-SG treatments had three-year net returns
of $2,035, $1,665, and $1,495 per hectare, respectively. The differences were primarily
due to the high cost of establishing big bluestem and switchgrass monocultures and the
pasture management constraints that limit the grazing of these pastures before mid-June
and after early-September. These management constraints reduce the pasture area
available for grazing on the CS-BBS and CS-SG treatments by 33% in the spring and
occasionally during the last part of the grazing season, resulting in the accumulation of
fewer animal grazing days. While producers in Southern Michigan could reduce the risk
of running out of mid-summer pasture by adopting pasture systems similar to the C S-
BBS or CS-SG treatments, the management constraints of these systems will frequently
induce. an opportunity cost (i.e., the annual loss of animal grazing days in the spring and

frequent loss in the fall) that seems to outweigh the value of reducing that risk.

121

INTRODUCTION

During the mid-summer months of most years, Michigan producers who keep
livestock on pasture are faced with a choice: to remove their livestock from dormant
summer pastures to locations where they are fed stored fOrage, or to allow their livestock
to continue grazing. The choice to leave livestock on pasture during periods of
inadequate pasture productivity can result in overgrazing. Overgrazing often reduces
pasture productivity (Coleman, 2007) resulting in weed invasion (Hazell, 1967), and
therefore may require pasture renovation or replanting. In southwest Lower Michigan,
spring and fall cool season grass-legume growth is excellent due to mild temperatures
and adequate rainfall, but drought stress during the summer months often causes these
grasses and legumes to become dormant (Laude, 1953). Several state universities in the
North Central Region promote the use of native perennial warm season grasses, such as
switchgrass and big bluestem as summer forage in some circumstances (Anderson, 2000;
Bamhart, 1994a; Bartholomew, 1995). These grasses have the potential to provide large
volumes of high quality mid-summer forage when cool season grasses and legumes are
less productive (Balasko, 2003). While both Species are indigenous to Michigan, no
grazing research has been conducted to determine if using these species in a pasture
system is a viable option. Beyond their capacity to produce large quantities of biomass,
switchgrass and big bluestem grow well in marginal soils and can provide excellent cover
for nesting birds (George et al., 1979; Tober, 1992) and for other types of wildlife and
can reduce the loss of sediment, nitrogen, and phosphorus in areas prone to erosion

(Blanco-Canqui, 2004). Switchgrass (Ma, 2000) and other native tallgrass prairie species

122

are believed to sequester large amounts of carbon in the soil, and grasslands in general
present great aesthetic appeal (Keeney, 2007), particularly in the summer and fall.

The Michigan Hay and Grazing Council has described alternative forage research as one
of its top priorities (Lindquist, 2002). Both the Michigan Department of Natural
Resources (Sargent, 1999) and the USDA Natural Resource Conservation Service
(USDA-NRCS, 1996) have encouraged Michigan graziers to include native warm season
grasses in their grazing systems. However, graziers may be reluctant to use these grasses
without knowledge of their performance in Michigan.

‘Optimal' management of grazing systems

‘Optimal’ management of a perennial pasture can be deﬁned as: pasture
management which allows the livestock in the grazing system to maintain or exceed
established (i.e. normal) levels of weight gain or maintenance and stocking rate for that
livestock class, without causing undue pasture degradation. This deﬁnition assumes that
the grazing management occurs within the normal range of environmental conditions for
that particular location.

Livestock producers are not just interested in adopting the system with the most
potential to maximize net proﬁt per hectare; rather, they are interested in maximizing net
return per acre while keeping their level of risk in balance (Parch, 1989). This balance
applies not only to ﬁnding the ‘optimal’ stocking rate, but also to the pasture species that
a producer decides to adopt, Since seasonal environmental dynamics inﬂuence the
productivity of a pasture over time (Bamhart, 1999) and will affect what constitutes a

low- to moderate-risk stocking rate at different points throughout the year. Further, the

123

optimal management of a particular grazing system will vary with plant species (and
combinations), environmental conditions, and class of livestock.
Economic comparisons of grazing systems
The net proﬁtability of a system is deﬁned as the difference between the gross revenue
and the total costs of production (Parch, 1989). For the purpose of economic
comparisons, the variable costs of production are included in calculations while ﬁxed
costs of production (e.g., depreciation of fencing and watering systems) are often not
included, as long as those costs are equivalent among treatments being compared (Black,
2007). The basis for calculating gross revenue generated from a grazing system in a
_ particular environment depends on the type of livestock production operation being
modeled and the types of challenges that livestock production systems in that
environments face. Parameters used to calculate gross revenue for these systems can
include, average daily gain (ADG) (Parch, 1989), or the total amount of milk or meat
produced per hectare during a given period of time. Units such as “animal grazing days”
per hectare are also sometimes used. However, when the sizes of experimental animals
are variable, the amount of forage dry-matter consumed in one day of grazing is also
variable. Because of this, it is helpful to standardize the animals mathematically. An
‘animal unit’ is deﬁned as, “one mature, non-lactating bovine weighing 500 kg and fed at
maintenance level, or the equivalent, expressed as (weight)°'75, in other kinds or classes of
animals” (Allen, 1991). This terminology allows units such as animal unit days (AUD)
on pasture, and animal unit grazing days (AUGD) per hectare.

Given that net proﬁt is the difference between gross revenue and total costs of

production, stocker-steer pasture managers have two main goals:

124

1. maximizing animal gain per hectare while balancing risk
2. minimizing costs without compromising the sustainability of the system

While ‘true’ net proﬁtability is the difference of gross revenue and the total cost
of production, both ﬁxed and variable, ﬁxed costs (e.g., depreciation of fencing and
watering systems) are often excluded from comparisons of grazing systems as long as
those costs are equivalent among treatments being compared (Black, 2007). In pasture
research, when forage Species are being compared, those activities that differ among
treatments are included in calculations (Black, 2007). These activities and expenses can
include use of equipment and labor and purchasing pesticides, chemicals, seed, and other
necessary inputs.
Determining variable costs

Farm operators Often hire custom Operators to complete certain types of work;
rates for particular custom farm Operations vary from region to region (Ward, 2006).
Publications of custom farm rates are often a starting place for negotiations between
equipment owner/operators and farm managers, and it is common for land-grant
cooperative extension services to publish these rates. These rates are sometimes
published as an average with a range of one standard deviation (Ward, 2006). To
standardize chemical input costs to make comparisons between different production
systems, researchers may compile price lists for inputs such as pesticides and associated
adjuvants (Sprague, 2007). Some variable costs, such as the average cost of pasture seed,
are not readily available, and would be highly subjective if it existed, due to the

variability in the performance of different forage varieties. Thus, cost estimates for less

125

common, proprietary, variable quality, or price-volatile inputs sometimes must be
obtained from reputable vendors.

Our objective was to compare the proﬁtability of three pasture systems over three
years of experimentation: a conventional cool-season grass and legume grazing system
and two systems which were similar except for the substitution of big bluestem or
switchgrass monocultures for one-third of the pasture acreage.

MATERIALS AND METHODS

This experiment was conducted from 2003 through 2005 at W. K. Kellogg
Biological Station in Hickory Comers, M1, on Kalamazoo series (ﬁne-loamy, mixed,
mesic Typic Hapludalfs) soils. F oUr replications of three treatments were assigned to the
experimental area using a completely randomized design. The experimental layout is
described in Figure A] of Appendix I.

The three treatments included:

I. A cool-season grass-legume pasture representative of pastures found throughout
the Lower Peninsula of Michigan. These pastures were comprised primarily of
perennial ryegrass (Lolium perenne), quackgrass (A gropyron repens), alfalfa
(Medicago sativa), white clover (Trifolium repens), red clover (T rifolium
pretense), orchardgrass. (Dactylis glomerata), and tall fescue (F estuca
arundinacea).

2. An integrated cool-season grass-legume and big bluestem pasture. One-third of
the system was a monoculture of big bluestem that was planted in 2002, and two-
thirds was composed of the cool-season grasses and legumes listed above

(Bamhart, 1994a; Bamhart, 1994b; Undersander, 2002).

126

3. An integrated cool-season grass-legume and switchgrass pasture. One-third of the
system was a monoculture of switchgrass that was planted in 2002, and two-thirds
was composed of the cool-season grasses and legumes listed above (Bamhart,
19943; Bamhart, l994b; Undersander, 2002). i

All pastures had an area of 1.6 ha with the exception of two which, because of the
shape of the experimental area and existing permanent fencing, differed somewhat. The
cool-season portions of all treatments predated this experiment. The major cool season
species included: perennial ryegrass, Kentucky bluegrass, white clover, quackgrass, and
alfalfa. In the planning stages of this experiment, differences in species abundance
between replications were considered negligible and blocking was not considered. After
the experiment began, it became apparent that the variability in ‘alfalfa abundance’ was
signiﬁcant and should be considered a blocking effect. All pastures which were planted
with alfalfa in 1994 were considered were included in ‘alfalfa abundant’ block. The
pasture layout with respect to experimental treatments and alfalfa abundance is described
in Figure A2 of Appendix I.

Pastures were managed as uniformly as possible, but each was managed separately
and optimally (Bransby, 1989). When grazing was initiated each spring, cattle on all
treatments were given access to the entire cool-season portion of their respective
pastures, with the exception of the fourth replication of the CS-Only treatment, where a
concurrent experiment was being carried out which required that livestock be excluded
from a portion of the pasture at certain points of time. As the spring ﬂush of cool-season
pasture growth intensiﬁed, pastures were “staged” (Barker, 1999): cattle were limited to

one half, and later, one quarter of the cool-season portion of their treatments at which

127

point rotational grazing began and continued until the end of the season. During the ﬁnal
grazing event of the season, the cattle were again given access to the entire cool-season
portion of their respective pastures. In 2003, the big bluestem portion of the CS-BBS
treatments was split into three parts each of which was grazed in rotation (the switchgrass
portion of the CS-SG treatment was not grazed in 2003). In 2003, residue of refused big
bluestem forage was mowed after grazing as needed in order to make the pasture height
more uniform at subsequent grazing events. However, it was noted that the small size of
the subdivisions within the big bluestem pasture resulted in excessive trampling of forage
which caused a high proportion of plants to re-grow from the crown rather from
intemodes, thereby increasing the amount of time required for re-growth. Further, the
tractor tire tracks left from mowing caused the crushed plants to re-grow from the crown;
this re—growth was lush and tended to be grazed preferentially by the livestock. Thus, in
2004 and 2005 the big bluestem and switchgrass portions Of the CS-BBS and CS-SG
treatments were not subdivided or mowed and each of these portions was left undivided
and grazed as one of the rotations.

Decisions about timing and duration of rotations were dictated by current and
anticipated pasture forage availability rather than using strict dates or interval lengths.
Variables such as soil moisture, physiological and re-growth stages of forage species,
typical seasonal weather patterns, and weather forecasts were used to make rotation
decisions. After the spring ﬂush, the grazing intervals were usually seven days for the
CS-BBS and CS-SG treatments; CS-only treatments were usually rotated every ten days
and consisted of four subdivisions. Environmental data is described in Figures A.3-A.5

of Appendix I.

128

Mid-season grazing pressure was adjusted using the put-and-take methodl of
pasture stocking (Mott, 1960). ‘Tester steers’ were left on the same replication of the
same treatment for the entire grazing season. Put-and-take steers were added to or
removed from a particular replication to increase or reduce the grazing pressure in order
to:

- optimize utilization Of pasture forage
o prevent the accumulation low-quality forage
- meet but not exceed the current or anticipated forage dry matter production

Within an individual pasture, animals moved to another paddock when one or
more of the following was true: 1) paddocks forward in the rotations were ,about'to
become over-mature; 2) animals or pasture might be harmed by low forage levels in the
current rotation; 3) rapid pasture growth indicated that shorter intervals were needed in
the immediate future to prevent accumulation of low-quality forage which would likely
be refused.

All animals were weighed twice before the experiment began in order to establish
a more accurate beginning weight (Koch et al., 195 8). After the weighing events, the
steers were placed on pasture and were weighed once every four weeks (see Table 3.1).
Weights recorded for each steer on consecutive dates within year were averaged to
establish beginning and ending weights for each animal with the exception of individual
replications of treatments that were terminated early due to poor pasture conditions.

Tester animals from treatment replications that had to be terminated early due to pasture

 

' See Technical Note 1 in Appendix III

129

conditions were weighted twice consecutively after being taken from their respective
pastures.

Weighing of the steers every four weeks was primarily intended to monitor
animal health. At the end of the season, the animals were removed from the pasture and
fed hay for seven days and weighed on the eighth and ninth days after being removed
from the pasture. The mean of these weighing events represented the end-of-season
weight and was used to calculate the seasonal gain for each steer (Buskirk, 2003).

Gain per hectare was calculated by multiplying the number of animal grazing
days per hectare accumulated on each pasture throughout each grazing season by the
ADG of the testers on that pasture during that grazing season:

Total weight gain per hectare = (animal grazing days per hectare) X (average daily gain)

Animals that remained on the same pasture throughout the entire grazing seasons
are referred to as ‘testers,’ while those that are used to adjust grazing pressure are referred
to as put-and-take animals (PT). The original protocol called for a minimum of four
testers for each treatment; in certain circumstances, however, it was necessary to reduce
the number to two or three, in order to continue collecting data without degrading the
pasture. Other pasture and animal performance data collected include average daily gain
(ADG) and accumulated animal unit grazing days (AUGD) per hectare. Statistical
analysis for all parameters was performed using PROC MIXED in SAS (SAS Institute,
2000-2003).

All forage samples collected in this study were placed in paper bags and dried for
a minimum of 48 hours at 43°C2, and then weighed. When grazing was initiated each

year, dry matter availability was determined by harvesting plant tissue from within six

 

2 See Technical Note 2 in Appendix III

130

0.25 m2 quadrats. Each time the rotation sequence brought steers to a cool-season portion
of their respective treatment, plants within the quadrats were clipped at a height of
approximately 5 cm. When steers were moved into the big bluestem or switchgrass
portions of treatments, plants within quadrats were harvested at approximately 13 cm.
When rotations began, dry matter availability was determined using a rising-plate meter
(Gabriels, 1993; Harmoney et al., 1997; Michell, 1982) unless conditions (lodging of
forage, rain) were not appropriate, in which case quadrat clippings were used instead. In
2003 and 2004, yield for the big bluestem and switchgrass monocultures yield was
determined using the rising-plate meter when conditions (precipitation, wind) permitted;
in 2005, quadrats were used exclusively.

ln-house modiﬁcations3 of the Modiﬁed Van Soest method (Stern, 1991) were
used to determine neutral detergent ﬁber and acid detergent ﬁber concentrations. The
Hach modiﬁed Kjeldahl procedure (Hach, 1985; Watkins, 1987) was used to determine
total N concentration in the forage. For a full description of this data, see Chapter 2 of
this dissertation. In addition to forage quality and forage dry matter presented, data was
collected to describe pasture botanical composition over the course of the experiment".
Economic Comparison

The objective of this experiment was to compare the proﬁtability of the three
grazing systems over three years of experimentation. The correlated hypotheses were:

I. When considering all inputs and outputs, the initial cost of establishing the warm
season grass-integrated grazing systems will be offset by their higher productivity

by the end of the third year.

 

3 See Technical Note 4 in Appendix III
4 See Appendix 11

I31

2. Big bluestem-integrated grazing systems will not be signiﬁcantly more proﬁtable
than switchgrass-integrated grazing systems by the third year of experimentation
Total animal weight gain per hectare was the primary parameter used to calculate net
proﬁtability of each treatment. Total animal weight gain per hectare per year is a
function of average daily gain per animal and the number of animal grazing-days
accumulated per hectare per year (i.e., stocking rate) (Parch, 1989):
Total weight gain per hectare = (animal grazing days per hectare) X (average daily gain)
Average daily gain (ADG), and accumulated AUGD per hectare and total weight gain per
hectare are discussed in more detail in Chapter 3 of this dissertation.
Several assumptions were used to make economic comparisons.
0 Because there is no treatment x year interaction, the calculated average for gain per
hectare for each treatment is a valid estimate for all years
0 Fixed expenses that are identical for all treatments are excluded from the net income
calculations (Black, 2007)
o Expenses related to treatment are included in the calculation (Black, 2007). These
expenses include:
0 Custom rates for ﬁeld operations including: no-till planting, pesticide
application, mowing, and fertilizer application
0 Pesticide costs
0 Fertilizer costs
Because Michigan has no recent estimates for custom rates for ﬁeld operations,
the rates used to estimate variable costs are based on data from Ohio in 2006 (Ward,

2006). Pesticide costs are based on data provided my Michigan State University

132

(Sprague, 2007) for most pesticides; Jorgensen Farm Elevator (Jorgensen, 2007) for 2,4-
D ester; and a BASF website (BASF, 2005) for imazapic. Niagara big bluestem and
Cave-in-Rock seed prices are from Ernst Conservation Seed (Ernst, 2007).

RESULTS AND DISCUSSION

Steer gain per hectare in CS-Only system was signiﬁcantly higher than the CS-
BBS and CS-SG treatments (Table 3.4). Also, the CS-BBS treatment yielded more Steer
gain per hectare than the CS-SG treatment. If the switchgrass portion of pasture had not
failed to establish by 2003, it is likely that there would not be a signiﬁcant difference in
gain per hectare between the CS-BBS and CS-SG treatments.

The difference in net income between the CS-Only and the CS-BBS treatments is
$370 per hectare over three years (Table 3.9); the higher cost of establishing the big
bluestem portion of the CS-BBS treatment accounts for $72 of this difference (Table 3.8).
On a year-to-year basis, the CS-BBS and CS-SG treatments are less expensive to
maintain than the CS-Only treatment because the big bluestem and switchgrass portions
of those treatments are not mowed and due to the single (rather than double) application
Of urea on the big bluestem and switchgrass portions of those treatments (Tables 3.5-3.8).
The failure of the switchgrass pasture to establish before 2003 accounts for much of the
difference between the CS-BBS and CS-SG treatments.

Because total weight gain per hectare is a ﬁinction of ADG (Table 3.2) and the
number of grazing days accumulated per hectare per year, total weight gain can be
improved by either increasing ADG or total number of animal grazing days, or both.
During the beginning of each grazing season, the number of AUGD accumulated by the

CS-BBS and CS-SG do not differ signiﬁcantly from each other, but the CS-Only

133

treatment accumulated signiﬁcantly more AUGD than either of the other treatments
(Table 3.3). This is because entire pasture area of the CS-Only treatment was available to
graze beginning in late-April or early-May, whereas the big bluestem and switchgrass
portions of their respective treatments were not generally ready to be grazed until mid-
June (Bamhart, 1994a; Bartholomew, 1995). By mid-June of 2004 and 2005, the big
bluestem and switchgrass portions of the CS-BBS and CS-SG treatments were ready to
graze, and the difference between the numbers of AUGD accumulated by treatment
diminish or disappear altogether until late-summer. By early-September, grazing on the
big bluestem and switchgrass portions of their respective treatments was terminated for
the season (Bamhart, 1994a; Bartholomew, 1995). Thus, the pasture area available for
grazing on those treatments was essentially reduced by 33% compared to the CS-only
treatment, until the end of the grazing season for the cool-season pasture. Differences in
ADG among systems were not nearly as pronounced as differences in accumulated
AUGD per hectare (Tables 3.2 and 3.3). The major loss in economic performance from
the CS-SG and CS-BBS systems results from the lower number of AUGD accumulated
in the spring and fall of the grazing seasons due to the management requirements of the
switchgrass and big bluestem components.

This experiment compared the three treatments with the assumption that excess
forage from the C3 portion would not be harvested for hay; rather, the stocking rate
varied according to current and anticipated pasture productivity. In practice, many
graziers have more acreage than their livestock can effectively graze in the spring, and

they often set aside some of their pasture to harvest one or two cuttings of bay for sale or

feeding at a later time. After the ﬁrst or second cutting has been harvested, the pastures

134

that had been set aside for hay production can begin to be grazed. Some producers,
however, would prefer to keep their livestock on pasture as much as possible and would
prefer to purchase bay of known quality rather than be forced to produce hay — especially
if they have limited time, equipment, or labor resources. Many have suggested that
warm-season grasses can help accomplish this (Balasko, 2003; Bamhart, 1994a;

Peterson, 2007). However, in all years Of this study, the big bluestem and switchgrass
were ready for grazing before the C3 pasture species declined in productivity. In this

situation, Anderson (2000) recommends that mixed or solid stands of switchgrass be

grazed when they are ready, regardless Of whether the cool-season pastures are still
productive, suggesting that it is better is to harvest or stockpile excess C3 forage and to

graze the big bluestem and switchgrass while their forage volume and quality are optimal.
This was practiced in this experiment, but it adds more complexity than most graziers are
accustomed to in traditional grazing systems.

Moore et al. (2004) suggest that in a Situation where a producer does not want to
harvest excess pasture forage for hay, they might consider devoting part of their pasture
acreage to switchgrass or big bluestem. In fact, a producer could reduce the risk of
running out of mid-summer pasture using this strategy in Southern Michigan, but the
management constraints of these grasses induce an opportunity cost (i.e., the annual loss
of grazing days in the spring and fall) that seems to outweigh the potential for reducing
risk. Ultimately, producer’s willingness to set aside one third of his pasture acreage for
hay production or warm-season grass pasture will be related to their:

0 capacity (equipment, storage, time) to harvest quality hay from a traditional

grazing system.

135

o willingness to increase the complexity of their system in order to keep their
livestock on pasture.

0 assessment of the severity of the summer slump on their pastures and the
capacity of warm-season grasses to complement the distribution of the C3
pasture productivity.

0 opinion of how well the forage quality matches the class Of livestock on their

pastures.

136

Table 3.1. Weighing events by yearl.

 

 

2003 2004 2005
28-Apr 1' 26-Apr ZO-Apr
29-Apr 27-Apr Zl-AprI
28-May3‘, 25-May: 19-May
24 June}: 22 June}: 16 JuneI
22-JulyI 20-July’; 14-July
l9-Aug l4-Aug l l-Augi
I 7-Sep 22-Sep 8-Sep
l 8-Sep 23-Sep 16-Sep
l 7-Sep

 

lWeights from italicized dates were used to establish beginning- and end- of season

weights.
ilndicates dates of application of ivermectin pour-on

137

Table 3.2 Average daily gain of steers by treatment and year at time intervals throughout
the 2003-2005 grazing seasons.

 

 

 

Time 1-2 Time 2-3 Time 3-4 Time 4-5

Year Treatment ADG SE ADG SE ADG SE ADG SE
kg day'l

2003 CS-BBS 1.72 a 0.09 1.38 a 0.09 - 0.79 a* 0.06 0.98 a“ 0.08
2003 CS-Only 1.81 a 0.07 1.50 a 0.08 0.69 a* 0.06 0.99 a* 0.07
2003 CS-SG 1.72 a 0.08 1.48 a 0.08 0.72 a 0.06 1.09 a* 0.09
2004 CS-BBS 1.72 a 0.05 1.32 a 0.06 1.1 1 a* 0.04 0.46 b* 0.05
2004 CS-Only 1.84 a 0.06 1.34 a 0.06 0.99 ab* 0.05 0.73 a* 0.06
2004 CS-SG 1.75 a 0.05 1.26 a 0.06 0.96 b* 0.04 0.57 b 0.05
2005 CS-BBS 1.81 a* 0.06 1.48 a* 0.06 0.79 a* 0.05 0.91 b“ 0.06
2005 CS-Only 1.52 b* 0.06 1.32 ab* 0.07 0.72 ab* 0.05 1.10 a* 0.06
2005 CS-SG 1.44 b* 0.06 1.29 b* 0.06 0.65 b* 0.04 0.83 b* 0.05

 

Means within year with different letters are signiﬁcantly different (p < 0.05).
*Indicates an interaction with alfalfa abundance during that time period

138

 

Table 3.3. Animal unit grazing days accumulated by treatment and year at intervals
throughout the 2003-2005 grazing seasons.

 

 

 

 

Year Treatment Time 1 Time 2 Time 3 Time 4 Time 5 Time 6 Season
AUGD/ha
2003 CS-BBS 65.1 b 42.2 b 54.0 a 44.1a 34.8 a 0.0 a 240.2 b
2003 CS-Only 80.9 a 58.8 a 56.9 a 47.3 a 34.6 a 2.5 a 281.0 a
2003 CS-SG 61.0 b 42.9 b 41.1 b 28.5 b 23.0 b 0.0 a 196.5 c
2004 CS-BBS 72.8 b 76.2 b 88.7 a 74.2 a 71.1 a 52.7 b 435.6 b
2004 CS—Only 95.5 a 97.9 a 88.9 a 78.6 a 77.5 a 87.8 a 526.1 a
2004 CS-SG 74.9 b 75.9 b 91.8 a 72.1 a 74.5 a 46.0 b 435.1 b
2005 CS-BBS 62.9 b 58.4 b 67.0 a 76.0 a 70.6 a 0.0 a 335.0 a
2005 CS-Only 81.1 a 74.4 a 64.8 a 61.2 b 58.8 b 0.0 a 340.3 a
2005 CS-SG 62.3 b 55.9 b 67.2 a 74.3 a 61.6 ab 0.0 a 321.3 a

 

Means within the same column and year with different letters are signiﬁcantly different

(p < 0.05). The standard error for CS-BBS = 3.7; CS-Only = 3.7; CS-SG = 3.2.

139

Table 3.4. Animal gain ha.l by treatment.
Treatment Gain (kg/ha)1' SE

 

 

CS-BBS 624.5 b 16.2
CS-Only 722.0 a 16.2
CS-SG 572.9 c 14.0

 

IMeans with different letters are signiﬁcantly different (p < 0.05).

140

Costs of Three Years of CS Pasture Maintenance

Table 3.5. The cost of three years Of cool-season grass and legume pasture maintenance.

 

Cost
Time Activity Area Rate: Per hectare Per 0.67 ha
Year 1 spring fertilizer 1 171 $34.1 1 $22.86
custom fertilizer application 1 ' $10.75 $7.20
summer fertilizer 1 171 $34.1 1 $22.86
custom fertilizer application 1 $10.75 $7.20
mowing 1 $25.95 $17.38
Year 1 Total $115.67 $77.50
Year 2 spring fertilizer 1 171 $34.11 $22.86
custom fertilizer application 1 $10.75 $7.20
summer fertilizer 1 171 $34.1 1 $22.86
custom fertilizer application 1 $10.75 $7.20
mowing 1 $25.95 $17.38
Year 2 Total $115.67 $77.50
Year 3 spring fertilizer 1 171 $34.11 $22.86
custom fertilizer application 1 $10.75 $7.20
summer fertilizer 1 171 $34.1 1 $22.86
custom fertilizer application 1 $10.75 $7.20
mowing 1 $25.95 $17.38
Year 3 Total $115.67 $77.50
3-Year Total $347.01

 

141

l

IFertilizer rates are in kilograms of urea per hectare, with a cost of $0.44 kg“ .

Table 3.6. The cost of establishment of big bluestem pasture and three years of

 

maintenance.
Costs of BBS Pasture Establishment and Two Years of Maintenance
Cost
Time Activity Area Rate: Per hectare Per 0.33 ha
Year Prior custom herbicide application 0.33 $13.22 $4.36
glyphosate 0.33 1.08 $14.60 $4.82
2,4-D ester 0.33 0.84 $8.52 $2.81
dicamba 0.33 0.54 $26.47 $8.74
AMS+NIS 0.33 $8.43 $2.78
custom no—till planting 0.33 $30.39 $10.03
seed 0.33 10.1 $161.60 $53.33
Year 1 custom herbicide application 0.33 $13.22 $4.36
glyphosate 0.33 1 .08 $14.60 $4.82
imazapic 0.33 0.07 $34.19 $11.28
AMS+NIS 0.33 $8.43 $2.78
Year-1 Total $333.68 $110.11
Year 2 custom herbicide application 0.33 $13.22 $4.36
glyphosate 0.33 1 .08 $14.60 $4.82
2, 4-D ester 0.33 0.84 $8.52 $2.81
dicamba 0.33 0.54 $26.47 $8.74
AMS+NIS 0.33 $8.43 $2.78
spring fertilizer 0.33 171 $34.11 $1 1.26
custom fertilizer application 0.33 $10.75 $3.55
Year-2 Total $116.10 $38.31
Year 3 spring fertilizer 0.33 171 $34.11 $11.26
custom fertilizer application 0.33 $10.75 $3.55
Year-3 Total $44.86 $14.80
3-Year Total $494.64 $163.23

IHerbicide rates are in kilograms active ingredient (a.i.) per hectare; seed rates are in
units of kilograms per hectare; fertilizer rates are in kilograms of urea per hectare.

I42

Table 3.7. The cost of establishment of switchgrass pasture and three years of
maintenance.

 

Costs of SG Pasture Establishment and Three Years of Maintenance

 

Cost

 

Time Activity Area Rate: Per hectare Per 0.33 ha
Year Prior custom herbicide application 0.33 $13.22 $4.36
glyphosate 0.33 1.08 $14.60 $4.82
2,4-D ester 0.33 0.84 $8.52 $2.81
dicamba 0.33 0.54 $26.47 $8.74
AMS+NIS 0.33 $8.43 $2.78
Year 1 custom herbicide application 0.33 $13.22 $4.36
glyphosate 0.33 1.08 $14.60 $4.82
imazethapyr 0.33 $38.63 $12.75
AMS+NIS 0.33 $8.43 $2.78
custom no-till planting 0.33 $30.39 $10.03
seed 0.33 10.1 $90.00 $29.70
Year 1 custom insecticide application 0.33 $13.22 $4.36
carbaryl 0.33 1 .68 $26.69 $8.81
carbaryl 0.33 1 .68 $26.69 $8.81
Year-1 Total $333.12 $109.93
Year 2 custom herbicide application 0.33 $13.22 $4.36
glyphosate 0.33 1.08 $14.60 $4.82
2, 4-D ester 0.33 0.84 $8.52 $2.81
dicamba 0.33 0.54 $26.47 $8.74
AMS+NIS 0.33 $8.43 $2.78
spring fertilizer 0.33 171 $34.11 $11.26
custom fertilizer application 0.33 $10.75 $3.55
Year-2 Total $116.10 $38.31
Year 3 spring fertilizer 0.33 171 $34.11 $11.26
custom fertilizer application 0.33 $10.75 $3.55
Year-3 Total $44.86 $14.80
3-Year Total $494.08 $163.05

 

. . . . . . . -I . . -1
IHerbIcrde rates are m kg active Ingredient (a.r.) ha ; seed rates are m units of kg ha '

. . . . -l
fertilizer rates are In kilograms of urea ha .

143

Table 3.8. Summary of maintenance and establishment costs of treatments.
CS-BBS Expenses
Year 1* Year 2 Year 3
CS portion $77.50 $77.50 $77.50
BBS portion $110.11 $38.31 $14.80
CS-BBS Year Total $187.61 $115.81 $92.30

 

 

3-Year Total $395.73

CS-Only Expenses
Year 1 Year 2 Year 3
CS-Only $115.67 $115.67 $115.67

 

3-Year Total $347.01

CS-SG Exgnses
Year 1* Year 2 Year 3
CS portion $77.50 $77.50 $77.50
SG portion $109.93 $38.31 $14.80
CS-SG Year Total $187.43 $115.81 $92.30

 

3-Year Total $395.54
*Also includes establishment costs from year prior to initiation of grazing.

 

144

Table 3.9. Summary of gross income, expenses, and net income of treatments by year.
Year Treatment gain ﬁg/ha) Value per kg Gross Expenses Net

 

 

1 08-888 624.5 $1.10 $687 $188 $499
1 CS—Only 722.0 $1.10 $794 $116 $678
1 CS-SG 572.9 $1.10 $630 $187 $443
2 CS-BBS 624.5 $1.10 $687 $116 $571
2 CS-Only 722.0 $1.10 $794 $116 $678
2 CS-SG 572.9 $1.10 $630 $1 16 $514
3 CS-BBS 624.5 $1.10 $687 $92 $595
3 CS-Only 722.0 $1.10 $794 $116 $678
3 CS-SG 572.9 $1.10 $630 $92 $538
3-Year Total CS-BBS 1873.4 $1.10 $2,061 $396 $1,665
3-Year Total CS-Only 2165.9 $1.10 $2,382 $347 $2,035
3-Year Total CS—SG 1718.8 $1.10 $1,891 $396 $1,495

 

145

LITERATURE CITED

Allen, V.G. 1991. Terminology for grazing lands and grazing animals Pocahontas Press,
Blacksburg, VA.

Anderson, B. 2000. Grazing management on warm season grasses [Online]
http://agebbmissouri.edu/mfgc/ZOOOmtgjwarmgrass.htm (veriﬁed 03/22/2006).

Balasko, J.A., and C. J. Nelson. 2003. Grasses for northem areas, p. 125-147, In R. F.
Barnes, Nelson, J. C., Moore, K. J., and M. Collins, ed. Forages: an introduction
to grassland agriculture, Vol. I.

Barker, J .M., Buskirk, D. D., Ritchie, H. D., Rust, S. R., Leep, H. R., and K. J. Barclay.
1999. Intensive grazing management of smooth bromegrass with or without
alfalfa or birdsfoot trefoil: heifer performance and sward characteristics. The
Professional Animal Scientist 15:130-135.

Bamhart, S.K. l994a. Warm-season grasses for hay and pasture - Pm-569. Iowa St. U.
Ext.

Bamhart, S.K. l994b. Warm-season grasses for hay and pasture - Pm-569, pp. 4, In 1. S.
U. Exension, (ed.). Iowa St. U. Ext.

Bamhart, SK. 1999. Selecting forage species - Pm 1792 [Online]. Available by Iowa St.
U. Ext. http://www.extcnsion.iastate.edu/Publications/PM 1 792.@‘ (veriﬁed
03/22/2006).

Bartholomew, H.M., Sulc, R.M., Hendershot, R., and J. Cline. 1995. Perennial warm
season grasses for Ohio [Online]. Available by Ohio St. U. Ext. Dept. of Hort. and
Crop Sci. http://ohioline.osu.edu/agf—fact/OOZZ.html (veriﬁed 03/22/2006).

BASF. 2005. Costs vs. beneﬁts: examining the numbers [Online]
http://www.vmanswers.com/printerfriendly.aspx?pid=898 (veriﬁed April 18).

Black, J .R. 2007. Professor of Agricultural Economics, Michigan State University, In D.
J. Hudson, (ed.), Mason, Ml.

Blanco-Canqui, H., Gantzer, C. J ., Anderson, S. H., Alberts, E. E., and A. L. Thompson.
2004. Grass barrier and vegetative ﬁlter strip effectiveness in reducing runoff,

sediment, nitrogen, and phosphorus loss. Soil Science Society of America:l670-
1678.

146

Bransby, DJ. 1989. Compromises in the design and conduct of grazing experiments, p.
136, In G. C. Marten, ed. Grazing research: design, methodology, and analysis.
CSSA Spec. Pub. No. 16. Crop Science Society of America and the American
Society of Agronomy, Madison.

Buskirk, D. 2003. Associate Professor of Animal Sciences, Michigan State University.

Coleman, S.W., Sollenberger, LE. 2007. Plant-Herbivore Interactions, p. 123-136, In R.
F. Barnes, Nelson, J. C., Moore, K. J., and M. Collins, ed. Forages: the science of
grassland agriculture, Vol. 2, 6th ed.

Ernst. 2007. Prices [Online]. Available by Ernst Conservation Seeds
http://wwwernstseedcom. (veriﬁed April 18).

 

Gabriels, P.C.J., and J .V. Van Den Berg. 1993. Calibration of two techniques for
estimating herbage mass. Grass and Forage Science 48:329-335.

George, J .R., Hintz, R. L, Moore, K. J., Bamhart, S. K., and D. R., Buxton. 1996. Steer
response to rotational or continuous grazing on switchgrass and big bluestem
pastures, p. 150-154, In M. J. Williams, ed. 1996 American Forage and Grassland
Council Proceedings, Vancouver.

George, RR, A.L. Farris, C.C. Schwartz, D.D. Humburg, and J .C. Coffey. 1979. Native
prairie grass pastures as nest cover for upland birds. Wildlife Society Bulletin 7:4-
9.

Harmoney, K.R., K.J. Moore, J.R. George, BC. Brummer, and LR. Russell. 1997.
Determination of pasture biomass using four indirect methods. Agronomy Journal
89:665-672.

Hazell, DB. 1967. Effect of Grazing Intensity on Plant Composition, Vigor, and
Production. Journal of Range Management 20:249-252.

Jorgensen. 2007. Price List, In D. J. Hudson, (ed.). Jorgensen Farm Elevator, Mason, Ml.

Keeney, D.R., Sanderson, M. A. 2007. Forages and the environment, p. 167-176, In R. F.
Barnes, Nelson, J. C., Moore, K. J., and M. Collins, ed. Forages: the science of
grassland agriculture, Vol. 2.

Koch, R.M., E.W. Schleicher, and V.H. Arthaud. 1958. The Accuracy of Weights and
Gains of Beef Cattle. Journal of Animal Science 17:604-61 1.

Laude, HM. 1953. The nature of summer dormancy in perennial grasses. Botanical
Gazettez284-292.

147

Lindquist, J. 2002. 2002 Michigan Hay and Grazing Council Industry Priorities [Online]

http://www.greeen.msu.edu/ind.priorities/2002haygrazingPriorities.pdf (veriﬁed
3/25/06).

 

Ma, 2., Wood, W., and D. I. Bransby. 2000. Soil management impacts on soil carbon
sequestration by switchgrass. Biomass and Bioenergy 181469-477.

Michell, P. 1982. Value of a rising-plate meter for estimating herbage mass of grazed
perennial ryegrass-white clover swards. Grass and Forage Science 37:81-87.

Mott, GO. 1960. Grazing pressure and the measurement of pasture production. Proc VIII
Int. Grassl. Cong.:606-611.

Parch, L.D., and Torell, L. A. 1989. Economic considerations in grazing research, p. 109-
125, In G. C. Marten, ed. CSSA Special Publication Number 16. Grazing
Research: Design, Methodology, and Analysis. CSSA-ASA, Madison.

Peterson, P.R., Rayburn, E. B., Cropper, J. B., and Belesky, D. P. 2007. Perennial warm-
season grasses, In E. B. Rayburn, ed. Forage utilization for pasture-based
livestock production. Natural Resource, Agriculture, and Engineering Service,
Ithaca, NY.

Sargent, M.S., Carter, K. S., eds. 1999. Grassland management, p. 297 Managing
Michigan wildlife: 3 landowners guide. Michigan United Conservation Clubs,
East Lansing.

SAS Institute, Inc. 2000-2003. SAS. Release 8 and 9. SAS Institute, Inc., Cary, NC.

Sprague, C. 2007. Professor of Crop and Soil Science, Michigan State University.

Tober, D.A., and A. D. Chamrad. 1992. Warm-season grasses in the northern great
plains. Rangelands 14:227-230.

Undersander, D. 2002. Conversation about using warm-season grasses as a component of
grazing systems.

USDA-NRCS. 1996. Establishing warm-season grasses, pp. 3, In USDA-NRCS, (ed.)
Hay and pasture management. USDA-NRCS, Columbia, Missouri.

Ward, B. 2006. Ohio Farm Custom Rates - 2006.

148

Chapter 4
RELATIVE ABUNDANCE OF GRASSLAND BIRDS IN COOL-SEASON
PASTURES AND GRAZING SYSTEMS INTEGRATED WITH SWITCHGRASS
AND BIG BLUESTEM
ABSTRACT

This experiment compares the relative abundance of grassland birds observed within each
of three treatments of a grazing trial that was conducted in Southwest-Lower Michigan.
The three treatments included: cool-season grass and legume pastures (CS-Only);
integrated big bluestem and cool-season grass andlegume pastures (CS-BBS); and
integrated switchgrass and cool-season grass and legume pastures (CS-SG). The CS-
BBS pastures had an average of 2.6 birds per transect, while the CS-Only and CS-SG
treatments had 1.9 and 1.7, respectively, and the differences were not statistically
signiﬁcant. While the combined area of the big bluestem and switchgrass monocultures
comprised about 23% of the experimental area, only 8% of the birds counted were within
those pastures. This may be due to: insufﬁcient plot size; intensive forage residue
management; slow spring growth of switchgrass and big bluestem; the lack of fencing
(serving as perches) within switchgrass and big bluestem monocultures and the presence
of subdividing fences in the cool-season portions of the treatments; and/or the lack of
plant structure heterogeneity in the switchgrass or big bluestem monocultures. While
periodic mowing is realistic part of pasture management, grazing and mowing activity
can be very disruptive to nesting birds, and probably confounded the objective. Studies
designed to describe the effect of big bluestem or switchgrass on grassland birds
populations in integrated grazing systems should begin earlier in the spring, on a much

larger experimental area, and include a treatment where the pastures are not mowed.

149

INTRODUCTION

The post-settlement decline of grassland bird populations

The populations of many Midwestern grassland bird species have declined
significantly since pre-settlement times. Many have attempted to describe the nature and
causes of these declines (Sauer et al., 1996) and possible solutions to the problem. The
North American Breeding Bird Survey has been used extensively to describe population
trends and to deduce possible causes for population changes. Although useful for some
types of population analysis, correlation data from the Breeding Bird Survey is not useful
for identifying precise causes of population changes (Sauer et al., 1996). Rather than
. yielding simple solutions, the grassland wildlife research of the past thirty years has
produced management principles which have the potential to mitigate or improve the
avian productivity of grassland ecosystems and landscapes that contain cover with
structure similar to native grasslands. There have also been many attempts to design land
management protocols that reﬂect the principles derived from such research, especially
by the USDA-NRCS and state departments of natural resources. Examples of such
programs include the Conservation Reserve Program (CRP), Grassland Reserve Program
(GRP), and the Wetland Reserve Program (WRP). The likelihood of adoption and
survival of government programs such as the Grassland Reserve Program by producers
depends on how producers perceive the compatibility of their objectives with those of
required practices as well as the agency promoting the program.

One grassland management practice that has been promoted by government
agencies is the use of native warm season grasses such as switchgrass (Panicum

virgatum) and big bluestem (Andropogon gerardii) as a dual purpose crop (George et al.,

150

1979). Implicit in these recommendations is that these grasses provide adequate forage
quantity and quality for certain classes of livestock and, if managed appropriately,
adequate habitat for grassland bird species.

In Michigan, approximately 360,000 hectares are grazed annually (Matthews,
2003) and another 410,000 hectares are harvested for hay (MASS, 2002). Keeping
livestock on pasture offers several advantages over conﬁnement feeding including lower
time, labor and capital inputs; superior manure distribution; improved animal health;
lower energy use per pound of animal produced; and reduced land degradation.

During the mid-summer months of most years, Michigan producers who keep
livestock on pasture are faced with a choice: to remove their livestock from dormant
summer pastures to locations where they are fed stored forage, or to allow their livestock
to continue grazing. Overgrazing often reduces pasture productivity (Coleman, 2007)
resulting in weed invasion (Hazell, 1967), and therefore may require renovation or
replanting. In Michigan, spring and fall cool season grass-legume growth is excellent
due to mild temperatures and adequate rainfall, but the hot, dry summer months cause
most of these grasses and legumes to become dormant. Several state universities in the
North Central Region promote the use of native perennial warm season grasses, such as
switchgrass and big bluestem as summer forage. These grasses have the potential to
provide large volumes of high quality mid-summer forage when the cool season grasses
and legumes are dormant. While both of these species are indigenous to Michigan, no
grazing research has been conducted to determine if using these species in a pasture

system is a viable option. Beyond their capacity to produce large quantities of biomass,

151

switchgrass and big bluestem grow well in marginal soils and are excellent nesting and
holding habitat for wild game and song birds.

The Michigan Hay and Grazing Council has described alternative(Lindquist,
2002) forage research as one of its top priorities (Lindquist, 2002). Both the Michigan
Department of Natural Resources (Sargent, 1999) and the USDA Natural Resource
Conservation Service (USDA-NRCS, 1996) have encouraged Michigan graziers to
include native warm season grasses in their grazing systems. However, graziers may be
reluctant to use these grasses without knowledge of their performance in Michigan. This
project was designed to demonstrate how livestock perform when these grasses are
integrated into the typical grazing systems in Southern Michigan. This project was
designed to demonstrate how livestock perform when these grasses are integrated into the
proposed grazing system models in Southern Michigan.
Warm season grasses in Michigan forage systems

In Southern Michigan, cool season grass and legume pastures often are grazed
beginning in late-April; hay harvest often begins in mid-to late-May; this timing is
particularly detrimental to nesting grassland birds (F rawley and Best, 1991). However,
native warm season grasses have a very different seasonal growth dynamics, reaching
optimal quality and yield in early- to mid-June in Southern Michigan. Further, because
under optimal summer forage management the grazing of native warm-season grasses
should stop when the canopy height is reduced to 20 cm (Peterson, 2007), more cover is
left for wildlife aﬁer each grazing cycle. While these facts seem to suggest that the use of

native warm season grasses as a major component of grazing systems could solve

152

multiple problems simultaneously, until now no speciﬁc research has been done on this
topic in Michigan.
Introduction to switchgrass and big bluestem

Switchgrass and big bluestem are both erect-growing native perennial warm
season grasses. They have C4 metabolism, so peak biomass production occurs at a leaf
temperature of 37°C, in contrast to cool season grasses which peak between 20 and 25°C
and dramatically decline at temperatures above 27°C (Nelson, 1995). This physiological
characteristic makes warm season grasses most productive in mid-summer. Big
bluestem, switchgrass, and indiangrass have been used extensively in US. government
programs such as the Soil Bank and the Conservation Reserve Program since the 1930’s,
resulting in several million hectares being replanted with either mixtures or monocultures
of these grasses (Moser, 1995). The extensive use of these grasses in the Conservation
Reserve Program demonstrates their contribution to grassland ecosystems and their role
in providing wildlife habitat (USDA-NRCS, 1999; USDA-NRCS, 2007). The ecological
importance of these grasses is widely recognized (Harvey, 2000; Henning, 1993; USDA-
NRCS, 1999) and they can be planted along railroad right-of-ways and roadsides, near
waterways, for wildlife cover (Moser, 1995), for erosion control (Blanco-Canqui, 2004;
Rankins, 2001), and can function well as vegetative conservation buffers to prevent soil
pesticide loss via surface water (Rankins, 2001). Both big bluestem and switchgrass
tolerate a soil pH of 4.5 and 4.9, respectively (Duke, 1978) and thus can be planted in
areas that are not typically used for crop production. Switchgrass and big bluestem both
have extensive root systems that penetrate the soil to depths of more than two meters

(Weaver, 1954) and show great potential for carbon sequestration. Beyond having vast

153

potential for wildlife habitat, erosion control, and carbon sequestration, these grasses are
recommended by many research and extension institutions for summer grazing. Mitchell
(1996) suggests that summer grazing is more efﬁcient if separate pastures of warm and
cool season grasses are maintained; placing grazing animals in the warm season pastures
during mid-summer, and then returning them to cool season pastures after cool season
pasture recovery. Henning (1993) points out that a combination of cool season and warm
season grass pastures can provide a more constant supply of high-quality feed through the
summer than either cool or warm season grass pastures can provide alone. Bamhart
(1994) indicates that pasture efﬁciency may be increased by converting one-fourth to
one-third of the cool season grass pasture acreage to a warm season grass pasture to be
grazed at different times during the grazing season. This allows the cool season grasses
to be grazed in the spring and early summer, occasionally throughout the summer, and in
the fall, while utilizing the warm season grass pasture more intensively during periods of
low cool season grass productivity. Also, this strategy allows greater rest periods for cool
season pastures, which will increase their vigor and productivity for late summer through
early fall grazing.

Recommendations for inclusion of warm season grasses in grazing systems are
supported by numerous studies in many states. Krueger and Curtis (1979) conducted a
study in South Dakota comparing switchgrass, big bluestem, indiangrass (Sorghastrum
nutans (L.) Nash), and sideoats grama (Bouteloua curtipendula) for mid-summer grazing
and concluded that switchgrass and big bluestem are useful species for beef production in
July and August. (George, 1996) demonstrated that grazing either big bluestem or

switchgrass pastures in mid-summer can result in impressive steer weight gains (1.42 and

154

1.1 1 kg day.l , respectively) and that rotational grazing management is superior to

continuous grazing management for both switchgrass and big bluestem. In an Iowa
study, (Moore, 2001) concluded that grazing systems in which cool season grasses and
warm season grasses are grazed sequentially can improve seasonal productivity.

Switchgrass matures earlier in the summer than big bluestem and the forage
quality of big bluestem does not drop off as rapidly as it does with switchgrass (Moser,
1995). Both switchgrass and big bluestem pastures must be managed differently than
cool season grass pastures. Neither species tolerates close grazing and both require rest
periods of 21 to 45 days between grazing events, depending on environmental conditions
(Anderson, 2000; Henning, 1993; Mitchell, 1996).

Objective and Hypothesis:

Objective three, as stated in the introductory chapter of this dissertation, is: to
describe songbird species diversity in the respective pasture systems, describing apparent
relationships between species spatial distribution and type of pasture system. The
associated hypothesis is that pasture systems that include switchgrass and big bluestem
will have a higher songbird species diversity, nesting success, and evidence greater use
by raptors feeding on mice and other rodents.

LITERATURE REVIEW

The most useful habitats for grassland wildlife include old ﬁelds, lightly or
moderately, grazed pastures, fallow ﬁelds, meadows, and grass and grass/legume hay that
is harvested late (Sample and Mossman, 1997). These vegetation cover types are
becoming rare as such land is either converted to more conventionally m'anaged forage

production, row crops, residential development, or subdivided recreational land. While

155

habitat fragmentation is linked to avian population changes, research in this area can be
complicated due to confounding factors such as land use change (Donovan and F lather,
2002). The complexity of the decline of many native Midwestern avian species seems to
be surpassed only by the difﬁculty of ﬁnding realistic and practical ways to slow or
reverse the trend. Some of the most difﬁcult aspects of ecological restoration are deciding
which species are of primary concern, which land should be targeted for change, who
should manage that restoration, and how the landowner should be compensated.
The relationship of cropping systems to bird populations

State level Breeding Bird Survey data from between 1980 and 1998 indicates that
grassland bird species are declining and that this is due in part to changes in agricultural
land use, such as a decline in range land and a decline in the use of cover crops (Murphy,
2003). While cropping systems vary by region and by individual farm, research on the
inﬂuence of these cropping systems on avian populations has demonstrated major
weaknesses of these systems for sustaining avian population and diversity. In a study
conducted in Iowa, (Patterson and Best, 1996) identiﬁed the nests of sixteen bird species
in CRP ﬁelds and two in row crop ﬁelds — the vesper sparrow (Pooecetes gamineus) and
the horned lark (Eremophila alpestris). In both habitat types, predation was the primary
cause of nest loss. Bird use of strip intercropping systems was also evaluated in Iowa
(Stallman and Best, 1996). Stallman and Best concluded that strip cropping increases
species diversity and abundance but nest failure is very high and the system ends up
being an ecological trap. Farmers using strip cropping systems frequently use mechanical

methods of weed control, resulting in the destruction of many nests; using herbicides

156

rather than mechanical weed control would likely result in higher avian productivity
(Stallman and Best, 1996).

Hedge rows between crop ﬁelds are commonly thought of as the primary source
of wildlife habitat in crop land. Some ecologists have suggested that hedge rows
frequently ﬁinction as ecological traps. Warner (1994) suggests that such conclusions are
simplistic because at least two scenarios can improve the usefulness of hedge row cover
by birds: heterogeneity of cover type and connection of the linear habitat to the
surrounding landscape. Nest densities and species survival were highest on plots where
there were heterogeneous cover types. The connection of the linear habitat to the
surrounding landscape improved the nest density and species diversity of the linear
habitat. Research conducted in the 19403 suggests that whether they function as
ecological traps or not, some species are certainly attracted to hedge rows, while others
are not (Good and Dambach, 1943).

Government sponsored conservation programs

The USDA Conservation Reserve Program and Canada’s Permanent Cover
Program (PCP) represent two major federal efforts to conserve easily degradable
agricultural land with corollary beneﬁts of improved habitat for grassland bird species.
In the CRP, producers are paid to take crop land out of production and plant and manage
species of plants known to provide cover for wildlife. After studying land enrolled in the
PCP, (McMaster and Davis, 2001) concluded that PCP acreage is characterized by more
dense and heterogeneous vegetative structure than cropland and that nine out of ten
studied bird species were present at higher frequencies than in cropland. The importance

of habitat quality is underscored by (Fahrig, 2001) who reports that up to 58% less habitat

157

area is needed for species persistence if a low quality habitat matrix is converted to one of
high quality.

While the CRP has resulted in hundreds of thousands of acres being set aside for
conservation purposes its impact has been isolated to certain regions or small tracts
within regions. One limitation that applies to all conservation programs is that of tract
size. A unit of cover that has excellent plant and residue structure for a bird species
might still be refused by that species due to the insufﬁcient area of the otherwise
acceptable habitat (Vickery et al., 1994). For example, grasshopper sparrows
(A mmodramus savannarum) and vesper sparrow preferentially utilize smaller strips of
nesting territory whereas bobolinks (Do/ichonyx oryzivorus), red-winged blackbirds
(A gelaius phoeniceus) , and meadowlarks (Sturnella spp.) prefer larger tracts over narrow
strips. Pheasants (Phasianus spp.), quail (Coturnicops spp.), and dicksessels (Spiza
Americana) do not demonstrate strong preferences for larger or smaller tracts of nesting
territory (Good and Dambach, 1943).

The CRP offers breeding habitat for species that have had decreasing population
trends and could actually reverse those trends (Johnson and Schwartz, 1993; Veech,
2006). Although the CRP has probably not yet reversed the overall decline of most
species of the indigenous grassland bird population of the United States, it is clearly
superior bird habitat compared to row crop ﬁelds. A 1997 report suggests that CRP has
between 1.5 and 10 times greater bird species abundance, about three times more species
of nesting birds, and 13.5 times more nests than row crop ﬁelds (Best et al., 1997).

While annual crops such as corn and soybeans clearly do not provide adequate

cover for most birds, the shortcomings go beyond excessive or insufﬁcient biomass.

158

Total bird abundance is associated with vertical density, litter cover, litter depth, ratio of
grass to forb cover, and bare ground cover, although the relationships can be positive or
negative, depending on bird species. For example, bobolink abundance is negatively
correlated with vertical density and positively correlated with percent litter cover, but
some species prefer structural complexity (Delisle and Savidge, 1997; Sample and
Mossman, 1997). The requirement for structural complexity and varying forms of
structural complexity is addressed at some level in the CRP by the two different types of
seed mixtures that landowners can plant. The CPl option requires the landowner to plant
cool-season grasses and legumes, while the CP2 optiOn requires them to plant native
warm season grasses — each which have different vegetative characteristics. Research
conducted in southeast Nebraska suggests that whereas total bird abundance may not
differ between CPl (cool-season grasses and legumes) and CP2 (native warm season
grasses), particular species demonstrate preferences for vegetative characteristics of one
over the other (Delisle and Savidge, 1997). I

Although the Permanent Cover Program of Canada differs from the CRP of the
United States, it also provides more dense and heterogenous vegetative structure than
cropland (McMaster and Davis, 2001 ). Whether part of a government conservation
program or a private effort to improve bird habitat to promote avian diversity, vegetative
structural diversity should be promoted within and among particular blocks of habitat, the
landscape, natural divisions, and even within the states (Sample and Mossman, 1997).
Ecological principles with implications for grassland management

The differences in avian productivity between land devoted to row crops and

grasslands might lead some to deduce that removing row-crop acreage from production

159

will necessarily improve habitat for grassland birds. However, land at certain stages of
succession can be very poor habitat for nesting grassland birds due to the uncontrolled
growth of woody species which provide structure for predators and parasitic species such
as hawks and brown-headed cowbirds (Molothrus ater) (Johnson and Temple, 1990).

A study of ﬁve grassland bird species nesting in tallgrass prairie fragments in
Minnesota revealed that nest predation and parasitism was greatest within 45 meters of a
wooded edge, when woody vegetation had encroached, and when it had been three or
more years since the vegetation had been burned. Where grassland bird management is a
priority, tracts of habitat should be large, far from wooded edges, and burned on a regular
basis (Johnson and Temple, 1990). Data from a study conducted in Michigan shows that _
CRP ﬁelds that were one or two years old and characterized by a combination of forbs
and bare ground had the greatest diversity and relative abundance of avian species; older
CRP ﬁelds that were characterized by deeper litter cover and grasses had the highest level
of avian productivity (Millenbah et al., 1996). Management protocols for landowners
who have enrolled in the CRP reflect the need to manage their acreage in a manner that
will optimize the heterogeneity of the landscape in terms of forb, grass, and bare ground
cover, and litter cover and depth. Such protocols include land management obligations
such as mowing or burning the accumulated biomass every several years, which reduces
or eliminate the encroachment of woody species. Robel et al. (1998) suggest that burning
CRP land every 2-3 years is sufﬁcient to prevent encroachment by woody plant species
and that, although spring burning of CRP ﬁelds reduces bird nest numbers and total avian
abundance in the season after burning, avian species richness and nesting success are not

different than in unburned CRP ﬁelds.

160

Delisle and Savidge (1997) point out that late-season burning or mowing can
decrease the value of CRP ﬁelds as winter cover. Askins (2001) uses a different
approach to arrive at similar conclusions to the studies above. He suggests that before
settlement, many grassland and shrubland species in eastern North America depended on
natural disturbances such as wildﬁres to create appropriate habitat. Early post-settlement
forms of agriculture simulated many of those disturbances, and these species have
declined with the decline of those forms of agriculture (Askins, 2001). Because pre-
settlement habitats experienced varying degrees of disturbances by phenomena such as
wildﬁres, land enrolled in programs such as the CRP should experience periodical
disturbances by ﬁre or mowing in order to meet the goal of habitat restoration (Askins,
2001)

Habitat area requirements vary by bird species. Habitat that is appropriate for a
given species due to the plant and residue structure may be refused by certain species of
birds due to the overall area of the otherwise acceptable habitat (Good and Dambach,
.1943; Herkert, 1994). Herkert observed that distributions of 8 of 15 bird species studied
in habitat fragments were affected by habitat area. Six of the species were affected by
habitat structure only, and the dicksissel was affected by neither structure nor fragment
size. Herkert ﬁirther observed that among ﬁve of the area-sensitive species, the minimum
area requirement ranged from 5 to 55 hectares. Good and Dambach (1943) also noted
that grasshopper sparrows and vesper sparrow preferentially utilize smaller strips of
nesting territory whereas bobolinks, red-winged blackbirds, and meadowlarks prefer
larger tracts over narrow strips. Pheasants, quail, and dicksessels do not demonstrate

strong preferences for larger or smaller tracts of nesting territory (Good and Dambach,

161

1943). Altemately, grassland birds may settle in suitable microenvironments in
otherwise less acceptable habitats (Sample and Mossman, 1997). Research conducted in
grassland barren sites in Maine has shown that the optimal tract size for maximum
grassland bird diversity is 200 ha. Upland sandpipers '(Bartramia longicauda) had the
greatest requirement for land area, requiring 200 ha to reach 50% incidence; the savannah
sparrow (Passerculus sandwichensis) reached 50% incidence at 10 ha. Field sparrows
(Spizella pusilla) incidence was not strongly associated with tract size, while edge species
such as the song sparrow was negatively correlated with grassland tract size (Vickery et
aL,1994)

The major causes of nesting failure in CRP-type habitat is predation (Best et al.,
1997); in roadsides adjacent to row crop ﬁelds, mowing, plant lodging, weather, and
cowbird parasitism can also contribute to nest failure (Camp and Best, 1994). Because
not all land is enrolled in the CRP, the inﬂuence of “edges” on bird populations near the
borders of a particular tract of CRP land is inevitable. Some have contended that edges
of such habitat can function as ecological traps. Ratti and Reese (1988) used artiﬁcial
bird nest containing Japanese quail (Coturnixjaponica) eggs to test the ecological trap
hypothesis as described by Gates and Gysel (Gates, 1978; Ratti and Reese, 1988). Their
research afﬁrmed the conclusions of Gates and Gysel and suggested that critical factor
determining whether an edge functions as a sink is the abruptness of the edge, contending
that a feathered edge provides superior nesting habitat due to its vegetative complexity
and its reduction of predator efﬁciency (Ratti and Reese, 1988). A study of ﬁve
grassland bird species nesting in tallgrass prairie fragments in Minnesota revealed that

nest predation and parasitism was greatest within 45 meters of a wooded edge, when

162

woody vegetation had encroached, and when it had been three or more years since the
vegetation had been burned. The authors concluded that where grassland bird
management is a priority, tracts of habitat should be large, far from wooded edges, and
burned on a regular basis (Johnson and Temple, 1990).
Ecological principles with implications for pasture and hay land management
Hay ﬁelds and pastures are well known havens for certain species of grassland

birds, although they differ in harvest timing and vegetative structure. Studies conducted
on pastures and hay land enrolled in the Permanent Cover Program conﬁrmed that ﬁelds
that are hayed or grazed have signiﬁcantly different vegetative structure and bird
community composition but similar species richness or evenness (McMaster and Davis,
2001). F urther, 70% of common bird species were detected at higher frequencies in hay
and pasture land than in crOp land (McMaster and Davis, 2001). Some species rely on
hay ﬁelds for their survival. For example, bobolinks thrive in large, old, late-mown hay
ﬁelds, and their population currently depends on hay ﬁelds since the original Midwest
prairies have mostly disappeared. In a study conducted in west-central New York,
(Bollinger et al., 1990) estimated that 74% of the bobolinks in the area studied nested in
hay ﬁelds.

Whereas predation is the greatest cause of nest failure in CRP land (Best et al.,
1997), mowing is the greatest cause of mortality of grassland birds such as bobolinks and
pheasants in hay land (Bollinger et al., 1990; Dale et al., 1997; Warner and Etter, 1989)
and it should be discouraged in areas where nesting success is critical (Camp and Best,
1994; Frawley and Best, 1991). Grazing intensity and patterns and residue control affect

the suitability of pasture habitat for particular bird species. Birds that like short

163

vegetation will be more prevalent in intensely grazed pastures (Sample and Mossman,

1997). Heavily grazed pastures probably do not offer habitat suitable to grassland bird
species that require intermediate or tall vegetation (McMaster and Davis, 2001).

The type of grazing that different species of grassland. birds will tolerate/prefer varies:

0 Light grazing: bobolink, eastern meadowlark (Sturnella magna). dickcissel, and

northern harrier (Circus cyaneus)

0 Moderate grazing: upland sandpiper, savanna and grasshopper sparrows, western

meadowlark (Sturnella neg/ecta), brewers blackbird (Euphagus cyanocephalus)

o Heavily grazed — horned lark (Eremophilia alpestris) and killdeer (Charadrius

vociferus) (Sample and Mossman, 1997)

It is clear that land managed for hay or pasture offers signiﬁcantly different structure
and levels of disturbance than idle land. Frawley and Best (1991) concluded that mowing
ﬁelds for hay and songbird reproduction are not compatible practices, while
acknowledging that management does make a difference, noting that early mowing is far
worse for songbird populations than late mowing (F rawley and Best, 1991). Pulliam
argues that reproductive surpluses from source habitats overﬂow into sink habitats, which
are areas that give the appearance of good habitat, but which have a fundamental
characteristic that prevents the majority of the population from carrying out successful
reproductive cycles (Pulliam, 1988). In the case of pastures and hay production ﬁelds
where mowing, trampling, predation, and parasitism are common, it is obvious that while
many nesting birds are present, the habitat is probably functioning as a sink.

Some level of nest disturbance is inevitable in all haying and grazing systems, which

raises the question about the overall avian productivity of these habitats and the abilities

164

and strategies of different bird species to recover from those disturbances. Whereas
mowing cannot avoid severely disrupting the majority of nests (Bollinger et al., 1990),
grazing is less systematic and at certain intensities allows a range of nesting success
(Temple et al., 1999). In addition to the degree of disruption caused by mowing and
grazing, a short harvest interval can often prevent successful ﬂedging (Bollinger et al.,
1990; Temple et al., 1999). Studies of grassland birds in Saskatchewan suggest that that
productivity of savanna sparrows can decline by up to 80% after mowing (Dale et al.,
1997)

In research conducted in Wisconsin, investigators observed that when hen pheasants
were disturbed and subsequently renested, they usually did so in a different type of cover
than they initially nested in and chose a location about 400 m from their ﬁrst nest
(Dumke, 1979). Several management practices to improve the nesting success of
grassland birds have been suggested. Research conducted in Iowa suggests that properly
managed native warm season grasses can provide both forage for livestock and suitable
habitat for upland birds (George et al., 1979). These grasses mature later in the spring,
providing the optimal quality and quantity for harvest at a much more optimal time for
nesting birds. Another management practice that could be used is establishing tracts of
undisturbed cover suitable for nesting (“refuge”) near or within large hay ﬁelds could
greatly increase the success of re—nesting efforts of grassland birds disturbed by haying or
grazing aetivity (Temple et al., 1999; Warner and Etter, 1989). Others have suggested
that harvesting individual ﬁelds every other year, leaving the remainder idle for three
years of more would increase the productivity of grassland bird species that prefer dense

cover (Dale et al., 1997). A common practice for governmental agencies overseeing

165

conservation programs has been to allow harvesting land enrolled in those practices only
after a certain calendar date (Dierberger, 2008). Because ﬂedging dates vary among
years, using calendar dates to establish harvest dates may not be appropriate (Dale et al.,
1997)

Speciﬁc grazing practices vary by region, livestock class, and producer. One
question often raised is whether continuous or rotational grazing is more beneﬁcial to
avian species. While this may vary by region, research conducted in Oklahoma
demonstrated that even low stocking densities signiﬁcantly disturb grassland birds; nests
are particularly vulnerable to trampling at grazing densities of 10 head per hectare or
more (Jensen et al., 1990). In native rangeland characteristic of the southwest where
stocking densities are typically low, short duration grazing does not result in signiﬁcantly
more nest trampling than continuous grazing (Koerth et al., 1983). Conclusions from
research conducted to describe the relationship of stocking densities to nest destruction
by trampling suggest that stocking densities of 10 head of cattle per hectare are
particularly disruptive to ground nesting birds (Jensen et al., 1990). Sample and
Mossman also concluded that due to nest trampling and short pasture rest periods,
intensive rotational grazing is not beneﬁcial to most grassland bird species (Sample and
Mossman, 1997).

Temple et al. (1999) compared three types of systems to see how they differed in
grassland bird abundance, diversity, and nesting success: ungrazed grasslands,
continuously grazed pastures, and rotationally grazed pastures. Signiﬁcant differences
existed for all three of the measured parameters. Density, diversity, and nesting success

were greatest on the ungrazed grassland treatment. Density and diversity were

166

intermediate on rotationally grazed pastures and lowest on continuously grazed pastures.
Nesting success was lowest on rotationally grazed pastures and highest on ungrazed
grasslands. Even though a “pro-bird” rotational grazing system with a 1:2 ratio of refuge
to grazed land is better than the typical forms of rotational or continuous grazing, the
annual losses of some bird species associated with the non-refuge portion of the system
are too great to be offset by the refuge portion of the system (Temple et al., 1999). Paine
et a1. (1996) conducted an experiment to show the effect of rotation frequency and
stocking density on nest trampling. The stocking densities and durations included 8 head
per hectare for 7 days, 15 head per hectare for 4 days, and 60 head per hectare for 1 day;
trampling damage averaged 75 percent of the nests for all treatments. Further, nest
trampling was inversely related to vegetation height-density, and percent ground cover.
Research-based recommendations for grassland bird habitat management

Most recommendations for grassland bird habitat management fall into one of
three categories, 1) protocol suggestions for land managed speciﬁcally for bird habitat,
and 2) creation and management of refuge areas near hay land and pasture, and 3) pasture
and hay land management strategies to improve nesting success of grassland birds
Management of plant and residue structure by disturbances and planting

F rawley and Best (1991) claim that mowing and songbird reproduction are
mutually exclusive. Native prairie grasses that are managed for bird nesting habitat
should be disturbed as little as possible to maximize cover for nesting during the
following spring (George et al., 1979). This management recommendation is qualiﬁed
by others who indicate that periodic disturbance (i.e., mowing or burning) is a part of

good grassland bird habitat management (Johnson and Temple, 1990) and can help

167

provide suitable habitat while controlling excessive residue and the encroachment of
woody plants (George et al., 1979). While periodic disturbance is good, (Robel et al.,
1998) point out that it should not be excessive; adverse impacts on grassland birds that
rely on CRP land for nesting habitat can be reduced by avoiding annual burning.
(Millenbah et al., 1996) suggest that every 3-5 years, CRP ﬁelds should be manipulated
to provide a range of successional stages in order to optimize avian productivity,
diversity, and abundance. Dale et al. (1997) advocate harvesting individual ﬁelds every
other year at a date later than is typical for hay production in order to provide habitat and
increase the productivity of grassland bird species that prefer dense cover. Mowing
should be discouraged in areas where nesting success is critical (Camp and Best, 1994).
When mowing dates must be established as part of a protocol, early cutting dates are
more detrimental to nesting success than late cutting dates (Frawley and Best, 1991).
When mowing right-of—ways, avoid unnecessary disturbance of vegetation suitable for
nesting grassland birds by mowing only the shoulder of the roadside (Camp and Best,
1994). Camp and Best (1994) go on to suggest that native grasses should be included in
the seeding mixtures used for road side right-of-ways in order to attract grassland birds
and that roadside vegetation should be burned periodically to increase plant vigor and
structural heterogeneity.
Creation and management of refuge areas near hay land and pasture

Several authors advocate the creation and maintenance of undisturbed cover
suitable for nesting (i.e., “refuge”) near or within large hay ﬁelds in order to increase the
success of resting efforts of grassland birds disturbed by haying or grazing activity

(Temple et al., 1999; Warner and Etter, 1989). Where grassland bird management is a

168

priority, tracts of habitat should be large and far from wooded edges (Johnson and
Temple, 1990). The topic of landowner compensation for management of grassland bird
habitat is complex because the optimal program would be simple, effective, ﬂexible,
‘fair,’ and would reward positive results. Musters (2001) suggest that producers be paid
for nesting success rather than for following a protocol of government prescribed
practices or for production losses. Because the optimal tract size for maximum grassland
bird diversity is 200 ha (Vickery et al., 1994), perhaps there should be a premium value
associated with such large tracts enrolled in conservation programs. Also, because of the
large degrees of variability in environmental conditions in different zones of activity,
local representative of governmental conservation agencies should be given great latitude
in adjusting land management protocols and compensation packages.
Pasture and hay land management strategies to improve nesting success of grassland
birds

Camp and Best (1994) agree, saying that mowing shouldbe discouraged in areas
where nesting success is critical. In order to increase grassland bird nesting success in
Saskatchewan, (Dale et al., 1997) suggest delaying the ﬁrst cutting until July 15 or later.
George et al. (1979) suggest that in south-central Iowa, native warm-season grasses could
be managed as nesting cover and forage if hay harvest began in early-July, grazing began
in July—August, or if the stand was harvested for seed in September, as long as the
particular activity did not reduce the residue height below 20-25 cm. For the conditions
found in west-central New York, (Bollinger et al., 1990) suggest that bobolinks can
beneﬁt from management practices in which hay ﬁelds only are harvested every two _or

three years and not before mid-July in the year of harvest. Jensen et al. (1990) point out

169

that when nesting areas must be grazed, stocking intensity must be kept at levels
consistent with land management objectives, understanding that higher stocking densities
are more disruptive to ground nesting birds. Many of the management protocols that
promote nesting success and increase avian productivity result in decreased forage
quality due to advanced plant maturity advanced maturity stage (Collins, 2003) and
decreased forage yield, as would be the case if the ﬁeld was only harvested after July 15
every third year.

The initial objective of this study was to describe songbird species diversity in the
respective pasture systems, describing apparent relationships between species spatial
distribution and type of pasture system.

MATERIALS AND METHODS
Grazing Experiment

This experiment was conducted at W. K. Kellogg Biological Station in Hickory
Comers, M1, on Kalamazoo series (ﬁne-loamy, mixed, mesic Typic Hapludalfs) soils.
Four replications of three treatments were assigned to the experimental area using a
completely randomized design. The experimental area covered 19.4 ha, and was
rectangular with dimensions of 268 m by 789 m, with 0.70 ha missing from the southeast
comer. The long axis of the pasture had a northfsouth orientation. The layout is
described in Figure A.1 of Appendix 1. Both permanent and temporary fencing materials
were used. The perimeter fence and ﬁve east-west fences were made of high-tensile wire
and wooden fence posts. One semi-permanent fence divided the experimental area in

half from north to south and was constructed of ﬁberglass posts, steel comer braces, and

170

aluminum wire and polywire. Inner temporary fences were constructed of plastic step-in
posts and polywire.
The three treatments included:

1. A cool season grass-legume pasture representative of pastures found throughout
the Lower Peninsula of Michigan. These pastures were comprised primarily of
perennial ryegrass (Lolium perenne), quackgrass (Agropyron repens), alfalfa
(Medicago sativa), white clover ( T rifolium repens), red clover (Trifolium
pratense), orchardgrass (Dactylis glomerata), and tall fescue (F estuca
arundinacea).

2. An integrated switchgrass and cool season grass-legume pasture. One-third of the
system is a monoculture of switchgrass that was planted in 2002, and two-thirds is
composed of the cool season grasses and legumes listed above (Bamhart, 1994;
Undersander, 2002).

3. An integrated big bluestem and cool season grass-legume pasture. One-third of
the system is a monoculture of big bluestem that was planted in 2002, and two-
thirds is composed of the cool season grasses and legumes listed above (Bamhart,
I994; Undersander, 2002).

All pastures had an area of 1.6 hectares with the exception of two which, because of
the shape of the experimental area and existing permanent fencing, differed somewhat.
The cool-season portions of all treatments predated this experiment. The major cool
season species included: perennial ryegrass, Kentucky bluegrass, white clover,
quackgrass, and alfalfa. In the planning stages of this experiment, differences in species

abundance between replications were considered negligible and blocking was not

171

considered. After the experiment began, it became apparent that the variability in ‘alfalfa
abundance’ was signiﬁcant and should be considered a blocking effect. All pastures
which were planted with alfalfa in 1994 were considered were included in ‘alfalfa
abundant’ block. The pasture layout with respect to experimental treatments and alfalfa
abundance is described in Figure A2 of Appendix I.

The animals used in this study were Holstein steers that weighed approximately
240 kg at the beginning of the grazing season. Pastures were managed as uniformly as
possible, but each was managed separately and optimally (Bransby, 1989). When
grazing was initiated each spring, cattle on all treatments were given access to the entire
cool-season portion of their respective pastures, with the exception of the fourth
replication of the C S-Only treatment, where a concurrent experiment was being canied
out which required that livestock be excluded from a portion of the pasture at certain
points of time. As the spring ﬂush of cool-season pasture growth intensiﬁed, pastures
were “staged” (Barker, 1999): cattle were limited to one half, and later, one quarter of the
cool-season portion of their treatments at which point rotational grazing began and
continued until the end of the season. During the ﬁnal grazing event of the season, the
cattle were again given access to the entire cool-season portion of their respective
pastures. In 2003, the big bluestem portion of the CS-BBS treatments was split-into three
parts each of which was grazed in rotation (the switchgrass portion of the CS-SG
treatment was not grazed in 2003). In 2003, residue of refused big bluestem forage was
clipped after grazing as needed in order to make the pasture height more uniform at
subsequent grazing events. However, it was noted that the small size of the subdivisions

within the big bluestem pasture resulted in excessive trampling of forage which caused a

172

high proportion of plants to re-grow from the crown rather from intemodes, thereby
increasing the amount of time required for regrowth. Further, the tractor tire tracks left
from mowing caused the crushed plants to regrow from the crown; this regrowth was lush
and tended to be grazed preferentially by the livestock. Thus, in 2004 and 2005 the WSG
portions of the CS—BBS and CS-SG treatments were not subdivided or clipped and each
WSG portions was left undivided and grazed as one of the rotations.

Decisions about timing and duration of rotations were dictated by current and
anticipated pasture forage availability rather than using strict dates or interval lengths.
Variables such as soil moisture, physiological and re-growth stages of forage species,
typical seasonal weather patterns, and weather forecasts were used to make rotation
decisions. After the spring ﬂush, the grazing intervals were usually seven days for the
CS-BBS and CS-SG treatments; CS-only treatments were usually rotated every ten days
and consisted of four subdivisions.

Mid-season grazing pressure was adjusted using the put-’and—take methodI of
pasture stocking (Mott, 1960). ‘Tester steers’ were left on the same replication of the
same treatment for the entire grazing season. Put-and-take steers were added to or
removed from a particular replication to increase or reduce the grazing pressure in order
to:

- optimize utilization of pasture forage
o prevent the accumulation low-quality forage

- meet but not exceed the current or anticipated forage dry matter production

 

'_ See Technical Note 1 in Appendix III

173

Within an individual pasture, animals were rotated when one or more of the
following was true: 1) paddocks forward in the rotations were about to become over-
mature; 2) animals or pasture might be harmed by low forage levels in the current
rotation; 3) rapid pasture growth indicated that shorter intervals were needed in the
immediate future to prevent accumulation of low-quality forage which would likely be
refused.

The anticipated date for termination of grazing was October 1, but pasture
conditions including plant stress, drought, and forage dry matter availability were the
primary considerations for termination of grazing for the season (Bransby, 1989; Leep,
2003). Dates of grazing and termination dates for each year are listed in Table 1.2 of
Chapter 1 of this dissertation.

Bird Study

Data was collected from transects (Hostetler, 2006) walked on June 4, June 14, and
July 2 of 2004 and July 9 of 2005. Transects had an east-west orientation and were 50 m
wide, centered within the experimental units. The locations and species of birds were
determined by sight and/or sound and recorded. A zig-zag pattern was used within each
transect and only birds present within the transect were counted. On each of the four
dates, surveys began approximately 30 minutes after sunrise, and took between 69 and
121 minutes to complete. All sample dates were at least partly sunny, calm, and between
12 and 21 oC.

Rodents were not recorded as suggested by the hypothesis. Birds species observed

during counting events included savanna sparrow, bobolink, red-winged blackbird,

174

Canada geese, common grackle, house sparrows, barn swallows, cliff swallows, phoebe,
and American robin, were not always included in the data set.

0 Birds that were on the fence between experimental units (individual pastures) or

on the perimeter fence were not counted. 1

o If a species was observed less than six times throughout the study, they were not

included in the data set

0 Although nest-defending behavior was occasionally noted it was not included in

the analysis

0 Canada geese were excluded from the analysis

The bird species that were present with sufﬁcient frequency according to the above
criteria were the savanna sparrow, red-winged blackbird, bobolink, and barn swallow.
Because of the small size of the experimental units, even these species were present in
low numbers, so their numbers were combined for data analysis. Table 4.1 lists the bird
species found within the respective treatments during the observation periods.

Other species observed in, around, and over the experimental area during the course
of the experiment include: cattle egret (Bubulcus ibis), Canada goose (Branta
Canadensis), red-tailed hawk.(Buteojamaicensis), sandhill crane (Grus Canadensis),
killdeer, snipe (Gallinago spp.), phoebe (Sayornis spp.), horned lark, eastern bluebird
(Sialia sialis), American robin (Turdus migratorius), common grackle (Quiscalus
quiscula), house sparrow (Passer domesticus), tree swallow (T achycineta bicolor),
mallard duck (Anas platyrhynchos), European starling (Sturnus vulgaris), American

goldfmch (Carduelis tristis), American crow (Corvis brachyrhnchos), mourning dove

175

(Zeniada macroura), chipping sparrow (Spizella passerine). ﬁeld sparrow, kingbird
(T yrannus spp.), and rock dove (Columbia liva)
Statistical Analysis

Because the low bird numbers and the sub-optimal size of the experimental area,
it was not realistic to attempt to statistically evaluate the original hypothesis. Rather, the
total numbers of birds observed within each treatment (using the three criteria listed
above) were compared.

PROC GLIMMIX of Statistical Analysis Systems (SAS) Version 9.1.3 was used
to conduct analysis of variance tests for the bird population data. The following data
transformation was used: log (total + 1), where ‘total’ represented the combined number
of savanna sparrows, red-winged blackbirds, bobolinks, and barn swallows observed in a
particular experimental unit on one of the four sampling dates. This data transformation
was necessary to normalize the distribution of residuals. Bird counts were taken on day
of year (DOY) 155, 165, and 184 (June 4, June 14, and July 2, respectively) of 2004 and
DOY 190 (July 9) of 2005.

Treatment, year, and the interaction of treatment and year were evaluated in the
statistical model. Year and the treatment by year interaction were not statistically
signiﬁcant, but the effect of treatment was statistically signiﬁcant. Thus, the statistical
model for total bird population includes the effect of treatment only. Day of year was
treated as a random effect.

The full model for forage dry matter offered is:
Yi = u + A1 + ei

where:

176

. . . .th .
y; IS the average number of birds observed in 1 treatment on any sampling date.

and:

p is the overall mean

A; is a ﬁxed effect, the ith treatment (CS-BBS, CS-Only, CS-SG)
e; is the error term

RESULTS AND DISCUSSION

While the overall model was statistically signiﬁcant, the only comparison that
approached statistical signiﬁcance (p = 0.07) was between the CS-BBS and CS-SG
treatments. Each CS-BBS pasture had an average of 2.6 birds per transect, while the CS—
Only and CS-SG treatments had 1.9 and 1.7, respectively (see Table 4.2). Environmental
conditions are described in Figures A.3-A.5 of Appendix I.

If the experimental area had been optimally designed for bird research, a less
stringent p-value might be appropriate. The greatest weakness of the bird portion of this
study was the size of the experimental units. Vickery et al. (1994) point out that even if a
unit of cover has excellent plant and residue structure for a bird species, it might still be
refused by that species due to the insufﬁcient area of the otherwise acceptable habitat.
Another difﬁculty was that replications of different treatments were immediately adjacent
to each other and it is very likely that the sphere of inﬂuence of one treatment overlapped
into adjacent treatments or that the territory of one bird might overlap into as many as
four different experimental units, given that the size of red-winged blackbird and

savannah sparrow territories can both exceed 1 ha (Albers, 1978; Weins, 1973).

177

Very few birds were documented in the big bluestem or switchgrass portions of
their respective treatments. Excluding the birds observed between experimental units or
on the perimeter fence, a total of 96 birds were counted within the experimental area. Of
these, only eight were observed in the big bluestem and switchgrass pastures, combined.
While the combined area of the big bluestem and switchgrass pastures comprised about
23% of the experimental area, only 8% of the birds counted were within those pastures.
There may be several reasons for this beyond insufﬁcient size. The ﬁrst is that residue
was always burned prior to the next grazing season, which left little or no cover suitable
for early-spring nesting (George et al., 1979). Big bluestem and switchgrass residue was
burned in the spring of 2004, prior to initiation of active growth and again in the fall of
2004 (rather than in the spring of 2005). Second, spring growth of switchgrass and big
bluestem is slow and little cover or structure is created until much later than in the cool-
season portions of the respective treatments. Third, because the switchgrass and big
bluestem pastures were managed with the goal of creating/maintaining a monoculture,
there tended to be less plant structural heterogeneity (McMaster and Davis, 2001) than in
the cool-season portions of the experimental area. The periodic mowing and moderate
grazing intensity probably attracted savannah sparrows and bobolinks, although it
probably reduced nesting success for those and other bird species (Bollinger et al., 1990;
Dale et al., 1997; Warner and Etter, 1989). Albers (1978) noted that red-winged
blackbirds abandon their territories within 48 hours of mowing, and this is likely true of
other species. It is possible that an experimental unit that is preferred by one or several
bird species could be mowed and the adult birds with territories within that area could

quickly move to adjacent or distant experimental units which may or may not be a

178

replication of the same treatment. This clearly would confound treatment effects.

Fourth, the disparity in the number of birds counted in cool-season pastures compared
with the number observed in switchgrass and big bluestem pastures could be partially due
to the presence of subdividing temporary fence (functioning as perches) in the cool-
season pastures and the absence of those fences in the big bluestem and switchgrass
pastures. Finally, the intersection of the big bluestem and switchgrass pastures with the
cool-season portions of the treatments also presented an abrupt edge which may have
increased bird predation (Gates, I978; Ratti and Reese, 1988). Anecdotally, nesting
activity was noted beneath the temporary fences, probably the result of re-nesting activity
following nest destruction due to grazing or mowing; these areas tended to be less
disturbed by grazing and mowing activities.

The earliest date that surveys began in either year was June 4. Bird studies that
hope to be able to describe something about the seasonal dynamics of bird populations
and/or nesting success typically begin by early-May; examples of this can be found in
many other studies of grassland bird populations including: DeLislie and Savidge (1997),
Frawley and Best (1991), Herkert (1994), and Millenbah et al. (1996). While beginning
the surveys earlier in the spring almost certainly would have demonstrated greater bird
species richness and the total number of birds observed in each transect, the small size of
the study area still would have greatly limited the interpretation of the superior dataset.

CONCLUSION

A signiﬁcant difference in the number of birds among the pasture treatments was

not found. The size and design of the greater grazing experiment was not suitable for

rigorous research on the effect of native warm season grasses in grazing systems on

179

grassland bird species. While periodic mowing is realistic agronomic management,
grazing (Temple et al., 1999) and mowing activity (Bollinger et al., 1990; Frawley and
Best, 1991) can be very disruptive to nesting birds, and probably confounded the
objective. Studies designed to describe the effect of big bluestem or switchgrass on
grassland birds populations in integrated grazing systems should begin earlier in the
spring, on a much larger experimental area, and include a treatment where the pastures

are not mowed.

180

Table 4.1. Birds observed within each treatment during observation periods.

 

 

Treatment

Bird Species CS-BBS CS-Only CS-SG
american robin X

barn swallow X X X
bobolink X X X
Canada goose X

common grackle X

house sparrow X
red winged blackbird X X X
savanna sparrow X X X

 

181

Table 4.2. Average number of birds per transect as affected by treatment.

 

Treatment Average Bird Count“ SE
CS-BBS 2.6 a 0.3
CS-Only 1.9 a 0.3

CS-SG 1.7 a _ 0.3

 

*Numbers within the column with the same letter are not statistically different (PS 0.05)

182

LITERATURE CITED

Albers, RH. 1978. Habitat selection by breeding red-winged blackbirds. Wilson Bulletin
90:619-634. ~

Anderson, B. 2000. Grazing management on warm season grasses Missouri Forage and
Grassland Council. University of Missouri.

Askins, R.A. 2001. Sustaining biological diversity in early successional communities: the
challenge of managing unpopular habitats. Wildlife Society Bulletin 29:407-412.

Barker, J.M., Buskirk, D. D., Ritchie, H. D., Rust, S. R., Leep, H. R., and K. J. Barclay.
1999. Intensive grazing management of smooth bromegrass with or without
alfalfa or birdsfoot trefoil: heifer performance and sward characteristics. The
Professional Animal Scientist 15:130-135.

Bamhart, SK. 1994. Warm-season grasses for hay and pasture - Pm-569. Iowa St. U.
Ext.

Best, L.B., H.I. Campa, K.E. Kemp, R.J. Robel, M.R. Ryan, J.A. Savidge, H.P. Weeks,
Jr., and SR. Winterstein. 1997. Bird abundance and nesting in CRP ﬁelds and
cropland in the Midwest: a regional approach. Wildlife Society Bulletin 25:864-
877.

Blanco-Canqui, H., Gantzer, C. J., Anderson, S. H., Alberts, E. E., and A. L. Thompson.
2004. Grass barrier and vegetative ﬁlter strip effectiveness in reducing runoff,

sediment, nitrogen, and phosphorus loss. Soil Science Society of Americazl670-
1678.

Bollinger, E.K., P.B. Bollinger, and TA. Gavin. 1990. Effects of hay cropping on eastern
populations of the bobolink. Wildlife Society Bulletin 18:142-150.

Bransby, DJ. 1989. Compromises in the design and conduct of grazing experiments, p.
136, In G. C. Marten, ed. Grazing research: design, methodology, and analysis.
CSSA Spec. Pub. No. 16. Crop Science Society of America and the American
Society of Agronomy, Madison.

Camp, M., and LB. Best. 1994. Nest density and nesting success of birds in roadsides
adjacent to rowcrop ﬁelds. American Midland Naturalist 131:347-358.

Coleman, S.W., Sollenberger, LE. 2007. Plant-Herbivore Interactions, p. 123-136, In R.

F. Barnes, Nelson, J. C., Moore, K. J., and M. Collins, ed. Forages: the science of
grassland agriculture, Vol. 2, 6th ed.

183

Collins, M., and J. 0. Fritz. 2003. Forage Quality, p. 363-390, In R. F. Barnes, Nelson, J.
C., Collins, M., and Moore, K. J., ed. Forages: an introduction to grassland
agriculture, Vol. 1. Iowa State University Press, Ames.

Dale, B.C., P.A. Martin, and PS. Taylor. 1997. Effects of hay management on grassland
songbirds in Saskatchewan. Wildlife Society. Bulletin 25:616-626.

Delisle, J.M., and J.A. Savidge. 1997. Avian use and vegetation characteristics of
conservation reserve program ﬁelds. Journal of Wildlife Management 61 :31 8-
325.

Dierberger, B. 2008. Harvesting land enrolled in conservation programs using calendar-
date. Telephone conversation, In D. J. Hudson, (ed.), Mason, Ml.

Donovan, T.M., and CH. F lather. 2002. Relationships among North American songbird

trends, habitat fragmentation, and landscape occupancy. Ecological Applications
12:364-374.

Duke, J.A. 1978. The quest for tolerant germplasm, p. 161, In M. Stelly, ed. Crop
tolerance to suboptimal land conditions. American Society of Agronomy, Crop
Science Society of America, Soil Science Society of America, Madison.

Dumke, RT, and Pils, C. M. 1979. Renesting and dynamics of nest site selection by
Wisconsin pheasants. Journal of Wildlife Management 43:705-716.

F ahrig, L. 2001. How much habitat is enough? Biological Conservation 100:65-74.

Frawley, 8.1., and LB. Best. 1991. Effects of mowing on breeding bird abundance and
species composition in alfalfa ﬁelds. Wildlife Society Bulletin 19:135-142.

Gates, J.E., and L. W. Gysel. 1978. Avian nest dispersion and ﬂedging success in ﬁeld-
forest ecotones. Ecology 59:871-883.

George, J.R., Hintz, R. L., Moore, K. J., Bamhart, S. K., and D. R., Buxton. 1996. Steer
response to rotational or continuous grazing on switchgrass and big bluestem

pastures, p. 150-154, In M. J. Williams, ed.’ 1996 American Forage and Grassland
Council Proceedings, Vancouver.

George, RR, A.L. Farris, C.C. Schwartz, D.D. Humburg, and J.C. Coffey. 1979. Native
prairie grass pastures as nest cover for upland birds. Wildlife Society Bulletin 7:4-
9.

Good, BE, and CA. Dambach. 1943. Effect of land use practices on breeding bird
populations in Ohio. Journal of Wildlife Management 7:291-297.

184

Harvey, R.G. 2000. Establishing prairie plants for CRP or wildlife habitat with herbicides
[Online]. Available by University of Wisconsin

http://ipcm.wisc.edu/uw_weeds/extension/articles/prarplntcrphtm (veriﬁed
03/22/2006).

Hazell, DB. 1967. Effect of Grazing Intensity on Plant Composition, Vigor, and
Production. Journal of Range Management 20:249-252.

Henning, J.C.A.p.G.B.b., indiangrass, and switchgrass. U of Missouri-Columbia. 1993.
Big bluestem, indiangrass, and switchgrass - G4673 [Online]. Available by
University of Missouri

http://muextension.missouri.edu/explore/agguides/crops/g04673.htm (veriﬁed
03/22/06).

Herkert, J .R. 1994. The effects of habitat fragmentation on Midwestern grassland bird
communities. Ecological Applications 4:461 -471 .

Hostetler, M.E., and M. B. Main. 2006. WEC155: Florida monitoring program: transect
method for surveying birds, pp. 8, In E. Wildlife, and Conservation, (ed.).
University of Florida Cooperative Extension Service, Institute of Food and
Agricultural Sciences, University of Florida.

Jensen, H.P., D. Rollins, and R.L. Gillen. 1990. Effects of cattle stock density on
trampling loss of simulated ground nests. Wildlife Society Bulletin 18:71-74.

Johnson, DH, and MD. Schwartz. 1993. The Conservation Reserve Program and
grassland birds. Conservation Biology 7:934-937.

Johnson, R.G., and SA. Temple. 1990. Nest predation and brood parasitism of tallgrass
prairie birds. Journal of Wildlife Management 54:106-1 11.

Koerth, B.H., W.M. Webb, F.C. Bryant, and F .S. Guthery. 1983. Cattle trampling of

simulated ground nests under short duration and continuous grazing. Journal of
Range Management 36:385-386.

Leep, R. 2003. Forage Specialist and Assistant Professor of Crop and Soil Sciences,
Michigan State University.

Lindquist, J. 2002. 2002 Michigan Hay and Grazing Council Industry Priorities [Online]
http://www.greeen.msu.edu/indpriorities/2002havgrazingPriorities.pdf (veriﬁed
3/25/06).

MASS. 2002. Speciﬁed crops by acres harvested: 2002 and 1997. Michigan Agricultual
Statistics Service.

Matthews, V. 2003. Deputy State Statistician, Michigan Agricultural Statistics Service.

185

McMaster, D.G., and SK. Davis. 2001. An evaluation of Canada's Permanent Cover
Program: Habitat for grassland birds? Journal of Field Ornithology 72:195-210.

Millenbah, K.F., S.R. Winterstein, H.I. Campa, L.T. Furrow, and RB. Minnis. 1996.
Effect of conservation reserve program ﬁeld age on avian relative abundance,
diversity, and productivity. Wilson Bulletin 108:760-770.

Mitchell, R., Moser, L., Anderson, B., and S. Waller. 1996. Switchgrass and big bluestem
for grazing and hay, G94-1198-A [Online] expired (veriﬁed 01/29/2003).

Moore, K.J., Hintz, R. L., Wiedenhoeft, M.H., Brummer, EC, and J. R. Russel. 2001.
Sequential grazing systems for beef cattle production: A.S. Leaﬂet R1749
[Online]. Available by Iowa State University

http://www.extension.iastate.edques/ansci/beefreports/asl 1 749.pdf (veriﬁed
03/22/2006).

Moser, LE, and K. P. Vogel. 1995. Switchgrass, big bluestem, and indiangrass, In R. F.
Barnes, Darrell, A.M., and OJ. Nelson, ed. Forages: An introduction to grassland
agriculture, Vol. 1, Fifth ed. Iowa State UP, Ames.

Mott, GO. 1960. Grazing pressure and the measurement of pasture production. Proc VIII
Int. Grassl. Cong.:606-61 1.

Murphy, MT. 2003. Avian population trends within the evolving agricultural landscape
of eastern and central United States. The Auk 120:20-34.

Nelson, C]. 1995. Photosynthesis and carbon metabolism, p. 31-43, In D. R.F., A.M.,
and OJ. Nelson, ed. Forages: an introduction to grassland agriculture, Vol. 1, Fifth
ed. Iowa State UP, Ames.

Patterson, MP, and LB. Best. 1996. Bird abundance and nesting success in Iowa CRP
ﬁelds: the importance of vegetation structure and composition. American Midland
Naturalist 135:153-167.

Peterson, P.R., Rayburn, E. B., Cropper, J. B., and Belesky, D. P. 2007. Perennial warm-
season grasses, In E. B. Rayburn, ed. Forage utilization for pasture-based
livestock production. Natural Resource, Agriculture, and Engineering Service,
Ithaca, NY.

Pulliam, HR. 1988. Sources, sinks, and population regulation. The American Naturalist
132:651-661.

Rankins, A.J., Shaw, D. R., and Boyette, M. 2001. Perennial grass ﬁlter strips for
reducing herbicide losses in runoff. Weed Science 49:647-651.

186

Ratti, J .T., and KP Reese. 1988. Preliminary test of the ecological trap hypothesis.
Journal of Wildlife Management 53:484-491.

Robel, R.J., J.P. Hughes, S.D. Hull, K.E. Kemp, and D.S. Klute. 1998. Spring burning:
Resulting avian abundance and nesting in Kansas CRP. Journal of Range
Management 51:132-138. -

Sample, D.W., and MJ. Mossman. 1997. Managing habitat for grassland birds: a guide
for Wisconsin Wisconsin Department of Natural Resources, Madison.

Sargent, M.S., Carter, K. S., eds. 1999. Grassland management, p. 297 Managing
Michigan wildlife: a landowners guide. Michigan United Conservation Clubs,
East Lansing.

Sauer, J .R., G.W. Pendleton, and B.G. Peterjohn. 1996. Evaluating causes of population
change in North American insectovorous songbirds. Conservation Biology
10:465-478.

Stallman, HR, and LB. Best. 1996. Bird use of an experimental strip intercropping
system in northeast Iowa. Journal of Wildlife Management 60:354-362.

Temple, S.A., B.M. Fevold, L.K. Paine, DJ. Undersander, and D.W. Sample. 1999.
Nesting birds and grazing cattle: accommodating both on midwestem pastures.
Studies in Avian Biology 19:196-202.

Undersander, D. 2002. Conversation about using warm—season grasses as a component of
grazing systems.

USDA-NRCS. 1996. Establishing warm-season grasses, pp. 3, In USDA-NRCS, (ed.)
Hay and pasture management. USDA-NRCS, Columbia, Missouri.

USDA-N RCS. 1999. Job sheet for native warm season grass establishment as wildlife
habitat (645) [Online]
http://www.sc.nrcs.usda.gov/ﬁles/pubs/nativegrassiobsheetpdf (veriﬁed
03/22/2006).

USDA-NRCS, 2007. Georgia Job sheet (645): Native grasses for wildlife habitat, pp. 3.

Veech, J.A. 2006. A comparison of landscapes occupied by increasing and decreasing
populations of grassland birds. Conservation Biology 20: 1422-1432.

Vickery, P.D., M.L.J. Hunter, and SM. Melvin. 1994. Effects of habitat area on the
distribution of grassland birds in Maine. Conservation Biology 8: 1087-1097.

Warner, RE, and S.L. Etter. 1989. Hay cutting and the survival of pheasants: a long-
terrn perspective. Journal of Wildlife Management 53:455-461.

187

Weaver, J.E. 1954. North American Prairie Johnsen Publishing Company, Chicago.

Weins, J.A. 1973. Interterritorial habitat variation in grasshopper and savannah sparrows.
Ecology 54:877-884.

188

Appendix 1: Experimental Design and Environmental data

Figure A. 1. Positioning of treatment replications within the experimental area.

CS-0n1y(1)

CS-BBS (1)

08-30 (3)

CS-Only (3)

csass (3)

CS-BBS (4)

CS-SG (1)

CS-SG (2)

CS-Only (2)

CS-BBS (2)

cs-se (4)

CS-Only (4)

 

I: Cool season grasses and
legumes

I = Big bluestem

. = Switchgrass

189

Figure A.2. Pasture layout with respect to alfalfa abundance

Alfalfa Abundant CS-On1y(1) CS-SG (1) Alfalfa Abundant l
_)

 

Alfalfa not Abundan CS-BBS (1) CS-SG (2) Alfalfa "01 Abundant
_) (-

Alfalfa Abundant (35-36 (3) (IS-Only (2)
_) I = C 001 season grasses and

legumes

I = Big bluestem
Alfalfa not Abundant
(_

I = Switchgrass

Alfalfa Abundant
Alfalfa not Abundar; CS-Only(3) 08-888 (2)

Alfalfa not Abundant
Alfalfa not Abundant C553?» (3) CS-SG (4) (—
_)

Alfalfa Abundant
e

CS-Only (4)

Alfalfa Abundant (ZS-335 (4)
9

190

Figure A.3. Precipitation by month for Hickory Comers, M1 (KBS-LTER, 2005).

1
1
1
1 30 1
1

1 A 25 1
1 5 ‘
1 :201 I17YrAvg
1 E»
1 s 15 1 , .2003
1 .5 1
1 1310 1 132004

'5 1
1 '5. 51 E12005
~ 1 111

01 1 1 ' ‘ 1 1 1 1 ’

1 , , v 's\ A, e \g 9‘ s s s s
1 a" a" a“ s‘ a” 50° 9* 3 a“ on" a“ s“
1 to ‘0’ S} I. Y' 6* 0‘ A“ c.
1 8 Qt (ORR 0 Q2.
1 Month

191

Figure A.4. Mean air temperature by month for Hickory Comers, MI (KBS—LTER,

2005)

Temperature (Degrees Celsius)

 

192

(I17 YrAvg

I 2003
El 2004
D 2005

 

Figure A.5. Maximum air temperature by month for Hickory Comers, MI (KBS-LTER,
2005).

 

1 35

1 30

1

1 25 _

1 20 '17 Yr Avg
.2003

El 2004

1

1

1

1

1

1

1

1

10 [1] I] 1:12005 1
5 1
1

0 FUﬂ I " 1 1‘ 1 l] 1
1

1

1

Temperature (Degrees Celsius)
G

 

m

 

 

-5 i d ‘ '3 s 2. a ‘9 6' v 6‘ 6
«r r» <9 5 ‘9 5‘ 9 ~o *0 *0 *0
0° .003 ‘8‘" YR V” 3° S ed: 66‘ 0x9 06‘ e
a Q“ o A 25‘
(oz)

$3 Q 90 Q

193

LITERATURE CITED

KBS-LTER. 2005. The Kellogg Biological Station Long-Term Ecological Research
Climate Database. Michigan State University Board of Trustees.

194

Appendix II: Botanical Composition

Fifty nine plant species were found in pastures throughout the experiment (see
Tables B. 1-B.2). Pasture botanical composition is described in terms of species and plant
category over time. The four categories used to describe pasture plants include: forage
grass, forage legume, palatable weed, and undesirable weed. In this context, ‘palatable’
means that when a given plant species is in a vegetative stage the plant is not selectively
refused by grazing animals (e. g., dandelion). “Undesirable” pasture plants are either
clearly selectively refused plants (e.g., bull thistle) or those that are generally known to
be eaten rarely by cattle (e. g., corn Speedwell).
Botanical Composition by Experimental Treatment: Pasture Plant Categories

Forage grass DM percent remained at similar levels among treatments over time
(Figure B.l).
Forage legume DM percent differences among treatments were greatest at the time of the
ﬁrst data collection in 2003 (Figure 8.2). Clover was no-tilled twice into the CS portion
of all treatments in 2003. At the time of the ﬁrst botanical composition assessment in
2004, the legume percentage of forage DM was similar among treatments and remained
similar for the duration of the experiment.

Palatable pasture weed dry matter ﬂuctuated between 3% and 12% throughout the
experiment; trends were similar among treatments (Figure B.3). Treatment differences of
percent pasture DM made up of undesirable pasture weeds were greatest in 2003 and

tended to converge over time, although seasonal ﬂuctuations were evident (Figure 8.4)

195

Treatment botanical composition by plant type
Weed dynamics in particular agricultural systems are inﬂuenced by the patterns of
disturbance of plants and soil. In the case of grazing systems which range from
continuous to intensive rotational in nature, speciﬁc weed species can either be favored or
selected against. For example, Canada thistle is favored by continuously grazing
pastures, while high-intensity, low-frequency grazing causes the frequency of Canada
thistle (De Bruijn, 2006). Altemately, excessive grazing under particular conditions can
cause other weed species to proliferate (Harker, 2000). In all treatments of this study
three trends emerged from botanical composition data relative to plant type:
0 Grass and legume dry matter percent were relatively stable (Figures B.1-B.2)
o Palatable weeds were stable or increased slightly over the course of the experiment
(Figure 3.3)

o Undesirable weeds declined from 2003 to 2005 (Figure B.4).
Botanical Composition by Experimental Treatment: Major Pasture Plant Species

Orchardgrass, tall fescue, and white clover were no-till seeded in the CS portions
of all treatments in 2003. Tables B.3-B.5 show the contribution of major plant species to
botanical composition of each treatment over time. In each treatment, the percent dry
matter contributed by orchardgrass increased from August of 2004 through July of 2005.
For pasture plant species that are very sensitive to drought conditions, true differences in
botanical composition due to pasture management should be more evident after periods
of adequate soil moisture. The DM contributed to each treatment by white clover

increased markedly from May 2003 to June 2004 and May 2005 while the DM

contributed by alfalfa declined over those same years and months. In each treatment,

196

perennial ryegrass was the major forage grass species, while quackgrass contributed
signiﬁcantly to each treatment at different times.

Prior to the assignment of treatment replications to the subdivisions within the
area allotted for the experiment, differences in forage species abundance among the
experimental units were considered negligible. After the experiment began, it became
apparent that ‘alfalfa abundance’ should be considered as a blocking effect. The result of
the random assignment of treatments to the available pasture areas irrespective of alfalfa
abundance resulted in the CS-BBS treatment only being assigned to one replication with
‘abundant alfalfa’ while three of the CS-Only and two of the CS-SG replications had
abundant alfalfa (Appendix I, Figure A.2). Thus, when botanical composition by
treatment is considered, the CS-Only treatment clearly contains a higher percentage of
legume (primarily alfalfa) than the CS-BBS treatment. The disparity of legume content
between treatments was recognized in early-Spring of 2003, and white clover was planted
(no-till) into the living sod the CS portion of each treatment. The percentages of legumes
in each of the treatments converge in June of 2004 and remain roughly parallel after that
point.

WSG botanical composition

The big bluestem and switchgrass monocultures generally remained at or above
80% crop dry matter at the height harvested (Figure B.5). Crabgrass growth below the
sampling height was often dense in the switchgrass portion of the CS-SG treatment.
Anecdotally, the steers preferred crabgrass over switchgrass, and the crabgrass
contributed signiﬁcantly to the quantity and the quality of the forage consumed. In 2003

and 2004, there was signiﬁcant competition from quackgrass in the switchgrass and big

197

bluestem portions of their respective treatments; this problem was mitigated by an
application of glyphosate to dormant big bluestem and switchgrass monocultures in April

2005 (See Chapter 1 Materials and Methods).

198

Table B. 1. Latin and common names of forage and palatable non-forage species found in
the CS portions of all treatments throughout the experiment.

 

 

Latin Name Common Name Type"
Agropyron repens (L.) Beauv quackgrass FG
Agrostis stolonifera L. var. palustris (Huds.) Farw. bentgrass FG
Bromus inermis Leyss. smooth bromegrass FG
Dactylis glomerata L. orchardgrass FG
Festuca arundinacea Schreb. tall fescue FG
Lolium perenne L. perennial ryegrass FG
Lolium perenne L. x Schedonorus pratensis (Huds.) Beauv. festulolium FG
Phleum pratense L. timothy FG

Poa pratensis L. Kentucky bluegrass F G
Lotus corniculatus L. birdsfoot trefoil FL
Medicago lupulina L. black medic FL
Medicago sativa L. alfalfa FL
Trifolium ambiguum Bieb. kura clover FL
Trifolium hybridum L. alsike clover FL
Trifolium pretense L. red clover FL
Trifolium repens L. white clover FL
Digitaria ischaemum (Schreb. ex Schweig.) Schreb. ex Muhl smooth crabgrass . PG
Digitaria sanguine/is (L.) Scop. large crabgrass - PG
Eleusine indica L. Gaertn. goosegrass PG
Poa annua L. - annual bluegrass PG
Stellaria media (L) Vill. common chickweed PB
Taraxacum ofﬁcinale Weber dandelion PB
Chenopodium album L. lambsquarters PB
Cerastium vulgatum L. mousear chickweed PB
Amaranthus retroﬂexus L. redroot pigweed PB
Silene Iatifolia Poir. White campion PB

 

*FG: forage grass; FL: forage legume; PG: palatable grass weed; PB: palatable broadleaf weed

199

Table 82. Latin and common names of undesirable plant species found in the CS

portions of experimental pastures

 

Latin Name

Common Name

 

Erigeron annuus(L) Pers.
Plantago an’stata Michx.

Plantago major L.

Cirsium vulgare(Savi) Tenore
Cirsium arvense (L.) Scop.
Bromus tectorum L.

Malva neglecta Wallr.

Ambrosia artemisiifolia L.

Rumex acetosella L.

Veronica arvensis L.

Oxalis comiculataL.

Rumex crispus L.

Panicum dichotomiﬂorum Michx.
Thlaspi arvense L.

Glechoma hederacea L.

Lamium amplexicauleL.

Conyza canadensis (L.) Cronq.
Plantago lanceolala (L.)

Veronica persica Poir.

Matricaria matricaroides (Less) Porter
Lactuca serriola L.

Polygonum arenastrum Jord. ex Bor.
Portulaca oleracea L.

Daucus carota L.

Potentilla norvegica L.

Erigeron slrigosus Muhl. Ex Willd.
Capes/la bursa-pastoris (L.) Medic.
Centaurea maculosaLam.

Rhus or Toxicodendron
Sisymbrium altissimum L.
Amaranthus albusL.

Several unknown species
Abutilon theophrasti Medic.
Barbarea vulgaris R. Br.

Oxalis stricta L.

annual ﬂeabane
bracted plantain
broadleaf plantain
bull thistle

canada thistle
cheat/downy brome
common mallow
common ragweed
common red sorrel
corn Speedwell
creeping woodsorrel
curly dock

fall panicum

ﬁeld pennycress
ground ivy

henbn

marestail
narrowleaf plantain
persian Speedwell
pineapple weed
prickly lettuce
prostrate knotweed
purslane

Queen Anne's lace
rough cinquefoil
rough ﬂeabane
shepherd's purse
spotted knapweed
sumac

tumble mustard
tumble pigweed
unknown
velvetleaf

yellow rocket
yellow woodsorrel

 

Figure B. l. Forage grass percentage of total CS pasture dry matter from 2003-2005.

 

1
Forage Grass Dry Matter Dynamics by Treatment

 

 

 

 

 

 

 

 

3:» 90*

E 8“ +Cs-BBs
f 70 '- '— +CS—Only
g 60 +C$SG

Q)

2 50 —— — —‘———-__Mf-_ .__

“r 40 . 1 .k .

Sb“ \B" \Q" \Q"

‘5 ‘5 hr
\% \§ \Q
a” «\‘E

at“ 0‘” 0““ 0”“

 

Date

Percent forage grass is relatively constant over time and changes are roughly parallel
among treatments.

201

Figure 8.2. Forage legume percentage of total CS pasture dry matter from 2003-2005.

, 7 _,

Forage Legume Dry Matter Dynamics by Treatment

llcs-BBS
+ CS-Only
,Tfsﬁﬁ

 

 

202

 

Figure 8.3. Palatable weed percentage of total CS pasture dry matter from 2003-2005.

Palatable Weed Dry Matter Dynamics by Treatment

 

 

 

g 15
g 10 «JCS-Bus
E +CS-Only
u .+CS-SG
: 5 . ~ — ~
3
’5
G- 0 _

90'

Date

 

203

 

 

Figure B.4. Undesirable weed percentage of total CS pasture dry matter from 2003-2005.

 

 

Undesirable Weed Dry Matter Dynamics by Treatment
3
3:: 10
a 1 - , ,- L W
E 2“ _ , - * ,, +CS-BBS
E 4 __ +CS-Only
E 2 — .+CS-§G
a; 0 ' l T ’T l
a. ‘3 ‘3 ‘9
o 6‘5 (oh e e e e
«\ it \ s\
W \ \ '1; ‘1- \ x
a «1i \ go
Date

 

 

204

Table 8.3. Contribution of major pasture plant species to botanical composition of the
cool-season portion of the CS-BBS treatment.

 

 

Species Date 5/26/03 8/18/03 6/18/04 8/27/04 5/24/05 7/18/05 10/11/05
Kentucky bluegrass avg 4.6 4.2 5.6 4.2 9.3 3.9 11.7
SE 1.6 1.8 1.7 1.7 2.9 2.5 4.2
orchardgrass avg 0.8 0.7 0.9 2.0 5.7 17.2 16.9
SE 0.1 0.0 0.5 0.7 1.7 3.5 2.9
perennial ryegrass avg 45.2 43.9 48.3 46.9 45.6 46.0 36.3
SE 3.5 5.9 2.5 2.4 0.8 4.2 2.2
quackgrass avg 19.2 20.3 15.3 10.2 7.5 3.8 12.2
SE 1.9 2.0 3.3 2.4 2.4 1.6 2.1
alfalfa avg 8.0 10.9 3.2 5.7 1.6 5.0 5.1
SE 7.6 9.3 3.0 4.2 1.3 3.9 3.5
white clover avg 8.6 4.2 16.9 14.1 19.0 7.5 8.5
SE 1.5 0.5 2.7 3.9 3.5 2.6 3.2
dandelion avg 2.5 3.0 1.2 3.6 2.2 4.1 6.2
SE 1.0 0.5 0.2 0.2 0.3 0.4 1.6
curly dock avg 3.3 2.7 1.0 3.3 1.4 1.1 0.7
SE 1.1 0.9 0.3 0.9 0.4 0.3 0.2
Type
forage legume avg 16.9 22.0 26.4 26.9 26.0 23.1 14.4
SE 6.2 6.9 2.9 2.9 2.9 2.6 5.5
forage grass avg 69.5 69.1 69.4 63.2 68.3 70.8 78.3
SE 4.1 6.0 1.9 3.7 2.6 2.7 4.8
palatable weeds avg 8.1 4.8 3.0 6.2 3.8 4.7 6.3
SE 1.6 0.4 1.1 1.4 0.6 0.3 1.6
undesirable weeds avg 5.5 4.1 1.2 3.7 1.9 1.3 1.0
SE 1.1 1.2 0.3 0.7 0.5 0.3 0.3

 

205

Table B.4. Contribution of major pasture plant species to botanical composition of the
CS-only treatment

 

 

Species Date 5/26/03 8/18/03 6/18/04 8/27/04 5/24/05 7/18/05 10/11/05
Kentucky bluegrass avg 0.6 1.0 1.7 2.1 2.2 2.0 2.2
SE 0.2 0.5 1.1 0.5 1.0 0.4 0.4
orchardgrass avg 0.3 2.1 . 2.4 3.4 4.8 10.4 8.2
SE 0.0 1.0 1.5 2.1 2.2 1.8 0.7
perennial ryegrass avg 50.8 50.7 53.8 50.9 51.7 51.8 45.4
SE 2.7 2.9 1.8 2.2 3.8 3.3 3.3
quackgrass avg 12.0 10.3 9.9 7.9 8.7 5.8 11.0
SE 2.0 1.4 0.9 1.4 2.6 1.9 1.8
alfalfa avg 15.1 18.6 8.9 8.6 7.0 9.6 7.5
SE 4.9 5.8 2.9 2.9 2.4 3.2 2.3
white clover avg 11.4 7.3 16.3 13.6 16.3 8.6 7.1
SE 2.3 2.5 1.6 2.8 2.3 1.5 2.4
dandelion avg 3.4 7.0 3.1 6.9 4.4 6.2 12.4
SE 1.4 2.3 0.6 1.0 0.6 0.6 2.2
curly dock avg 1.7 0.9 0.9 1.1 0.6 0.7 0.9
SE 0.6 0.6 0.2 0.5 0.3 0.2 0.4
Type
forage legume avg 26.8 28.2 27.2 25.5 26.0 21.8 17.7
SE 2.7 4.7 3.0 3.5 2.1 2.3 2.5
forage grass avg 63.2 62.9 67.3 64.5 67.3 70.5 68.5
SE 1.8 3.6 2.2 3.5 1.9 2.2 2.6
palatable weeds avg 6.1 7.2 4.2 8.0 5.6 6.6 12.5
SE 1.3 2.4 0.8 0.9 0.6 0.6 2.2
undesirable weeds avg 3.9 1.7 1.3 2.0 1.0 1.0 1.3
SE 0.8 0.8 0.3 0.6 0.4 0.3 0.6

 

206

Table B.5. Contribution of major pasture plant species to botanical composition of the
cool seasonponion of the CS-SG treatment.

 

 

Species/type Date 5/26/03 8/18/03 6/18/04 8/27/04 5/24/05 7/18/05 10/11/05

Kentucky bluegrass avg 1.3 1.5 1.1 2.1 1.8 4.6 3.9
SE 0.8 0.4 0.4 0.5 0.9 1.7 0.8

orchardgrass avg 0.9 0.3 0.7 2.2 2.9 6.3 9.8
SE - 0.1 0.4 0.7 1.4 1.4 1.8
perennial ryegrass avg 53.5 55.5 53.7 54.2 56.1 56.3 50.3
SE 7.5 6.6 6.7 7.2 3.8 3.2 3.9

quackgrass avg 10.5 8.9 10.1 6.6 8.3 3.6 6.3
SE 3.1 1.5 2.7 1.8 2.6 1.2 1.7

alfalfa avg 13.3 13.3 6.1 5.4 4.8 8.5 4.7
SE 6.3 6.1 3.0 2.7 2.7 4.5 2.4
white clover avg 10.3 10.3 19.9 17.2 16.7 9.2 10.9
SE 1.3 2.2 3.8 4.4 2.2 1.7 3.7
dandelion avg 3.4 4.1 2.8 5.5 3.4 5.7 10.2
SE 1.1 0.8 0.6 0.8 0.6 1.8 1.4

curly dock avg 1.0 1.2 0.2 1.3 1.0 0.5 0.6
SE 0.3 0.4 0.1 0.4 0.1 0.2 0.2
- forage legume avg 23.8 25.9 29.1 25.7 24.5 21.4 17.2
SE 5.2 5.4 4.7 5.4 2.4 1.1 4.6
forage grass avg 65.3 65.9 65.6 65.6 69.2 71.0 71.3
- SE 5.4 5.2 4.4 5.4 2.9 2.8 5.1
palatable weeds avg 7.9 4.6 6.3 6.3 4.9 6.1 10.4
SE 0.8 0.9 0.5 0.9 0.7 2.0 1.4

undesirable weeds avg 3.0 3.6 2.5 2.5 1.4 1.4 1.2
SE 0.5 0.8 0.3 0.6 0.3 0.4 0.5

total avg 100 100 100 100 100 100 100

 

207

Figure B.5. Crop percentage of big bluestem and switchgrass pasturesfrom the CS-BBS
and CS-SG treatments, respectively.

 

 

 

 

 

 

 

 

 

 

100
95 LLL-LLL .-__L.--L ..._L_.LL_ LL ...L L L L LL _
E 90 _.LL- L- ”f":- -L L.. L..-.L LL .L. L
g 85 L1,... -1 _- _. .. W - _ - 1--...-1 ,
1.. + big bluestem
U 80 LL. ...L-L.- r L- LL LL L L L1 . 1
*5 —l— swrtchgrass
Q) 75 .L-. LL- -L....___. . _ .LLLL- - .L LL L L -.'-.-LL L L L. - - W
2 ..
a: 70 L- LL .._-__....-_. ...-......L L- .. L. L-. -L L. L L L _ L
65 - LLL L L.- .L.-. LL . .L-. L. __..L L L L
60 1 1 1 . 1
”b ”5 ”5 ”5 ”b b. b. b: ‘9 ‘9 ‘3
\9 1,)“ $9 \° \9 9 g9 ()9 \9 \9 \9
ﬁsdﬁﬁiyfowee *3

 

Date

 

208

LITERATURE CITED

De Bruijn, S.L., and E. W. Bork. 2006. Biological control of Canada thistle in temperate

pastures using high density rotational cattle grazing. Biological Control 36:305-
315.

Hill, GM. 2004. Designating 'testers' in the put-and-take method adjusting grazing
pressure, pp. Telephone conversation, In D. J. Hudson, (ed.), Hickory Comers,
Ml.

KBS-LTER. 2005. The Kellogg Biological Station Long-Term Ecological Research
Climate Database. Michigan State University Board of Trustees.

Stern, M.D., and M.I. Endres. 1991. Laboratory manual: research techniques in ruminant
nutrition University of Minnesota, Department of Animal Science.

- OTHER RESOURCES USED

USDA. No date. Plant Search [Online]. http:ﬂﬂantsusda.gov/java/nameSearch (veriﬁed
05/23/08)

MDNR. No date. Plant species: rough cinquefoil (potentilla norvegica). [Online].
http://www.michigan.gov/dnr/0,l607,7-153-10370 12146 12213-36280--
00.html (veriﬁed 05/23/08). '

 

209

Appendix 111: Technical Notes

Technical Note I:

Testers were assigned differently in each year, butwithin each year they were assigned in
the same manner for each treatment.

2003: Testers were not assigned at the beginning of the season. At the beginning of the
grazing season all animals were put on their respective pasture as described in Chapter 3.
As the grazing season progressed and pasture growth declined and the animals increased
in size, the smaller animals were removed as indicated by pasture condition indicators.
Steers that remained on the trial all summer were designated ‘testers’. This minimized
social disturbance resulting from taking a more dominant animal from the pastures.
2004: The poor gainers were taken off ﬁrst. Some of the low gains seemed anomalous
and it was believed that they would severely, artificially, and adversely affect the mean
ADG for the replication. The steers used in this experiment were from an auction and
were presumably of the Holstein breed. While most of them appeared to be Holsteins
and were relatively uniform in size, there seemed to be some variability in age and
previous treatment. To avoid biasing the averages, we continued to take the lesser
gainers off, assigning them ‘put-and-take’ status in order to avoid confounding the
averages. Hill (2004) indicated that using the top number of animals (the same proportion
of lower gainers removed for each before the average is calculated) is defensible in order
to remove bias against a particular system.

2005: Testers were assigned beforehand, with the understanding that animals that turned

out to be anomalous could be re-classiﬁed (i.e., ‘tester’ or ‘put-and-take’) if necessary.

210

Technical Note 2:

On one occasion, one or more forage samples were dried at a different research farm

where the plant tissue oven-driers were set at 60°C. (rather than 43°C) for a minimum of

48 hours and then weighed.

Technical Note 3:

On one occasion 80 rising-plate meter readings were recorded taken rather than 100; on

all other occasions that the rising plate meter was used, deviation from the protocol of

collecting 100 rising-plate meter readings were small and associated with worker error

Technical Note 4:

In house modiﬁcations of the Modiﬁed Van Soest method described in the laboratory

manual of the Department of Animal Science, University of Minnesota (Stern, 1991)

were used to determine neutral detergent ﬁber and acid detergent ﬁber concentrations:

0 263.0 g of EDTA was used rather than 236.0 g.

0 Acid detergent ﬁber analysis: ﬁber samples were each rinsed once with acetone rather
than twice.

0 Weighing ash: after placing crucibles in the muffle furnace for ﬁve hours at 500 °C,
the crucibles were allowed to cool and then placed in drying oven at 105 °C for a

minimum of 12 hours and then weighed.

211

LITERATURE CITED

Hill, G.M. 2004. Designating 'testers' in the put-and-take method adjusting grazing
pressure, pp. Telephone conversation, In D. J. Hudson, (ed.), Hickory Comers,
MI.

Stern, M.D., and MJ. Endres. 1991. Laboratory manual: research techniques in ruminant
nutrition University of Minnesota, Department of Animal Science.

212

    

 
 

«111111111111: