FORAGE QUALITY, YIELD, CONDENSED TANNIN CONCENTRATION, SOIL
RESPIRATION, AND ROOT MORPHOLOGY OF BIRDSFOOT TREFOIL-TALL FESCUE
MIXTURES
By
Molly Suzanne Kreykes

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Crop and Soil Sciences ̶̶ Master of Science
2017

ABSTRACT
FORAGE QUALITY, YIELD, CONDENSED TANNIN CONCENTRATION, SOIL
RESPIRATION, AND ROOT MORPHOLOGY OF BIRDSFOOT TREFOIL-TALL FESCUE
MIXTURES
By
Molly Suzanne Kreykes
Growing birdsfoot trefoil (Lotus corniculatus L., BFT) in mixtures with tall fescue [Schedonorus
arundinaceus (Schreb.) Dumort] may help livestock neutralize toxic alkaloids in endophyteinfected tall fescue, but little is known about condensed tannin (CT) concentrations of such
mixtures. Therefore, I evaluated forage yield and quality, CT concentration, soil microbial
activity, and root morphology of BFT-tall fescue mixtures grazed by sheep in Lansing, MI.
Mixtures contained eight BFT cultivars with a range of CT concentrations and four tall fescue
cultivars varying in endophyte infection status. The BFT proportion of forage mixtures declined
from > 95% to < 10% from 2015-2016 due to reduced stand density of BFT. Mixtures with
Oberhaunstädter BFT expressed greater CT concentrations, with levels from 7.5 to 31.7 g kg -1,
and 70% greater stand density in 2016 than all other mixtures. Nutritive value and pregrazing
forage yield was always adequate for ewes at maintenance, regardless of CT concentration or
endophyte type. Therefore, high-tannin Oberhaunstädter in mixtures with endophyte-free tall
fescue proves most suitable for forage production in south-central Michigan. Belowground
processes were influenced more by variability between grazing seasons than by CT or endophyte
type; cumulative C mineralization was greater in 2015 whereas root traits associated with
superior resource acquisition (root length density, specific root length, root surface area density)
were greater in 2016. This provides insight for future studies to further evaluate the effects of
secondary compounds in forage crops on belowground dynamics.

ACKNOWLEDGEMENTS

I would like to my express my gratitude and respect to my advisor, Dr. Kimberly Cassida,
for her guidance, support, and patience over the last two years. I would also like to thank my
committee members, Dr. Philip Robertson and Dr. Richard Ehrhardt, for their time and expertise.
Thank you to the members of the forage lab group for their help on all aspect of this
research. I would especially like to thank my lab technician, Joe Paling. He has always been an
advocate for my success and a source of encouragement; this project could not have been
completed without him. Gratitude is also expressed to Dr. Edzard van Santen for his help with
data analysis. I greatly appreciate the time and energy he dedicated towards my development as
a data analyst and researcher.
I feel very lucky to have been a part of the department of Plant, Soil, and Microbial
Sciences and would like to thank all of the faculty and staff members for their support during my
time at Michigan State University. I would also like to thank the MSU Agronomy Farm for the
use of their equipment and facilities and to the MSU Sheep Teaching and Research Center for
the use of the animals used in this study.
A special thank you to my family for their unending love and support during this process.
I also extend my thanks to all of my friends and loved ones who have continuously supported
and encouraged me to pursue my dreams.

iii

TABLE OF CONTENTS

LIST OF TABLES .......................................................................................................................vi
LIST OF FIGURES .....................................................................................................................viii
KEY TO ABBREVIATIONS ......................................................................................................ix
CHAPTER 1 ................................................................................................................................
LITERATURE REVIEW ............................................................................................................1
BIRDSFOOT TREFOIL ..................................................................................................1
Adaptation and morphology of birdsfoot trefoil ..................................................1
Productivity and establishment of birdsfoot trefoil .............................................2
Benefits of condensed tannins in birdsfoot trefoil ...............................................5
TALL FESCUE ...............................................................................................................8
The endophyte-plant relationship ........................................................................8
Strategies for managing the tall fescue-endophyte symbiosis .............................11
Morphology and agronomic value of tall fescue .................................................13
BIRDSFOOT TREFOIL-TALL FESCUE MIXTURES .................................................14
Benefits of birdsfoot trefoil-tall fescue mixtures .................................................14
Effects of birdsfoot trefoil-tall fescue mixtures on belowground dynamics........15
LITERATURE CITED ....................................................................................................19
CHAPTER 2 ................................................................................................................................
FORAGE QUALITY, YIELD, CONDENSED TANNIN CONCENTRATION, AND
PERSISTENCE OF BIRDSFOOT TREFOIL-TALL FESCUE MIXTURES UNDER GRAZING
IN MICHIGAN ............................................................................................................................28
ABSTRACT .....................................................................................................................28
INTRODUCTION ...........................................................................................................29
MATERIALS AND METHODS .....................................................................................33
Site description.....................................................................................................33
Experimental design.............................................................................................33
Plot establishment ................................................................................................34
Grazing periods ....................................................................................................35
Botanical composition .........................................................................................35
Birdsfoot trefoil stand density ..............................................................................36
Condensed tannin and forage quality analyses ....................................................36
Forage yield and herbage utilization ....................................................................37
Grazing preference ...............................................................................................37
Statistics ...............................................................................................................38
RESULTS ........................................................................................................................39
Weather ................................................................................................................39
Botanical composition .........................................................................................39
Birdsfoot trefoil stand density ..............................................................................41
Condensed tannin concentration ..........................................................................41
iv

Forage quality ......................................................................................................42
Forage yield and herbage utilization ....................................................................44
Grazing preference ...............................................................................................46
DISCUSSION ..................................................................................................................46
Botanical composition and persistence ................................................................47
Condensed tannin concentration ..........................................................................48
Forage quality ......................................................................................................52
Forage yield and herbage utilization ....................................................................54
Grazing preference ...............................................................................................56
CONCLUSIONS..............................................................................................................56
APPENDIX ......................................................................................................................58
LITERATURE CITED ....................................................................................................80
CHAPTER 3 ................................................................................................................................
SOIL RESPIRATION AND ROOT MORPHOLOGY OF BIRDSFOOT TREFOIL-TALL
FESCUE MIXTURES .................................................................................................................87
ABSTRACT .....................................................................................................................87
INTRODUCTION ...........................................................................................................88
MATERIALS AND METHODS ....................................................................................91
Experimental design.............................................................................................91
Soil sampling and analyses ..................................................................................91
Soil respiration .....................................................................................................92
Root sampling and processing .............................................................................93
Root analyses .......................................................................................................94
Statistics ...............................................................................................................95
RESULTS ........................................................................................................................96
Morphological components .................................................................................96
Soil respiration .....................................................................................................97
Root morphology .................................................................................................97
DISCUSSION ..................................................................................................................100
Soil respiration .....................................................................................................100
Root morphology .................................................................................................102
CONCLUSIONS..............................................................................................................106
APPENDIX ......................................................................................................................108
LITERATURE CITED ....................................................................................................121

v

LIST OF TABLES

Table 2.1. Grazing start and end date, grazing duration, rest period, grazing animals, and
stocking density (animal mass per unit land) of birdsfoot trefoil-tall fescue mixtures over eight
grazing periods from 2015-2016 in Lansing, MI .........................................................................59
Table 2.2. Analysis of variance for botanical composition, forage biomass, and grazing
preference of birdsfoot trefoil (BFT)-tall fescue (TF) mixtures over eight grazing periods (GP)
from 2015-2016 in Lansing, MI...................................................................................................60
Table 2.3. Proportion of weed biomass in birdsfoot trefoil (BFT)-tall fescue (TF) mixtures prior
to the November 2015 grazing period (GP) and all GPs in 2016 in Lansing, MI .......................61
Table 2.4. Botanical composition of birdsfoot trefoil (BFT)-tall fescue (TF) mixtures prior to the
November 2015 grazing period (GP) and all GPs in 2016 in Lansing, MI .................................62
Table 2.5. Spring stand density of birdsfoot trefoil (BFT) in mixtures with tall fescue in April
2015, 2016, and 2017 in Lansing, MI ..........................................................................................63
Table 2.6. Analysis of variance for condensed tannin (CT) concentrations and nutritive value of
birdsfoot trefoil (BFT)-tall fescue (TF) mixtures over eight grazing periods (GP) from 2015-2016
in Lansing, MI..............................................................................................................................64
Table 2.7. Condensed tannin concentrations of birdsfoot trefoil (BFT)-tall fescue mixtures prior
to each grazing period in 2015 and 2016 in Lansing, MI. ...........................................................65
Table 2.8. Condensed tannin (CT) concentrations and nutritive value averaged across all
birdsfoot trefoil-tall fescue mixtures prior to each grazing period in 2015 and 2016 in Lansing,
MI. ................................................................................................................................................66
Table 2.9. Crude protein (CP) and rumen undegradable protein (RUP) concentrations of
birdsfoot trefoil (BFT)-tall fescue mixtures prior to each grazing period in 2015 and 2016 in
Lansing, MI ..................................................................................................................................67
Table 2.10. Neutral detergent fiber (NDF) and acid detergent fiber (ADF) concentrations of the
birdsfoot trefoil (BFT)-tall fescue mixtures prior to each grazing period in 2015 and 2016 in
Lansing, MI ..................................................................................................................................69
Table 2.11. Neutral detergent fiber digestibility (NDFD) and lignin concentrations of birdsfoot
trefoil (BFT)-tall fescue mixtures averaged over eight grazing periods from 2015-2016 in
Lansing, MI ..................................................................................................................................71
Table 2.12. Relationship between forage dry weight (y) in Mg ha-1 and forage height (x) obtained
from rising plater meter (RPM) readings using the linear model y=a+bx for each grazing period
in 2015 and 2016 in Lansing, MI .................................................................................................72

vi

Table 2.13. Pregrazing forage yield of birdsfoot trefoil (BFT)-tall fescue (TF) mixtures prior to
the June and November 2015 grazing periods (GPs) and all GPs in 2016 in Lansing, MI .........73
Table 2.14. Herbage utilization of birdsfoot trefoil (BFT)-tall fescue (TF) mixtures prior to the
June and November 2015 grazing periods (GPs) and all GPs in 2016 in Lansing, MI ...............75
Table 2.15. Sheep preference of birdsfoot trefoil (BFT) and tall fescue (TF) in mixed plots after
each grazing period (GP) in 2016 in Lansing, MI .......................................................................77
Table 3.1. Condensed tannin (CT) concentrations of birdsfoot trefoil (BFT)-tall fescue mixtures
averaged over two years (2015-2016) in Lansing, MI .................................................................109
Table 3.2. Analysis of variance for soil respiration and root traits of birdsfoot trefoil (BFT)-tall
fescue (TF) mixtures in 2015 and 2016 in Lansing, MI ..............................................................110
Table 3.3. Pearson correlation coefficients between morphological components and all variables
for birdsfoot trefoil-tall fescue mixtures ......................................................................................111
Table 3.4. Cumulative CO2 respiration over the 32-d incubation for birdsfoot trefoil (BFT)-tall
fescue mixtures averaged over two years (2015-2016) in Lansing, MI ......................................112
Table 3.5. Root length density of birdsfoot trefoil (BFT)-tall fescue mixtures in 2015 and 2016 in
Lansing, MI ..................................................................................................................................113
Table 3.6. Root traits averaged across all birdsfoot trefoil-tall fescue mixtures in 2015 and 2016
in Lansing, MI..............................................................................................................................114
Table 3.7. Root traits of birdsfoot trefoil (BFT)-tall fescue (TF) mixtures averaged over two
years (2015-2016) in Lansing, MI ...............................................................................................115
Table 3.8. Specific root length and root dry weight for birdsfoot trefoil-tall fescue (TF) mixtures
averaged over two years (2015 and 2016) in Lansing, MI ..........................................................116
Table 3.9. Analysis of variance for relative diameter class length across 35 diameter classes for
birdsfoot trefoil (BFT)-tall fescue (TF) mixtures in 2015 and 2016 in Lansing, MI...................117

vii

LIST OF FIGURES

Figure 2.1. (A) Monthly minimum, maximum and average temperature (°C) in Lansing, MI from
2015-2016 with the 30-year average for comparison. (B) Total monthly precipitation (mm) in
Lansing, MI from 2015-2016 with the 30-year average for comparison .....................................78
Figure 2.2. Proportion of tall fescue (TF) biomass (% of dry matter) in birdsfoot trefoil-TF
mixtures prior to the November 2015 grazing period (GP) and all GPs in 2016 in Lansing, MI
(least square means ± SE). Linear contrasts are based on TF endophyte type, and values indicate
numeric differences between contrast groups. Endophyte-infected (E+) TF in linear contrasts
included KY31, endophyte-free (E-) included KY32 and Martin II, and novel endophyte-infected
included Martin II Protek. When no asterisks are present, a significant endophyte effect was not
observed (P > 0.05) ......................................................................................................................79
Figure 3.1. Botanical composition (% of dry matter) of birdsfoot trefoil-tall fescue mixtures in
2015 and 2016 in Lansing, MI (least square means ± SE). Values within each species with
different lowercase letters differ by the PDIFF option in the LSMEANS statement (P < 0.10)
(SAS Institute, Cary, NC) (SAS Institute, Cary, NC) ..................................................................118
Figure 3.2. Cumulative C mineralization over the 32-d incubation of birdsfoot trefoil-tall fescue
mixtures in 2015 and 2016 in Lansing, MI (least square means ± SE). Values with different
lowercase letters differ by the PDIFF option in the LSMEANS statement (P < 0.10) (SAS
Institute, Cary, NC) ......................................................................................................................119
Figure 3.3. Relative diameter class length (rDCL) distribution across 35 diameter classes as a
proportion of total root length of birdsfoot trefoil-tall fescue mixtures in 2015 and 2016 in
Lansing, MI (least square means ± SE). Values within each diameter class with different
lowercase letters differ by the PDIFF option in the LSMEANS statement (P < 0.10) (SAS
Institute, Cary, NC). .....................................................................................................................120

viii

KEY TO ABBREVIATIONS
ADF – Acid Detergent Fiber
BFT – Birdsfoot Trefoil
CP – Crude Protein
CT – Condensed Tannin
DM – Dry Matter
GDD – Growing Degree Days
GP – Grazing Period
NDF – Neutral Detergent Fiber
NDFD – In Vitro Neutral Detergent Fiber Digestibility
NIRS – Near Infrared Reflectance Spectroscopy
RLD – Root Length Density
RPM – Rising Plate Meter
RSAD – Root Surface Area Density
RUP – Rumen Undegradable Protein
RVD – Root Volume Density
SRL – Specific Root Length
TF – Tall Fescue
WFPS – Water-Filled Pore Space

ix

CHAPTER 1
LITERATURE REVIEW
BIRDSFOOT TREFOIL
Adaptation and morphology of birdsfoot trefoil
Birdsfoot trefoil (Lotus corniculatus L., BFT) is a perennial forage legume in the genus
Lotus, a highly diversified group consisting of more than 180 species (Díaz et al., 2005). Native
to the Mediterranean basis, BFT is considered to have the greatest agronomic importance and
widest distribution of all the cultivated Lotus species (Díaz et al., 2005; Seaney and Henson,
1970). Main regions where Lotus species, predominately BFT, are grown for forage production
include South America, North America, and Europe (Díaz et al., 2005). Birdsfoot trefoil was
likely introduced into the United States as seed contamination from Europe and is now used
mainly for pasture and hay production in north-central and northeastern regions with temperate
and humid climates (Díaz et al., 2005; Seaney and Henson, 1970).
Birdsfoot trefoil exhibits a high degree of genetic variation with phenotypical differences
observed for size, shape, color, pubescence, and growth habit among cultivars (Duke, 1981). For
growth habit, plants are grouped into two main categories: prostrate or upright. Cultivars with
prostrate stems are more winter hardy and better utilized for grazing, whereas cultivars with
upright stems tend to be used for hay production (Barnes et al., 2003). Cultivars with upright
stems also tend to produce more palatable forage compared to cultivars with prostrate growth
(McGraw et al., 1989).
Birdsfoot trefoil has pentafoliate leaves with three leaflets attached to the end of the
petiole and two attached at the base (Barnes et al., 2003; Undersander et al., 1993). They range in
shape from round to oblanceolate and are attached on opposite sides of the stem (Seaney and

1

Henson, 1970). Plants require a minimum daylight length of 14 hours to flower and set seed
(MacAdam et al., 2006), producing inflorescences with two to eight flowers ranging in color
from light to dark yellow, occasionally with either orange or red stripes (Seaney and Henson,
1970; Undersander et al., 1993). Each flower produces a cluster of seed pods that resemble a
bird’s ̶̶foot, ̶̶giving ̶̶rise ̶̶to ̶̶the common name (Seaney and Henson, 1970; Undersander et al.,
1993). As pods mature, they turn dark brown and shatter once ripe, dispersing 15-20 seeds per
pod (Seaney and Henson, 1970). Birdsfoot trefoil also has self-reseeding characteristics,
therefore stands can be maintained naturally if plants are allowed to set seed (Blumenthal and
McGraw, 1999; MacAdam et al., 2006)
In regards to root morphology, BFT has a strong, well-developed taproot with numerous
lateral branches (Barnes et al., 2003; Seaney and Henson, 1970). It has a relatively shallow
rooting depth compared to other forage legumes, rooting to an average 1 m whereas alfalfa
(Medicago sativa L.) can reach up to 1.5 m (MacDonald, 1946). Therefore, BFT is more
persistent on shallow soils and can tolerate flooding and soil heaving better than alfalfa, but tends
to be less drought-tolerant than other forage legumes (MacAdam et al., 2006; Seaney and
Henson, 1970).

Productivity and establishment of birdsfoot trefoil
Birdsfoot trefoil is suitable for many soil types and will grow better than other forage
legumes on infertile, acidic or poorly drained soils (Duke, 1981; Seaney and Henson, 1970).
However when grown on fertile land, other forages such as alfalfa will often outperform BFT.
Grabber et al. (2014) compared forage yield between alfalfa and 14 BFT cultivars at four
locations throughout the United States, reporting total seasonal dry matter yield for alfalfa was

2

about 1.5-times greater than the highest performing BFT cultivars under all environments.
Therefore, forage producers in the United States tends to grow alfalfa in preference to BFT on
their fertile soils (Blumenthal and McGraw, 1999; MacAdam et al., 2006).
Variability in yield also exists based on the geographic origin of BFT cultivars. In the
United Kingdom, Marley et al. (2006) reported European ̶̶cultivars ̶̶(‘Oberhaunstädter’ ̶̶and
‘Lotar’) had greater yields after two harvest years than BFT cultivars from other geographic
origins. Similarly, in the United States Grabber et al. (2014) reported greatest yield and plant
vigor for European BFT cultivars, ̶̶specifically ̶̶‘Lotar’, ̶̶while ̶̶McGraw et al. (1989) reported BFT
cultivars originating from the United States and Canada yielded more than cultivars from other
geographic origins. These studies suggest that geographic origin of BFT is an important factor
affecting productivity and competitiveness across different growing environments.
In North America, BFT is a minor forage crop because it is challenging to grow, with
slow, unreliable establishment and poor persistence (MacAdam et al., 2006). In Canada,
Chapman et al. (2008) reported BFT exhibited slower establishment and was a weaker
competitor compared to alfalfa or forage chicory (Cichorium intybus L.), resulting in excessive
weed growth during the establishment year. This in turn reduced plant vigor and caused a high
level of winter kill, which further reduced the persistence of BFT stands (Chapman et al., 2008).
Similarly, Sleugh et al. (2000) reported BFT-grass mixtures had the most significant yield
reduction compared to other binary legume-grass mixtures, attributing this to reduced BFT vigor
over time.
Variability in seeding rates is one factor that contributes to unreliable BFT establishment.
Recommended seeding rates vary, ranging from 1 to 12 kg ha -1, because BFT plants have hard,
relatively small seeds (Blumenthal and McGraw, 1999). Root and crown disease is another factor

3

restricting the persistence of BFT stands. Comparing BFT and alfalfa, Nelson and Smith (1968)
reported that the dry weight increase for BFT throughout the growing season was mainly from
top growth production whereas alfalfa also increased in root and crown weight. In their
experiment, lack of root growth in BFT plants was caused by root and crown rotting diseases,
resulting in reduced persistence and productivity of BFT stands (Nelson and Smith, 1968).
Fusarium wilt disease is particularly problematic, especially in warm and humid climates
(MacAdam et al., 2006). One way to overcome this is to plant BFT cultivars resistant to the root
disease. In a greenhouse trial, Zeiders and Hill (1988) were able to significantly improve
resistance to Fusarium wilt and crown rot in two BFT populations using recurrent phenotypic
selection. More recently, Miller-Garvin et al. (2011) improved resistance of BFT populations by
35% in the greenhouse, which were then tested under field conditions and again expressed
greater yield and longer persistence than other cultivars (Miller-Garvin et al., 2011).
Low midseason carbohydrate reserves are the final major factor contributing to poor
persistence of BFT plants. Compared to forage legumes such as alfalfa that undergo a cyclical
trend in carbohydrates reserves, BFT plants store small amounts of reserves in their roots during
the summer (Barnes et al., 2003; Undersander et al., 1993). Therefore, grazing management is
critical when using BFT as a forage crop in pastures. Alison and Hoveland (1989) reported
greater midseason carbohydrate levels when forage stubble was left at a 10 cm height compared
to 3 cm. This is because BFT regrows from the axillary buds positioned high on the stem (Nelson
and Smith, 1968). Therefore, BFT is better adapted to situation in which defoliation is frequent
but not low, such as grazed pastures, to maintain a higher amount of residual leaf area and active
axillary buds needed for successful regrowth (Blumenthal and McGraw, 1999; Nelson and
Smith, 1968).

4

Benefits of condensed tannins in birdsfoot trefoil
Birdsfoot trefoil is a high quality forage capable of supporting livestock performance
superior to alfalfa and other forage legumes in part due to the presence of condensed tannins
(CTs) (Grabber et al., 2014; Waghorn, 2008). Condensed tannins, also known as
proanthocyanidins, are naturally occurring plant secondary compounds that bind to proteins
(Mueller-Harvey, 2006; Waghorn, 2008). The binding potential of CTs impact digestion, the
nutritive value of forages, and anthelmintic properties in ruminant livestock, which improves
animal performance and overall productivity of livestock operations (Waghorn, 2008).
Condensed tannins can be found in numerous forage legumes. In temperate legumes, they
are often restricted to seed coats or flowers, however CTs need to be expressed in the herbage for
maximal benefits; BFT, big trefoil (Lotus pedunculatus Cav.), sainfoin (Onobrychis viciifolia
Scop.) and sulla (Hedysarum coronarium L.) all express tannins in their herbage (Waghorn,
2008). In BFT, CT concentrations vary by multiple factors. Accessions with different geographic
origins have been shown to express concentrations ranging from 0 to 132 g kg-1 DM for 97
diverse accession, as reported by Roberts et al. (1993). Herbage CT content tends to be the
lowest in cultivars originating from North America and Asia Minor, intermediate in European
cultivars, and highest for cultivars originating from the Mediterranean region (Steiner et al.,
2001). Condensed tannin concentrations also vary throughout the growing season. Levels tends
to increase with plant maturity (Cassida et al., 2000) and early-maturing cultivars have been
shown to produce greater concentrations than later-maturing cultivars (Grabber et al., 2014).
Others have reported that concentrations are greater in the spring and early summer than the fall
(Gebrehiwot et al., 2002; Roberts et al., 1993) and that concentrations fluctuate depending on the
growing environment (Grabber et al., 2014). Additionally, CT concentrations vary throughout

5

the plant, with flowers and leaves expressing greater concentrations than stems (Gebrehiwot et
al., 2002; Marshall et al., 2008).
Historically, CTs have ̶̶been ̶̶recognized ̶̶as ̶̶an ̶̶“anti-nutritional” ̶̶factor ̶̶in ̶̶livestock ̶̶
production (McMahon et al., 2000). This is because if present in high concentrations, > 55 g kg-1
DM, tannins can depress feed intake, inhibit rumen fiber digestion, and reduce overall ruminant
livestock performance (Aerts et al., 1999; Min et al., 2003). Others report concentrations
exceeding 40 to 50 g kg-1 DM can cause problems (McMahon et al., 2000). However,
concentration rarely reach these levels in BFT and at low- to moderate-CT concentrations, the
CTs in BFT provide a multitude of nutritional, health, and environmental benefits. These benefits
stem from a pH-dependent reaction between the CTs in BFT and plant proteins (McMahon et al.,
2000). Jones and Mangan (1977) were the first to report the tannin-plant protein complex is
stable and insoluble at pH between 3.5 and 7.0, but dissociates and solubilizes at a pH < 3.5 or
pH > 8.5. This allows CTs to bind to proteins in the rumen and reduce microbial proteolysis,
increasing flow of plant proteins to the lower pH of the abomasum for greater amino acid
absorption in the small intestine (Waghorn et al., 1987; Wang et al., 1996). Waghorn et al.
(1987) reported apparent absorption of essential amino acids in the small intestine was 62%
greater in sheep consuming BFT with 22 g CT kg-1 DM than sheep receiving intraruminal
polyethylene glycol (PEG) infusion, which prevents CTs from binding proteins. Therefore,
improved ruminant performance associated with BFT can be attributed to higher efficiency of
feed utilization due to improved amino acid absorption (Waghorn, 2008; Waghorn et al., 1987).
Additionally, the CTs in BFT are known to prevent bloat in ruminants (Díaz et al., 2005;
McMahon et al., 2000).

6

While the optimal concentration for maximal benefits is still unknown (McMahon et al.,
2000), several authors have suggested that CT concentrations < 50 g kg-1 DM prove valuable
(Aerts et al., 1999; Min et al., 2003; Mueller-Harvey, 2006). Others have reported concentrations
as low as 5 g kg-1 DM can impart health and production benefits (Broderick et al., 2017; Li et al.,
1996).
Aside from the nutritional and health benefits associated with BFT, CTs also have the
potential to reduce the environmental impact associated with ruminant livestock production.
Globally, ruminants are responsible for 28% of the total anthropomorphic methane (CH4)
emissions, producing approximately 80 million tonnes of CH4 annually (Beauchemin et al.,
2008). The protein-binding capability of CTs have been shown to reduce CH4 and ammonia
(NH3) production in the rumen and NH3 excretion in urine (Misselbrook et al., 2005; Waghorn,
2008). Woodward et al. (2004) found the CTs in BFT lowered CH4 in dairy cattle by 13% while
Noviandi et al. (2014) reported increasing BFT proportions in mixtures with tall fescue had
beneficial effects on ruminal fermentation by producing more propionate and less CH4 and NH3N. In addition, Williams et al. (2010) found that BFT reduced in vitro concentration and flow of
NH3-N compared to alfalfa, suggesting the CTs in BFT can increase N utilization and reduce
excretions of nitrogenous waste.
Due to the benefits listed above, livestock often exhibit significant production
improvements when fed BFT-containing diets. In a two-year grazing study, MacAdam et al
(2011) reported cattle average daily gain (ADG) on non-tannin cicer milkvetch (Astragalus cicer
L.), low-tannin BFT, and medium-tannin BFT were 1.18, 1.30, and 1.52 kg d -1, respectively.
Additionally, they reported that during periods of very high productivity, cattle grazing BFT
gained up to 2.3 kg d-1, which is approaching feedlot rates of gain (MacAdam et al., 2011).

7

Yields of milk and true protein were also highest for dairy cows when fed BFT silage with the
highest CT (16 g kg-1 DM) compared to the lowest (8 g kg-1 DM) CT concentrations or alfalfa
(Hymes-Fecht et al., 2013). Similarly, greater microbial protein yield and N utilization has been
observed when dairy cows were fed a total mixed ration with BFT hay compared to alfalfa
(Christensen et al., 2015). Most recently, Broderick et al. (2017) reported that BFT silage with
low levels of CTs (5 g kg-1 DM) resulted in greater milk production and protein efficiency in
dairy cows compared to alfalfa silage or BFT diets with higher CT concentrations. Utilization of
BFT may also increase liveweight gain and wool production in sheep (Ramírez-Restrepo et al.,
2004). Doran-Browne et al. (2015) modelled the use of BFT in Australian lamb operations,
reporting that over a 10-year period, farms grazing BFT were able to generate up to $50 ha-1 in
profit due to increased liveweight gain and fecundity. However, they also noted that profitability
would depend strongly on pasture establishment and persistence of BFT stands.

TALL FESCUE
The endophyte-plant relationship
Tall fescue [Schedonorus arundinaceus (Schreb.) Dumort.] is a perennial, cool-season
grass native to temperate and cool climates throughout Europe, North Africa, and West and
Central Asia (Gibson and Newman, 2001). It is an aggressive, competitive bunchgrass that forms
densely tillered clumps in forage stands (Barnes et al., 2003; Beuselinck et al., 1992). It is also
capable of producing short rhizomes, improving its ability to persist and spread laterally (Barnes
et al., 2003; Porter, 1958).
Tall fescue was introduced in the United States in the late 1800s but did not gain
popularity until the 1940s with the discovery ̶̶of ̶̶‘Kentucky ̶̶31’ ̶̶(KY31) (Ball et al., 1993). In

8

1931, seed from a vigorous tall fescue ecotype was discovered on a farm in Menifee County,
Kentucky; this ecotype was released in 1943 as the original KY31 cultivar (Ball et al., 1993).
The new cultivar quickly gained popularity by farmers in the late 1940s and 1950s due to ease of
establishment, a wide range of adaptation, pest resistance, and a longer grazing season than other
grass cultivars (Stuedemann and Hoveland, 1988). For these reasons, KY31 became the
predominant cool-season forage grass in the United States, occupying an estimated 19 million ha
of pasture (Bouton, 2007).
The majority of the tall fescue grown in the United States, including KY31, is infected
with the endophytic fungus, Neotyphodium coenophialum [(Morgan-Jones and Gams) Glenn,
Bacon, and Hanlin] (Bouton, 2007). In the 1960s, people began to notice animal health and
performance problems associated with the increased use of tall fescue in pastures (Ball et al.,
1993). In the mid-1970s, Bacon et al. (1977) found that pastures in the United States with severe
animal disorders were 100% infected with a fungal endophyte, reporting that the toxicity of tall
fescue was due to infection by Epichloe typhina; the fungal endophyte has since been
reclassified as N. coenophialum.
While endophyte-infected (E+) tall fescue is famed for its superior agronomic
performance, N. coenophialum produces ergot alkaloids that harm animal health and
performance (Bacon et al., 1986; Stuedemann and Hoveland, 1988). Ergovaline, the main ergot
alkaloid produced by the endophyte, is responsible for causing fescue toxicity in livestock
(Bacon et al., 1986). Referred ̶̶to ̶̶as ̶̶“fescue ̶̶toxicosis” or ̶̶“summer ̶̶slump”, symptoms include
reduced weight gain and milk production, reproductive difficulties, gangrene on the extremities,
elevated temperatures, rough hair coat, reduced feed intake, and increased respiration rates
(Bacon et al., 1986; Bouton, 2007; Stuedemann and Hoveland, 1988). Fescue toxicosis has had a

9

significant economic impact, resulting in annual losses estimated to be over $600 million
(Bouton, 2007; Fribourg and Waller, 2005). This issue is particularly problematic in the southern
and western portions of the tall fescue adaptation zone (Malinowski and Belesky, 2006).
The symbiosis between N. coenophialum and tall fescue is considered mutualistic;
mutualism involves an exchange of benefits between partners in which the benefits must
outweigh the costs for the host and the symbiont (Schardl et al., 2004). For the tall fescueendophyte symbiosis, N. coenophialum gains nutrition, shelter, and a means of propagation
while tall fescue gains enhanced ecological fitness for growth and persistence (Belesky and
West, 2009; Schardl et al., 2004).
The lifecycle of N. coenophialum follows the growth and reproduction of tall fescue (Ball
et al., 1993). When a tall fescue seed germinates, the mycelium of the endophyte grows and
moves through the intercellular space between the plant cells, colonizing the lateral buds then the
inflorescences (Ball et al., 1993; Schardl et al., 2004). Within the plant, concentration of the
endophyte varies, with the highest concentrations expressed in the leaf sheaths and seeds,
moderate concentrations in crown and stems, and low concentrations in leaf blades and roots
(Siegel et al., 1984). The fungus reproduces asexually and is transmitted solely by seed (Ball et
al., 1993; Schardl et al., 2004). Transmission is very effective, therefore nearly all seedlings that
germinate contain the fungus (Ball et al., 1993; Schardl et al., 2004). In stored seeds, the fungus
can remain viable for up to one year, after which the fungus dies and seed becomes endophyte
free (E-) (Ball et al., 1993).
Research has reported few differences in productivity between E+ and E- tall fescue in
more northern latitudes (Brink and Casler, 2012), therefore it is not recommended to plant E+
tall fescue in Michigan (Cassida et al., 2016). However, endophyte infection is needed to ensure

10

productivity in regions that face more extreme environmental stresses (Belesky and West, 2009;
Bouton et al., 1993). In a greenhouse study, Arachevaleta et al. (1989) reported E+ plants
subjected to drought stress had 50% greater herbage mass and more rapid regrowth after harvest
compared to E- plants. Similarly, Nagabhyru et al. (2013) reported E+ plants recovered better
than E- plants after exposure to drought stress and accumulated more metabolites early in the
stress exposure. These findings illustrate the necessity of the endophyte to ensure plant
persistence in harsh environments, such as under severe drought.

Strategies for managing the tall fescue-endophyte symbiosis
A variety of strategies are used by producers to overcome the livestock problems
associated with E+ tall fescue. One approach is to avoid grazing during periods when alkaloid
levels are greatest. Ergovaline concentrations tend to peak in the spring due to the formation of
ergovaline-rich seed heads, followed by a decline in the summer then increasing again in the fall
(Rottinghaus et al., 1991). However, Rogers et al. (2011) reported that when tall fescue was kept
in a vegetative stage throughout the growing season, low ergovaline concentrations were
sustained throughout the early spring and summer. The toxic effect of ergot alkaloids can also be
reduced by planting tall fescue in mixtures with other grasses or legumes. Hoveland et al. (1981)
reported ADG in Northern Alabama were greater on ladino white clover (Trifolium repens L.)tall fescue and BTF-tall fescue mixtures than on pure stands of E+ tall fescue during spring
grazing.
Another strategy is to replace toxic E+ tall fescue with cultivars that contain non-ergot
alkaloid producing endophytes, which are non-toxic to livestock. Referred to as novel
endophytes, these endophyte strains are capable of maintaining stress tolerance and improved

11

plant persistence traditionally associated with E+ tall fescue while minimizing detrimental
impacts on livestock productivity (Malinowski and Belesky, 2006). Bouton et al. (2002)
evaluated five non-toxic N. coenophialum strains and found that the best cultivar-strain
combination, Jesup (AR542), had greater yield and stand survival than the E- cultivars; this
combination was later released as Jesup-MaxQ. A number of other novel endophyte-infected
cultivars have since been released with different cultivar-strain combinations.
Both livestock and agronomic advantages and disadvantages have been reported with the
use of novel endophyte-infected tall fescue. Bouton et al. (2002) reported ADG of lambs
consuming novel endophyte-infected tall fescue was 57% greater than gains on E+ tall fescue,
while Hopkins et al. (2010) reported lambs grazing tall fescue infected with four different novel
endophyte strains gained on average more than twice that of lambs grazing KY31. Parish et al.
(2003) reported cattle grazing E- and novel endophyte-infected tall fescue exhibited superior
growth, greater dry matter intake, and lower rectal temperatures (a symptom of fescue toxicosis)
than cattle grazing E+ tall fescue. Johnson et al. (2012) similarly reported steers grazing E- and
two novel endophyte-infected cultivars had lower rectal temperatures and 20% greater daily
gains compared to steers grazing KY31.
However, novel endophyte-infected cultivars may not be as suitable to stressful
environments or marginal fertility compared to E+ cultivars (Malinowski and Belesky, 2006). In
Kentucky, lower carrying capacity has been reported for novel endophyte-infected pastures
compared to KY31 pastures (Johnson et al., 2012). Additionally in Georgia, only one of three
novel endophyte-infected cultivar-strain combinations reached a yield comparable to the E+
version (Bouton et al., 2002). Therefore, careful grazing management and selection of the best

12

novel endophyte strain-cultivar combination is needed to ensure novel endophyte-infected tall
fescue persists throughout the growing season in the southern regions of the United States.

Morphology and agronomic value of tall fescue
Tall fescue has many desirable agronomic characteristics, including the ability to adapt to
a variety of environments (Burns and Chamblee, 1979). It can tolerate acidic soil, drought, heat
stress, periodic flooding, insects, and grazing better than many other cool-season grasses (Barnes
et al., 2003; Belesky and West, 2009; Burns and Chamblee, 1979). However, a common
criticism of tall fescue is its lack of palatability. Leaves have a rough margin with a prominent
midrib, making it tougher than other forage grasses (Barnes et al., 2003). Leep et al. (2002)
reported tall fescue mixtures in Michigan tended to be less palatable than other cool-season
grasses. Their study used E- cultivars, suggesting the lower grazing preference observed with tall
fescue was due to the grass species rather than an endophyte effect (Leep et al., 2002). While tall
fescue is often viewed as a low-quality forage due to poor palatability and issues associated with
the toxic fungal endophyte, it in fact has a relatively high nutritive value (Lacefield et al., 2003).
Similar to other cool-season grasses, most of the growth of tall fescue occurs in the spring
followed by a period of semi-dormancy during the summer months and increased production
during fall and early winter (Burns and Chamblee, 1979). However, it is able to produce more
dry matter in the late summer and fall than other forages because of its wide temperature
adaptation (Wolf et al., 1979).
Due to superior yield and nutritive value, tall fescue can support livestock performance
similar to other cool-season grasses if managed to prevent fescue toxicosis. Schaefer et al. (2014)
reported pastures containing E- tall fescue had greater forage availability than meadow fescue

13

[Schedonorus pratensis (Huds.) P. Beauv] with slightly lower steer ADG. However, tall fescue
pastures had greater productivity, therefore gain per ha were equal for the two grasses (Schaefer
et al., 2014). Additionally, Burns and Fisher (2010) reported that while steers consumed more
orchardgrass (Dactylis glomerata L.) than novel endophyte-infected tall fescue, forage
digestibility was greater for tall fescue, therefore animal daily responses were expected to be
similar for the two grasses.

BIRDSFOOT TREFOIL-TALL FESCUE MIXTURES
Benefits of birdsfoot trefoil-tall fescue mixtures
Grass-legume mixtures are commonly used in pastures to enhance forage production and
forage quality. In such mixtures, grasses with fibrous root systems provide soil organic matter,
protect against soil erosion, and are more resistant to grazing damage than legumes, while
legumes improve neutral detergent fiber (NDF), crude protein (CP), and the seasonal distribution
of forage yield (MacAdam et al., 2006; Sleugh et al., 2000). For example, forage legumes
interseeded into grass pastures have been shown to increase dry matter yield and CP by 94% and
111%, respectively, compared to cool-season grass monocultures (Taylor and Allinson, 1983).
The compatibility of BFT and tall fescue has received increased attention due to the
superior agronomic value and improved livestock performance unique to both forage species.
Birdsfoot trefoil is rarely used as a monoculture due to high plant mortality associated with root
and crown disease, but rather is best when planted with a perennial companion grass, such as tall
fescue, that can utilize the nitrogen (N) fixed by BFT and fill the stand where weeds might
otherwise grow (MacAdam et al., 2006; Undersander et al., 1993). Tall fescue is considered by
some to be the most compatible forage grasses to be planted with BFT in the United States

14

(Beuselinck et al., 1992; Blumenthal and McGraw, 1999). When seeded together, yield and
quality has been shown to be greater than pure tall fescue receiving low levels of N (Beuselinck
et al., 1992). Similarly, Taylor and Allinson (1983) reported BFT-tall fescue mixtures produce
greater yield and CP than other grass-legume mixtures. Birdsfoot trefoil-tall fescue mixtures
have also been shown to improve ruminal fermentation and nutrient utilization more so than
other binary legume-grass mixtures (Noviandi et al., 2014).
In addition, forage mixtures containing CTs may neutralize the toxicity of E+ tall fescue.
Lyman et al. (2012) reported a substantial increase in grazing on both high-alkaloid tall fescue
and high-alkaloid reed canarygrass (Phalaris arundinacea L.) when cattle first grazed either
high-tannin BFT or high-saponin alfalfa. Similarly, lambs supplemented with BFT prior to
feeding tall fescue subsequently consumed more high-alkaloid tall fescue than un-supplemented
lambs or lambs supplemented with alfalfa (Owens et al., 2012). In additions, lambs infused with
a commercial tannin mixture have been shown to graze more high-alkaloid tall fescue than lambs
that were not infused (Lisonbee et al., 2009). These findings suggest that secondary compounds
can influence animal foraging behavior and that CTs may bind to the alkaloids in E+ tall fescue,
helping to neutralize the alkaloid toxicity.

Effect of birdsfoot trefoil-tall fescue mixtures on belowground dynamics
There has been increased interest in understanding how grazing and pasture management
influences soil carbon (C) and fine root dynamics (Franzluebbers and Stuedemann, 2015;
Greenwood and Hutchinson, 1998; Pucheta et al., 2004). However, less is known about how
aboveground secondary compounds in forage crops can influence belowground processes in
pasture soils.

15

There is evidence the endophyte infection in tall fescue may over time influence soil
microbial activity. Soil CO2 respiration, or C mineralization, is an important biological indicator
of soil quality for sustaining productive plant growth (Karlen et al., 1997). Increases in soil
organic C and N concentrations along with a reduction in C mineralization have been reported in
tall fescue pastures with high endophyte infection compared to pastures with low endophyte
infection (Franzluebbers et al., 1999; Franzluebbers and Stuedemann, 2005). Short- and longterm exposure of E+ leaves decomposing in the soil have also been shown to reduce the rate of C
mineralization compared to E- leaves (Franzluebbers and Hill, 2005). These studies suggest that
E+ tall fescue may suppress soil microbial activity, leading to a greater accumulation of soil
organic C and N. However, some studies have reported higher rates of CO2 respiration associated
with E+ compared to E- tall fescue (Van Hecke et al., 2005). Others have also reported that the
aboveground endophytes in tall fescue can alter root exudate composition which in turn may
influence soil processes either directly by altering the inputs to C and N pools or indirectly by
altering the structure of soil microbial communities (Guo et al., 2015). Therefore, it is not clear
exactly how endophyte type influences soil microbial activity, particularly in pasture soils.
Polyphenols may also play a role in nutrient cycling through their interaction with
mineralization processes (Hättenschwiler and Vitousek, 2000). Tannin structures are chemically
very diverse but can be divided into two main classes, condensed and hydrolysable, both of
which are able to bind proteins (Mueller-Harvey, 2006). They have received special attention in
forest ecosystems, as reviewed by Kraus et al. (2003), for their ability to influence belowground
dynamics including litter decomposition, nutrient cycling, N mineralization, and microbial
activity. However, because tannins are complex polyphenols that are chemically heterogeneous,
the consensus as to how tannins interact with belowground processes to effect soil respiration

16

varies in the literature (Fierer et al., 2001; Kraus et al., 2003; Kraus et al., 2004). Some report
that tannins may reduce microbial activity, as reviewed by Scalbert (1991), by inhibiting
extracellular microbial enzymes, depriving substrate needed for microbial growth, or through a
direct action on microbial metabolism. In contrast, others have reported that tannin additions in
forest soils can increase C respiration and decrease N mineralization by acting as a microbial C
source (Kraus et al., 2004). Fierer et al. (2001) similarly reported that additions of low molecular
weight tannins in forest soils can increase respiration by acting as a source of C for microbial
growth while additions of higher molecular weight tannins can prove toxic by binding substrates
needed for microbial growth. Madritch et al. (2007) argues that the inhibitorty action of tannins
is only evident when there is a high demand for N in forest soils, such as when large amounts of
leaf letter are present. Despite evidence that tannins interact with belowground processes in
forest soils, little to no research has been conducted looking at the effect of the CTs in BFT on
soil microbial activity.
Greater root production has been reported for E+ compared to E- tall fescue
(Franzluebbers, 2006). However, this is not consistent across all studies (Van Hecke et al.,
2005). In addition, differences in root biomass have been reported between novel endophyteinfected strain-cultivar combinations (Guo et al., 2015). Unfortunately, solely measuring root
mass does not allow for an accurate interpretation of root function. Fine roots represent up to
95% of the total root system length in many agriculture crops and forest species and are
historically defined as < 2 mm in diameter, however most are actually much smaller (Böhm,
1979; Zobel, 2008). Fine roots are particularly important because they serve as the major
pathway for water and nutrient absorption (Eissenstat, 1992). Unfortunately, the small size of
fine roots makes them challenging to study and characterization of root systems is time-

17

consuming, labor intensive, and often prone to inaccuracy (Costa et al., 2000; Wang and Zhang,
2009; Zobel et al., 2007). Therefore, more research is needed on fine roots to better understand
how root dynamics differ among forage crops and to determine whether secondary compounds
can influence root morphology.

18

LITERATURE CITED

19

LITERATURE CITED

Aerts, R.J., T.N. Barry, and W.C. McNabb. 1999. Polyphenols and agriculture: Beneficial effects
of proanthocyanidins in forages. Agric. Ecosyst. Environ. 75: 1–12.
Alison, M.W., and C.S. Hoveland. 1989. Birdsfoot trefoil management. I. Root growth and
carbohydrate storage. Agron. J. 81: 739–745.
Arachevaleta, M., C.W. Bacon, C.S. Hoveland, and D.E. Radcliffe. 1989. Effect of the tall
fescue endophyte on plant response to environmental stress. Agron. J. 81: 83–90.
Bacon, C.W., P.C. Lyons, J.K. Porter, J.D. Robbins. 1986. Ergot toxicity from endophyteinfected grasses: A review. Agron. J. 78: 106–116.
Bacon, C.W., J.K. Porter, J.D. Robbins, and E.S. Luttrell. 1977. Epichloe typhina from toxic tall
fescue grasses. Appl. Environ. Microbiol. 34: 576–581.
Ball, D.M., J.F. Pedersen, and G.D. Lacefield. 1993. The tall-fescue endophyte. Am. Sci. 81:
370–379.
Barnes, R. F., Nelson, C. F., Collins, M., and K. J., Moore. 2003. Forages. An Introduction To
Grassland Agriculture, 6th ed Iowa State Press, Ames, Iowa.
Beauchemin, ̶̶K.A., ̶̶M. ̶̶Kreuzer, ̶̶F. ̶̶O’Mara, ̶̶and ̶̶T.A. ̶̶McAllister. ̶̶2008. ̶̶Nutritional ̶̶management ̶̶
for enteric methane abatement: A review. Aust. J. Exp. Agric. 48: 21–27.
Belesky, D.P., and C.P. West. 2009. Abiotic stresses and endophyte effects. In: H. A. Fribourg,
D. B. Hannaway, and C. P. West, editors, Tall Fescue for the Twenty-First Century.
Agron. Monogr. 53rd ed. ASA, CSSA, and SSSA, Madison, WI. p. 49–64.
Beuselinck, P.R., D.A. Sleper, S.S. Bughrara, and C.A. Roberts. 1992. Effect of mono and mixed
culture of tall fescue and birdsfoot trefoil on yield and quality. Agron. J. 84: 133–137.
Blumenthal, M.J., and R.L. McGraw. 1999. Lotus adaptation, use, and management. In: P. R.
Beuselinck, editor, Trefoil: The Science and Technology of Lotus. CSSA Special
Publication 28. ASA, CSSA, Madison, WI. p. 97–119.
Böhm, W. 1979. Methods of Studying Root Systems. Springer Verlag, New York.
Bouton, J. 2007. The economic benefits of forage improvement in the United States. Euphytica
154: 263–270.
Bouton, J.H., R.N. Gates, D.P. Belesky, and M. Owsley. 1993. Yield and persistence of tall
fescue in the southeastern coastal plain after removal of its endophyte. Agron. J. 55: 52–55.

20

Bouton, J.H., G. C. M. Latch, N.S. Hill, C.S. Hoveland, M.A. McCann, R.H. Watson, J.A.
Parish, L.L. Hawkins, and F.N. Thompson. 2002. Reinfection of tall fescue with nonergot alkaloid producing endophytes. Agron. J. 94: 567–574.
Brink, G.E., and M.D. Casler. 2012. Yield and nutritive value differences among cool-season
grasses. Forage and Grazinglands.: doi:10.1094/FG-2012-0619-01-RS.
Broderick, G.A., J.H. Grabber, R.E. Muck, and U.C. Hymes-Fecht. 2017. Replacing alfalfa
silage with tannin-containing birdsfoot trefoil silage in total mixed rations for lactating
dairy cows. J. Dairy Sci. 100: 1–15.
Burns, J.C., and D.S. Chamblee. 1979. Adaptation. In: R. C. Bukner and L. P. Bush, editors, Tall
Fescue. Agron. Monogr. 20th ed. ASA, CSSA, and SSSA, Madison, WI. p. 9–30.
Burns, J.C., and D.S. Fisher. 2010. Intake and digestibility of improved selections of tall fescue
and orchardgrass hays. Crop Sci. 50: 419–426.
Cassida, K.A., T.S. Griffin, J. Rodriguez, S.C. Patching, O.B. Hesterman, and S.R. Rust. 2000.
Protein degradability and forage quality in maturing alfalfa, red clover, and birdsfoot
trefoil. Crop Sci. 40: 209–215.
Cassida, K., J. Paling, and C. Kapp. 2016. 2016 Michigan forage variety test report: Forage
factsheet #17-01. http://forage.msu.edu/wp-content/uploads/2014/07/2014-ForageVariety-Test-REPORT_Final.pdf. (accessed 19 April 2017).
Chapman, G., E. Bork, N. Donkor, and R. Hudson. 2008. Forage yield and quality of chicory,
birdsfoot trefoil, and alfalfa during the establishment year. Open Agric. J. 2: 68–74.
Christensen, R.G., S.Y. Yang, J.-S. Eun, A.J. Young, J.O. Hall, and J.W. MacAdam. 2015.
Effects of feeding birdsfoot trefoil hay on neutral detergent fiber digestion, nitrogen
utilization efficiency, and lactational performance by dairy cows. J. Dairy Sci. 98: 7982–
7992.
Costa, C., L.M. Dwyer, R.I. Hamilton, C. Hamel, L. Nantais, and D.L. Smith. 2000. A sampling
method for measurement of large root systems with scanner-based image analysis.
Agron. J. 92: 621–627.
Díaz, P., O. Borsani, and J. Monza. 2005. Lotus-related species and their agronomic importance.
In: A. J. Marquez, editor, Lotus Japonicus Handbook. Springer, Dordrecht, The
Netherlands. p. 25–37.
Doran-Browne, N., R. Behrendt, R. Kingwell, and R. Eckard. 2015. Modelling the potential of
birdsfoot trefoil (Lotus corniculatus) to reduce methane emissions and increase
production on wool and prime lamb farm enterprises. Anim. Prod. Sci. 55: 1097–1105.

21

Duke, J. A.1981. Handbook of Legumes of World Economic Importance. Plenum Press, NY.
p.125-128.
Eissenstat, D.M. 1992. Costs and benefits of constructing roots of small diameter. J. Plant Nutr.
15:763–782.
Fierer, N., J.P. Schimel, R.G. Cates, and J. Zou. 2001. Influence of balsam poplar tannin
fractions on carbon and nitrogen dynamics in Alaskan taiga floodplain soils. Soil Biol.
Biochem. 33: 1827–1839.
Franzluebbers, A.J. 2006. Short-term responses of soil C and N fractions to tall fescue endophyte
infection. Plant Soil 282: 153–164.
Franzluebbers, A.J., N. Nazih, J.A. Stuedemann, J.J. Fuhrmann, H.H. Schomberg, and P.G.
Hartel. 1999. Soil carbon and nitrogen pools under low- and high-endophyte-infected tall
fescue. Soil Sci. Soc. Am. J. 63: 1687–1694.
Franzluebbers, A.J., and J.A. Stuedemann. 2005. Soil carbon and nitrogen pools in response to
tall fescue endophyte infection, fertilization, and cultivar. Soil Sci. Soc. Am. J. 69: 396–
403.
Franzluebbers, A.J., and N.S. Hill. 2005. Soil carbon, nitrogen, and ergot alkaloids with shortand long-term exposure to endophyte-infected and endophyte-free tall fescue. Soil Sci.
Soc. Am. J. 69: 404–412.
Franzluebbers, A.J., and J.A. Stuedemann. 2015. Does grazing of cover crops impact
biologically active soil carbon and nitrogen fractions under inversion or no tillage
management? J. Soil Water Conserv. 70: 365–373.
Fribourg H.A. and J.C. Waller. 2005. Neotyphodium research and application in the USA. In:
C.A. Roberts, C.P. West, and D.E. Spiers, editors, Neotyphodium in Cool-Season
Grasses. Blackwell Publishing, Ames, Iowa. p. 3–22.
Gebrehiwot, L., P.R. Beuselinck, and C.A. Roberts. 2002. Seasonal variations in condensed
tannin concentration of three Lotus species. Agron. J. 94: 1059–1065.
Gibson, D.J., and J.A. Newman. 2001. Festuca arundinacea Schreber (F. elatior L. ssp.
arundinacea (Schreber) Hackel). J. Ecol. 89: 304–324.
Grabber, J.H., H. Riday, K.A. Cassida, T.C. Griggs, D.H. Min, and J.W. MacAdam. 2014. Yield,
morphological characteristics, and chemical composition of European- and
Mediterranean-derived birdsfoot trefoil cultivars grown in the colder continental United
States. Crop Sci. 54: 1893–1901.

22

Greenwood, K.L., and K.J. Hutchinson. 1998. Root characteristics of temperate pasture in New
South Wales after grazing at three stocking rates for 30 years. Grass Forage Sci. 53: 120–
128.
Guo, J., R.L. McCulley, and D.H. McNear. 2015. Tall fescue cultivar and fungal endophyte
combinations influence plant growth and root exudate composition. Front. Plant Sci. 6: 113.
Hättenschwiler, S., and P.M. Vitousek. 2000. The role of polyphenols in terrestrial ecosystems
nutrient cycling. Tree 15: 238–243.
Hopkins, A.A., C.A. Young, D.G. Panaccione, W.R. Simpson, S. Mittal, and J.H. Bouton. 2010.
Agronomic performance and lamb health among several tall fescue novel endophyte
combinations in the south-central USA. Crop Sci. 50: 1552–1561.
Hoveland, C.S., R.R. Harris, E.E. Thomas, E.M. Clark, J.A. McGurie, J.T. Eason, and M.E. Ruf.
1981. Tall fescue with ladino clover or birdsfoot trefoil as pasture for steers in Northern
Alabama. Ala. Agric. Exp. Stn. Bull. No. 515. Ala. Agric. Exp. Stn., Auburn.
Hymes-Fecht, U.C., G.A. Broderick, R.E. Muck, and J.H. Grabber. 2013. Replacing alfalfa or
red clover silage with birdsfoot trefoil silage in total mixed rations increases production
of lactating dairy cows. J. Dairy Sci. 96: 460–469.
Johnson, J.M., G.E. Aiken, T.D. Phillips, M. Barrett, J.L. Klotz, and F.N. Schrick. 2012. Steer
and pasture responses for a novel endophyte tall fescue developed for the upper transition
zone. J. Anim. Sci. 90: 2402–2409.
Jones, W.T., and J.L. Mangan. 1977. Complexes of the condensed tannins of sainfoin
(Onobrychis viciifolia Scop.) with fraction 1 leaf protein and with submaxillary
mucoprotein, and their reversal by polyethylene glycol and pH. J. Sci. Food Agric. 28:
126–136.
Karlen, D.L., M.J. Mausbach, J.W. Doran, R.G. Cline, R.F. Harris, and G.E. Schuman. 1997.
Soil quality: A concept, definition, and framework for evaluation (A guest editorial). Soil
Sci. Soc. Am. J. 61: 4–10.
Kraus, T.E., R.A., Dahlgren, and R.J. Zasoski. 2003. Tannins in nutrient dynamics of forest
ecosystems - a review. Plant Soil 256: 41–66.
Kraus, T.E.C., R.J. Zasoski, R.A. Dahlgren, W.R. Horwath, and C.M. Preston. 2004. Carbon and
nitrogen dynamics in a forest soil amended with purified tannins from different plant
species. Soil Biol. Biochem. 36: 309–321.
Lacefield, G.D., J.C. Henning, and T.D. Phillips. 2003. Tall fescue. Univ. of Kentucky Coop.
Extension Service Pub. AGR-59.

23

Leep, R., P. Jeranyama, D. Min, T. Dietz, S. Bughrara, and J. Isleib. 2002. Grazing effects on
herbage mass and composition in grass – birdsfoot trefoil mixtures. Agron. J. 94: 1257–
1262.
Li, Y-G., G. Tanner, and P. Larkin. 1996. The DMACA–HCl protocol and the threshold
proanthocyanidin content for bloat safety in forage legumes. J. Sci. Food Agric. 70: 89101.
Lisonbee, L.D., J.J. Villalba, and F.D. Provenza. 2009. Effects of tannin on selection by sheep of
forages containing alkaloids, tannins and saponins. J. Sci. Food Agric. 89: 2668–2677.
Lyman, T.D., F.D. Provenza, J.J. Villalba, and R.D. Wiedmeier. 2012. Phytochemical
complementarities among endophyte-infected tall fescue, reed canarygrass, birdsfoot
trefoil and alfalfa affect cattle foraging. Animal 6: 676–682.
MacAdam, J.W., T.C. Griggs, P.R. Beuselinck, and J.H. Grabber. 2006. Birdsfoot trefoil, a
valuable tannin-containing legume for mixed pastures. Forage and Grazinglands.:
doi:10.1094/FG-2006-0912-01-RV.
MacAdam, J.W., R.E. Ward, T.C. Griggs, B.R. Min, and G.E. Aiken. 2011. C and S blood:
Average daily gain acid composition of cattle grazing the nonbloating legumes birdsfoot
trefoil and cicer milkvetch in the Mountain West. Prof. Anim. Sci. 27: 574–583.
MacDonald, H.A. 1946. Birdsfoot trefoil (Lotus corniculatus L.), its characteristics and
potentialities as a forage legume. Cornell University Agric. Exp. Stn. Mem. 261. Cornell
Univ., Ithaca, NY.
Madritch, M.D., L.M. Jordan, and R.L. Lindroth. 2007. Interactive effects of condensed tannin
and cellulose additions on soil respiration. Can. J. For. Res. 37: 2063–2067.
Malinowski, D.P., and D.P. Belesky. 2006. Ecological importance of Neotyphodium spp. grass
endophytes in agroecosystems. Grassl. Sci. 52: 14.
Marley, C.L., R. Fychan, and R. Jones. 2006. Yield, persistency and chemical composition of
Lotus species and varieties (birdsfoot trefoil and greater birdsfoot trefoil) when harvested
for silage in the UK. Grass Forage Sci. 61: 134–145.
Marshall, A., D. Bryant, G. Latypova, B. Hauck, P. Olyott, P. Morris, and M. Robbins. 2008. A
high-throughput method for the quantification of proanthocyanidins in forage crops and
its application in assessing variation in condensed tannin content in breeding programs
for lotus corniculatus and lotus uliginosus. J. Agric. Food Chem. 56: 974–981.
McGraw, R.L., P.R. Beuselinck, and G.C. Marten. 1989. Agronomic and forage quality
attributes of diverse entries of birdsfoot trefoil. Crop Sci. 29: 1160–1164.

24

McMahon, L.R., T.A. McAllister, B.P. Berg, W. Majak, S.N. Acharya, J.D. Popp, B.E.
Coulman, Y. Wang, and K.-J. Cheng. 2000. A review of the effects of forage condensed
tannins on ruminal fermentation and bloat in grazing cattle. Can. J. Plant Sci. 80: 469–
485.
Miller-Garvin, J.E., J.L. Hansen, N.J. Ehlke, Y.A. Papadopoulos, R.R. Smith, G.C. Bergstrom,
M.J. Wunsch, D.W. Kalb, B.P. Tillapaugh, J.L. Crawford, and D.R. Viands. 2011.
Improving birdsfoot trefoil for resistance to Fusarium wilt. Crop Sci. 51: 585–591.
Min B.R, T.N. Barry, G.T. Attwood, and W.C. McNabb. 2003. The effect of condensed tannins
on the nutrition and health of ruminants fed fresh temperate forages: a review. Anim Feed
Sci Technol 106: 3–19.
Misselbrook, T.H., J.M. Powell, G.A. Broderick, and J.H. Grabber. 2005. Dietary manipulation
in dairy cattle: laboratory experiments to assess the influence on ammonia emissions. J.
Dairy Sci. 88: 1765–1777.
Mueller-Harvey, I. 2006. Unravelling the conundrum of tannins in animal nutrition and health. J.
Sci. Food Agric. 86: 2010-2037.
Nagabhyru, P., R.D. Dinkins, C.L. Wood, C.W. Bacon, and C.L. Schardl. 2013. Tall fescue
endophyte effects on tolerance to water-deficit stress. BMC Plant Biol. 13: 1–17.
Nelson, C.J., and D. Smith. 1968. Growth of birdsfoot trefoil and alfalfa. II. Morphological
development and dry matter distribution. Crop Sci. 8: 21–25.
Noviandi, C.T., K. Neal, J.-S. Eun, M.D. Peel, B.L. Waldron, D.R. ZoBell, and B.R. Min. 2014.
Comparison of alfalfa, birdsfoot trefoil, and cicer milkvetch in combination with 25, 50,
or 75% tall fescue in a continuous-culture system. Prof. Anim. Sci. 30: 23–32.
Owens, J., F.D. Provenza, R.D. Wiedmeier, and J.J. Villalba. 2012. Supplementing endophyteinfected tall fescue or reed canarygrass with alfalfa or birdsfoot trefoil increases forage
intake and digestibility by sheep. J. Sci. Food Agric. 92: 987–992.
Parish, J.A., M.A. McCann, R.H. Watson, N.N. Paiva, C.S. Hoveland, A.H. Parks, B.L.
Upchurch, N.S. Hill, and J.H. Bouton. 2003. Use of nonergot alkaloid-producing
endophytes for alleviating tall fescue toxicosis in stocker cattle. J. Anim. Sci. 81: 2856–
2868.
Porter, H.L. 1958. Rhizomes in tall fescue. Agron. J. 50: 493–494.
Pucheta, E., I. Bonamici, M. Cabido, and S. Díaz. 2004. Below-ground biomass and productivity
of a grazed site and a neighbouring ungrazed exclosure in a grassland in central
Argentina. Austral Ecol. 29: 201–208.

25

Ramírez-Restrepo, C.A., T.N. Barry, N. López-Villalobos, P.D. Kemp, and W.C. McNabb.
2004. Use of Lotus corniculatus containing condensed tannins to increase lamb and wool
production under commercial dryland farming conditions without the use of
anthelmintics. Anim. Feed Sci. Technol. 117: 85–105.
Rogers, W.M., Roberts, C.A., J.G. Andrae, D.K. Davis, G.E. Rottinghaus, N.S. Hill, R.L.
Kallenbach, and D.E. Spiers. 2011. Seasonal fluctuation of ergovaline and total ergot
alkaloid concentrations in tall fescue regrowth. Crop Sci. 51: 1291–1296.
Roberts, C.A., P.R. Beuselinck, M.R. Ellersieck, D.L. Davis, and R.L. McGraw. 1993.
Quantification of tannins in birdsfoot trefoil. Crop Sci. 33: 675–679.
Rottinghaus, G.E., G.B. Garner, C.N. Cornell, and J.L. Ellis. 1991. HPLC method for
quantitating ergovaline in endophyte-infected tall fescue: Seasonal variation of
ergovaline levels in stems with leaf sheaths, leaf blades and seed heads. J Agric Food
Chem 39: 112–115.
Scalbert, A. 1991. Antimicrobial properties of tannins. Phytochemistry 30: 3875–3883.
Schaefer, M.R., K.A. Albrecht, and D.M. Schaefer. 2014. Stocker steer performance on tall
fescue or meadow fescue alone or in binary mixture with white clover. Agron. J. 106:
1902–1910.
Schardl, C.L., A. Leuchtmann, and M.J. Spiering. 2004. Symbioses of grasses with seedborne
fungal endophytes. Annu. Rev. Plant Biol. 55: 315–40.
Seaney, R.R., and P.R. Henson. 1970. Birdsfoot trefoil. Adv. Agron. 22: 119–157.
Siegel, M.R., M.C. Johnson, D.R. Varney, W.C. Nesmith, R.C. Buckner, L.P. Bush, P.B. Burrus
II, T.A. Jones, and J.A. Boling. 1984. A fungal endophyte in tall fescue: Incidence and
dissemination. Phytopathology 74: 932–937.
Sleugh, B., K.J. Moore, J.R. George, and E.C. Brummer. 2000. Binary legume–grass mixtures
improve forage yield, quality, and seasonal distribution. Agron. J. 92: 24-29.
Steiner, J.J., P.R. Beuselinck, S.L. Greene, J.A. Kamm, J.H. Kirkbride, and C.A. Roberts. 2001.
A description and interpretation of the NPGS birdsfoot trefoil core subset collection.
Crop Sci. 41: 1968–1980.
Stuedemann, J.A., and C.S. Hoveland. 1988. Fescue endophyte: History and impact on animal
agriculture. J. Prod. Agric. 1: 39–44.
Taylor, R.W., and D.W. Allinson. 1983. Legume establishment in grass sods using minimumtillage seeding techniques without herbicide application: Forage yield and quality. Agron.
J. 75: 167–172.

26

Undersander, D., L. Greub, R. Leep, P. Beuselinck, J. Wedberg, D. Smith, K. Kelling, J. Doll, D.
Cosgrove, C. Grau, S. Peterson, M. Wipfli, and J. English. 1993. Birdsfoot trefoil for
grazing and harvested forage. North Cent. Reg. Ext. Publ. 474. Cooperative Extension
Service, University of Wisconsin, Madison, WI.
Van Hecke, M.M., A.M. Treonis, and J.R. Kaufman. 2005. How does the fungal endophyte
Neotyphodium coenophialum affect tall fescue (Festuca arundinacea) rhizodeposition
and soil microorganisms? Plant Soil 275: 101–109.
Waghorn, G. 2008. Beneficial and detrimental effects of dietary condensed tannins for
sustainable sheep and goat production-progress and challenges. Anim. Feed Sci. Technol.
147: 116–139.
Waghorn, G.C., M.J. Ulyatt, A. John, and M.T. Fisher. 1987. The effect of condensed tannins on
the site of digestion of amino acids and other nutrients in sheep fed on Lotus corniculatus
L. Br. J. Nutr. 57: 115–126.
Wang, Y., G.C. Waghorn, W.C. McNabb, T.N. Barry, M.J. Hedley, and I.D. Shelton. 1996.
Effect of condensed tannins in Lotus corniculatus upon the digestion of methionine and
cysteine in the small intestine of sheep. J. Agric. Sci. 127: 413–421.
Wang, M.B, and Q. Zhang. 2009. Issues in using the WinRhizo to determine physical
characteristics of plant fine roots. Acta Ecol. Sin. 29: 136–138.
Williams, C.M., J.-S. Eun, C.M. Dschaak, J.W. MacAdam, B.R. Min, and A.J. Young. 2010.
Case study: In vitro ruminal fermentation characteristics of birdsfoot trefoil (Lotus
corniculatus L.) hay in continuous cultures. Prof. Anim. Sci. 26: 570–576.
Wolf, D.D., R.H. Brown, and R.E. Blaser. 1979. Physiology of growth and development. In:
R.C. Bukner and L.P. Bush, editors, Tall Fescue. Agron. Monogr. 20. ASA, CSSA, and
SSSA, Madison, WI. p. 75-92.
Woodward, S.L., G.C. Waghorn, and P.G. Laboyrie. 2004. Condensed tannins in birdsfoot trefoil
(Lotus corniculatus) reduce methane emissions from dairy cows. Proc. New Zeal. Soc.
Anim. Prod. 64: 160–164.
Zeiders, K.E., and R.R. Hill. 1988. Measurement of resistance to fusarium wilt/root and crown
rot in birdsfoot trefoil populations. Crop Sci. 28: 468–474.
Zobel, R.W. 2008. Hardware and software efficacy in assessment of fine root diameter
distributions. Comput. Electron. Agric. 60: 178–189.
Zobel, R.W., T.B. Kinraide, and V.C. Baligar. 2007. Fine root diameters can change in response
to changes in nutrient concentrations. Plant Soil 297: 243–254.

27

CHAPTER 2
FORAGE QUALITY, YIELD, CONDENSED TANNIN CONCENTRATION, AND
PERSISTENCE OF BIRDSFOOT TREFOIL-TALL FESCUE MIXTURES UNDER
GRAZING IN MICHIGAN
ABSTRACT
Birdsfoot trefoil (Lotus corniculatus L., BFT), a perennial forage legume containing proteinbinding condensed tannins (CTs) that enhance ruminant livestock production, has the capacity to
yield productive pastures when grown with a perennial companion grass. The objective of this
study was to identify nutritious and productive BFT-tall fescue [Schedonorus arundinaceus
(Schreb.) Dumort] mixtures with moderate- to high-CT concentration that would be suitable for
forage production in south-central Michigan. Eight BFT cultivars with a range of CT
concentrations were established during 2014 in Lansing, MI, and four tall fescue cultivars
varying in endophyte infection status were overseeded into stands in 2015. Mixtures were grazed
by sheep, with frequency of grazing determined by forage growth rate, for a total of eight highintensity, short-duration grazing periods from 2015-2016. The proportion of BFT in mixtures
declined from > 95% in 2015 to < 10% in 2016 due to reduced vigor of BFT. Spring stand
density of BFT plants declined over years, however high-tannin Oberhaunstädter expressed 70%
greater density in 2016 than all other cultivars (P < 0.01), suggesting improved persistence after
one grazing season. Variance in crude protein (CP), neutral detergent fiber (NDF), acid detergent
fiber (ADF), neutral detergent fiber digestibility (NDFD), lignin, and CT concentration was
influenced by botanical changes over the grazing seasons as well as differences among BFT
cultivars. Condensed tannin concentrations for mixtures with high-tannin BFT ranged from 7.5
to 31.7 g kg-1 DM, while CT concentrations ranged from 3.2 to 21.6 g kg-1 DM for all other BFT
mixtures. In order for BFT-tall fescue mixtures to provide production and health benefits for

28

ruminant livestock through the entire grazing season, at least 20% high-tannin BFT should be
present in grass mixtures. Alternatively, at least 40% medium- to low-tannin BFT in grass
mixtures will provide benefits during spring and summer grazing periods. Pregrazing forage
yield exceeded 2.7 Mg ha-1 for all grazing periods and nutritive value was always adequate for
ewes at maintenance, regardless of CT concentration or endophyte type. Therefore, high-tannin
Oberhaunstädter in mixtures with endophyte-free tall fescue will yield the most productive,
persistent, and nutritious forage mixture suitable for grazing in south-central Michigan while
minimizing tall fescue seed cost and maximizing the CT-induced benefits for ruminant livestock
health and performance.

INTRODUCTION
High costs for grain and nitrogen (N) fertilizer, along with an interest in reducing the
environmental impact of livestock production, has led to a renewed interest in using N-fixing
forage legumes in grass-legume pasture mixtures as an alternate feed source for ruminant
livestock. While not considered a major forage legume in the United States, birdsfoot trefoil
(Lotus corniculatus L., BFT) produces high quality forage with livestock performance that is
often superior to alfalfa (Medicago sativa L.) and other forages (Grabber et al., 2014). This is
primarily related to the presence of condensed tannins (CTs) in the aboveground plant parts
(Grabber et al., 2014; Waghorn, 2008).
Condensed tannins are secondary plant compounds that bind to proteins, providing a
multitude of unique benefits for ruminant nutrition, health, and environmental sustainability
(Mueller-Harvey, 2006; Waghorn, 2008). These phenolic compounds improve ruminant
performance due to a pH-dependent tannin-plant protein complex, reducing rumen proteolysis

29

and increasing protein flow to the lower digestive tract, which improves amino acid absorption in
the small intestine for more efficient feed utilization (Waghorn et al., 1987; Wang et al., 1996).
Studies show grazing BFT increases lamb growth and wool production (Ramírez-Restrepo et al.,
2004) and average daily gains on BFT can approach feedlot rates of gain (MacAdam et al.,
2011). Additional benefits stemming from the CTs in BFT include the prevention of pasture
bloat (McMahon et al., 2000), a reduction in internal parasitic infection in sheep (Marley et al.,
2003), and a reduction in greenhouse gas emissions through lower enteric methane emissions
(Woodward et al., 2004) and reduced urinary N excretion in cattle (Misselbrook et al., 2005).
Despite the performance, health, and environmental benefits, BFT production in the
United States is limited because it proves challenging to grow with slow, unreliable
establishment, poor persistence and lower yield compared to alfalfa when grown on fertile soils
(Blumenthal and McGraw, 1999; MacAdam et al., 2006). While models show lamb operations
grazing BFT can gain up to $50 ha-1 in profit from increased liveweight gain and fecundity,
profitability is highly dependent on establishment and persistence of BFT stands (Doran-Browne
et al., 2015).
To overcome these undesirable characteristics, BFT can be grown in mixtures with
perennial grasses. Birdsfoot trefoil is more productive when seeded with a companion grass
because high plant mortality associated with root and crown disease creates bare spots in the
stands conducive to invasion by weeds if no grass is present (MacAdam et al., 2006;
Undersander et al., 1993) . These types of pasture mixes could provide a solution to BFT
production challenges if the right combination of species and specific cultivars are identified.
One option is to use tall fescue [Schedonorus arundinaceus (Schreb.) Dumort.] as a
companion grass in BFT-grass mixtures. Tall fescue is a highly productive, widely adapted cool-

30

season forage grass, occupying an estimated 19 million ha of pasture in the United States
(Bouton, 2007). The majority of tall fescue pastures support ‘Kentucky ̶̶31’ ̶̶(KY31) ̶̶infected ̶̶with ̶̶
the endophyte Neotyphodium coenophialum [(Morgan- Jones and Gams) Glen Bacon and
Hanlin]. The fungus grows symbiotically with the grass to enhance yield, persistence, and
reproduction; this is particularly important in the southeastern United States to ensure the grass
remains productive in hot and dry climates (Belesky and West, 2009; Bouton et al., 1993). Tall
fescue has not traditionally been a major forage in Michigan due to its relatively poor winter
hardiness compared to other grasses, despite the fact that it is a high yielding grass in forage
variety testing (Cassida et al., 2016). However, as the climate warms, tall fescue has a high
potential to attract interest from Michigan producers.
The major disadvantage associated with tall fescue is that the toxic wild-type fungal
endophyte produces ergot alkaloids to the detriment of livestock health and performance.
Ergovaline, the main ergot alkaloid produced by the endophyte, is responsible for reduced
livestock performance, including lower liveweight gain and reduced reproductive efficiency,
commonly ̶̶referred ̶̶to ̶̶as ̶̶“fescue ̶̶toxicosis” ̶̶or ̶̶“summer ̶̶slump” ̶̶(Bacon et al., 1986; Stuedemann
and Hoveland, 1988). Annual losses from fescue toxicosis in the United States are estimated to
be over $600 million (Bouton, 2007; Fribourg and Waller, 2005). While there have been
introductions of endophyte-free (E-) and low-alkaloid novel endophyte-infected tall fescue
cultivars, E+ tall fescue remains the predominant type grown throughout the United States
because of its superior productivity under diverse environmental conditions.
Evidence suggests that combining tall fescue with BFT can neutralize the toxicity of E+
tall fescue. Lambs receiving intraruminal infusions of CTs consumed more E+ tall fescue relative
to lambs that were not infused (Lisonbee et al., 2009). Similarly, lambs supplemented with BFT

31

consumed more E+ tall fescue than un-supplemented lambs or lambs supplemented with alfalfa
(Owens et al., 2012). Birdsfoot trefoil has also been shown to change it chemical composition
when grown in mixtures with tall fescue, expressing greater CT concentrations in mixtures with
tall fescue than when grown as a BFT monoculture (Wen et al., 2003). These findings offer the
opportunity for combining forages with complementary secondary compounds to produce a high
yielding, nutritious forage mixtures with superior animal performance. Therefore, I hypothesize
that productive and persistent moderate- to high-tannin BFT grown in mixtures with E+ tall
fescue will increase ruminant productivity directly, because legumes have a lower fiber content
that is more rapidly digestible than grasses, and indirectly, through binding the alkaloids
produced by E+ tall fescue.
The aim of this study was to identify nutritious and productive BFT-tall fescue mixtures
with moderate- to high-CT concentration that would be suitable for forage production in southcentral Michigan. Thirty-two BFT-tall fescue mixtures with varying CT concentrations and
endophyte types were evaluated under grazing by sheep over two grazing seasons in Lansing,
Michigan. The objectives were to evaluate (i) botanical changes over the grazing season and (ii)
BFT spring stand density to determine the most persistence mixtures under grazing. We also
measured (iii) CT concentration and (iv) forage quality in spring and after grazing regrowth to
identify nutritious BFT-tall fescue mixtures with optimal CT concentrations. Additionally, (v)
pregrazing forage yield, (vi) herbage utilization and (vii) grazing preference were evaluated to
determine the most suitable mixture for grazing in south-central Michigan.

32

MATERIALS AND METHODS
Site description
The experiment was conducted at the Michigan State University (MSU) Agronomy Farm
(Lansing, ̶̶MI; ̶̶42°42’N, ̶̶84°28’W, ̶̶elevation ̶̶262 ̶̶m). ̶̶Plots were on a Capac loam (fine-loamy,
mixed, active, mesic aquic Glossudalfs) that was previously in alfalfa. Initial soil fertility values
in 2014 were: soil pH: 7.1, P: 22 ppm, K: 99 ppm, Mg: 204 ppm, and Ca: 1163 ppm. Monthly
weather data were obtained from a weather station located within 1 km of the site (Michigan
State University EnviroWeather, 2017) and growing degree days (GDD) were calculated from
the same weather site starting March 1st using 5°C base temperature (MAWN, 2017). 30-year
averages were gathered by the National Oceanic and Atmospheric Administration at a site in
Lansing, MI (NOAA, 2017).

Experimental design
Treatments were arranged in a randomized complete block split-plot design with two
replications. The main plot consisted of four tall fescue cultivars that varied in endophyte type
while the sub-plot included eight BFT cultivars with varying genetic expression of CTs for a
total of 32 BFT-tall fescue treatment combinations. The study included the following tall fescue
cultivars: E+ ̶̶‘Kentucky ̶̶31’ (KY31) infected with the wild-type toxic endophyte, E- ‘Kentucky ̶̶
32’ ̶̶(KY32), ̶̶novel ̶̶endophyte-infected ̶̶‘Martin ̶̶II ̶̶Protek’ ̶̶infected ̶̶with ̶̶a ̶̶non-toxic endophyte
(Olson et al., 2012), and E- ‘Martin ̶̶II’. ̶̶These ̶̶cultivars ̶̶were ̶̶chosen ̶̶based ̶̶on ̶̶presumed ̶̶alkaloid ̶̶
among endophyte types, ranging from high-alkaloid for E+ tall fescue, low-alkaloid for novel
endophyte-infected tall fescue, and no alkaloids for E- tall fescue. The BFT cultivars were
selected for a range of CT concentrations and represent all commercially available cultivars

33

available in the United States as of early 2014: ‘Oberhaunstädter’ from Germany; ‘Bull’,
‘Bruce’, and ‘AC Langille’ from Canada; ‘Pardee’, a generic BFT seed (‘Common’), and two
sources of ‘Norcen’ from the United States (Grabber et al., 2015; Grant, 2004; Papadopoulos et
al., 2008).

Plot establishment
2014 served as the experimental establishment year for this study. The plot area was
prepared with conventional tillage using a moldboard plow (20-cm depth), followed by disking
and cultipacking. Immediately before planting, BFT seed was inoculated with Rhizobium loti. On
11 July 2014, a plot planter with a cone seeder (Carter Manufacturing, Brookston, IN) was used
to seed BFT at a rate of 6.7 kg ha-1 into the prepared seedbed. Plot dimensions were 2.3 x 4.0 m.
Grass weeds were controlled by sethoxydim herbicide (0.21 kg a.i. ha-1; Poast Plus, BASF,
Research Triangle Park, NC) mixed with a crop oil concentrate (2.33 L ha -1) in August 2014 and
broadleaf weeds were controlled by hand weeding within the plot area. In October 2014, plots
were fertilized with 47 kg ha-1 of P2O5 and 141 kg ha-1 of K2O. On 24 March 2015, BFT stands
were overseeded with tall fescue at 12.3 kg ha -1. Dry conditions and minimal freeze-thaw cycles
after frost seeding resulted in poor tall fescue establishment. Therefore, tall fescue was reseeded
on 18 August 2015 using a no-till drill (Great Plains Ag, Salina, KS) oriented perpendicular to
BFT rows at a seeding rate of 12.3 kg ha -1. In 2015 and 2016, invasion by Canada thistle
(Cirsium arvense L.) was managed by spot spraying with glyphosate (N-phosphonomethyl
glycine, 2.5 kg a.e. ha-1) as needed using a modified hood to keep the herbicide on the target
plant and minimize overspray onto surrounding herbage.

34

Grazing periods
Grazing animals used in this study included Suffolk (mean body weight=100 kg) and
Dorset-Polypay crossbred ewes (mean body weight=68 kg) from the MSU Sheep Teaching and
Research Center (Table 2.1). The entire plot area (approximately 40 x 40-m including alleys and
borders) was fenced with electric netting attached to a battery-powered fence charger and
animals had access to the entire plot area. Frequency of grazing was determined by forage
growth rates. Grazing began when mean forage dry matter reached a threshold of approximately
3.5 to 4.5 Mg ha-1 and ended when the average residual sward height was approximately 10 cm.
No grazing occurred during the establishment year in 2014. Plots were grazed three times in
2015 and five times in 2016 for a total of eight grazing periods (Table 2.1). Plots were clipped to
a 10-cm stubble height after grazing, if necessary, to remove remaining reproductive stems that
were not consumed.

Botanical composition
Before each grazing period, representative forage samples were clipped to a 10-cm height
from the four edges of each plot. Botanical composition was determined by hand sorting fresh
samples into respective forage species, weeds, or dead components. Individual components were
then dried at 60°C in a forced-air oven for 72 h and weighed. In 2015, botanical composition
samples from the first two grazing periods were not collected because grass and weed content
was negligible; all plots during these grazing periods were > 95% BFT.

35

Birdsfoot trefoil stand density
Spring BFT plant counts were obtained prior to each grazing season on 24 April 2015
and 20 April 2016, as well as on 25 April 2017. One 0.6 x 0.6-m quadrat was placed in the
center of each plot and individual BFT plants were counted to assess overwinter survival and
persistence.

Condensed tannin and forage quality analyses
Before each grazing period, representative forage samples were clipped to a 10-cm height
from the four edges of each plot. Samples were dried at 60°C in a forced-air oven for 72 h,
ground to pass through a 4-mm mesh screen in a Wiley mill (Wiley Laboratory Mill, Model 4,
Auther H. Thomas Co., Philadelphia, PA), and reground to pass through a 1-mm mesh screen in
a cyclone mill (UDY Manufacturing, Fort Collins, CO).
Mixed samples were analyzed for CT concentration at Utah State University following
the butanol-HCl-acetone method of Grabber et al. (2013). Tannin standards for the
spectrophotometric analysis were isolated from a single sample of Oberhaunstädter cv. of L.
corniculatus following the methods described by Hagerman (2011).
Forage quality was determined via near-infrared reflectance spectroscopy (NIRS) using a
scanning monochromator (Model 6500, FOSS NIRSystems, Inc., Laurel, MD) and chemometrics
software (WinISI II, ver. 1.5, Infrasoft International LLC, State College, PA). A commercial
mixed hay equation (16mh50-2.eqa, NIRS Forage Feed and Testing Consortium, Winfield, WI)
from a calibration set containing multiple grass and legume forage species was used to predict
nutritive value of each mixed sample. The distribution and boundaries of BFT-tall fescue sample
spectra followed the population structure of the spectra in the calibration set, so no additional

36

chemical analysis was needed. Mixed samples were analyzed for crude protein (CP), neutral
detergent fiber (NDF), acid detergent fiber (ADF), lignin, rumen undegradable protein (RUP),
and in vitro neutral detergent fiber digestibility (NDFD). Digestible NDF was calculated as:
NDFD = [(dNDF48/NDF) x 100] where dNDF48 is in vitro NDF digestion (as a % of DM) for
48 h.

Forage yield and herbage utilization
Forage yield and herbage utilization was estimated nondestructively using a calibrated
rising plate meter (RPM, Jenn Quip model, Feilding, New Zealand) (Sanderson et al., 2001).
Throughout the grazing season, RPM measurements were taken before and after each grazing
period. On each sampling date, herbage mass was estimated from the mean of 30 RPM
measurements within each plot. An additional six to eight calibration samples were collected
throughout the entire plot area to convert forage height to forage biomass. For each calibration
sample, a single rising plate measurement was manually recorded and the forage under the disk
was clipped to ground level. Calibration samples were then dried at 60°C in a forced-air oven for
72 h and weighed to develop a linear regression curve (Scrivner et al., 1986). Herbage utilization
was calculated as the difference between pre- and post-grazing RPM measurements (Leep et al.,
2002).

Grazing preference
In 2016, grazing preference was determined by visually assessing the proportion of tillers
and shoots grazed immediately after each grazing period. For each plot, BFT and tall fescue
plants were visually assessed by two trained observers using a 0 (0% of tillers grazed) to 9

37

(100% of tillers grazed) scale. The two scores were averaged to produce a sample value for
grazing preference.

Statistics
Data were analyzed using mixed models procedures as implemented in PROC MIXED
of SAS 9.4 (SAS Institute Inc., Cary, NC). The analysis was based on a randomized complete
block design with a split-plot treatment arrangement. Tall fescue (TF) cultivars were randomly
assigned to main plots and BFT cultivars to sub-plot. Fixed effects were TF, BFT, grazing period
(GP), and block (Dixon, 2017). The block x TF interaction term was the scale-appropriate
random effect to evaluate the effect of TF cultivar, whereas the block x TF x BFT interaction
was the scale-appropriate error term for the BFT and the TF x BFT interaction effects. The
model was analyzed as repeated across grazing periods using the spatial power covariance
structure to account for unevenly spaced intervals. Denominator degrees of freedom were
calculated using Kenward-Rogers option. Linear contrasts were constructed to compare
significant treatment means and adjusted using the simulation-based method (Edwards and
Berry, 1987). Contrasts were grouped based on presumed differences in alkaloid toxicity among
tall fescue endophyte types and measured differences in CT content among BFT cultivars. Tall
fescue contrasts included: (i) E+ (KY31) vs. E- (KY32 and Martin II), (ii) Novel (Martin II
Protek) vs. E-, and (iii) E+ vs. novel. The BFT contrasts included: (iv) high-tannin
(Oberhaunstädter) vs. all others, (v) high- vs. low-tannin (AC Langille, Bruce, Common, Norcen
1, Norcen 2), (vi) high- vs. medium-tannin (Bull and Pardee), and (vii) medium- vs. low-tannin.
Differences among grazing periods were analyzed using the PDIFF function of the LSMEANS

38

command with the SIMULATE adjustment to compare least square means. Significance was
declared at P < 0.05 unless otherwise stated.

RESULTS
Weather
In 2015 and 2016, mean monthly air temperatures at the field site was consistent with the
30-year average (Figure 2.1A). Mean air temperature between grazing periods was also normal
for the region, with highest temperatures occurring in mid- to late-summer. Cumulative growing
degree days (GDD) between March and November varied slightly between years, with 159 fewer
GDD in 2015 compared to 2016 (2532 and 2691, respectively).
Precipitation also varied during and between years (Figure 2.1B). There was less total
precipitation in 2015 (686 mm) and 2016 (695 mm) compared to the 30-year average (817 mm).
In 2015, conditions were very dry from February through April (36 mm total), with 119 mm less
precipitation than the 30-year average. June was the wettest month of the year (145 mm),
followed by a dry July (57 mm) and increased precipitation in August (117 mm). In 2016, mean
monthly precipitation was above average in March (90 mm), followed a dry period from May
through July (145 mm total), while August was the wettest month of the year (147 mm).

Botanical composition
There was a TF x GP (P < 0.001) interaction for the proportion of tall fescue in forage
mixtures (Table 2.2). This was driven by large variability among grazing periods related to
seasonal differences and changes in the total proportion of tall fescue in mixtures over the two
grazing seasons (Figure 2.2). Therefore, no significant differences among endophyte groups in

39

the contrasts were detected (P > 0.05). This also occurred with TF x GP interactions for BFT (P
< 0.05) and dead (P < 0.05) botanical components (Table 2.2). For the weed component, the TF
x GP interaction (P < 0.001, Table 2.2) was driven by greater proportion of weeds in E+ tall
fescue mixtures compared other tall fescue mixtures during the second grazing season (Table
2.3). In 2016, the proportion of weed in E+ tall fescue mixtures was greater than mixtures with
E- cultivars or novel endophyte-infected tall fescue in May and June. A similar trend was
observed in October, with a greater proportion of weeds in E+ tall fescue mixtures than mixtures
with E- cultivars (Table 2.3).
The main effect of GP was highly significant for BFT, tall fescue, and dead botanical
components (P < 0.001, Table 2.2). Averaged across all BFT-tall fescue mixtures, the proportion
of BFT in forage mixtures declined from > 95% for the first two grazing periods in 2015 to
66.0% by November 2015, with the proportion of tall fescue in mixtures increasing to 30.5%
(Table 2.4). In 2016, the proportion of BFT again declined from August through October, with
mixtures averaging 9.2% BFT and 88.2% tall fescue by the final grazing period in October 2016.
Mixtures with high-tannin BFT averaged 20.1 ± 5.75% BFT during this grazing period compared
to 7.7 ± 5.75% for mixtures with all other BFT cultivars, however differences were not
statistically significant (P > 0.05). Variability in the dead botanical component was driven by a
substantial amount of leftover litter in August 2016 (10.8%) from the previous grazing periods
(Table 2.4). In 2016, plots were clipped to a height of 10-cm after the May and June grazing
periods to remove all remaining reproductive stems. Therefore, the significant dead component
was caused by top-clipping rather than a biologically important treatment effect.

40

Birdsfoot trefoil stand density
There was a BFT x year interaction for spring BFT stand density (P < 0.01). Large
differences were detected between the high-tannin BFT cultivar and all other BFT cultivars in
2016, the spring following the first grazing season (Table 2.5). Stand density for the high-tannin
BFT cultivar was 130 plants m-2, 33 and 47 more plants m-2 than low- and medium-tannin
cultivars, respectively.

Condensed tannin concentration
There was a BFT x GP interaction for CT concentration (P < 0.001, Table 2.6). The
interaction was driven by differences between mixtures with medium- and low-tannin BFT
cultivars across grazing periods (Table 2.7). In 2015, mixtures with medium-tannin cultivars
expressed 2.4 and 4.7 g kg-1 DM more CT than low-tannin BFT mixtures in June and August. In
2016, mixtures with medium-tannin BFT cultivars expressed greater CT concentrations than
mixture with low-tannin BFT cultivars in June, while no differences were detected between
mixtures with low- and medium-tannin BFT cultivars for the other grazing periods (Table 2.7).
Across all grazing periods, mixtures with the high-tannin BFT cultivar consistently
expressed greater CT concentrations than mixtures with all other BFT cultivars (Table 2.7).
These mixtures had on average over 1.5-times greater CT concentrations than low-tannin BFT
mixtures. Across both years, August had the widest range in CT concentration between high- and
low-tannin BFT mixtures. Concentrations then declined in the fall of both years, however CT
content was still greater for high-tannin BFT mixtures than mixtures with all other BFT cultivars
in the fall of 2015 and 2016 (P < 0.001, Table 2.7).

41

Condensed tannin content also varied seasonally (P < 0.001, Table 2.6), with
concentrations peaking in the summer and dropping in the fall (Table 2.8). Averaged across all
BFT-tall fescue mixtures, August 2015 had the overall greatest CT concentration (19.9 g kg -1
DM), however June of both years also had relatively high concentrations (12.5 and 13.7 g kg-1
DM in 2015 and 2016, respectively). Concentrations then dropped in the fall for both years, with
the lowest concentrations produced in October 2016 (3.9 g kg-1 DM).

Forage quality
There was a TF x GP interaction for all forage quality components (Table 2.6); however,
lack of differences among endophyte groups in the contrasts indicates these interactions were
caused by large variability among grazing periods. There was also a BFT x GP interaction for CP
(P < 0.001), RUP (P < 0.05), NDF (P < 0.05), and ADF (P < 0.01) forage quality components
(Table 2.6). The CP interaction was driven by a shift from low to high CP from 2015 to 2016 for
the high-tannin BFT mixtures (Table 2.9). In 2015, CP concentrations for mixtures with hightannin BFT were 148 and 160 g kg-1 DM in August and November, respectively; 23 and 15 g kg-1
DM less than CP than mixtures with all other BFT cultivars for the respective grazing periods.
By the final grazing period in October 2016, CP for mixtures with high-tannin BFT was 216 g
kg-1 DM, 21 g kg-1 DM more CP than mixtures with all other BFT cultivars (Table 2.9).
Concentrations of CP also varied greatly across grazing periods (P < 0.001, Table 2.6), with the
greatest levels produced in May and September 2016 when averaged across all BFT-tall fescue
mixtures (219 and 231 g kg-1 DM, respectively, Table 2.8). These CP concentrations were
greater than all concentrations produced in 2015.

42

The BFT x GP interaction for RUP was driven by a shift from high to low RUP
concentrations for mixtures with high-tannin BFT across grazing seasons (Table 2.9). In 2015,
RUP for high-tannin BFT mixtures were 322 and 339 g kg-1 CP in June and August,
respectively; this was 33 and 53 g kg-1 CP more RUP than mixtures with all other BFT cultivars
for the respective grazing periods. By the final grazing period in October 2016, RUP for hightannin BFT mixtures was 364 g kg-1 CP, 30 g kg-1 CP less RUP than mixtures with all other BFT
cultivars. Concentrations of RUP also varied greatly across grazing periods (P < 0.001, Table
2.6). Averaged across all BFT-tall fescue mixtures, RUP concentrations were greater in October
2016 than for all other grazing periods (Table 2.8).
Similar to the protein components, BFT x GP interactions for NDF and ADF were again
driven by mixtures with high-tannin BFT (Table 2.10). In November 2015, NDF for high-tannin
BFT mixtures was 351 g kg-1 DM, 41 g kg-1 DM more NDF than mixtures with all other BFT
cultivars. This was reversed in 2016; NDF for high-tannin BFT mixtures were 384 and 419 g kg1

DM in August and October, respectively, which was 37 and 48 g kg-1 DM less NDF than

mixtures with all other BFT cultivars for the respective grazing periods. A similar trend was
observed for ADF (Table 2.10). Concentrations of ADF for high-tannin BFT mixtures was 262 g
kg-1 DM in November 2015, 37 g kg-1 more ADF than mixtures with all other BFT cultivars.
This was again reversed in 2016; ADF for high-tannin BFT mixtures were 282 and 273 g kg-1
DM in August and October 2016, which was 19 and 20 g kg-1 less ADF than mixtures with all
other BFT cultivars for the respective grazing periods (P < 0.10).
Both NDF and ADF varied seasonally (P < 0.001, Table 2.6), showing similar trends
when averaged across all BFT-tall fescue mixtures. Concentrations of NDF were high for the
first grazing period in 2015 (478 g kg-1 DM in June), low for the last grazing period in 2015 and

43

first grazing period in 2016 (315 and 281 g kg-1 DM in November and May, respectively) and
moderate to high for the remaining mid- and late-season grazing periods (Table 2.8). For ADF,
concentrations again were greatest for the first grazing period in 2015 (393 g kg-1 DM in June),
low for the last grazing period in 2015 and first grazing period in 2016 (230 and 199 g kg-1 DM
in November and May, respectively) and moderate for the remaining grazing periods (Table 2.8).
For the NDFD and lignin content, there were BFT and GP main effects (P < 0.001, Table
2.6). Averaged across all grazing periods, NDFD for high-tannin BFT mixtures (481 g kg-1
NDF) was 59 and 49 g kg -1 NDF lower than NDFD for low- and medium-tannin mixtures,
respectively (Table 2.11). Additionally, mixtures with medium-tannin cultivars had on average
11 g kg-1 NDF less NDFD than mixtures with low-tannin cultivars (P < 0.10). Averaged across
all BFT-tall fescue mixtures, NDFD was repeatedly greater in 2016 than 2015. In 2016, high
NDFD concentrations were produced in May (685 g kg-1 NDF), September (676 g kg -1 NDF),
and October (681 g kg -1 NDF), while in 2015, concentrations were greatest in November (560 g
kg-1 NDF) (Table 2.8).
For lignin, high-tannin BFT mixtures expressed greater lignin content than mixtures with
all other BFT cultivars (Table 2.11). Average concentration was 52 g kg -1 DM; this was 4 and 3
g kg-1 DM more lignin than mixtures with low- and medium-tannin cultivars, respectively.
Opposite to the seasonal variation of NDFD, lignin content was greater for all grazing periods in
2015 than those in 2016 when averaged across all BFT-tall fescue mixtures (Table 2.8).

Forage yield and herbage utilization
High winds and heavy rain in August 2015 immediately prior to the second grazing
period lodged the forage and resulted in a poor rising plate meter calibration equation (Table

44

2.12); therefore, this grazing period was removed from analysis. There were BFT x GP (P <
0.001), TF x GP (P < 0.001), and BFT x TF (P < 0.05) interactions for pregrazing forage yield
(Table 2.2). The BFT x GP interaction was driven by differences in pregrazing yield for mixtures
with high-tannin BFT across grazing seasons (Table 2.13). For the first grazing period in June
2015, high-tannin BFT mixtures yielded 6.62 Mg ha-1; this was 0.55 and 0.37 Mg ha-1 less than
mixtures with low- and medium-tannin cultivars, respectively. By September 2016, high-tannin
BFT mixtures yielded 0.39 Mg ha-1 more than mixtures with medium-tannin cultivars. For the TF
x GP interaction, no difference among endophyte groups in contrasts were detected (P > 0.05,
Table 2.13), indicating no major biological importance of endophyte type on pregrazing yield.
The same was true for BFT x TF interactions (data not shown).
Pregrazing forage yield also varied seasonally when averaged across all BFT-tall fescue
mixtures (P < 0.001, Table 2.2). Yield varied greatly in 2015 (Table 2.13), with the highest and
lowest yield produced during the first and last grazing periods (7.06 and 2.90 Mg ha-1 in June
and November, respectively). In 2016, pregrazing yield was greater for spring and early summer
grazing periods compared to the fall.
Herbage utilization, calculated as the differences between pregrazing and postgrazing
yield estimates, had BFT x GP (P < 0.01) and TF x GP (P < 0.001) interactions (Table 2.2). In
August 2016, utilization was greater for mixtures with high-tannin BFT (31.8%) compared to
mixtures with all other BFT cultivars (22.9%) (Table 2.14). A similar trend was observed in
September 2016, with greater herbage utilization for mixtures with high-tannin BFT (32.3%)
compared to mixtures with all other BFT cultivars (24.8%). In October 2016, herbage utilization
in mixtures with E+ tall fescue (28.0%) was greater than in mixtures with E- cultivars (19.2%) or
novel endophyte-infected tall fescue (18.8%) (Table 2.14). Averaged across all BFT-tall fescue

45

mixtures, herbage utilization varied seasonally (P < 0.01, Table 2.2). Utilization was greatest in
May 2016 (32.3%) and ranged from 21.3% to 28.1% for all other grazing periods (Table 2.14).

Grazing preference
Throughout the 2016 grazing season, sheep showed distinct preference among tall fescue
cultivars, as shown by the TF x GP interaction (P < 0.001, Table 2.2). This interaction was
driven by greater preference towards E+ tall fescue than either E- cultivars or novel endophyteinfected tall fescue during the final grazing period in October (Table 2.15). Overall preference
for tall fescue also varied throughout the season, with the lowest scores detected during the earlyspring grazing period (May, 51-60%).
Sheep showed no preference among BFT cultivars (P > 0.05), however overall preference
varied throughout the season (P < 0.05, Table 2.2). Grazing preference for BFT was lower
during mid-summer (August, 61-75%) compared to the other grazing periods (Table 2.15).
Overall sheep preference was also greater for BFT than tall fescue throughout the season,
ranging from 61-99% and 51-90% for BFT and tall fescue, respectively (Table 2.15).

DISCUSSION
Before discussing specifics of the data, it is necessary to discuss the study design because
it influences the interpretation of this data. This study was an add-on to an existing conventional
forage breeding project in which BFT populations will be selected from the forage mixture
where they will ultimately be used in. Therefore, all plots were mixtures of BFT and tall fescue
with no individual species monocultures.

46

Botanical composition and persistence
One of the major limitations commonly associated with BFT is poor stand persistence
(Wen et al., 2002). The proportion of BFT in forage mixtures steadily declined for all cultivars
over the two grazing seasons, falling from > 95% to 66% in 2015 and from 57% to < 10% in
2016. Numerous studies have reported a similar decline in the BFT fraction of grass-legume
mixtures over time (Hopkins et al., 1996; Leep et el., 2002; Sleugh et al., 2000; Wen et al.,
2002). This is likely due to reduced legume vigor from grazing pressure along with increased
competitiveness of grass in the fall when growing conditions become more favorable (Leep et
el., 2002). The number of grazing periods across the two grazing seasons, increasing from three
in 2015 to five in 2016, may have also influenced botanical changes. Other work has found the
decline in the number of BFT plants increases with the number of cuts (McKenzie et al., 2004).
The severe decline in the proportion of BFT in mixtures from August through October in 2016
may also be due to low midseason storage of root and crown carbohydrates typical of BFT plants
(Undersander et al., 1993) as well as dry conditions and with inconsistent rainfall in early 2016.
Regarding spring stand densities, the high-tannin BFT cultivar expressed higher plant densities
after one grazing season than other BFT cultivars used in the study. Others have similarly
reported higher plant populations for Oberhaunstädter (Marley et al., 2006), suggesting improved
persistence of this BFT cultivar. However, this advantage did not last after two grazing seasons.
Tall fescue endophyte infection is associated with greater tall fescue persistence (Bouton
et al., 1993; Malinowski and Belesky, 2006). However in this study, mixtures with E+ tall fescue
had a greater proportion of weeds in 2016, suggesting slower spring growth for E+ tall fescue
compared to the other tall fescue endophyte types, which created an environment conducive to

47

invasion by weeds. Therefore, endophyte infection is not needed for weed prevention in more
northern latitudes with less severe environmental conditions.
Botanical changes over the grazing seasons were similar for E- and novel endophyteinfected tall fescue. Seeds for E- tall fescue can cost 15-20% more than E+ seeds, while novelendophyte seeds may cost up to three times as much (Burdine, 2007). Therefore, E- cultivars are
a more economically efficient option to use a companion grass in BFT-tall fescue mixtures in
south-central Michigan than novel endophyte-infected tall fescue. This is why it is often
recommended to use E- tall fescue over E+ or novel endophyte-infected tall fescue in Michigan
(Cassida et al., 2016). However, this is likely not the case in southern regions of the United
States that rely on either E+ or novel endophyte-infected tall fescue to insure long-term
persistence of forage stands (Bouton et al., 1993; Hopkins et al., 2010; Malinowski and Belesky,
2006). Because this study was an add-on to an existing multi-state conventional forage breeding
project, endophyte infection may be needed at other locations to ensure tall fescue persistence in
mixtures with BFT. This is where the potential of high-tannin BFT to neutralize the toxicity of
E+ tall fescue is most valuable (Lyman et al., 2012; Owens et al., 2012).

Condensed tannin concentration
Optimal concentration of CTs for improved ruminant health and performance varies in
the literature. Several authors have suggested concentrations < 50 g kg -1 DM can exert beneficial
effects (Aerts et al., 1999; Min et al., 2003; Mueller-Harvey, 2006), while others have reported
concentrations as low as 5 g kg -1 DM can prove beneficial (Broderick et al., 2017; Li et al.,
1996). Concentrations > 55 g kg-1 DM have been shown to decrease forage digestibility, depress
feed intake, and reduce overall ruminant performance (Aerts et al., 1999; Min et al., 2003), while

48

others argue concentrations ranging from 40 to 50 g kg-1 DM can cause problems (McMahon et
al., 2000). In addition, CT content has been shown to be greater when BFT is grown in mixtures
with tall fescue than as a BFT monoculture (Wen et al., 2003). However, concentrations in BFT
usually range from 10 to 40 g kg -1 DM (MacAdam and Villalba, 2015) and never exceeded 32 g
kg-1 DM in mixtures throughout the study. Therefore, CT content was not high enough to reduce
feed intake or digestibility; however, concentrations varied by cultivar and across the grazing
seasons which has important implications for ruminant livestock performance.
There is a lack of knowledge regarding the CT content for many of the existing BFT
cultivars used in the United States. In this study, CT concentrations were greatest for mixtures
with Oberhaunstädter, intermediate to low depending on the grazing period for Pardee and Bull
mixtures, and lowest for mixtures with AC Langille, Bruce, Common, and both sources of
Norcen. Norcen typically contains very low concentrations of CTs (Grabber et al., 2014; Marley
et al., 2006; Marshall et al., 2008) which is consistent with the findings in this study. Pardee has
been reported to contain intermediate concentrations (Grabber et al., 2014) which was observed
in this study for both Pardee and Bull mixtures in June of both years and in August 2015.
Concentration for these grazing periods ranged from 12 g kg -1 DM to 21 g kg-1 DM. Grabber et
al. (2014) reported CT content of Pardee averaged 24.4 g kg-1 DM across 10 location-harvest
environments; the slightly lower concentrations observed in this study is likely due to a dilution
from the increased proportion of tall fescue in swards after reseeding in August 2015. In 2016,
lack of differences in CT concentrations betweeen mixtures with medium-tannin and low-tannin
BFT cultivars for all grazing periods except June can also be explained by the steady increase in
the proportion of tall fescue during these grazing periods. The exception to this in June 2016

49

coincides with higher CT content detected for cuttings made during warmer months (Cassida et
al., 2000; Gebrehiwot et al., 2002).
Results suggest there are minimal differences in CT concentrations among the mediumand low-tannin BFT cultivars used in this study when grown in mixtures with tall fescue,
regardless of endophyte type, which can be seen after August 2015. The exception to this is seen
with peak concentrations during warmer months. However, CT content was > 5 g kg-1 DM which
has been suggested to be high enough for bloat prevention and beneficial milk yield responses
(Broderick et al., 2017; Li et al., 1996) for all grazing periods except October 2016. Therefore,
low- and medium-tannin BFT cultivars are capable of providing benefits for ruminant health and
performance during spring and summer grazing periods when they constitute > 40% of grass
mixtures.
Intermediate CT concentrations have been reported for Oberhaunstädter (Grabber et al.,
2014; Marshall et al., 2008), however some studies have shown concentrations as low as Norcen
(MacAdam et al., 2011). In this study, CT concentrations of mixtures with Oberhaunstädter were
greater than mixtures with all other BFT cultivars. Concentrations ranged from 17 g kg-1 DM to
31 g kg-1 DM for all spring and summer grazing periods which is well within the reported range
to be beneficial for growth, milk yields, and improved amino acid absorption in the small
intestine (Hymes-Fecht et al., 2013; Ramírez-Restrepo et al., 2004; Waghorn et al., 1987).
Condensed tannin concentrations of high-tannin BFT mixtures decreased in the fall of both
grazing seasons, with a 52% decline in 2015 from August to November and a 33% decline in
2016 from June to October. This follows the same seasonal trend as mixtures with medium- and
low-tannin BFT cultivars, however the magnitude was larger for mixtures with high-tannin BFT.
Similar findings were reported by Roberts et al. (1993) who reported a 40% decline in CT

50

content from July to September for high-tannin cultivars, while low-tannin cultivars showed little
to no seasonal variation. However, in this study CT content for mixtures with high-tannin BFT
was still > 5 g kg-1 DM for fall grazing periods. Assuming CT concentrations > 5 g kg -1 DM can
provide bloat and performance benefits (Broderick et al., 2017; Li et al., 1996), results suggest
high-tannin BFT can exert beneficial effects for ruminant livestock throughout the entire grazing
season when they constitute > 20% of grass mixtures.
Various factors likely contributed to the greater CT content observed in mixtures with
Oberhaunstädter. From visual observations, Oberhaunstädter was the first cultivar to flower,
hence the earliest maturing. Condensed tannin content is positively associated with plant
maturity, with early maturing cultivars expressing higher CT concentrations than later maturing
cultivars (Grabber et al., 2014). Flowers and leaves also express greater CT concentrations than
stems, both of which visually appeared to be more abundant in Oberhaunstädter (Gebrehiwot et
al., 2002; Marshall et al., 2008). In addition, this cultivar was the only European cultivar used in
this study, which tend to express greater CT concentrations than cultivars originating from North
America (Marley et al., 2006; Steiner et al., 2001).
Some of the differences among studies on reported CT concentrations likely relates to the
large variability in CT content across difference growing environments (Grabber et al., 2014).
Consequently, continued work is needed to overcome this variability so that BFT can be grown
with a reliable CT concentration across different environments. Findings from this study suggest
that of the eight cultivars tested in this study, grazing high-tannin Oberhaunstädter in mixtures
with tall fescue during the spring and summer will produce forage with the greatest CT
concentration in south-central Michigan and therefore likely provide maximal benefits for
ruminant livestock nutrition, health, and performance.

51

Forage quality
Concentrations of CP were lower in 2015 than 2016 when a greater proportion of BFT
was present in mixtures, likely related to the fact that CP in BFT decreases with maturity
(Cassida et al., 2000). In 2015, grazing periods were less frequency and intensity compared to
those in 2016. Therefore, plants were much more mature, which can explain the lower CP
concentrations observed in this study throughout the 2015 grazing season.
Mixtures with high-tannin Oberhaunstädter, an early maturing cultivar, had lower CP
concentrations in August and November 2015 compared to mixtures with all other BFT cultivars.
Other work has similarly found early maturing cultivars tend to express lower concentrations of
CP and greater concentrations of CTs (Grabber et al., 2015) due to an inverse relationship
between CTs and CP (Miller and Ehlke, 1996). However, Grabber et al. (2014) noted that all
cultivars, even those with high CTs, produce forage with adequate levels of CP. This agrees
with the findings in this study; 148 g kg-1 DM was the lowest level of CP produced, which is still
adequate lambs post-weaning and non-lactating ewes on pasture. In 2016, CP concentrations for
high-tannin BFT mixtures were comparable to all other mixtures, even those with Norcen which
has been reported to have a relatively a high CP concentration (Grabber et al., 2014). By the final
grazing period in October 2016, mixtures with high-tannin BFT produced more CP than mixtures
with all other BFT cultivars. This may be because high-tannin Oberhaunstädter numerically
contained more BFT than other cultivars during spring 2016 stand counts, resulting in less grass
and therefore higher CP concentrations; however this difference in botanical composition was
not statistically significant. Differences in botanical composition was likely harder to detect than
differences in forage quality because forage samples were separated for individual species
components by hand whereas nutritive value was estimated digitally using NIRS. Therefore,

52

spring BFT stand density data provides an alternative mode for evaluating plant composition
differences among BFT cultivars.
In 2015, mixtures with high-tannin BFT exhibited greater RUP concentrations in June
and August than mixtures with all other BFT cultivars. Cultivars with high CT concentrations
tend to have a high RUP content (Grabber et al., 2015). This is because the tannin-protein
complex reduces protein degradation in the rumen, increasing dietary protein flow to the lower
digestive tract (Waghorn et al., 1987). However, no differences in RUP were detected in 2016
until the final grazing period in October, where mixtures with high-tannin BFT expressed a
lower RUP content than all other mixtures despite having a greater CT concentration.
Concentrations of RUP were greatest for all BFT-tall fescue mixtures during this grazing period,
suggesting an environmental factor may be driving the differences observed in mixtures with
high-tannin BFT. These findings were not expected and warrants further investigation to better
understand how seasonal variability and CT concentrations may influence RUP concentrations in
BFT-tall fescue mixtures.
For the fiber components, ADF and NDF were greater in November 2015 but lower in
August and October 2016 for mixtures with high-tannin BFT compared to mixtures with all other
BFT cultivars, while no differences were detected during other grazing periods. Other work has
similarly found no consistent relationships between CT in BFT and NDF, ADF or lignin
concentrations (Grabber et al., 2014; Miller and Ehlke, 1996). Grazing frequency and intensity
likely had a larger influence on fiber components than did CT concentrations. Forage mixtures
were very mature for the first grazing period in 2015, resulting in high ADF and NDF and lower
NDFD concentrations. In contrast, spring grazing in 2016 occurred one month earlier, leading to
overall lower NDF and ADF and higher NDFD concentrations for all mixtures.

53

Across both grazing seasons, mixtures with high-tannin BFT had lower NDFD and
greater lignin concentrations than mixtures with all other BFT cultivars. High CT concentrations
can reduce forage digestibly, however CT concentrations in this study were not within the
reported range for causing fiber problems (McMahon et al., 2000; Min et al., 2003). Higher fiber
and lower digestibility detected for mixtures with high-tannin Oberhaunstädter was likely due to
the fact that this cultivar was the first to flower, therefore more mature than other BFT cultivars.
High concentrations of NDFD are important for forage quality because the digestibility of the
cell wall is linked to higher ruminant production, including improved intake and milk yield (Oba
and Allen, 1999). However, it should be noted that no cultivars produced forage with undesirably
high levels of ADF and NDF or low levels of NDFD (NRC, 2001).

Forage yield and herbage utilization
The rising plate meter used in this study was calibrated frequently over the grazing
season to increase the accuracy of estimating forage yield (Ferraro et al., 2012). However, error
likely still occurred, particularly due to extreme weather events. A large wind and rain storm
immediately prior to the sheep arriving in August 2015 severely lodged the forage, which caused
inaccurate pregrazing rising plate meter readings, forcing yield data for this grazing period to be
removed from analysis. The rising plate meter used in this study was satisfactory as it was
mainly used to determine when forage yield reached the grazing threshold, however results are
strictly an estimate of available forage throughout the grazing season, not an exact measurement
of forage yield.
In 2015, high-tannin BFT mixtures had lower pregrazing forage yield for the first grazing
period than mixtures with all other BFT cultivars, suggesting slower early season growth. High-

54

tannin Oberhaunstädter plants expressed a shorter, more dense growth habit compared to other
cultivars which likely contributed to lower RPM readings. However, yields of mixtures with
high-tannin BFT were comparable for all other grazing periods and was greater than mediumtannin BFT mixtures in September 2016. This is consistent with other studies that have shown
greater yields of Oberhaunstädter after multiple harvest years compared to other BFT cultivars
(Marley et al., 2006). Herbage utilization was also greater for mixtures with high-tannin BFT
from August through September in 2016, suggesting high-tannin Oberhaunstädter mixtures were
more palatable and with superior forage quality, allowing sheep to consume a larger amount of
forage. This is consistent with the high CP and low ADF and NDF concentrations observed for
high-tannin BFT mixtures during the final grazing period in 2016.
No differences in pregrazing forage yield were detected among tall fescue cultivars based
on endophyte type. Similarly in Wisconsin, little differences in productivity have been reported
between E+ and E- tall fescue (Brink and Casler, 2012). Long-term forage variety testing has
also shown E- and novel endophyte-infected tall fescue cultivars are just as productive as E+ tall
fescue in Michigan (Cassida et al., 2016). Therefore as stated earlier, endophyte infection is not
needed to ensure persistence or productivity of tall fescue in forage mixtures in south-central
Michigan or regions with similar environmental conditions.
No differences in herbage utilization were detected among tall fescue endophyte types
until the final grazing period in October 2016. Greater utilization was observed for mixtures with
E+ tall fescue than mixtures with all other endophyte types, likely related to higher stocking
density and a longer grazing duration in October 2016 compared to other grazing periods (Table
2.1). It is also possible that consumption of high-tannin BFT may have increased utilization of
the high-alkaloid tall fescue cultivar (Lyman et al., 2012; Owens et al., 2012) since mixtures with

55

high-tannin BFT averaged approximately 20% BFT compared to 8% BFT for mixtures with all
other BFT cultivars during this grazing period; however, botanical differences were not
statistically significant. In addition, lack of interaction between CT concentration and endophyte
type suggest more work is needed to confirm the hypothesis that high-tannin BFT cultivars in
combination with E+ tall fescue will produce highly productive, persistence pastures by binding
to and neutralizing the toxic alkaloids in E+ tall fescue.

Grazing preference
Sheep tend to preferentially graze legumes over grasses (Rutter, 2006). In 2016, sheep
preferred grazing BFT over tall fescue, however no BFT-tall fescue treatment combination had a
significant effect on grazing preference. Similar to herbage utilization, we found that sheep
preferred E+ tall fescue more than all other endophyte types by the final grazing period in
October 2016. Because no differences in fiber and digestibility were detected among endophyte
groups in the contrasts, difference in preferences among cultivars for this grazing period was
likely due to a greater stocking density and longer grazing duration compared to the other
grazing periods in 2016 (Table 2.1).

CONCLUSIONS
In this study, the botanical composition of all mixtures changed to be grass dominated
due to the reduced vigor of BFT swards over time, declining from > 95% BFT in 2015 to < 10%
in 2016. While the proportion of BFT decline by 75% for high-tannin BFT mixtures after two
grazing seasons, CT content remained > 5 g kg-1 DM. Therefore, 20% high-tannin BFT in grass
mixtures can provide production and health benefits for ruminant livestock throughout the entire

56

grazing season, while 40% low- to medium-tannin BFT in grass mixtures will provide benefits
during spring and summer grazing. There were also no clear interactions between CT
concentration and endophyte type. Therefore, the hypothesis that high-tannin BFT cultivars in
combination with E+ tall fescue will produce highly productive and persistence pastures by
binding with alkaloids to neutralize their toxicity cannot be accepted from the results of this
study.
Overall, the high-tannin German cultivar, Oberhaunstädter, proves the most suitable BFT
cultivar for grazing based on superior agronomic performance and nutritive value and can be
grown successfully in mixtures with tall fescue, regardless of the endophyte type. However,
planting novel endophyte-infected tall fescue will cost more money without providing greater
benefits than E- tall fescue. Therefore, E- tall fescue is the most economically efficient
endophyte type to grow as a companion grass for high-tannin Oberhaunstädter. Such forage
mixtures will produce highly productive, persistent, and nutritious forage that would be suitable
for grazing in south-central Michigan.

57

APPENDIX

58

APPENDIX
CHAPTER 2 TABLES AND FIGURES
Table 2.1. Grazing start and end date, grazing duration, rest period, grazing animals, and
stocking density (animal mass per unit land) of birdsfoot trefoil-tall fescue mixtures over eight
grazing periods from 2015-2016 in Lansing, MI.
Rest
Stocking
Grazing
Start
End
Duration
Year
period Grazing animals
density
period
date
date
(h)
(d) †
(kg ha-1)
Suffolk and Dorset2015
June
17 June 18 June
29
108
20650
Polypay ewes
Aug.
11 Aug. 14 Aug.
68
53
Suffolk ewes
10625
2016

Nov.

10 Nov. 12 Nov.

50

87

Dorset-Polypay ewes

12750

May

17 May

20 May

71

77

Dorset-Polypay ewes

11613

June

20 June

24 June

87

30

Dorset-Polypay ewes

8925

Aug.

1 Aug.

4 Aug.

71

37

Dorset-Polypay ewes

11050

Sept.

6 Sept.

9 Sept.

74

32

Dorset-Polypay ewes

10200

Oct.
10 Oct. 14 Oct.
102
30
Dorset-Polypay ewes
13175
†Calculated as number of days between grazing end date to the following grazing period start
date, starting March 1 of each year.

59

Table 2.2. Analysis of variance for botanical composition, forage biomass, and grazing
preference of birdsfoot trefoil (BFT)-tall fescue (TF) mixtures over eight grazing periods (GP)
from 2015-2016 in Lansing, MI.
Grazing
Botanical composition
Forage biomass
preference†
Pregrazing
Herbage
Source
BFT TF Weeds Dead
forage yield utilization‡
BFT
TF
Block
ns§
ns
ns
ns
ns
ns
ns
ns
TF
ns
ns
***
ns
ns
ns
ns
***
BFT
ns
ns
ns
ns
ns
*
ns
ns
BFT x TF
ns
ns
ns
ns
*
ns
ns
ns
GP
*** ***
**
***
***
**
*
**
TF x GP
*
***
***
*
***
***
*
***
BFT x GP
ns
ns
ns
ns
***
**
ns
ns
BFT x TF x GP
ns
ns
ns
ns
ns
ns
ns
ns
*Significant at the 0.05 probability level.
**Significant at the 0.01 probability level.
***Significant at the 0.001 probability level.
†Observations in 2016 only.
‡Calculated as 100 x (pregrazing forage yield-postgrazing forage yield)/pregrazing forage yield
(Leep et al., 2002).
§ns, nonsignificant.

60

Table 2.3. Proportion of weed biomass in birdsfoot trefoil (BFT)-tall fescue (TF) mixtures prior
to the November 2015 grazing period (GP) and all GPs in 2016 in Lansing, MI.
2015†
2016
Cultivar
Nov.
May
June
Aug.
Sept.
Oct.
-------------------------------------% DM-----------------------------------Tall fescue
2.5
18.8
9.1
0.9
5.4
5.6
KY31
6.3
8.6
6.0
1.2
1.7
1.3
KY32
2.8
6.1
2.8
0.2
2.2
1.7
Martin II Protek
2.4
7.3
3.6
0.2
2.2
0.9
Martin II
1.35
1.35
1.35
1.35
1.35
1.35
SE‡
Contrasts§
-1.8
10.8***
4.3*
0.2
3.4
4.5*
E+1 vs. E-2
3
-1.5
-1.8
-2.0
-0.5
0.2
0.6
Novel vs. E-0.3
12.7***
6.3**
0.7
3.2
3.9
E+ vs. novel
†No plant separations were conducted for June or August GPs because mixtures were > 95%
BFT due to poor TF establishment.
‡Standard error of the least square means.
§Linear contrasts based on TF endophyte type. Values indicate numeric differences between
contrast groups.
1E+, endophyte-infected (KY31).
2E-, endophyte-free (KY32 and Martin II).
3Novel endophyte-infected (Martin II Protek).

61

Table 2.4. Botanical composition of birdsfoot trefoil (BFT)-tall fescue (TF) mixtures prior to the
November 2015 grazing period (GP) and all GPs in 2016 in Lansing, MI.
2015†
2016
Nov.
May
June
Aug.
Sept.
Oct.
SE§
-------------------------------------% DM------------------------------------Birdsfoot trefoil 66.0a‡
56.6ab
63.4a
53.5ab
42.0b
9.2c
3.65
Tall fescue
30.5c
33.2c
30.7c
35.1c
54.7b
88.2a
3.74
Weeds
3.5b
10.2a
5.4ab
0.6b
2.9b
2.4b
0.94
Dead
0.0b
0.0b
0.0b
10.8a
0.5b
0.2b
0.32
†No plant separations were conducted for June or August GPs because mixtures were > 95%
BFT due to poor TF establishment.
‡a-c: values within rows followed by different lowercase letters differ by the PDIFF option in the
LSMEANS statement (P < 0.05) (SAS Institute, Cary, NC).
§Standard error of the least square means.

62

Table 2.5. Spring stand density of birdsfoot trefoil (BFT) in mixtures with tall fescue in April
2015, 2016, and 2017 in Lansing, MI.
Cultivar
2015
2016
2017
-2
Birdsfoot trefoil
------------plants m -----------52
130
57
Oberhaunstädter
58
86
35
Bull
56
80
27
Pardee
80
96
34
AC Langille
73
84
39
Bruce
77
115
39
Common
72
98
45
Norcen 1
56
91
39
Norcen 2
13.1
13.1
13.1
SE†
Contrasts‡
High-tannin1 vs. all others2
-15
37**
20
3
High- vs. low-tannin
-19
33**
18
4
High- vs. medium-tannin
-4
47**
26
Medium- vs. low-tannin
-15
-14
-8
**Significant at the 0.01 probability level.
†Standard error of the least square means.
‡Linear contrasts based on BFT condensed tannin concentration. Values indicate numeric
differences between contrast groups.
1Oberhaunstädter.
2
Bull, Pardee, AC Langille, Bruce, Common, Norcen 1 and Norcen 2.
3AC Langille, Bruce, Common, Norcen 1 and Norcen 2.
4Bull and Pardee.

63

Table 2.6. Analysis of variance for condensed tannin (CT) concentrations and nutritive value of
birdsfoot trefoil (BFT)-tall fescue (TF) mixtures over eight grazing periods (GP) from 2015-2016
in Lansing, MI.
Source
CT
CP†
NDF
ADF
Lignin
RUP
NDFD
Block
ns‡
ns
ns
ns
ns
ns
ns
TF
ns
ns
ns
ns
ns
ns
ns
BFT
***
ns
ns
ns
***
ns
***
BFT x TF
ns
ns
ns
ns
ns
ns
ns
GP
***
***
***
***
***
***
***
TF x GP
***
***
**
***
***
***
**
BFT x GP
***
***
*
**
ns
*
ns
BFT x TF x GP
ns
ns
ns
ns
ns
ns
ns
*Significant at the 0.05 probability level.
**Significant at the 0.01 probability level.
***Significant at the 0.001 probability level.
†CP, crude protein; NDF, neutral detergent fiber; ADF, acid detergent fiber; RUP, rumen
undegradable protein; NDFD, NDF digestibility.
‡ns, nonsignificant.

64

Table 2.7. Condensed tannin concentrations of birdsfoot trefoil (BFT)-tall fescue mixtures prior to each grazing period in 2015 and
2016 in Lansing, MI.
2015
2016
Cultivar
June
Aug.
Nov.
May
June
Aug.
Sept.
Oct.
-1
Birdsfoot trefoil
--------------------------------------------g kg DM-------------------------------------------------Oberhaunstädter
21.6
31.7
16.7
17.0
22.2
20.0
16.9
7.5
Bull
13.0
21.6
8.2
9.2
16.4
9.8
8.3
3.6
Pardee
12.9
21.6
8.3
10.0
14.5
10.7
8.3
3.2
AC Langille
10.4
16.7
6.8
7.6
11.2
9.4
7.3
3.6
Bruce
10.7
16.6
6.8
7.5
10.5
8.7
7.1
3.2
Common
10.2
16.5
7.1
8.1
12.8
8.3
7.7
3.4
Norcen 1
10.7
17.5
7.2
8.1
11.7
7.7
7.3
3.3
Norcen 2
11.0
17.2
6.6
7.0
10.5
8.2
7.0
3.4
SE†
1.10
1.10
1.10
1.10
1.10
1.10
1.10
1.10
Contrasts‡
High-tannin1 vs. all others2 10.3*** 13.4***
9.4***
8.7***
9.7***
11.1***
9.3***
4.2***
High- vs. low-tannin3
11.0*** 14.8***
9.8***
9.3***
10.8***
11.6***
9.6***
4.2***
4
High- vs. medium-tannin
8.6***
10.1***
8.4***
7.4***
6.7***
9.7***
8.6***
4.1**
Medium- vs. low-tannin
2.4*
4.7***
1.3
2.0
4.1***
1.8
1.0
0.0
*Significant at the 0.05 probability level.
**Significant at the 0.01 probability level.
***Significant at the 0.001 probability level.
†Standard error of the least square means.
‡Linear contrasts based on BFT condensed tannin concentration. Values indicate numeric differences between contrast groups.
1Oberhaunstädter.
2
Bull, Pardee, AC Langille, Bruce, Common, Norcen 1 and Norcen 2.
3AC Langille, Bruce Common, Norcen 1 and Norcen 2.
4
Bull and Pardee.

65

Table 2.8. Condensed tannin (CT) concentrations and nutritive value averaged across all
birdsfoot trefoil-tall fescue mixtures prior to each grazing period in 2015 and 2016 in Lansing,
MI.
Grazing
Period
CT
CP†
NDF
ADF
Lignin
RUP
NDFD
-1
-1
--------------------g kg DM ---------------------g kg CP-- --g kg-1 NDF-2015
June
12.5bc‡
169c
478a
393a
68a
293bc
414c
Aug.
19.9a
168c
399bcd
328b
65ab
293bc
376c
Nov.

8.5d

173c

315e

230d
60b
250c
560b
2016
May
9.3cd
219a
281e
199e
32e
276bc
685a
June
13.7b
178c
397cd
296c
46d
282bc
428c
Aug.
10.4bcd
177c
416abc
298bc
53c
318b
418c
Sept.
8.7cd
231a
343de
234d
33e
308bc
676a
Oct.
3.9e
198b
461ab
290c
35e
390a
681a
SE§
0.76
3.7
12.0
5.5
1.2
11.0
12.7
†CP, crude protein; NDF, neutral detergent fiber; ADF, acid detergent fiber; RUP, rumen
undegradable protein; NDFD, NDF digestibility.
‡a-e: values within columns followed by different lowercase letters differ by the PDIFF option in
the LSMEANS statement (P<0.05) (SAS Institute, Cary, NC).
§Standard error of the least square means.

66

Table 2.9. Crude protein (CP) and rumen undegradable protein (RUP) concentrations of birdsfoot trefoil (BFT)-tall fescue mixtures
prior to each grazing period in 2015 and 2016 in Lansing, MI.
2015
2016
Cultivar
June
Aug.
Nov.
May
June
Aug.
Sept.
Oct.
-1
Birdsfoot trefoil
CP (g kg DM)
Oberhaunstädter
159
148
160
218
178
186
240
216
Bull
168
166
171
211
183
168
225
197
Pardee
166
171
169
219
177
174
231
192
AC Langille
168
175
174
218
173
178
229
198
Bruce
170
174
178
218
173
176
225
192
Common
175
174
174
224
181
180
238
199
Norcen 1
172
166
180
222
176
177
229
192
Norcen 2
172
168
181
223
183
175
234
196
SE‡
5.6
5.6
5.6
5.6
5.6
5.6
5.6
5.6
Contrasts§
High-tannin1 vs. all others2
-12
-23***
-15*
-1
0
10
10
21***
3
High- vs. low-tannin
-13
-23***
-17*
-3
0
9
9
20**
4
High- vs. medium-tannin
-8
-21**
-10
3
-3
15
13
21**
Medium- vs. low-tannin
-5
-3
-8
-6
3
-6
-3
-1
Birdsfoot trefoil
RUP (g kg-1 CP)
Oberhaunstädter
322
339
266
287
290
306
292
364
Bull
284
287
252
285
268
332
318
397
Pardee
291
288
245
275
279
323
314
397
AC Langille
314
291
254
275
290
320
312
384
Bruce
289
278
243
273
292
318
316
395
Common
287
285
241
268
269
309
294
394
Norcen 1
281
288
244
272
291
320
310
396
Norcen 2
276
290
256
271
278
316
312
393
SE
13.9
13.9
13.9
13.9
13.9
13.9
13.9
13.9

67

Table 2.9 (cont’d)
2015
Aug.

2016
Aug.

Cultivar
June
Nov.
May
June
Sept.
Oct.
Contrasts
High-tannin vs. all others
33*
53***
18
13
9
-14
-19
-30*
High- vs. low-tannin
32*
53***
18
15
6
-11
-17
-29†
High- vs. medium-tannin
34*
52***
18
7
16
-22
-24
-33†
Medium- vs. low-tannin
-2
1
1
8
-11
11
7
5
*Significant at the 0.05 probability level.
**Significant at the 0.01 probability level.
***Significant at the 0.001 probability level.
†Significant ̶̶at ̶̶the ̶̶0.10 ̶̶probability ̶̶level.
‡Standard error of the least square means.
§Linear contrasts based on BFT condensed tannin concentration. Values indicate numeric differences between contrast groups.
1Oberhaunstädter.
2Bull, Pardee, AC Langille, Bruce, Common, Norcen 1 and Norcen 2.
3AC Langille, Bruce, Common, Norcen 1 and Norcen 2.
4Bull and Pardee.

68

Table 2.10. Neutral detergent fiber (NDF) and acid detergent fiber (ADF) concentrations of the birdsfoot trefoil (BFT)-tall fescue
mixtures prior to each grazing period in 2015 and 2016 in Lansing, MI.
2015
2016
Cultivar
June
Aug.
Nov.
May
June
Aug.
Sept.
Oct.
-1
Birdsfoot trefoil
NDF (g kg DM)
Oberhaunstädter
471
400
351
286
401
384
321
419
Bull
481
398
311
295
377
438
355
467
Pardee
482
390
317
281
397
423
349
470
AC Langille
498
396
321
283
412
418
353
459
Bruce
485
402
302
285
406
422
353
471
Common
473
394
315
270
385
407
326
464
Norcen 1
467
405
300
279
410
419
351
472
Norcen 2
464
407
300
272
389
420
339
468
SE‡
15.7
15.7
15.7
15.7
15.7
15.7
15.7
15.7
Contrasts§
High-tannin1 vs. all others2
-7
1
41*
5
4
-37*
-25
-48**
High- vs. low-tannin3
-6
-1
43**
8
0
-33
-23
-47**
4
High- vs. medium-tannin
-10
6
36
-3
13
-46*
-31
-49**
Medium- vs. low-tannin
4
-7
7
10
-13
14
7
2
-1
Birdsfoot trefoil
ADF (g kg DM)
Oberhaunstädter
386
327
262
204
300
282
232
273
Bull
395
327
224
205
287
309
238
293
Pardee
395
322
232
199
297
302
236
293
AC Langille
408
327
232
199
303
297
236
289
Bruce
399
332
220
201
299
306
239
293
Common
391
325
234
194
290
292
227
290
Norcen 1
385
333
222
198
301
299
238
295
Norcen 2
383
334
215
194
289
299
229
292
SE
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0

69

Table 2.10 (cont’d)
2015
Aug.

2016
Aug.

Cultivar
June
Nov.
May
June
Sept.
Oct.
Contrasts
High-tannin vs. all others
-7
-2
37***
5
5
-19†
-3
-20†
High- vs. low-tannin
-7
-3
38***
7
3
-17
-2
-20†
High- vs. medium-tannin
-9
3
35***
2
8
-24*
-5
-21
Medium- vs. low-tannin
2
-6
3
5
-5
7
3
1
**Significant at the 0.01 probability level.
***Significant at the 0.001 probability level.
†Significant at the 0.10 probability level.
‡Standard error of the least square means.
§Linear contrasts based on BFT condensed tannin concentration. Values indicate numeric differences between contrast groups.
1Oberhaunstädter.
2Bull, Pardee, AC Langille, Bruce, Common, Norcen 1 and Norcen 2.
3AC Langille, Bruce, Common, Norcen 1 and Norcen 2.
4Bull and Pardee.

70

Table 2.11. Neutral detergent fiber digestibility (NDFD) and lignin concentrations of birdsfoot
trefoil (BFT)-tall fescue mixtures averaged over eight grazing periods from 2015-2016 in
Lansing, MI.
NDFD
Lignin
Cultivar
-1
--g kg NDF-- --g kg-1 DM-Birdsfoot trefoil
481
52
Oberhaunstädter
532
49
Bull
526
49
Pardee
534
49
AC Langille
541
48
Bruce
537
48
Common
539
49
Norcen 1
549
48
Norcen 2
9.7
0.9
SE‡
Contrasts§
High-tannin1 vs. all others2
-56***
4***
3
High- vs. low-tannin
-59***
4***
High- vs. medium-tannin4
-49***
3***
Medium- vs. low-tannin
-11†
1
***Significant at the 0.001 probability level.
†Significance at the 0.10 probability level.
‡Standard error of the least square mean.
§Linear contrasts based on BFT condensed tannin concentration. Values indicate numeric
differences between contrast groups.
1Oberhaunstädter.
2Bull, Pardee, AC Langille, Bruce, Common, Norcen 1 and Norcen 2.
3AC Langille, Bruce, Common, Norcen 1 and Norcen 2.
4Bull and Pardee.

71

Table 2.12. Relationship between forage dry weight (y) in Mg ha-1 and forage height (x) obtained
from rising plater meter (RPM) readings using the linear model y=a+bx for each grazing period
in 2015 and 2016 in Lansing, MI.
Year
Grazing period
n
RPM equation
r2†
0.78
2015
June
16
y=3.25+0.30x
0.15
Aug.
12
y=8.18+0.14x
0.70
Nov.
9
y=1.10+0.25x
0.83
2016
May
12
y=2.22+0.21x
y=2.94+0.19x
0.37
June
12
y=2.29+0.41x
0.86
Aug.
12
y=1.11+0.28x
0.69
Sept.
12
y=1.02+0.34x
0.77
Oct.
12
2
†r , coefficient of determination.

72

Table 2.13. Pregrazing forage yield of birdsfoot trefoil (BFT)-tall fescue (TF) mixtures prior to the June and November 2015 grazing
periods (GPs) and all GPs in 2016 in Lansing, MI.
2015†
2016
Cultivar
June
Nov.
May
June
Aug.
Sept.
Oct.
-1
Birdsfoot trefoil
------------------------------------------------Mg ha -----------------------------------------------Oberhaunstädter
6.62
2.98
5.78
5.80
6.90
4.88
4.85
Bull
6.77
2.90
5.92
5.91
6.7
4.55
4.88
Pardee
7.22
2.94
5.81
5.84
6.53
4.44
4.72
AC Langille
7.44
2.94
5.75
5.73
6.57
4.65
4.91
Bruce
7.28
2.96
5.83
5.93
6.82
4.71
4.88
Common
7.05
2.83
5.73
5.86
6.47
4.61
4.82
Norcen 1
6.97
2.83
5.54
5.8
6.79
4.63
4.88
Norcen 2
7.14
2.85
5.72
5.67
6.65
4.62
4.85
SE‡
0.111
0.111
0.111
0.111
0.111
0.111
0.111
Contrasts§
High-tannin1 vs. all others2
-0.50***
0.09
0.03
-0.01
0.25
0.28
0.00
3
High- vs. low-tannin
-0.55***
0.10
0.07
0.01
0.24
0.24
-0.02
4
High- vs. medium-tannin
-0.37*
0.06
-0.08
-0.07
0.29
0.39*
0.05
Medium- vs. low-tannin
-0.18
0.04
0.15
0.08
-0.05
-0.15
-0.07
Tall fescue
KY31
7.02
2.92
5.74
5.89
6.64
4.74
5.04
KY32
7.06
2.83
5.55
5.67
6.58
4.29
4.56
Martin II Protek
7.05
2.79
5.71
5.86
6.59
4.70
4.67
Martin II
7.12
3.07
6.04
5.84
6.90
4.82
5.12
SE
0.101
0.101
0.101
0.101
0.101
0.101
0.101
Contrasts¶
E+5 vs. E-6
-0.07
-0.04
-0.05
0.13
-0.10
0.19
0.20
7
Novel vs. E-0.04
-0.16
-0.09
0.10
-0.14
0.14
-0.17
E+ vs. novel
-0.03
0.13
0.04
0.03
0.04
0.04
0.36

73

Table 2.13 (cont’d)
2015
Cultivar
GP main effect

June

Nov.

May

June

2016
Aug.

Sept.

Oct.

7.06a#
2.90e
5.76c
5.82c
6.68b
4.64d
4.85d
SE
0.066
0.066
0.066
0.066
0.066
0.066
0.066
***Significant at the 0.001 probability level.
†August 2015 GP was not included in analysis due to a poor rising plate meter calibration.
‡Standard error of the least square means.
§Linear contrasts based on BFT condensed tannin concentration. Values indicate numeric differences between contrast groups.
¶Linear contrasts based on TF endophyte type. Values indicate numeric differences between contrast groups.
#a-d: values within rows are followed by different lowercase letters differ by the PDIFF option in the LSMEANS statement (P < 0.05)
(SAS Institute, Cary, NC).
1Oberhaunstädter.
2Bull, Pardee, AC Langille, Bruce, Common, Norcen 1 and Norcen 2.
3AC Langille, Bruce, Common, Norcen 1 and Norcen 2.
4Bull and Pardee.
5E+, endophyte-infected (KY31).
6E-, endophyte-free (KY32 and Martin II).
7Novel endophyte-infected (Martin II Protek).

74

Table 2.14. Herbage utilization of birdsfoot trefoil (BFT)-tall fescue (TF) mixtures prior to the June and November 2015 grazing
periods (GPs) and all GPs in 2016 in Lansing, MI.
2015†
2016
Cultivar
June
Nov.
May
June
Aug.
Sept.
Oct.
--------------------------------------------------------%-------------------------------------------------------Birdsfoot trefoil
Oberhaunstädter
26.1
19.5
34.6
25.1
31.8
32.3
24.7
Bull
24.6
20.1
33.1
25.1
23.9
24.3
21.6
Pardee
29.4
21.1
31.3
24.2
23.8
25.0
20.8
AC Langille
31.5
23.0
32.2
22.7
20.1
23.4
17.6
Bruce
30.3
23.1
29.0
23.4
23.7
26.4
20.5
Common
27.9
22.3
33.3
24.0
20.2
27.0
22.3
Norcen 1
27.1
20.3
30.1
24.7
26.1
24.2
19.1
Norcen 2
28.1
21.9
34.5
21.5
22.3
23.6
23.9
SE‡
2.12
2.12
2.12
2.12
2.12
2.12
2.12
Contrasts§
High-tannin1 vs. all others2
-2.4
-2.2
2.6
1.5
8.9***
7.5**
3.8
High- vs. low-tannin3
-2.9
-2.6
2.8
1.9
9.3***
7.4**
4.0
4
High- vs. medium-tannin
-1.0
-1.1
2.3
0.5
7.9**
7.7**
3.5
Medium- vs. low-tannin
-2.0
-1.6
0.4
1.4
1.3
-0.3
0.5
Tall fescue
KY31
27.5
23.9
34.6
24.5
24.8
25.8
28.0
KY32
27.3
20.1
31.4
22.3
24.3
23.6
19.1
Martin II Protek
28.4
20.7
33.3
26.1
23.0
29.1
18.8
Martin II
29.2
21.0
29.7
22.4
23.9
24.5
19.3
SE
1.97
1.97
1.97
1.97
1.97
1.97
1.97
Contrasts¶
E+5 vs. E-6
-0.7
3.4
4.0
2.2
0.7
1.8
8.8**
7
Novel vs. E0.1
0.1
2.7
3.7
-1.0
5.0
-0.4
E+ vs. novel
-0.8
3.2
1.3
-1.5
1.8
-3.3
9.2*

75

Table 2.14 (cont’d)
2015
Cultivars
GP main effect

June

Nov.

May

June

2016
Aug.

Sept.

Oct.

28.1ab#
21.4b
32.3a
23.8b
24.0b
25.8ab
21.3b
SE
1.41
1.41
1.41
1.41
1.41
1.41
1.41
*Significant at the 0.05 probability level.
**Significant at the 0.01 probability level.
***Significant at the 0.001 probability level.
†Calculated as 100 x (pregrazing forage yield - postgrazing forage yield)/pregrazing forage yield (Leep et al., 2002); August 2015 GP
was not included in analysis due to a poor rising plate meter calibration.
‡Standard error of the least square means.
§Linear contrasts based on BFT condensed tannin concentration. Values indicate numeric differences between contrast groups.
¶Linear contrasts based on TF endophyte type. Values indicate numeric differences between contrast groups.
#a-d: values within rows are followed by different lowercase letters differ by the PDIFF option in the LSMEANS statement (P < 0.05)
(SAS Institute, Cary, NC).
1Oberhaunstädter.
2Bull, Pardee, AC Langille, Bruce, Common, Norcen 1 and Norcen 2.
3
AC Langille, Bruce, Common, Norcen 1 and Norcen 2.
4Bull and Pardee.
5E+, endophyte-infected (KY31).
6E-, endophyte-free (KY32 and Martin II).
7Novel endophyte-infected (Martin II Protek).

76

Table 2.15. Sheep preference of birdsfoot trefoil (BFT) and tall fescue (TF) in mixed plots after
each grazing period (GP) in 2016 in Lansing, MI.
Grazing preference†
May
June
Aug.
Sept.
Oct.
Tall fescue
Tall fescue score
KY31
5
7
7
6
7
KY32
6
7
7
7
7
Martin II Protek
5
7
7
6
5
Martin II
4
7
6
5
4
SE‡
0.3
0.3
0.3
0.3
0.3
Contrasts§
E+1 vs. E-2
0
0
0
0
1***
3
Novel vs. E0
0
0
0
0
E+ vs. novel
0
-1
0
-1
2***
GP main effect
Tall fescue
5d¶
7a
7ab
6bc
6cd
SE
0.2
0.2
0.2
0.2
0.2
Birdsfoot trefoil score
Birdsfoot trefoil
8a
7a
6b
7a
8a
SE
0.2
0.2
0.2
0.2
0.2
***Significant at the 0.001 probability level.
†Grazing preference assessed as visually removal of tillers for both BFT and TF separately after
all grazing periods in 2016 (0: 0%; 1: <10%; 2: 11-25%; 3: 26-40%; 4: 41-50%; 5: 51-60%; 6:
61-75%; 7:76-90%; 8: 91-99%; 9: 100%).
‡Standard error of the least square means.
§Linear contrasts based on TF endophyte type. Values indicate numeric differences between
contrast groups.
¶a-b: values within rows followed by different lowercase letters differ by the PDIFF option in the
LSMEANS statement (P < 0.05) (SAS Institute, Cary, NC).
1E+, endophyte-infected (KY31).
2E-, endophyte-free (KY32 and Martin II).
3Novel endophyte-infected (Martin II Protek).

77

40
A

Air temperature (°C )

30

Minimum temperature

Maximum temperature

Average temperature

30-year average

20

10

0

-10

-20 Jan -15
160

Feb -15 Mar-15 Ap r-15 May -15 Ju n -15

Ju l-15 Au g -15 Sep -15

B

Oct-15 No v -15 Dec-15

Actual

Jan -16 Feb -16 Mar-16 Ap r-16 May -16 Ju n -16

Ju l-16 Au g -16 Sep -16

Oct-16 No v -16 Dec-16

30-year average

140

Precipitation (mm)

120
100
80
60
40
20

Dec-16

Nov-16

Oct-16

Sep-16

Aug-16

Jul-16

Jun-16

May-16

Apr-16

Mar-16

Feb-16

Jan-16

Dec-15

Nov-15

Oct-15

Sep-15

Aug-15

Jul-15

Jun-15

May-15

Apr-15

Mar-15

Feb-15

Jan-15

0

Month-Year

Figure 2.1. (A) Monthly minimum, maximum and average temperature (°C) in Lansing, MI from
2015-2016 with the 30-year average for comparison. (B) Total monthly precipitation (mm) in
Lansing, MI from 2015-2016 with the 30-year average for comparison.

78

Tall fescue percentage of forage dry matter

100
90
80
70
60
50
40
30
20
10
0

KY31
Contrasts
E+ vs. ENovel vs. EE+ vs. novel

2015
Nov.
-0.2
-3.5
3.3

KY32
May
-20.9
-10.4
-10.5

Martin II Protek
2016
June
Aug.
5.6
-0.9
3.6
-4.6
2.0
3.7

Martin II
Sept.
-5.3
-11.0
5.7

Oct.
0.8
3.2
-2.3

Figure 2.2. Proportion of tall fescue (TF) biomass (% of dry matter) in birdsfoot trefoil-TF mixtures prior to the November 2015
grazing period (GP) and all GPs in 2016 in Lansing, MI (least square means ± SE). Linear contrasts are based on TF endophyte type,
and values indicate numeric differences between contrast groups. Endophyte-infected (E+) TF in linear contrasts included KY31,
endophyte-free (E-) included KY32 and Martin II, and novel endophyte-infected included Martin II Protek. When no asterisks are
present, a significant endophyte effect was not observed (P > 0.05).
79

LITERATURE CITED

80

LITERATURE CITED

Aerts, R.J., T.N. Barry, and W.C. McNabb. 1999. Polyphenols and agriculture: Beneficial effects
of proanthocyanidins in forages. Agric. Ecosyst. Environ. 75: 1–12.
Bacon, C.W., P.C. Lyons, J.K. Porter, and J.D. Robbins. 1986. Ergot toxicity from endophyteinfected grasses: A review. Agron. J. 78: 106–116.
Belesky, D.P., and C.P. West. 2009. Abiotic stresses and endophyte effects. In: H. A. Fribourg,
D. B. Hannaway, and C. P. West, editors, Tall Fescue for the Twenty-First Century.
Agron. Monogr. 53rd ed. ASA, CSSA, and SSSA, Madison, WI. p. 49–64.
Blumenthal, M.J., and R.L. McGraw. 1999. Lotus adaptation, use, and management. In: P. R.
Beuselinck, editor, Trefoil: The Science and Technology of Lotus. CSSA Special
Publication 28. ASA, CSSA, Madison, WI. p. 97–119.
Bouton, J. 2007. The economic benefits of forage improvement in the United States. Euphytica
154: 263–270.
Bouton, J.H., R.N. Gates, D.P. Belesky, and M. Owsley. 1993. Yield and persistence of tall
fescue in the southeastern coastal plain after removal of its endophyte. Agron. J. 85:52–
55.
Brink, G.E., and M.D. Casler. 2012. Yield and nutritive value differences among cool- season
grasses. Forage and Grazinglands.: doi:10.1094/FG-2012-0619-01-RS.
Broderick, G.A., J.H. Grabber, R.E. Muck, and U.C. Hymes-Fecht. 2017. Replacing alfalfa
silage with tannin-containing birdsfoot trefoil silage in total mixed rations for lactating
dairy cows. J. Dairy Sci. 100: 3548-3562.
Burdine, K. 2007. Can I afford to seed novel-endophyte tall fescue? Forages at KCA. Special
publication, KFGC 2007-1.
https://www.uky.edu/Ag/Forage/KCA2007%20Proceedings.pdf. (accessed 11 July 2017).
Cassida, K.A., T.S. Griffin, J. Rodriguez, S.C. Patching, O.B. Hesterman, and S.R. Rust. 2000.
Protein degradability and forage quality in maturing alfalfa, red clover, and birdsfoot
trefoil. Crop Sci. 40: 209–215.
Cassida, K., J. Paling, and C. Kapp. 2016. 2016 Michigan forage variety test report: Forage
factsheet #17-01. http://forage.msu.edu/wp-content/uploads/2014/07/2014-ForageVariety-Test-REPORT_Final.pdf. (accessed 19 April 2017).
Dixon, P.M. 2017. Should blocks be fixed or random? Proceedings ̶̶of ̶̶the ̶̶28’th ̶̶Annual ̶̶Kansas ̶̶
State University Conference on Applied Statistics in Agriculture.

81

Doran-Browne, N., R. Behrendt, R. Kingwell, and R. Eckard. 2015. Modelling the potential of
birdsfoot trefoil (Lotus corniculatus) to reduce methane emissions and increase
production on wool and prime lamb farm enterprises. Anim. Prod. Sci. 55: 1097–1105.
Edwards, D., and J.J. Berry. 1987. The efficiency of simulation-based multiple comparisons. Int.
Biometric Soc. 43: 913–928.
Ferraro, F.P., R.L.G. Nave, R.M. Sulc, and D.J. Barker. 2012. Seasonal variation in the rising
plate meter calibration for forage mass. Agron. J. 104: 1–6.
Fribourg H.A. and J.C. Waller. 2005. Neotyphodium research and application in the USA. In:
C.A. Roberts, C.P. West, and D.E. Spiers, editors, Neotyphodium in Cool-Season
Grasses. Blackwell Publishing, Ames, Iowa. p. 3–22.
Gebrehiwot, L., P. R. Beuselinck, and C. A. Roberts. 2002. Seasonal variations in condensed
tannin concentration of three Lotus species. Agron. J. 94: 1059-1065.
Grabber, J.H., W.K. Coblentz, H. Riday, T.C. Griggs, D.H. Min, J.W. MacAdam, and K.A.
Cassida. 2015. Protein and dry-matter degradability of European- and Mediterraneanderived birdsfoot trefoil cultivars grown in the colder continental USA. Crop Sci. 55:
1356-1364.
Grabber, J.H., H. Riday, K.A. Cassida, T.C. Griggs, D.H. Min, and J.W. MacAdam. 2014. Yield,
morphological characteristics, and chemical composition of European- and
Mediterranean-derived birdsfoot trefoil cultivars grown in the colder continental United
States. Crop Sci. 54: 1893–1901.
Grabber, J.H., W.E. Zeller, and I. Mueller-Harvey. 2013. Acetone enhances the direct analysis of
procyanidin- and prodelphinidin-based condensed tannins in Lotus species by the
butanol-HCl-iron assay. J. Agric. Food Chem. 61: 2669–2678.
Grant, W.F. 2004. List of Lotus corniculatus (Birdsfoot trefoil), L. pedunculatus (Big trefoil), L.
glaber (Narrowleaf trefoil) and L. subbiflorus cultivars. Part 1. Cultivars with known or
tentative country of origin. Lotus Newsl. 34: 12–26.
Hagerman, A.E. 2011. The tannin handbook. http://www.users.miamioh.edu/hagermae/
(accessed 11 October 2014).
Hopkins, A., T.M. Martyn, R.H. Johnson, R.D. Sheldrick, and R.L. Lavender. 1996. Forage
production by two Lotus species as influenced by companion grass species. Grass Forage
Sci. 51: 343–349.
Hopkins, A.A., C.A. Young, D.G. Panaccione, W.R. Simpson, S. Mittal, and J.H. Bouton. 2010.
Agronomic performance and lamb health among several tall fescue novel endophyte
combinations in the south-central USA. Crop Sci. 50: 1552–1561.

82

Hymes-Fecht, U.C., G.A. Broderick, R.E. Muck, and J.H. Grabber. 2013. Replacing alfalfa or
red clover silage with birdsfoot trefoil silage in total mixed rations increases production
of lactating dairy cows. J. Dairy Sci. 96: 460–469.
Leep, R., P. Jeranyama, D. Min, T. Dietz, S. Bughrara, and J. Isleib. 2002. Grazing effects on
herbage mass and composition in grass – birdsfoot trefoil mixtures. Agron. J. 94: 1257–
1262.
Li, Y.-G., G. Tanner, and P. Larkin. 1996. The DMACA–HCl protocol and the threshold
proanthocyanidin content for bloat safety in forage legumes. J. Sci. Food Agric. 70: 89–
101.
Lisonbee, L.D., J.J. Villalba, and F.D. Provenza. 2009. Effects of tannin on selection by sheep of
forages containing alkaloids, tannins and saponins. J. Sci. Food Agric. 89: 2668–2677.
Lyman, T.D., F.D. Provenza, J.J. Villalba, and R.D. Wiedmeier. 2012. Phytochemical
complementarities among endophyte-infected tall fescue, reed canarygrass, birdsfoot
trefoil and alfalfa affect cattle foraging. Animal 6: 676–682.
MacAdam, J.W., T.C. Griggs, P.R. Beuselinck, and J.H. Grabber. 2006. Birdsfoot trefoil, a
valuable tannin-containing legume for mixed pastures. Forage and Grazinglands.:
doi:10.1094/FG-2006-0912-01-RV.
MacAdam, J.M., and J.J. Villalba. 2015. Beneficial effects of temperate forage legumes that
contain condensed tannins. Agriculture 5: 475–491.
MacAdam, J.W., R.E. Ward, T.C. Griggs, B.R. Min, and G.E. Aiken. 2011. C and S blood:
Average daily gain acid composition of cattle grazing the nonbloating legumes birdsfoot
trefoil and cicer milkvetch in the Mountain West. Prof. Anim. Sci. 27: 574–583.
Malinowski, D.P., and D.P. Belesky. 2006. Ecological importance of Neotyphodium spp. grass
endophytes in agroecosystems. Grassl. Sci. 52: 1-14.
Marley, C.L., R. Cook, R. Keatinge, J. Barrett, and N.H. Lampkin. 2003. The effect of birdsfoot
trefoil (Lotus corniculatus) and chicory (Cichorium intybus) on parasite intensities and
performance of lambs naturally infected with helminth parasites. Vet. Parasitol. 112:
147–155.
Marley, C.L., R. Fychan, and R. Jones. 2006. Yield, persistency and chemical composition of
Lotus species and varieties (birdsfoot trefoil and greater birdsfoot trefoil) when harvested
for silage in the UK. Grass Forage Sci. 61: 134–145.

83

Marshall, A., D. Bryant, G. Latypova, B. Hauck, P. Olyott, P. Morris, and M. Robbins. 2008. A
high-throughput method for the quantification of proanthocyanidins in forage crops and
its application in assessing variation in condensed tannin content in breeding programs
for Lotus corniculatus and Lotus uliginosus. J. Agric. Food Chem. 56: 974–981.
MAWN (Michigan Automated Weather Network). 2017. Michigan State Univ.
https://mawn.geo.msu.edu/ (accessed 26 April 2017).
McKenzie, D.B., Y.A. Papadopoulos, and K.B. McRae. 2004. Harvest management affects yield
and persistence of birdsfoot trefoil (Lotus corniculatus L.) in cool summer climates. Can.
J. Plant Sci. 84: 525–528.
McMahon, L.R., T.A. McAllister, B.P. Berg, W. Majak, S.N. Acharya, J.D. Popp, B.E.
Coulman, Y. Wang, and K.-J. Cheng. 2000. A review of the effects of forage condensed
tannins on ruminal fermentation and bloat in grazing cattle. Can. J. Plant Sci. 80: 469–
485.
Michigan State University EnviroWeather. 2017. Weather station network. Michigan State Univ.
http://www.agweather.geo.msu.edu/mawn/ (accessed 12 January 2017).
Miller, P.R., and N.J. Ehlke. 1996. Condensed tannins in birdsfoot trefoil: Genetic relationships
with forage yield and quality in NC-83 germplasm. Euphytica. 92: 383–391.
Min, B.R., T.N. Barry, G.T. Attwood, and W.C. McNabb. 2003. The effect of condensed tannins
on the nutrition and health of ruminants fed fresh temperate forages: a review. Anim.
Feed Sci. Technol. 106: 3–19.
Misselbrook, T.H., J.M. Powell, G.A. Broderick, and J.H. Grabber. 2005. Dietary manipulation
in dairy cattle: Laboratory experiments to assess the influence on ammonia emissions. J.
Dairy Sci. 88: 1765–1777.
Mueller-Harvey, I. 2006. Unravelling the conundrum of tannins in animal nutrition and health. J.
Sci. Food Agric. 86: 2010-2037.
NRC (National Research Council). 2001. Nutrient Requirements of Dairy Cattle. 7th rev. Natl.
Acad. Sci., Washington, DC
NOAA (National Centers for Environmental Information). 2017. Data tools: 1981-2010 normals.
https://www.ncdc.noaa.gov/cdo-web/datatools/normals (accessed 12 January 2017).
Oba, M., and M.S. Allen. 1999. Evaluation of the importance of the digestibility of neutral
detergent fiber from forage: Effects on dry matter intake and milk yield of dairy cows. J.
Dairy Sci. 82: 589–596.
Olson, G.L., S.R. Smith, T.D. Phillips, G.D. Lacefield, and D.C. Ditsch. 2012. 2012 Tall fescue
and bromegrass report. Univ. of Kentucky Agric. Exp. Stn. PR-646, Lexington, KY.

84

Owens, J., F.D. Provenza, R.D. Wiedmeier, and J.J. Villalba. 2012. Supplementing endophyteinfected tall fescue or reed canarygrass with alfalfa or birdsfoot trefoil increases forage
intake and digestibility by sheep. J. Sci. Food Agric. 92: 987–992.
Papadopoulos, Y.A., K.B. McRae, D. Gehl, and S.A.E. Filmore. 2008. Bruce birdsfoot trefoil.
Can. J. Plant Sci. 88: 1093–1094.
Ramírez-Restrepo, C.A., T.N. Barry, N. López-Villalobos, P.D. Kemp, and W.C. McNabb.
2004. Use of Lotus corniculatus containing condensed tannins to increase lamb and wool
production under commercial dryland farming conditions without the use of
anthelmintics. Anim. Feed Sci. Technol. 117: 85–105.
Roberts, C.A., P.R. Beuselinck, M.R. Ellersieck, D.L. Davis, and R.L. McGraw. 1993.
Quantification of tannins in birdsfoot trefoil. Crop Sci. 33: 675–679.
Rutter, S.M. 2006. Diet preference for grass and legumes in free-ranging domestic sheep and
cattle: Current theory and future application. Appl. Anim. Behav. Sci. 97: 17–35.
Sanderson, M.A., C.A. Rotz, S.W. Fultz, and E.B. Rayburn. 2001. Estimating forage mass with a
commercial capacitance meter, rising plate meter, and pasture ruler. Agron. J. 93: 1281–
1286.
Scrivner, J.H., D.M. Center, and M.B. Jones. 1986. A rising plate meter for estimating
production and utilization. J. Range Manag. 39: 475–477.
Sleugh, B., K.J. Moore, J.R. George, and E.C. Brummer. 2000. Binary legume–grass mixtures
improve forage yield, quality, and seasonal distribution. Agron. J. 92: 24-29.
Steiner, J.J., P.R. Beuselinck, S.L. Greene, J.A. Kamm, J.H. Kirkbride, and C.A. Roberts. 2001.
A description and interpretation of the NPGS birdsfoot trefoil core subset collection.
Crop Sci. 41: 1968–1980.
Stuedemann, J.A., and C.S. Hoveland. 1988. Fescue endophyte: History and impact on animal
agriculture. J. Prod. Agric. 1: 39–44.
Undersander, D., L. Greub, R. Leep, P. Beuselinck, J. Wedberg, D. Smith, K. Kelling, J. Doll, D.
Cosgrove, C. Grau, S. Peterson, M. Wipfli, and J. English. 1993. Birdsfoot trefoil for
grazing and harvested forage. North Cent. Reg. Ext. Publ. 474. Cooperative Extension
Service, University of Wisconsin, Madison, WI.
Waghorn, G. 2008. Beneficial and detrimental effects of dietary condensed tannins for
sustainable sheep and goat production-progress and challenges. Anim. Feed Sci. Technol.
147: 116–139.

85

Waghorn, G.C., M.J. Ulyatt, A. John, and M.T. Fisher. 1987. The effect of condensed tannins on
the site of digestion of amino acids and other nutrients in sheep fed on Lotus corniculatus
L. Br. J. Nutr. 57: 115–126.
Wang, Y., G.C. Waghorn, W.C. McNabb, T.N. Barry, M.J. Hedley, and I.D. Shelton. 1996.
Effect of condensed tannins in Lotus corniculatus upon the digestion of methionine and
cysteine in the small intestine of sheep. J. Agric. Sci. 127: 413–421.
Wen, L., R.L. Kallenbach, J.E. Williams, C.A. Roberts, P.R. Beuselinck, R.L. McGraw, and
H.R. Benedict. 2002. Performance of steers grazing rhizomatous and nonrhizomatous
birdsfoot trefoil in pure stands and in tall fescue mixtures. J. Anim. Sci. 80: 1970-1976.
Wen, L., C.A. Roberts, J.E. Williams, R.L. Kallenbach, P.R. Beuselinck, and R.L. McGraw.
2003. Condensed tannin concentration of rhizomatous and nonrhizomatous birdsfoot
trefoil in grazed mixtures and monocultures. Crop Sci. 43: 302–306.
Woodward, S.L., G.C. Waghorn, and P.G. Laboyrie. 2004. Condensed tannins in birdsfoot trefoil
(Lotus corniculatus) reduce methane emissions from dairy cows. Proc. New Zeal. Soc.
Anim. Prod. 64: 160–164

86

CHAPTER 3
SOIL RESPIRATION AND ROOT MORPHOLOGY OF BIRDSFOOT TREFOIL-TALL
FESCUE MIXTURES
ABSTRACT
Many studies have explored the effect of secondary compounds in grass-legumes pasture
mixtures on aboveground productivity, yet our understanding of belowground dynamics remains
limited. The objective of this study was to determine how birdsfoot trefoil (Lotus corniculatus
L., BFT) grown in mixtures with tall fescue [Schedonorus arundinaceus (Schreb.) Dumort.] can
influence soil CO2 respiration and root morphology, and whether condensed tannin (CT)
concentration or endophyte type can influence such processes. Sixteen BFT-tall fescue mixtures
were grazed by sheep from 2015-2016 in Lansing, MI, including four BFT cultivars ranging
from low to high CT concentrations and four tall fescue cultivars varying in endophyte type.
Differences in soil CO2 respiration over the 32-d incubation related to CT concentration or
endophyte type were minimal. Large differences were detected between years, with greater
cumulative C mineralization detected after the first grazing season than the second. Root traits
also varied greatly between years. The second grazing season had root traits associated with
superior resource acquisition (greater root length density, specific root length, root surface area
density, and smaller root diameter), most of which were highly correlated to the proportion of
tall fescue in forage mixtures. Results suggests large differences between species which
translates to a competitive advantage for tall fescue when grown in mixtures with BFT. No clear
CT or endophyte effect were detected among root traits, therefore differences were derived more
from variance between species than secondary compounds.

87

INTRODUCTION
Birdsfoot trefoil (Lotus corniculatus L., BFT) is a valuable forage legume for mixed
pastures, in part because it contains condensed tannins (CTs) beneficial for ruminant nutrition,
health, and environmental sustainability (MacAdam et al., 2006; Waghorn, 2008). It can be
grown in mixtures with tall fescue [Schedonorus arundinaceus (Schreb.) Dumort.], a highly
productive perennial cool-season grass commonly infected with the endophyte Neotyphodium
coenophialum [(Morgan- Jones and Gams) Glen Bacon and Hanlin] which increases plant
persistence in hot and dry climates (Belesky and West, 2009). However, the fungus produces
toxic ergot alkaloids to the detriment of livestock health and performance, commonly referred to
as ̶̶“fescue ̶̶toxicosis” ̶̶or ̶̶“summer ̶̶slump” ̶̶(Stuedemann and Hoveland, 1988). There is evidence
that when BFT is fed with endophyte-infected (E+) tall fescue, the CTs in BFT may bind with
alkaloids to neutralize their toxicity (Lisonbee et al., 2009; Lyman et al., 2012; Owens et al.,
2012). This offers the opportunity for combining BFT and tall fescue to produce a high yielding,
nutritious, and persistent forage mixture (Chapter 2).
Additionally, tannins and alkaloids have received increased attention for their influences
on belowground processes. One of these processes is microbial respiration, an important
biological indicator of soil quality for sustaining productive plant growth (Karlen et al., 1997).
Research suggests that the CTs in BFT may have the potential to impact soil carbon (C) and
nitrogen (N) dynamics. In forest ecosystems, tannins have received special attention, as reviewed
by Kraus et al. (2003), for their ability to influence belowground dynamics including litter
decomposition, nutrient cycling, N mineralization, and microbial activity.
However, tannins are complex polyphenols, therefore the consensus as to how CTs
influence soil respiration varies in the literature due to a wide diversity in tannin chemistry

88

(Fierer et al., 2001; Kraus et al., 2003; Kraus et al., 2004). Kraus et al. (2004) reported tannin
amendments to forest soil either had no effect or increased CO2 respiration depending on the
species in which the tannins were purified from, indicating that tannins did not prove toxic to the
microbial community. However, Madritch et al. (2007) argues that tannins never enter soils
alone, therefore the overall effect of tannins on soil respiration is likely negative, despite
temporary increases in microbial activity observed from tannin amendments in laboratory
studies. Others argue the inhibitory mechanism of tannins is dependent on the molecular weight
(Fierer et al., 2001). Fierer et al. (2001) reported amendments with low molecular weight tannins
in forest soils acted as a source of C, increasing C mineralization, while amendments with higher
molecular weight tannins may be toxic, primarily by binding extracellular substrates and
therefore limiting C and N mineralization. Condensed tannins in BFT have a high molecular
weight (McMahon et al., 2000), therefore high-tannin BFT cultivars may reduce microbial
activity by binding substrate needed for microbial growth. Despite evidence that tannins can alter
the rate and amount of C respired in forest soils, studies quantifying the effect of CTs on
microbial activity in agricultural soils are limited.
There is also evidence that endophyte infection in tall fescue may over time influence soil
microbial activity. Previous work has shown microbial communities are inhibited under E+
compared to endophyte free (E-) tall fescue pastures, reducing C mineralization which allows
more soil organic C and N to accumulate (Franzluebbers et al., 1999; Franzluebbers and
Stuedemann, 2005). Similarly, short- and long-term exposure to E+ leaves decomposing in the
soil has been shown to reduce soil C mineralization rates (Franzluebbers and Hill, 2005).
However, this is not consistent across all studies (Franzluebbers, 2006). Additionally, no studies
to date have examined this interaction in forage mixtures with BFT.

89

In this study, I hypothesize that mixtures containing high-tannin will reduce C
mineralization more so than mixtures with low- or medium-tannin BFT based on the proteinbinding ability of tannins (Madritch et al., 2007). Additionally, I hypothesize that mixtures with
E+ tall fescue will increase total C and N concentrations and reduce soil microbial activity more
so than mixtures with E- tall fescue (Franzluebbers et al., 1999; Franzluebbers and Stuedemann,
2005). Therefore, high-tannin BFT in mixtures with E+ tall fescue are expected to have greater
soil C and N concentration and lower C mineralization compared to all other BFT-tall fescue
mixtures.
Root systems are also essential for maintaining soil biological activity and regulating
changes to soil quality (Zobel, 2005a). Fine roots are particularly important because they are the
major pathway for water and nutrient absorption necessary for plant growth (Eissenstat, 1992).
Historically defined as < 2 mm in diameter (Böhm, 1979), fine roots represent up to 95% of the
total root length in many agriculture crops and forest species (Zobel, 2008). Most fine roots are
actually much smaller, however root diameter, the metric that classifies fine roots, is rarely the
focus of root research (Zobel et al., 2007). This is because their small size makes them difficult
to study and characterization of root systems is time-consuming, labor intensive, and prone to
inaccuracy (Costa et al., 2000; Wang and Zhang, 2009; Zobel et al., 2007). Consequently, studies
are lacking on root morphology of forage crops, particularly under grazing conditions; therefore
it was unclear how CT concentration or endophyte type would influence such traits in this study.
The aim of this study was to explore the effects of BFT-tall fescue mixtures on
belowground dynamics and to gain insight as to whether CT concentration or endophyte type can
influence such processes. Soil respiration and root morphology of BFT-tall fescue mixtures that
were grazed by sheep and varied in CT concentration and endophyte type were evaluated from

90

2015-2016 in Lansing, Michigan. The objectives were to (i) measure total soil C and N
concentrations and 32-d cumulative C mineralization of BFT-tall fescue mixtures and (ii) gain
knowledge of fine root traits (root diameter, length, volume, surface area), root length diameter
distribution, and fine root dry weight of these forage mixtures to better understand the
belowground processes occurring under mixed BFT-tall fescue pastures.

MATERIALS AND METHODS
Data for this chapter was collected simultaneously with the study discussed in Chapter 2.
For specifics regarding the Michigan State University Agronomy Farm weather and site
description, plot establishment, grazing periods, botanical composition and CT analysis, refer to
Chapter 2 Materials and Methods.

Experimental design
Treatments were arranged in a randomized complete block split-plot design with two
replications. ̶̶Main ̶̶plot ̶̶consisted ̶̶of ̶̶four ̶̶tall ̶̶fescue ̶̶cultivars ̶̶(‘KY31’, ̶̶‘KY32’, ̶̶‘Martin ̶̶II ̶̶
Protek’, ̶̶‘Martin ̶̶II’) ̶̶while sub-plot consisted of a subset of four BFT cultivars
(‘Oberhaunstädter’, ̶̶‘Pardee’, ̶̶‘Bruce’, ̶̶‘Norcen’) ̶̶selected ̶̶for ̶̶a ̶̶range ̶̶of ̶̶CT ̶̶concentrations. This
study had total of 16 BFT-tall fescue treatment combinations.

Soil sampling and analyses
Soil cores were collected in November 2015 and September 2016 with 2.54-cm diameter
stainless steel soil probes. Ten 15-cm deep cores were randomly sampled, coarsely crumbled by
hand and composited within each plot. Samples were then air-dried at 25°C and stored in paper

91

bags until analysis. To determine total soil C and N concentrations, air-dried samples from 2016
were ground with a Vial rotator (SampleTek Model 200, Mavco Industries, Inc., Science Hill,
KY) and 15-20 mg of soil was weighed into 5 x 9-mm tin capsules (Costech Analytical
Technologies, Valencia, CA) using a microbalance (Sartorius AG, Göttingen, Germany).
Samples were then analyzed for total soil C and total soil N using an elemental analyzer (Model
No. Costech ECS 4010, Costech Analytical Technologies, Valencia, CA).

Soil respiration
Soil respiration was measured over a 32-d incubation. Air-dried soil was passed through a
4.75-mm sieve and large pieces of organic debris were removed by hand to obtain homogenous
samples (Franzluebbers, 2016). Ten grams of soil were weighed into 125-mL glass serum bottles
and adjusted to 50% water-filled pore space (WFPS) with deionized water (Haney and Haney,
2010). Jars were then sealed with a rubber septum lid. A needle and syringe was used to collect
5-mL gas samples through the septa lid. Gas samples were then analyzed for CO2 concentration
using an infrared gas absorption analyzer (LI-820, LI-COR Biosciences, Lincoln, NE).
On each sampling date, headspace gas was sampled twice: time-one (T1) and time-two
(T2). For the first CO2 measurement, T1 CO2 readings were taken immediately after the water
addition and T2 CO2 readings were taken after a period of 3 h. For the remained of the
incubation, lids were remove and headspace gas was released prior to T1 readings. Jars were
then re-capped and T1 CO2 readings were immediately collected. The T2 CO2 readings were then
taken after a period of 3-5 h. The rate of C respired was calculated as the difference between T2
and T1 CO2 readings on each sampling date. The T2 CO2 readings were measured at 3, 49, 96,
145, 195, 241, 337, 407, 503, 577, 698, and 745 h over the course of 32-d for a total of 12

92

measurement days. Cumulative C respiration over the 32-d incubation was calculated using the
average respiration rate between two sampling dates and applying it to the time elapsed between
those measurements (Tiemann and Billings, 2011).

Root sampling and processing
Roots were sampled from soil cores collected in November 2015 and September 2016
using a 5-cm diameter stainless steel probe attached to a sliding hammer (Art's Manufacturing
and Supply, Inc., American Falls, ID). Two samples were collected to a depth of 15-cm
randomly within each plot, avoiding plant crowns. Composited samples were kept moist in
plastic bags and stored at 4°C until processing.
A ̶̶hydropneumatic ̶̶root ̶̶elutriation ̶̶washing ̶̶machine ̶̶(Gillison’s ̶̶Variety ̶̶Fabrication, ̶̶
Benzonia, MI) was used to wash roots from the soil cores (Smucker et al., 1982). The elutriation
system separates all debris less dense than the mineral fraction of the soil, retaining 99.4% of
roots > 0.05 mm (Pietola and Smucker, 2006). Composited sample were evenly split between
four chamber tubes and washed onto a 500-µm sieve for 10 min to ensure a high recovery rate.
Roots were then transferred onto a 250-µm sieve and gently washed with water (Thivierge et al.,
2016). Roots and any corresponding debris were placed in a 1:5 methanol-water preservative
solution and stored at 4°C. After the initial washing, roots were separated from remaining soil
particles and organic debris by flotation and transferred onto a fine sieve (Bolinder et al., 2002).
Washed samples were again placed in a 1:5 methanol-water preservative solution and stored at
4°C until analysis.

93

Root analyses
For root analyses, washed samples were placed onto a 225 x 225-mm Plexiglas tray and
floated in a 3-4 mm deep layer of water to help untangle roots (Costa et al., 2000). Roots were
then separated with plastic forceps to minimize overlap and any remaining non-root debris was
removed with tweezers. Roots were scanned (Epson Perfection V700) at 47.24 p mm -1 (1200 dpi)
and stored as TIFF files. Actual scanning area was 90 x 90 mm at four non-overlapping sites on
the tray to avoid extremely large files (Zobel et al., 2013).
Image analysis of scanned roots was performed using WinRhizo software (Reg. ver.
2016a, Regent Instruments, Quebec City, Quebec, Canada), including total root length, total root
surface area, total root volume, and average root diameter. Settings for WinRhizo were diameter
interpolation, maximum diameter sensitivity, and a length/width ratio of 4:1 to distinguish root
from non-root material, rejecting the latter from analysis (Zobel et al., 2007; Thivierge et al.,
2016). After image analysis, all root samples were oven-dried at 55°C for 48 h to determine root
dry weight.
To produce a sample value, root parameters were summed across the four sub-images
(Zobel et al., 2013), except for root diameter which was averaged across sub-images. Assuming
random distribution of roots on the scanning tray, soil volume was divided by 0.64 (180 x 180
mm / 225 x 225 mm) to provide an estimate of root parameters per unit soil (Zobel et al., 2013).
This was then used to determine: root length density (RLD, total root length divided by soil
volume); root surface area density (RSAD, root surface area divided by soil volume); root
volume density (RVD, root volume divided by soil volume); and root dry weight (total root
weight divided by soil volume). Specific root length (SRL) was calculated as total root length
divided by total root dry weight.

94

The image analysis system also categorized roots according to their diameter, with 100
diameter classes from 0.000 to 2.095 mm in 0.021 mm increments, equivalent to the pixel size of
the scanning resolution (Zobel et al., 2013). Relative diameter class length (rDCL) was used to
determine the distribution of root length among root diameter classes, calculated as sum of the
root lengths within a diameter class divided by total root length (Zobel et al., 2007, 2013).
WinRhizo produces artifacts at the selected scanning resolution which can cause issues with the
interpretation of diameter class length results (Zobel, 2008). Therefore, a three cell traveling
mean was used to smooth the data prior to analysis, as proposed by Zobel et al. (2013), without
eliminating any patterns in the data.

Statistics
Results were analyzed as a randomized complete block design with a split-plot treatment
arrangement. Tall fescue (TF) cultivars were randomly assigned to main plots and BFT cultivars
to sub-plot. All data were analyzed in SAS (Version 9.3, SAS Institute, Inc., Cary, NC) using the
MIXED procedure with BFT, TF, year, and block as fixed effects (Dixon, 2017). The block x TF
interaction term was the scale-appropriate random effect to evaluate the effect of tall fescue
cultivar, whereas the block x TF x BFT interaction was the scale-appropriate error term for the
BFT and the TF x BFT interaction effects. The Kenward-Rogers method was used to calculate
degrees of freedom. Year was considered a repeated measure and the appropriate covariance
model (CS) was used. No year effect was used for soil C and N concentration because analyses
were only conducted on samples collected in 2016. Linear contrasts were constructed to compare
significant treatment means and adjusted using the simulation-based method (Edwards and
Berry, 1987). Contrasts were grouped based on presumed differences in alkaloid toxicity among

95

tall fescue endophyte types and measured differences in CT content among BFT cultivars. Tall
fescue contrasts included: (i) E+ (KY31) vs. E- (KY32 and Martin II), (ii) Novel (Martin II
Protek) vs. E-, and (iii) E+ vs. novel. The BFT contrasts included: (iv) high-tannin
(Oberhaunstädter) vs. all others, (v) high- vs. low-tannin (Bruce and Norcen), (vi) high- vs.
medium-tannin (Pardee), and (vii) medium- vs. low-tannin. Linear contrasts for significant BFT
x TF interactions included a combination of the seven main effect contrasts. Differences between
years were analyzed using the PDIFF function of the LSMEANS command with the
SIMULATE adjustment to compare least square means. Pearson correlation coefficients were
calculated between morphological components, averaged by year, and all variables using the
PROC CORR in SAS. Significance was declared at P < 0.10.

RESULTS
Morphological components
Morphological components, including botanical composition and CT concentrations,
were averaged across all grazing periods within each year to produce a sample value. The
proportion of BFT and tall fescue in mixtures varied by year (P < 0.001), while CT concentration
varied by year (P < 0.001) and BFT (P < 0.001) main effects. For the botanical components, the
proportion of BFT in mixtures declined from 67.1% in 2015 to 46.0% in 2016 while the
proportion of tall fescue in mixtures increased from 29.7% in 2015 to 47.7% in 2016 (Figure
3.1). Concentrations of CTs declined from 2015 to 2016 (Table 3.1). Mixtures with the hightannin BFT cultivar had a significantly greater CT concentration than mixtures with all other
BFT cultivars, while mixtures with the medium-tannin BFT cultivar had greater CT

96

concentrations than mixtures with low-tannin BFT cultivars (Table 3.1); this was expected as it
was the basis for the linear contrasts used in this study.

Soil respiration
Total soil C, total soil N, and soil C:N ratio were not influenced by BFT, TF, or BFT x
TF interactions (P > 0.10). Therefore, soil respiration data is presented on a per unit soil basis
rather than per unit soil organic C. Total soil C averaged 14.6 ± 0.95 g kg-1, total soil N averaged
2.8 ± 0.18 g kg-1, and C:N ratio averaged 5:1 across all BFT-tall fescue mixtures in 2016.
Cumulative CO2 respiration over the 32-day incubation varied by year (P < 0.001, Table
3.2) and was greater in 2015 than 2016 (Figure 3.2). Total C respired was positively correlated to
the proportion of BFT in mixtures and CT concentration over both years (Table 3.3). Cumulative
C mineralization also varied by BFT (P < 0.10, Table 3.2), however lack of differences among
CT groups in the contrasts indicates this was caused by a factor other than CT concentration
(Table 3.4).

Root morphology
For root length density, there were BFT x TF (P < 0.05) and BFT x year (P < 0.10)
interactions as well as TF (P < 0.05) and year (P < 0.10) main effects (Table 3.2). Among BFT
cultivars, no differences in root length density were detected in 2015, while in 2016, mixtures
with high-tannin BFT had smaller root length density than mixtures with all other cultivars,
particularly compared to mixtures with low-tannin BFT cultivars (Table 3.5). Root length density
for mixtures with high-tannin BFT was 8.93 cm cm-3, 2.34 cm cm-3 smaller than mixtures with
low-tannin BFT cultivars at the end of the second grazing season. Averaged across all BFT-tall

97

fescue mixtures, root length density was smaller in 2015 compared to 2016 (Table 3.6). Among
mixtures, root length density was 2.25 cm cm-3 smaller when the medium-tannin cultivar was
grown in mixtures with novel endophyte-infected tall fescue than in mixtures with E- cultivars
(Table 3.7). These findings are consistent with the positive correlation between root length
density and the proportion of tall fescue in mixtures as well as the negative correlation to the
proportion of BFT in mixtures and CT concentration over both years (Table 3.3).
For specific root length, there were year (P < 0.001) and TF (P < 0.05) main effects
(Table 3.2). Specific root length for mixtures with E+ tall fescue was 68.09 m g-1, 9.51 and 14.38
m g-1 greater than mixtures with E- cultivars or novel endophyte-infected tall fescue, respectively
(Table 3.8). There was also a positive correlation between specific root length and the
proportion of tall fescue in mixtures over both years (Table 3.3). Across all BFT-tall fescue
mixtures, specific root length was greater at the end of the second grazing season than the first,
averaging 41.03 and 78.45 m g-1 in 2015 and 2016, respectively (Table 3.6).
There were no differences in average root diameter for BFT or TF main effects (P >
0.10), however there was a year (P < 0.001) main effect (Table 3.2). An overall larger root
diameter was detected at the end of the first grazing season than the second, averaging 0.321 and
0.225 mm in 2015 and 2016, respectively (Table 3.6). Average root diameter was also positively
correlated to the proportion of BFT in mixtures and CT concentration over both years (Table
3.3).
No clear treatment effects among cultivars were detected for relative diameter class
length within root diameter classes, however there was a clear pattern between years (Table 3.9).
There was a larger peak in rDCL in 2016 compared to 2015, with > 75% of the root length
expressed in the diameter classes < 0.233 mm in 2016 (Figure 3.3). In contrasts, < 55% of the

98

root length in 2015 was expressed in these root diameter classes, while > 75% of root length was
expressed in diameter classes < 0.402 mm. For both years, > 95% of the total root length was
expressed in diameter classes < 0.741 mm.
There was a BFT x TF (P < 0.05) interaction for root surface area density as well as TF
(P < 0.05) and year (P < 0.10) main effects (Table 3.2). Among BFT-tall fescue mixtures, root
surface area density was greater when low-tannin BFT cultivars were grown in mixtures with
novel endophyte-infected tall fescue than in mixtures E+ or E- cultivars (Table 3.7). There was
also a negative correlation between root surface area density and CT concentration over both
years (Table 3.3). Across all mixtures, root surface area density was greater in 2016, averaging
0.66 and 0.74 cm2 cm-3 in 2015 and 2016, respectively (Table 3.6).
There was a BFT x TF (P < 0.05) interaction for root volume density as well as TF (P <
0.10) and year (P < 0.01) main effects (Table 3.2). Similar to root surface area density, root
volume density was greater when low-tannin cultivars were grown in mixtures with novel
endophyte-infected tall fescue than in mixtures with E+ or E- cultivars (Table 3.7). There was a
negative correlation between root volume density and the proportion of tall fescue in mixtures
over both years (Table 3.3). Root volume density was also greater in 2015 when less tall fescue
was present compared to 2016 (Table 3.6).
For root dry weight, there was a TF (P < 0.10) main effect (Table 3.2). Root dry weight
ranged from 0.56 to 0.73 mg cm-3, however no differences were detected among endophyte
groups in the contrasts (P > 0.10, Table 3.8). There were also no correlations between root dry
weight and any of the morphological components (P > 0.10, Table 3.3).

99

DISCUSSION
Before discussing specifics of the data, it is necessary to discuss the study design because
it influences the interpretation of this data. This study was an add-on to an existing conventional
forage breeding project in which BFT populations will be selected from the forage mixture
where they will ultimately be used in. Therefore, all plots were mixtures of BFT and tall fescue
with no individual species monocultures.

Soil respiration
The hypothesis of this study was high-tannin BFT in mixtures with E+ tall fescue would
inhibit soil microbial activity more so than other BFT-tall fescue mixtures. Our results showed
no differences in cumulative C mineralization over the 32-d incubation among BFT-tall fescue
mixtures. However, large differences were observed between years, with greater total C respired
in 2015 than in 2016. Cumulative C mineralization was positively correlated to the proportion of
BFT in mixtures and CT concentration over both years, both of which were greater in 2015,
suggesting the BFT fraction of mixtures may have enhanced microbial activity more so than tall
fescue. It is also possible that greater CO2 respiration in 2015 could have been influences by
elevated respiration from plowing in 2014 to prepare the plot site.
Differences in the frequency and severity of grazing periods between the two years may
have influenced differences detected for cumulative C mineralization over the 32-d incubation.
Mixtures were subjected to fewer, less intense grazing periods in 2015 compared to 2016 (Table
2.1, Chapter 2). Therefore, a larger amount of plant material on the soil surface could have
contributed to increased CO2-C evolved from the soil that was observed in 2015 (Robertson et
al., 1995).

100

However, neither endophyte type or CT content influenced total soil C and N
concentrations or cumulative C mineralization in our study. Previous work has shown endophyte
infection can increase total soil organic C and N and reduce C mineralization in tall fescue
pastures that ranged from 8 to 20 years old (Franzluebbers et al., 1999; Franzluebbers and
Stuedemann, 2005). This is because the toxic alkaloids deposited in long-term pasture soil via
plant litter, manure, and urine can reduce soil microbial activity, eventually leading to greater
accumulation of soil organic C and N (Franzluebbers et al., 1999; Franzluebbers and
Stuedemann, 2005). These studies were both conducted under long-term tall fescue monoculture.
Therefore, the inhibitory action of the toxic endophyte may have been diluted in our study in
BFT mixtures and more time may be needed to detect differences since C mineralization was
measured from mixtures three and thirteen months after tall fescue was reseeded in 2015 and
2016, respectively. Others have similarly found no endophyte effect on soil C fraction after a
relatively short-term (60 week) exposure to E+ and E- tall fescue (Franzluebbers, 2006). Future
studies could incorporate tall fescue leaves into the soil to speed up this process since leaf tissue
has high concentrations of C and alkaloids (Franzluebbers and Hill, 2005).
While C mineralization over the 32-d incubation varied by BFT, this appeared to be
driven by differences among cultivars rather than CT concentration. Bruce reduced C
mineralization more so than other BFT cultivars, while Norcen, with a comparable CT
concentration, did not. Therefore, we found no clear inhibitory action of CT on microbial
activity. Other studies have shown varying responses of microbial activity to tannin amendments
(Fierer et al., 2001; Kraus et al., 2004; Madritch et al. 2007). Tannin purified from different
forest species have been reported to range from no effect to doubling the amount of total C
respired (Kraus et al., 2004), while others have reported a negative effect of tannins on soil

101

respiration desipite a transient increase in soil respiration observed in laboratory studies
(Madritch et al., 2007). In addition, Fierer et al. (2001) reported high molecular weight tannins
reduced C and N mineralization through substrate binding (Fierer et al., 2001). Condensed
tannins in BFT have a high molecular weight (McMahon et al., 2000), therefore we would expect
high-tannin BFT to reduce microbial activity by binding substrate needed for microbial growth.
The lack of CT effect observed in our study may be due to a dilution from tall fescue in mixtures.
Birdsfoot trefoil roots need to turn over and release CTs into soil before respiration can be
effected, therefore it is possible that BFT plants did not have enough time to release a substantial
amount of CTs in the soil before tall fescue dominated mixtures. Future studies should
incorporate tannins purified from BFT into the soil at different rates, as is often done in forest
studies, to better understand how the CTs in BFT affect soil microbial activity in pasture soils.
Once we understand the mechanism as to how CT concentration and endophyte type can
independently mediate changes in microbial activity, purified tannin and tall fescue foliage
should be combined to contribute to our understanding of how these two species interact to affect
soil microbial activity in pasture soils.

Root morphology
Root length density is an important root trait for studying belowground dynamics because
water and nutrient acquisition is largely influenced by root length (Eissenstat, 1992). Van
Noordwijk (1983) reported that root length densities ranging from 1–5 cm cm-3 and 1–10 cm
cm-3 can supply plants with adequate water and phosphorus, respectively. Based on these ranges,
root length densities measured in our study were adequate for successful belowground water and
nutrient uptake in all mixtures over both years. Values increased in 2016, with all mixtures

102

expressing root length densities > 10 cm cm -3, with the except of mixtures with high-tannin BFT.
These values are within the range reported in long-term pastures (Greenwood and Hutchinson,
1998), suggesting root dynamics of 2016 were more comparable to permanent pastures.
In 2016, the lower root length density detected for mixtures with high-tannin BFT may in
part be due to less grass in mixtures with the high-tannin BFT cultivar; however, botanical
differences among cultivars were not statistically significant. Alternatively, the high-tannin BFT
cultivar had a greater numerical stand density in April 2016 compared to other BFT cultivars
(Table 2.5, Chapter 2), suggesting more BFT and less tall fescue in mixtures. Root length density
was also greater in 2016 when the proportion BFT in mixtures declined to < 50%. This root
parameter was highly correlated to the proportion of tall fescue in mixtures over both years,
suggesting greater root length density for tall fescue roots compared to BFT roots. Others have
similarly reported forage grasses express greater root length density than legumes (Gould et al.,
2016). The lower root length density detected in mixtures with medium-tannin BFT and novel
endophyte-infected tall fescue suggests this combination would have the lowest competitive
advantage compared to the other mixtures.
Specific root length, the ratio of root length to root weight, is one of the most commonly
used parameters used to determine root function because it can characterize the economic aspects
of the root system (Ostonen et al., 2007). Plants with a high specific root length have relatively
long, thin roots and are able to invest less biomass to produce root length and therefore may have
an advantage in water and nutrient uptake over those with low specific root length in extreme
environments (Eissenstat, 1992). Our results indicate that specific root length was greater for
mixtures with E+ tall fescue than mixtures with other endophyte types, suggesting a finer root
system which allows for more efficient resource acquisition. Others have similarly reported

103

greater specific root length for E+ compared to E- tall fescue (Malinowski and Belesky, 2000)
and that species with greater specific root length tend to express greater aboveground biomass
(Thivierge et al., 2016). This suggests a competitive advantage for E+ tall fescue compared to
other endophyte types related to root morphology.
Specific root length was also greater in 2016. Similar to root length density, this suggests
tall fescue roots have greater specific root length than BFT roots, which is consistent with
previous findings in that forage grasses tend to have a relatively greater specific root length
compare to forage legumes (Gould et al., 2016; Yan et al., 2015). Additionally, specific root
length and average root diameter have previously been reported to be highly correlated (Yan et
al., 2015). In our study, average root diameter was greater in 2015, when the proportion of BFT
in mixtures was > 65%, compared to 2016. This suggests tall fescue roots have a narrower
diameter than BFT roots. Others have similarly reported forage legumes tend to have a wider
root diameter than grasses (Yan et al., 2015), particularly BFT roots (Gould et al., 2016).
Distribution of root length among various classes of root diameter also differed between
years. While root length distribution of fine roots was measured for 100 class intervals up to
2.095 mm, distribution was only reported up to 35 class intervals, or 0.741 mm. This was
because these class intervals encompassed 95% of the total root length for both years. Other
work has similarly reported 90 to 95% of the root length in pastures is made up of roots < 0.6
mm in diameter (Zobel, 2005b). For wheat (Triticum aestivum L.) and oilseed crops, Liu et al.
(2010) reported 85% of the root length was contributed by fine roots < 0.40 mm in diameter.
Additionally, Zobel et al. (2006) reported almost 90% of the total root length in chicory
(Cichorium intybus L.) was contributed to by thin roots in the 0.28 mm diameter class.

104

Therefore, it is essential to study fine roots of grass-legume mixtures, since they ̶̶make ̶̶up ̶̶the ̶̶
majority ̶̶of ̶̶the ̶̶total ̶̶root ̶̶length, ̶̶to ̶̶best ̶̶understand ̶̶root ̶̶function. ̶̶
In this study, increased root length in smaller diameter classes was detected from 2015 to
2016, with > 75% of the root length expressed in the diameter classes < 0.402 mm and < 0.233
mm in 2015 and 2016, respectively. This may be due to botanical changes to a higher proportion
of tall fescue in 2016, which is consistent with the finer root system observed for tall fescue.
However, an experiment evaluating root length distribution of tall fescue and BFT roots grown
independently should be conducted to test this.
Root surface area density and root volume density, an alternative measurement of root
mass (Greenwood and Hutchinson, 1998), can also be used to explain root function. Both root
surface area and root volume density were greater when low-tannin BFT cultivars were grown in
mixtures with novel endophyte-infected tall fescue than in mixtures with E+ or E- cultivars,
while no differences were detected among mixtures with other BFT cultivars. This suggests that
if a low-tannin BFT cultivar is desired, novel endophyte-infected tall fescue would be the best
endophyte type to use as a companion grass because greater root absorptive surface and volume
would allow for superior water and nutrient absorption.
Root surface area density and root volume density also varied by year. Root surface area
density was greater in 2016, which was negatively correlated to CT concentration over both
years. In contrast, root volume density was greater in 2015 and was positively correlated to the
proportion of BFT in mixtures over both years. This suggests that tall fescue has a greater
absorptive surface than BFT, particularly when grown in mixtures with low-tannin BFT
cultivars, while BFT has a larger root volume than tall fescue.

105

Differences in the frequency and severity of grazing periods between the two years may
have also influenced differences in root traits (Table 2.1, Chapter 2). Grazing has been show to
increase root length, surface area, average diameter, and root volume compared to systems
without grazing (Von Linsingen Piazzetta et al., 2014). Therefore, greater root length density,
specific root length, and root surface area observed in 2016 could have additionally been
stimulated by the more frequent and intense grazing periods in 2016 compared to those in 2015.
While root weight tends to decline with an increase in grazing intensity (Schuster, 1964),
no differences in root weight were detected between years. In addition, no differences in root
weight were detected relating to endophyte type contrary to our expectation that E+ tall fescue
would have greater root dry weight than E- cultivars (Franzluebbers, 2006). More time may be
needed in order for plants to fully develop their root system before differences in root weight can
be detected.

CONCLUSIONS
This study provides preliminary insight as to how BFT-tall fescue mixtures varying in CT
concentrations and endophyte type interact to affect belowground processes. Botanical changes
over time had a large influence on soil CO2 respiration and various root traits, more so than did
the secondary compounds. Cumulative C mineralization was greater when a larger proportion of
BFT was present in mixtures. When a larger proportion of tall fescue was present in mixtures,
root length density, specific root length, and root surface area density were greater, while
average root diameter was narrower. These findings suggest the long, thin roots of tall fescue
enables more efficient water and nutrient uptake compared to uptake by BFT roots. This
competitive advantage may in part explain the decline in the BFT proportion of BFT-tall fescue

106

mixtures observed between years. Increased grazing intensity during the second grazing seasons
also likely contributed to differences in belowground dynamics. Further work is needed to
characterize the root systems of BFT and tall fescue grown in monocultures to better understand
how CT concentration and endophyte type might independently influence belowground
dynamics which will then allow us to determine how such processes change when the two
species are grown in mixtures.

107

APPENDIX

108

APPENDIX
CHAPTER 3 TABLES AND FIGURES
Table 3.1. Condensed tannin (CT) concentrations of birdsfoot trefoil (BFT)-tall fescue mixtures
averaged over two years (2015-2016) in Lansing, MI.
Cultivar
CT
-1 DM-------g
kg
Birdsfoot trefoil
Oberhaunstädter
20.0
Pardee
11.8
Bruce
9.4
Norcen
9.4
SE†
0.55
Contrasts‡
High-tannin1 vs. all others2
9.82***
3
High- vs. low-tannin
10.62***
High- vs. medium-tannin4
8.22***
Medium- vs. low-tannin
2.40***
Year
2015
15.1a§
2016
10.2b
SE
0.49
***Significance at the 0.001 probability level.
†Standard error of the least square means.
‡Linear contrasts based on BFT condensed tannin concentration. Values indicate numeric
differences between contrast groups.
§a-b: values followed by different lowercase letters differ by the PDIFF option in the LSMEANS
statement (P < 0.10) (SAS Institute, Cary, NC).
1Oberhaunstädter.
2Pardee, Bruce and Norcen.
3Bruce and Norcen.
4
Pardee.

109

Table 3.2. Analysis of variance for soil respiration and root traits of birdsfoot trefoil (BFT)-tall fescue (TF) mixtures in 2015 and 2016
in Lansing, MI.
Soil respiration
Root traits‡
Source
32-d CO2-C
RLD
SRL
Average root diameter
RSAD
RVD
Root dry weight
ns§
ns
ns
ns
ns
ns
ns
Block
ns
*
*
ns
*
†
†
TF
†
ns
ns
ns
ns
ns
ns
BFT
ns
*
ns
ns
*
*
ns
BFT x TF
***
†
***
***
†
**
ns
Year
ns
ns
ns
ns
ns
ns
ns
TF x Year
ns
†
ns
ns
ns
ns
ns
BFT x Year
ns
ns
ns
ns
ns
ns
ns
BFT x TF x Year
*Significant at the 0.05 probability level.
**Significant at the 0.01 probability level.
***Significant at the 0.001 probability level.
†Significance at the 0.10 probability level.
‡RLD, root length density; SRL, specific root length; RSAD, root surface area density; RVD, root volume density.
§ns, nonsignificant.

110

Table 3.3. Pearson correlation coefficients between morphological components and all variables for birdsfoot trefoil-tall fescue
mixtures.
Soil respiration
Root traits§
Average root
Root dry
Morphological components‡
32-d CO2-C
RLD
SRL
diameter
RSAD
RVD
weight
0.23†
-0.58***
-0.54***
0.68***
-0.19
0.31*
0.12
Birdsfoot trefoil proportion
-0.18
0.53***
0.50***
-0.63***
0.17
-0.30*
Tall fescue proportion
0.29*
-0.53***
-0.34**
0.39**
-0.31*
0.05
Condensed tannin concentration
*Significant at the 0.05 probability level.
**Significant at the 0.01 probability level.
***Significant at the 0.001 probability level.
†Significant ̶̶at ̶̶the ̶̶0.10 ̶̶probability ̶̶level.
‡Averaged across all grazing periods within each year to obtain sample value.
§RLD, root length density; SRL, specific root length; RSAD, root surface area density; RVD, root volume density.

111

-0.12
-0.06

Table 3.4. Cumulative CO2 respiration over the 32-d incubation for birdsfoot trefoil (BFT)-tall
fescue mixtures averaged over two years (2015-2016) in Lansing, MI.
Cultivar
32-d CO2-C
Birdsfoot trefoil
----mg kg-1 soil---Oberhaunstädter
599
Pardee
584
Bruce
503
Norcen
584
SE†
27.5
Contrasts‡
High-tannin1 vs. all others2
42
3
High- vs. low-tannin
56
4
High- vs. medium-tannin
15
Medium- vs. low-tannin
40
†Standard error of the least square means.
‡Linear contrasts based on BFT condensed tannin concentration. Values indicate numeric
differences between contrast groups.
1Oberhaunstädter.
2Pardee, Bruce and Norcen.
3Bruce and Norcen.
4Pardee.

112

Table 3.5. Root length density of birdsfoot trefoil (BFT)-tall fescue mixtures in 2015 and 2016 in
Lansing, MI.
Cultivar
Root length density
------cm cm-3 soil-----Birdsfoot trefoil
2015
2016
Oberhaunstädter
6.74
8.93
Pardee
6.62
10.7
Bruce
6.15
11.46
Norcen
6.66
11.07
SE†
0.535
0.535
Contrasts‡
High-tannin1 vs. all others2
0.26
-2.15**
High- vs. low-tannin3
0.33
-2.34**
High- vs. medium-tannin4
0.11
-1.77
Medium- vs. low-tannin
0.22
-0.56
**Significance at the 0.01 probability level.
†Standard error of the least square means.
‡Linear contrasts based on BFT condensed tannin concentration. Values indicate numeric
differences between contrast groups.
1Oberhaunstädter.
2Pardee, Bruce and Norcen.
3Bruce and Norcen.
4Pardee.

113

Table 3.6. Root traits averaged across all birdsfoot trefoil-tall fescue mixtures in 2015 and 2016
in Lansing, MI.
Year
Root traits†
Units
2015
2016
SE‡
-3
RLD
cm cm soil
6.64b§
10.54a
0.291
SRL
m g-1 root dry weight
41.03b
78.45a
2.103
Average diameter
mm
0.321a
0.225b
0.0052
2
-3
RSAD
cm cm soil
0.66b
0.74a
0.023
3
-3
RVD
cm cm soil
0.005a
0.004b
0.0002
Root dry weight
mg cm-3 soil
0.71a
0.57a
0.057
†RLD, ̶̶root ̶̶length ̶̶density; ̶̶SRL, specific root length; RSAD, root surface area density; RVD,
root volume density.
‡Standard error of the least square means.
§a-b: values within rows followed by different lowercase letters differ by the PDIFF option in
the LSMEANS statement (P < 0.10) (SAS Institute, Cary, NC).

114

Table 3.7. Root traits of birdsfoot trefoil (BFT)-tall fescue (TF) mixtures averaged over two
years (2015-2016) in Lansing, MI.
Cultivar
Birdsfoot trefoil-tall fescue
KY31
KY32
Martin II Protek
Martin II
SE‡
Contrasts§
E+4 vs. E-5
Novel6 vs. EE+ vs. novel

Oberhaunstädter
Pardee
Bruce Norcen
-3
Root length density (cm cm soil)
7.14
9.35
9.14
8.39
7.95
7.86
7.88
7.85
7.19
6.93
8.50
10.58
9.04
10.51
9.69
8.64
0.647
0.647
0.647
0.647
High-tannin1
Medium-tannin2
Low-tannin3
-1.36
0.16
0.25
-1.31
-2.25†
1.03
-0.05
2.42
-0.77
2
-3
Root surface area density (cm cm soil)
KY31
0.58
0.77
0.71
0.65
KY32
0.67
0.63
0.66
0.61
Martin II Protek
0.60
0.58
0.73
0.97
Martin II
0.73
0.85
0.79
0.73
SE
0.057
0.057
0.057
0.057
Contrasts
High-tannin
Medium-tannin
Low-tannin
E+ vs. E-0.13
0.04
-0.02
Novel vs. E-0.10
-0.16
0.15†
E+ vs. novel
-0.02
0.20
-0.16†
Root volume density (cm3 cm-3 soil)
KY31
0.004
0.005
0.004
0.005
KY32
0.005
0.004
0.004
0.006
Martin II Protek
0.005
0.005
0.005
0.005
Martin II
0.004
0.004
0.008
0.005
SE
0.0006
0.0006
0.0006 0.0006
Contrasts
High-tannin
Medium-tannin
Low-tannin
E+ vs. E-0.001
0.001
-0.000
Novel vs. E-0.001
-0.001
0.002*
E+ vs. novel
-0.000
0.001
-0.002*
*Significance at the 0.05 probability level
†Significance ̶̶at ̶̶the ̶̶0.10 ̶̶probability ̶̶level.
‡Standard error of the least square means.
§Linear contrasts based on BFT condensed tannin concentration and TF endophyte type. Values
indicate numeric differences between contrast groups.
1Oberhaunstädter.
2Pardee.
3Bruce and Norcen.
4E+, endophyte-infected (KY31).
5E-, endophyte-free (KY32 and Martin II).
6Novel endophyte-infected (Martin II Protek).

115

Table 3.8. Specific root length and root dry weight for birdsfoot trefoil-tall fescue (TF) mixtures
averaged over two years (2015 and 2016) in Lansing, MI.
Cultivar
Specific root length
Root dry weight
Tall fescue
---m g-1 root dry weight-----mg cm-3 soil--KY31
68.09
0.56
KY32
61.51
0.57
Martin II Protek
53.71
0.70
Martin II
55.66
0.73
SE‡
3.359
0.062
Contrasts§
E+1 vs. E-2
9.51†
-0.09
Novel3 vs. E-4.87
0.05
E+ vs. novel
14.38*
-0.14
*Significant at the 0.05 probability level.
†Significant ̶̶at ̶̶the ̶̶0.10 ̶̶probability ̶̶level.
‡Standard error of the least square means.
§Linear contrasts based on TF endophyte type. Values indicate numeric differences between
contrast groups.
1E+, endophyte-infected (KY31).
2E-, endophyte-free (KY32 and Martin II).
3Novel endophyte-infected (Martin II Protek).

116

Table 3.9. Analysis of variance for relative diameter class length across 35 diameter classes for birdsfoot trefoil (BFT)-tall fescue (TF)
mixtures in 2015 and 2016 in Lansing, MI.
Diameter class (mm)
Source
0.032 0.042 0.064 0.085 0.106 0.127 0.148 0.169 0.191 0.212 0.232 0.254
Block
ns‡
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
TF
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
BFT
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
*
BFT x TF
*
*
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
Year
ns
ns
ns
†
***
***
*
***
†
***
ns
***
TF x Year
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
BFT x Year
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
BFT x TF x Year
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
0.275 0.296 0.318 0.339 0.36 0.381 0.402 0.423 0.445 0.466 0.487 0.508
Block
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
TF
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
BFT
*
*
†
ns
ns
ns
ns
ns
ns
ns
ns
ns
BFT x TF
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
Year
***
***
***
***
***
***
***
***
***
***
***
†
TF x Year
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
BFT x Year
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
BFT x TF x Year
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
0.529 0.55 0.572 0.593 0.614 0.635 0.656 0.677 0.699 0.72 0.741
Block
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
TF
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
BFT
ns
ns
ns
ns
ns
†
†
*
*
**
*
BFT x TF
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
Year
†
†
†
†
†
*
*
***
***
***
*
TF x Year
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
BFT x Year
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
BFT x TF x Year
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
*Significant at the 0.05 (*), 0.01 (**), and 0.001 (***) probability level.
†Significant ̶̶at ̶̶the ̶̶0.10 ̶̶probability ̶̶level.
‡ns, nonsignificant

117

80
Birdsfoot trefoil

a

Tall fescue

Percentage of forage dry matter

70
60
a

b
50
40
b
30
20
10
0
2015

2016

Figure 3.1. Botanical composition (% of dry matter) of birdsfoot trefoil-tall fescue mixtures in
2015 and 2016 in Lansing, MI (least square means ± SE). Values within each species with
different lowercase letters differ by the PDIFF option in the LSMEANS statement (P < 0.10)
(SAS Institute, Cary, NC) (SAS Institute, Cary, NC).

118

640

32-d CO2-C (mg CO2-C kg-1 soil)

a
600

b

560

520

480

440
2015

2016

Figure 3.2. Cumulative C mineralization over the 32-d incubation of birdsfoot trefoil-tall fescue
mixtures in 2015 and 2016 in Lansing, MI (least square means ± SE). Values with different
lowercase letters differ by the PDIFF option in the LSMEANS statement (P < 0.10) (SAS
Institute, Cary, NC).

119

0.16

2015

2016
a

0.14

a

a
0.10

a
b

a
0.08

b
a
a

b
b

a

b

a
b

a

b

a

b

a

b

a

a
b b a b ab
0.741

b

0.720

a

b

0.699

a

b

0.677

a

b

0.656

a

0.635

b

0.614

a

0.593

b

0.572

b

a

0.550

b

a

0.529

b

a

0.508

b

a

0.487

b

a

0.466

b

a

0.445

b

a

0.423

0.275

a

0.402

a

0.381

b a

0.254

0.233

0.212

0.191

0.169

0.148

0.127

0.106

0.085

0.00

aa a

a

0.064

0.02

a

b

0.032

a

b

0.360

a

a

0.339

0.04

b

0.318

b

0.296

0.06

0.042

Relative diameter class length

0.12

Diameter class (mm)

Figure 3.3. Relative diameter class length (rDCL) distribution across 35 diameter classes as a proportion of total root length of
birdsfoot trefoil-tall fescue mixtures in 2015 and 2016 in Lansing, MI (least square means ± SE). Values within each diameter class
with different lowercase letters differ by the PDIFF option in the LSMEANS statement (P < 0.10) (SAS Institute, Cary, NC).

120

LITERATURE CITED

121

LITERATURE CITED

Belesky, D.P., and C.P. West. 2009. Abiotic stresses and endophyte effects. In: H. A. Fribourg,
D. B. Hannaway, and C. P. West, editors, Tall Fescue for the Twenty-First Century.
Agron. Monogr. 53rd ed. ASA, CSSA, and SSSA, Madison, WI. p. 49–64.
Böhm, W. 1979. Methods of Studying Root Systems. Springer Verlag, New York.
Bolinder, M.A., D.A. Angers, G. Bélanger, R. Michaud, and M.R. Laverdière. 2002. Root
biomass and shoot to root ratios of perennial forage crops in eastern Canada. Can. J. Plant
Sci. 82: 731–737.
Costa, C., L.M. Dwyer, R.I. Hamilton, C. Hamel, L. Nantais, and D.L. Smith. 2000. A sampling
method for measurement of large root systems with scanner-based image analysis.
Agron. J. 92: 621–627.
Dixon, P.M. 2017. Should blocks be fixed or random? Proceedings ̶̶of ̶̶the ̶̶28’th ̶̶Annual ̶̶Kansas ̶̶
State University Conference on Applied Statistics in Agriculture.
Edwards, D., and J.J. Berry. 1987. The efficiency of simulation-based multiple comparisons. Int.
Biometric Soc. 43: 913–928.
Eissenstat, D.M. 1992. Costs and benefits of constructing roots of small diameter. J. Plant Nutr.
15: 763–782.
Fierer, N., J.P. Schimel, R.G. Cates, and J. Zou. 2001. Influence of balsam poplar tannin
fractions on carbon and nitrogen dynamics in Alaskan taiga floodplain soils. Soil Biol.
Biochem. 33: 1827–1839.
Franzluebbers, A.J. 2006. Short-term responses of soil C and N fractions to tall fescue endophyte
infection. Plant Soil 282: 153–164.
Franzluebbers, A.J. 2016. Should soil testing services measure soil biological activity? Agric.
Environ. Lett. 1:150009.: doi:10.2134/ael2015.11.0009.
Franzluebbers, A.J., N. Nazih, J.A. Stuedemann, J.J. Fuhrmann, H.H. Schomberg, and P.G.
Hartel. 1999. Soil carbon and nitrogen pools under low- and high-endophyte-infected tall
fescue. Soil Sci. Soc. Am. J. 63: 1687–1694.
Franzluebbers, A.J., and J.A. Stuedemann. 2005. Soil carbon and nitrogen pools in response to
tall fescue endophyte infection, fertilization, and cultivar. Soil Sci. Soc. Am. J. 69: 396–
403.

122

Franzluebbers, A.J., and N.S. Hill. 2005. Soil carbon, nitrogen, and ergot alkaloids with shortand long-term exposure to endophyte-infected and endophyte-free tall fescue. Soil Sci.
Soc. Am. J. 69: 404–412.
Gould, I.J., J.N. Quinton, A. Weigelt, G.B. De Deyn, and R.D. Bardgett. 2016. Plant diversity
and root traits benefit physical properties key to soil function in grasslands. Ecol. Lett.
19: 1140–1149.
Greenwood, K.L., and K.J. Hutchinson. 1998. Root characteristics of temperate pasture in New
South Wales after grazing at three stocking rates for 30 years. Grass Forage Sci. 53: 120–
128.
Haney, R.L., and E.B. Haney. 2010. Simple and rapid laboratory method for rewetting dry soil
for incubations. Commun. Soil Sci. Plant Anal. 41: 1493–1501.
Karlen, D.L., M.J. Mausbach, J.W. Doran, R.G. Cline, R.F. Harris, and G.E. Schuman. 1997.
Soil quality: A concept, definition, and framework for evaluation (A guest editorial). Soil
Sci. Soc. Am. J. 61: 4-10.
Kraus, T.E., R.A., Dahlgren, and R.J. Zasoski. 2003. Tannins in nutrient dynamics of forest
ecosystems - a review. Plant Soil 256: 41–66
Kraus, T.E.C., R.J. Zasoski, R.A. Dahlgren, W.R. Horwath, and C.M. Preston. 2004. Carbon and
nitrogen dynamics in a forest soil amended with purified tannins from different plant
species. Soil Biol. Biochem. 36: 309–321.
Lisonbee, L.D., J.J. Villalba, and F.D. Provenza. 2009. Effects of tannin on selection by sheep of
forages containing alkaloids, tannins and saponins. J. Sci. Food Agric. 89: 2668–2677.
Liu, L., Y. Gan, R. Bueckert, K. Van Rees, and T. Warkentin. 2010. Fine root distributions in
oilseed and pulse crops. Crop Sci. 50: 222–226.
Lyman, T.D., F.D. Provenza, J.J. Villalba, and R.D. Wiedmeier. 2012. Phytochemical
complementarities among endophyte-infected tall fescue, reed canarygrass, birdsfoot
trefoil and alfalfa affect cattle foraging. Animal 6: 676–682.
MacAdam, J.W., T.C. Griggs, P.R. Beuselinck, and J.H. Grabber. 2006. Birdsfoot trefoil, a
valuable tannin-containing legume for mixed pastures. Forage and Grazinglands.:
doi:10.1094/FG-2006-0912-01-RV.
Madritch, M.D., L.M. Jordan, and R.L. Lindroth. 2007. Interactive effects of condensed tannin
and cellulose additions on soil respiration. Can. J. For. Res. 37: 2063–2067.
Malinowski, D.P., and D.P. Belesky. 2000. Adaptations of endophyte-infected cool-season
grasses to environmental stresses: Mechanisms of drought and mineral stress tolerance.
Crop Sci. 40: 923–940.

123

McMahon, L.R., T.A. McAllister, B.P. Berg, W. Majak, S.N. Acharya, J.D. Popp, B.E.
Coulman, Y. Wang, and K.-J. Cheng. 2000. A review of the effects of forage condensed
tannins on ruminal fermentation and bloat in grazing cattle. Can. J. Plant Sci. 80: 469–
485.
Ostonen, I., Ü. Püttsepp, C. Biel, O. Alberton, M.R. Bakker, K. Lõhmus, H. Majdi, D. Metcalfe,
A.F.M. Olsthoorn, A. Pronk, E. Vanguelova, M. Weih, and I. Brunner. 2007. Specific
root length as an indicator of environmental change. Plant Biosyst. 141: 426–442.
Owens, J., F.D. Provenza, R.D. Wiedmeier, and J.J. Villalba. 2012. Supplementing endophyteinfected tall fescue or reed canarygrass with alfalfa or birdsfoot trefoil increases forage
intake and digestibility by sheep. J. Sci. Food Agric. 92: 987–992.
Pietola, L., and A.J.M. Smucker. 2006. Elimination of non-root residue by computer image
analysis of very fine roots. Comput. Electron. Agric. 53: 92–97.
Robertson, F.A., R.J.K. Myers, and P.G. Saffigna. 1995. Respiration from soil and litter in a
sown perennial grass pasture. Aust. J. Soil Res. 33: 167–178.
Schuster, J.L. 1964. Root development of native plants under three grazing intensities. Ecol. Soc.
Am. 45: 63–70.
Smucker, A.J.M., S.L. McBurney, and A.K. Srivastava. 1982. Quantitative separation of roots
from compacted soil profiles by the hydropneumatic elutriation system. Agron. J. 74:
500–503.
Stuedemann, J.A., and C.S. Hoveland. 1988. Fescue endophyte: History and impact on animal
agriculture. J. Prod. Agric. 1: 39–44.
Thivierge, M.N., D.A. Angers, M.H. Chantigny, P. Seguin, and A. Vanasse. 2016. Root traits
and carbon input in field-grown sweet pearl millet, sweet sorghum, and grain corn.
Agron. J. 108: 459–471.
Tiemann, L.K., and S.A. Billings. 2011. Changes in variability of soil moisture alter microbial
community C and N resource use. Soil Biol. Biochem. 43: 1837–1847.
Van Noordwijk, M. 1983. Functional interpretation of root densities in the field for nutrient and
water uptake. In: Böhm W, Kutschera L, Lichtenegger E, editors. Root ecology and its
practical application, International Symposium Gumpenstein 1982. Bundesanstalt
Gumpenstein, Irdning. p. 207–226.
Von Linsingen Piazzetta, H., A. De Moraes, T.M.D. Ribeiro, I.E. Sandini, S.B.C. Lustosa, and
A. Pelissari. 2014. Grazing and nitrogen on the growth of roots in the mixture of oat and
ryegrass (In Spanish, with English abstract). Semin. Agrar. 35: 2749–2765.

124

Waghorn, G. 2008. Beneficial and detrimental effects of dietary condensed tannins for
sustainable sheep and goat production-progress and challenges. Anim. Feed Sci. Technol.
147: 116–139.
Wang, M.B, and Q. Zhang. 2009. Issues in using the WinRhizo to determine physical
characteristics of plant fine roots. Acta Ecol. Sin. 29: 136–138.
Yan, Z., R.A. Culvenor, R.E. Haling, A. Stefanski, M.H. Ryan, G.A. Sandral, D.R. Kidd, H.
Lambers, and R.J. Simpson. 2015. Variation in root traits associated with nutrient
foraging among temperate pasture legumes and grasses. Grass Forage Sci.: doi:10.
1111/gfs.12199.
Zobel, R.W. 2005a. Primary and secondary root systems. In: R.W. Zobel and S. F. Wright,
editors, Roots and Soil Management: Interactions Between Roots and the Soil. Agron.
Monogr. 48. ASA, CSSA, and SSSA, Madison, WI. p. 3-14.
Zobel, R.W. 2005b. Tertiary root systems. In: R.W. Zobel and S. F. Wright, editors, Roots and
Soil Management: Interactions Between Roots and the Soil. Agron. Monogr. 48. ASA,
CSSA, and SSSA, Madison, WI. p. 35–56.
Zobel, R.W. 2008. Hardware and software efficacy in assessment of fine root diameter
distributions. Comput. Electron. Agric. 60: 178–189.
Zobel, R.W., K.A. Cassida, T.B. Kinraide, J.M. Gonzalez, and J.G. Foster. 2013. Morphological
response of forage chicory fine roots to manipulated soil phosphorus levels. J. Plant Nutr.
36: 1901–1914.
Zobel, R.W., T.B. Kinraide, and V.C. Baligar. 2007. Fine root diameters can change in response
to changes in nutrient concentrations. Plant Soil 297: 243–254.
Zobel, R.W., G.A. Alloush, and D.P. Belesky. 2006. Differential root morphology response to no
versus high phosphorus, in three hydroponically grown forage chicory cultivars. Environ.
Exp. Bot. 57: 201–208.

125