s”, . ,
w......?9i$.x

 

ti... 1
‘3! 5.. . . v .

‘ ~11
: r1:

 

c. I.

half

. 1.1::

.. a. z.
Exit.

 

 

 

.3 I

.! l‘bn L‘IA

oi .Rvilultll!
‘3‘:- 1)..

: , i» .2
I! ‘~ {Owl}. 311.!
V J.3i..§|lt,.|:1 J.

. :...>
P Lu...
r) . J13...

 

 

“£5315

 

205' I LIBRARY
2 I Michigan State
University

 

 

 

This is to certify that the
dissertation entitled

INVESTIGATION OF SPECIFIC STEREOTYPIC BEHAVIORS
IN HORSES

presented by

CARISSA I.. WICKENS

has been accepted towards fulfillment
of the requirements for the

PhD. degree in Animal Science

 

 

m M
Major Profeésor’s Signature
‘I/‘Z‘Y/o C;
Date

 

MSU is an Afiimrative Action/Equal Opportunity Employer

—-—<--——----_-_--~-|-I-I“I-l-l-0-.-l-l-l-.—O-I-I-O--C-I'D-I-I’I-I-l-U-I-.-l—'. -— * 4 -

 

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

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 KllProj/Acc8Pres/ClRC/DateDue.indd

 

INVESTIGATION OF SPECIFIC STEREOTYPIC BEHAVIORS IN HORSES

By

Carissa L. Wickens

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Animal Science

2009

ABSTRACT
INVESTIGATION OF SPECIFIC STEREOTYPIC BEHAVIORS IN HORSES
By

Carissa L. Wickens

Gastrointestinal irritation has been implicated in crib-biting (CB) in horses.
Eighteen horses, 9 CB and 9 non crib-biting (N CB), were used to determine 1)
prevalence and severity of gastric mucosa] damage, and 2) effect of concentrate feeding
on circulating gastrin concentrations in CB and NCB horses. Endoscopic examinations
(EE) of the squamous mucosa were performed and gastric fluid sampled after 24-28 hr
feed removal. Three days afier EE, blood was collected at 1400 hrs on pasture, following
12-hr feed removal (0 min), and at 60 and 120 min afier consuming 1 kg of a pelleted
concentrate. There were no differences in the number (P > 0.05) and severity (P > 0.05)
of ulcers and prevalence of hyperkeratosis (P > 0.05) between CB and NCB. There was
no difference (P = 0.87) in gastric pH of CB compared to NCB (3.92 vs. 3.78,
respectively, SEM = 0.60). There was no effect of CB (P = 0.56) on serum gastrin
concentration (14.22 vs. 12.16 pg/ml for CB and NCB, respectively, SEM = 2.46) with
free access to hay and pasture. Concentrate feeding increased gastrin concentration (P <
0.01). Serum gastrin concentration within CB was greater at times 60 and 120 min
compared to 0 min (P < 0.01). Compared to 0 min, serum gastrin concentration in NCB
tended to differ at 60 minutes (P = 0.07) and was greater at 120 minutes (P < 0.05).
Serum gastrin concentration at 60 minutes was greater (P < 0.05) in CB compared to
NCB. Compared to NCB, there was a trend for greater serum gastrin concentration in CB

horses at 120 minutes post-concentrate feeding (P = 0.06). The results suggest gastric

mucosal damage is not associated with CB in mature horses maintained on pasture.
Greater gastrin response to concentrate feeding in CB may indicate altered
gastrointestinal function in CB, which could result in a more acidic gastric environment
following the consumption of concentrate feed.

Weaving and crib-biting behavior (WCB) are two of the most recognizable equine
stereotypies (repetitive, invariant behaviors with no apparent function), and are viewed as
a management and welfare concern. A web—based questionnaire was developed to
investigate WCB in Michigan horses. Responses from 293 individuals were received
representing a total of 2,181 horses. The percentage of horses exhibiting weaving (W)
and CB was 2.7% and 5.2%, respectively. The proportion of respondents attempting to
stop CB (81.0%) was greater (P < 0.01) compared to W (37.5%). Methods employed to
stop or reduce W included increased tum-out (21.4%) and provision of toys (14.3%).
Cribbing collars (77.6%), alteration of CB surfaces (55.3%), and increased tum-out
(54.1%) were used most frequently to stop or reduce CB. Many owners used a
combination of methods. A negative association was identified between hours of daily
tum-out and the probability of WCB (P < 0.01 ). Risk of CB was greater for horses with
visual contact only (Adjusted odds ratio (OR) = 5.61) and for horses without social
contact (Adjusted OR = 6.26) compared to horses allowed tum-out with conspecifics.
Contrary to previous studies, a larger daily amount of concentrate was associated with a
decrease in the odds of W (P < 0.05) and CB (P < 0.01). Michigan horse owners
expressed concern about WCB and attempted to control WCB. Additional research into
the risk factors associated with WCB is needed, as well as careful documentation of the

development of these behaviors in order to improve horse management and welfare.

ACKNOWLEDGEMENTS

First, I would like to express tremendous gratitude to my advisor, Dr. Steve
Bursian, for his overwhelming support of my research program and of both my
professional and personal development. I would also like to thank my co-advisor, Dr.
Camie Heleski, for her involvement in my degree program. Additionally, I would like to
thank my guidance committee members, Dr. Nathalie Trottier, Dr. Dave Beede, Dr. John
Kaneene, and Dr. Cindy McCall for the input of their time and valuable advice pertaining
to my research and writing. I am especially thankful for Dr. Cindy McCall’s involvement
on the committee. Her participation and guidance was crucial to the dissertation research
and to the overall success of my program. Thank you also to Dr. Zanella for his guidance
and support during the early part of my program. I wish to thank the Auburn University
Horse Unit and Auburn University Large Animal Teaching Hospital for providing the
animals, facilities, administrative and technical assistance needed to complete the crib-
biting and gastric ulceration study. I am sincerely appreciative of Dr. Reid Hanson’s
involvement and encouragement during the study conducted in collaboration with
Auburn University. Thank you to Karen Clark and Deborah Rusz of the MSU Office for
Survey Research for their assistance with the Michigan Horse Behavior Survey study.
Special appreciation is extended to Jane Link, Jim Liesman, Dave Main, Barb Sweeney,
Kim Dobson, and Kathy Tatro for all of their assistance. Thank you to the many faculty
members in the ANS Department, graduate, and undergraduate students who greatly
enhanced my graduate experience. Finally, I wish to express my deepest love and
appreciation to my husband, Edward, for his unyielding support and encouragement. This

work is dedicated to him and to our beautiful daughter, the joy of my life, Eileen.

iv

TABLE OF CONTENTS

List of Tables ......................................................................................... vi
List of Figures ........................................................................................ vii
List of Abbreviations .............................................................................. viii
Introduction ........................................................................................... 1

Chapter 1. Crib—biting Behavior in Horses: A Review

Abstract ..................................................................................... 10
1. Introduction ............................................................................ 11
1.1.Aim of the review ............................................................ 13
2. Neuroendocrine physiology and brain function in crib-biting horses .......... 14
3. Gastrointestinal irritation and crib-biting behavior ................................ 18
4. Findings of survey and epidemiological research on crib-biting behavior. ....21
5. Motivation to crib-bite and attempts to stop the behavior ....................... 28
Conclusion .................................................................................. 29
Literature cited ............................................................................. 31

Chapter 2. Gastric Ulceration in Mature Horses with History of Crib-biting

Abstract .................................................................................................................. 37
Introduction ................................................................................. 38
Materials and Methods ..................................................................... 41
Results ....................................................................................... 47
Discussion ..................................................................................... 50
Conclusion .................................................................................. 58
Literature Cited ............................................................................ 70

Chapter 3. Investigating Weaving and Crib-biting Behavior in the Michigan Horse
Population Via Web-based Survey Methodology

Abstract ..................................................................................... 74
Introduction ................................................................................. 75
Materials and Methods .................................................................... 78
Results ........................................................................................ 83
Discussion ................................................................................... 90
Conclusion ................................................................................. 102
Literature Cited ........................................................................... 121
Chapter 4. Summary and Conclusions .......................................................... 125

Table 2-1.

Table 2-2.

Table 2-3.

Table 2-4.

Table 2-5.

Table 3-1.

Table 3-2.

Table 3-3.

Table 3-4.

Table 3-5.

Table 3-6.

Table 37

LIST OF TABLES

Characteristics of study horses ....................................................... 60
Main ingredients of commercial pelleted diet ...................................... 61
Ulcer number and severity scores by behavior ..................................... 62
Presence of hyperkeratosis by behavior ............................................. 63
Blood chemistry and pH values by behavior ....................................... 64
Examples ofquestion types used104
Levels of the explanatory variables (housing and management factors). . . ...105

Percentage of operations by number of horses .................................... 106
Percentage of operations by housing category .................................... 107
Sex distribution of weaving and crib-biting horses (percentage of horses in

each category) in the present study ................................................. 108

Breed distribution of weaving and crib-biting horses compared to general
Michigan horse population (number of horses in each breed
category)........... .......................................................................................... 109

The univariate relationships between farm management factors and the risk of
performing weaving and crib-biting behavior as estimated based on logistic
regression .............................................................................. 110

Table 3-8. The multivariate relationships between farm management factors and the risk
of performing weaving and crib-biting behavior as estimated based on logistic
regression .............................................................................. 112

Table 3-9. Content themes from the qualitative assessment of owner responses to

questions related to horse history (e. g. early environment) and management
strategies... .................................................................................................... 1 13
Appendix A. Michigan horse behavior questionnaire in Microsoft Word format ........ 130

vi

LIST OF FIGURES

Figure 2-1. Schematic representation of equine stomach anatomy ........................... 65

Figure 2-2. Mean crib-biting frequency of CB horses observed for 24 hours on
pasture .................................................................................. 66

Figure 2-3. Mean crib-biting duration of CB horses observed for 24 hours on pasture. . .67

Figure 2-4. Serum gastrin concentrations in CB and NCB horses following free access
to Berrnudagrass hay and pasture ................................................... 68

Figure 2-5. Serum gastrin concentrations in CB and NCB horses following 12-hour feed
removal, 60 and 120 minutes after consuming 1 kg of the pelleted diet ...... 69

Figure 3-1. The percentage of respondents agreeing with three separate statements about
weaving behavior .................................................................... 115

Figure 3-2. The percentage of respondents agreeing with three separate statements about
crib-biting behavior .................................................................. 116

Figure 3-3. Apparent triggers to weaving and crib-biting as reported by owners. . . . . . ...1 17

Figure 3-4. Percentage of respondents attempting to stop horses from performing
weaving and crib-biting behavior .................................................. 118

Figure 3-5. Methods employed by farms in an attempt to stop or reduce weaving
behavior ........................................................................................................ 119

Figure 3-6. Methods employed by farms in an attempt to stop or reduce crib-biting
behavior .............................................................................. 120

vii

CB

DA

EE

MES

MSU

NCB

OR

OSR

SST

W

LIST OF ABBREVIATIONS

Crib-biting

Dopamine

Endoscopic examinations
Michigan Equine Survey
Michigan State University
Non crib-biting

Odds ratio

Office for Survey Research
P-value

Serum separator tube

Weaving

WCB Weaving and Crib-biting

viii

INTRODUCTION

Stereotypies are defined as repetitive, relatively invariant patterns of behavior
with no apparent goal or function (Mason, 1991). Development and continued
performance of stereotypic behavior has been linked to sub-optimal environments
(Odberg, 1987; Cooper and Albentosa, 2005). Specifically, stereotypic behavior can
develop within the following contexts: when an animal is unable to execute a behavior
pattern that it is highly motivated to perform, such as nesting or feeding behavior; when it
cannot escape or avoid a stressful or fearful situation; or when it is kept in confinement or
social isolation (Mason, 1991). Stereotypies have been observed in several species kept in
captivity (Mason, 1991; Mason and Rushen, 2006). The performance of stereotypic
behavior has been used as an indicator of poor welfare (Broom, 1983; Mason and
Latharn, 2004) although it is often difficult to determine whether the behavior is the result
of poor welfare in the past or due to current adverse conditions. Ethologists and welfare
scientists have employed a multidisciplinary approach to address questions related to
stereotypic behavior including the use of behavioral and physiological measures, as well
as application of epidemiological research methods.

Studies conducted in Canada (Luescher et al., 1991) and the United Kingdom
(McGreevy et al., 1995a) have reported that greater than 13% of domesticated horses
exhibit stereotypies. The primary classifications assigned to stereotypic behavior patterns
observed in domestic horses (Houpt and McDonnell, 1993; Mills, 2002) and captive wild
horses, e. g. Przewalski horse, (Boyd, 1986) are oral and locomotor. Crib-biting and
weaving behavior are two of the most widely recognized equine stereotypies (Kiley-

Worthington, 1983). Crib-biting is an example of an oral stereotypic behavior in which

the horse anchors its top incisor teeth on a fixed object (6. g. fence, stall or building
structures), pulls backward, contracting the neck muscles, and draws air into the cranial
esophagus emitting an audible grunt (McGreevy et al., 1995a,b). Weaving is a locomotor
stereotypic behavior pattern characterized by a lateral swaying movement in which the
head, neck, forequarters and sometimes hindquarters are engaged (McGreevy et al.,
1995c). The average prevalences of crib-biting and weaving behavior in Europe and
Canada are 4.1% and 3.3 %, respectively (Nicol, 1999a). In the United States, the
reported prevalence of crib-biting behavior is 4.4%. Crib-biting and weaving are
recognized as both a management and a welfare concern, and many owners attempt to
physically prevent horses from engaging in these behaviors (McGreevy and Nicol, 1998;
McBride and Long, 2001 ). The primary problem with physical prevention, e. g.
attempting to stop crib-biting using a cribbing collar or by removing crib-biting surfaces,
is that these approaches fail to address the underlying cauSes of the behavior and may
further reduce equine welfare (McBride and Cuddeford, 2001).

Although the underlying cause of crib-biting behavior remains to be elucidated,
there is some evidence to suggest a link between crib-biting behavior and gastrointestinal
irritation. Nicol (1999b) proposed that the behavior is an adaptive response to gastric
acidity and that the act of crib-biting may raise gastric pH as a result of increased flow of
alkaline saliva. A study conducted by Moeller et a1. (2008) demonstrated that crib-biting
behavior stimulates salivation, lending support to this hypothesis. Crib-biting behavior
recently has been associated with gastric ulceration in foals, with gastric ulceration and
inflammation present in 60% of cribbing foals compared to 20% of non-cribbing foals

(Nicol et al., 2002). In addition, crib-biting foals had greater severity of ulceration and

inflammation upon initial endoscopic examination (Nicol et al., 2002). In adult horses,
crib-biting has been associated with lower basal and post-feeding gastric pH (Lillie et al.,
2004) and long-term treatment with antacids has reduced the frequency of crib-biting
(Mills and Macleod, 2002). However, an association between crib-biting behavior and
gastric ulceration in mature horses has never been documented.

Few experimental studies have been conducted specifically to address weaving
behavior. However, the provision of mirrors in the stable reduced the performance of
weaving (McAfee et al., 2002) suggesting that this behavior may be performed in
response to social isolation. Stable designs that increased visual horizons, such as open
stable doors providing access to views of adjacent horses and surrounding fields also
reduced weaving behavior (Cooper et al., 2000). Ninomiya et a1. (2007) found that
weaving behavior was mainly observed before feeding, and that investigation of bedding
was more likely to follow eating. These results suggested that weaving also may be
performed in an attempt to cope with frustration associated with meal anticipation.

Survey studies in the UK and Canada have demonstrated an association between
various management practices and stereotypic behavior. For example, Waters et a1.
(2002) found young horses fed concentrate feed post-weaning to be 4 times more likely
to develop crib-biting behavior than foals not receiving concentrate. Management factors
associated with a reduced risk of stereotypic behavior include increasing forage intake,
allowing visual contact between stalled horses, and increasing the amount of time spent
outside the stable (McGreevy et al., 1995a,c; Redbo et al., 1998). Management factors

associated with an increased risk of stereotypic behavior include the feeding of

concentrates (Redbo et al., 1998; Waters et al., 2002; Bachmann et al., 2003) and
stabling/stalling of foals prior to weaning (Parker et al., 2008)

Certain breeds of horses may be more likely to exhibit stereotypic behavior than
others. This was demonstrated by Bachmann et a1. (2003) in a survey of stereotypic
behavior in Swiss horses in which Warmbloods and Thoroughbreds were at 1.8 and 3.1
times greater risk of performing stereotypic behavior, respectively, compared to other
breeds. Vecchiotti and Galanti (1986) also have suggested the involvement of a genetic
component in the performance of stereotypic behavior with the finding that certain
Thoroughbred lines were more susceptible to developing stereotypic behavior.
Stereotypic behavior has never been investigated in horses in Michigan. Specifically,
there is a paucity of information regarding the prevalence and owner perceptions of, and
risk factors associated with crib-biting and weaving behavior within the Michigan horse
population.

The overall goal of this dissertation is to provide a further understanding of crib-
biting and weaving behavior in horses. The Specific objectives of this research were to: 1)
determine prevalence and severity of gastric mucosal damage in mature crib-biting
horses, 2) determine the effect of feeding on circulating gastrin concentrations in crib-
biting and non crib-biting horses, 3) assess the perceptions of weaving and crib-biting
behavior held by Michigan horse owners, and 4) investigate risk factors associated with
these behaviors in the Michigan horse population. The first and second objectives are
addressed in Chapter 2 using video endoscopy and measurement of serum gastrin
concentrations in crib-biting and non crib-biting horses following free-access to

pasture/hay and in response to pelleted concentrate feeding. We hypothesized that horses

exhibiting crib-biting behavior would have a higher degree of gastric mucosal damage
and greater serum gastrin response to concentrate feeding compared to non crib-biting
horses.

The third and fourth objectives are addressed in Chapter 3 utilizing a web-based
questionnaire. It was hypothesized that Michigan horse owners would express concern
regarding weaving and crib-biting behavior and that the majority of owners attempt to
stop or reduce performance of these behaviors. Additionally, the author hypothesized that
certain housing and feeding strategies, for example, those which limit a horse’s time out
of the stable, limit contact with other horses, or limit foraging opportunities, would result
in an increased probability of horses exhibiting weaving and/or crib-biting behavior.
Some of the specific predictions were that increased turn-out would reduce the
probability of a horse being a weaver or crib-biter and that the probability of being a
weaver or crib-biter would be increased in horses with no visual or tactile contact with
other horses. Furthermore, those horses fed larger amounts of concentrate would be more
likely to weave or crib-bite.

The rationale for hypothesizing greater mucosal damage and gastrin response to
concentrate feeding in mature crib-biting horses is based on the associations between
crib-biting behavior, concentrate feeding, and gastrointestinal irritation identified in the
literature. The rationale behind the hypotheses and predictions related to horse owner
perceptions and risk factors associated with weaving and crib-biting behavior in
Michigan are based on findings of previous survey and epidemiological research.
Weaving behavior was included in the Michigan horse behavior survey study for two

reasons. First, in comparison with crib-biting behavior, the number of studies dedicated

specifically to weaving remains low and we wanted to add to the body of knowledge
regarding this locomotor stereotypy. Secondly, because information about stereotypic
behavior has never been collected within the Michigan horse industry, we were interested
in learning about more than just one behavior. However, crib-biting behavior is the main
focus of the dissertation. Thus, the purpose of Chapter 1 is to provide the reader with a
relatively comprehensive review of what is currently known about crib-biting behavior in
horses. Chapter 4 provides a summary of the main findings of the dissertation research

and implications for the equine industry.

Literature Cited

Albright, J.D., Mohammed, H.O., Heleski, C.R., Wickens, C.L., Houpt, K.A., 2009. Crib-
biting in US horses: Breed predispositions and owner perceptions of aetiology.
Equine Vet. J. 41, doi: 10.2746/042516409X3 72584.

Bachmann, I., Audigé, L., Stauffacher, M., 2003. Risk factors associated with
behavioural disorders of crib-biting, weaving and box-walking in Swiss horses.
Equine Vet. J. 35(2), 158-163.

Boyd, L., 1986. Behavior problems of equids in zoos. In: CroweIl-Davis, S.L., Houpt,
K.A., (Eds), Behavior, The Veterinary Clinics of North America: Equine Practice
2(3), W.B. Saunders, Philadelphia, pp. 653-664.

Broom, D.M., 1983. Stereotypies as animal welfare indicators. In: Indicators Relevant to
Farm Animal Welfare (Ed. by D. Schmidt), pp. 81-87. The Hague: Martinus
Nijhoff.

Cooper, J .J ., Albentosa, M.J., 2005. Behavioural adaptation in the domestic horse:
potential role of apparently abnormal responses including stereotypic behaviour.
Livest. Prod. Sci. 92, 177-182.

Cooper, J .J ., McDonald, L., Mills, D.S., 2000. The effect of increasing visual horizons on
stereotypic weaving: implications for the social housing of stabled horses. Appl.
Anim. Behav. Sci. 69, 67-83.

Houpt, K.A., McDonnell, SM. 1993. Equine stereotypies. Compendium Continuing
Education 15, 1265-1271.

Kiley-Worthington, M. 1983. Stereotypes (sic) in horses. Equine Practice 5(1), 34-40.

Lillie, HQ, 2004. Comparison of gastric pH in crib-biting and non crib-biting horses.
Master’s thesis, Auburn University, Auburn.

Luescher, U.A., McKeown, D.B., Halip, J ., 1991. Reviewing the causes of obsessive-
compulsive disorders in horses. Vet. Med. 86, 527-530.

McAfee, L.M., Mills, D.S., Cooper, J .J ., 2002. The use of mirrors for the control of
stereotypic weaving behaviour in the stabled horse. Appl. Anim. Behav. Sci. 78,
159-173.

Mason, G.J. 1991. Stereotypies: a critical review. Anim. Behav. 41:1015-1037.

Mason, G.J., Latham, NR, 2004. Can’t stop, won’t stop: is stereotypy a reliable animal
welfare indicator? Anim. Welf. 13, $57-$69.

Mason, G., Rushen, J ., 2006. A decade-or-more’s progress in understanding stereotypic
behaviour. In: Mason, G., Rushen, J. (Eds.), Stereotypic animal behavior:
Fundamentals and applications to welfare. CAB International, Wallingford, p. 5-6.

McBride, S.D., Cuddeford, D., 2001. The putative welfare-reducing effects of preventing
equine stereotypic behaviour. Anim. Welf. 10, 173-189.

McBride, S.D., Long, L., 2001. Management of horses showing stereotypic behaviour,
owner perception and the implications for welfare. Vet. Rec. 148, 799-802.

McGreevy, P.D., Nicol, C .J ., 1998. Prevention of crib-biting: a review. Equine Vet. J.
Suppl. 27, 35-38.

McGreevy, P.D., French, N.P., Nicol, C.J. 1995a. The prevalence of abnormal behaviours
in dressage, eventing and endurance horses in relation to stabling. Vet. Rec. 137,
36-37.

McGreevy, P., Richardson, J .D., Christine, J .N., Lane, J .G. 1995b. Radiographic and
endoscopic study of horses performing an oral based stereotypy. Equine Vet. J. 27,
92-95.

McGreevy, P., Nicol, C.J., Cripps, R, Green, L., French, N., 1995c. Management factors
associated with stereotypic and redirected behaviour in the thoroughbred horse.
Equine Vet. J. 27, 86-91.

Mills, D.S., 2002. Recent advances in the treatment of equine stereotypic behavior.
[Serial online] Animal Behavior Cognition and Welfare Group. University of
Lincoln. Available from:
http://research.vet.upenn.edu/HavemeyerEquineBehaviorLabHomePage/Reference
LibraryHavemeyerEquineBehaviorLab/HavemeyerWorkshops/HorseBehaviorand
Welfare 1 3 16June2002/HorseBehaviorandWe1fare2/RecentAdvancesintheTreatment
oquuineStereot/tabid/ 3 130/Default.aspx
[Accessed 2009 February 8].

Mills, D.S., Macleod, C.A., 2002. The response of crib-biting and windsucking in horses
to dietary supplementation with an antacid mixture. lppologia 13, 33-41.

Moeller, B.A., McCall, C.A., Silverman, S.J., McElhenney, W.H., 2008. Estimation of
saliva production in crib-biting and normal horses. J. Equine Vet. Sci. 28(2), 85-90.

Nicol, C.J., 1999a. Stereotypies and their relation to management. In: Harris, P.A.,
Gomarsall, G.M., Davidson, H.P.B., Green, R.E. (Eds.), Proceedings of the BEVA
Specialist Days on Behaviour and Nutrition. Newmarket, UK, Equine Vet. J ., 1 1-
l4.

Nicol, C.J., 1999b. Understanding equine stereotypies. Equine Vet. J. Suppl. 28, 20-25.

Nicol, C.J., Davidson, H.P.B., Harris, P.A., Waters, A.J., Wilson, A.D., 2002. Study of
crib-biting and gastric inflammation and ulceration in young horses. Vet. Rec. 151,
658-662.

Ninomiya, S., Sato, S., Sugawara, K., 2007. Weaving in stabled horses and its
relationship to other behavioural traits. Appl. Anim. Behav. Sci. 106, 134-143.

Odberg, F., 1987. The influence of cage size and environmental enrichment on the
development of stereotypies in bank voles. Behav. Proc. 14, 155-173.

Parker, M., Goodwin, D., Redhead, E.S., 2008b. Survey of breeders’ management of
horses in Europe, North America and Australia: Comparison of factors associated
with the development of abnormal behavior. App]. Anim. Behav. Sci. 114, 206-215.

Redbo, I., Redbo-Torstensson, P., 6dberg, F.O., Hedendahl, A., Holm, J ., 1998. Factors
affecting behavioural disturbances in race-horses. Anim. Sci. 66, 475-481.

Vecchiotti, G.G., Galanti, R. 1986. Evidence of heredity of cribbing, weaving and stall-
walking in Thoroughbred horses. Livest. Prod. Sci. 14, 91-95.

Waters, A.J., Nicol, C.J., French, NR, 2002. Factors influencing the development of
stereotypic and redirected behaviours in young horses: findings of a four year
prospective epidemiological study. Equine Vet. J. 34(6), 572-579.

CHAPTER 1

CRIB-BITING BEHAVIOR IN HORSES: A REVIEW

Abstract

During the past decade, stereotypic behavior in horses, specifically crib-biting
behavior, has received considerable attention in the scientific literature. Epidemiological
and experimental studies designed to investigate crib-biting behavior have provided
valuable insight into the prevalence, underlying mechanisms, and owner perceptions of
the behavior. The findings of these studies have demonstrated how the management of
horses can influence their behavior and well being. The work of previous authors also has
been vital in generating additional research questions and hypotheses related to crib-
biting. The findings of several survey and experimental studies are reviewed, with
emphasis on research conducted since the late 1990’s, in an attempt to provide the reader
with a relatively comprehensive look into what is currently known about crib-biting
behavior in horses. Knowledge deficiencies and areas for future research are identified.

Keywords: Horse, Behavior, Crib-biting, Welfare, Review

10

1. Introduction

Stereotypies are defined as repetitive, relatively invariant patterns of behavior
with no apparent goal or function (Mason, 1991). Development and continued
performance of stereotypic behavior have been linked to sub-optimal environments
(Odberg, 1987; Cooper and Albentosa, 2005). Specifically, stereotypic behavior can
develop within the following contexts: when an animal is unable to execute a behavior
pattern that it is highly motivated to perform, such as feeding behavior; when it cannot
escape or avoid a stressful or fearful situation; or when it is kept in confinement or social
isolation (Mason, 1991). It has been suggested that stereotypic behavior may serve as a
coping mechanism, functioning to reduce stress or to provide the animal with some form
of control over its environment (Mason, 1991; Cooper and Albentosa, 2005). The
presence of stereotypies has been used as an indicator of poor welfare (Broom, 1983;
Mason and Latham, 2004) although whether the welfare is currently poor or has simply
been poor in the past is more difficult to determine. Stereotypies have been observed in
several species (Mason, 1991) and in captive ungulates, performance of oral stereotypic
behavior is common (Bergeron et al., 2006; Mason and Rushen, 2006). Specific
examples include object-licking in giraffes, bar-biting and sham chewing in sows,
tongue-rolling in cattle, and crib-biting in horses (Mason and Rushen, 2006).

Horses exhibiting crib-biting behavior anchor their top incisor teeth on a fixed
object (e.g. fence, stall or building structures), pull backward, contract the neck muscles,
and draw air into the cranial esophagus emitting an audible grunt (McGreevy eta1.,
1995a,b; Dodman et al., 2005). The behavior is not known to occur in feral, free-ranging

horses, but is observed in domestic (Houpt and McDonnell, 1993; Mills, 2000) and

11

captive wild horses, e. g. Przewalski horse (Boyd, 1986). Performance of crib-biting
behavior has been reported to occupy from 15% (Nicol et al., 2002) up to 65%
(Bachmann et al., 2003a) of the daily time budget.

It is widely reported in the literature that crib-biting, and other stereotypic
behaviors, are viewed by owners as being problematic and undesirable (Kiley-
Worthington, 1983; Houpt and McDonnell, 1993; Nicol, 1999a; Mills, 2002). Crib-biting
behavior has been linked to unthrifiiness (weight loss and poor condition) in horses. This
is thought to be a result of increased energy expenditure and/or a decrease in the amount
of time spent eating and grazing during performance of the behavior (Houpt and
McDonnell, 1993; McGreevy and Nicol, 1998a). The behavior also has been associated
with excessive tooth wear (Owen, 1982; Boyd, 1986), which in severe cases may impair
the horse’s ability to graze or result in dental disease. Two recent studies have
demonstrated an association between epiploic foramen entrapment, a specific form of
colic, and crib-biting behavior (Archer et al., 2004, 2008). Despite the latter findings
however, evidence for direct negative consequences of crib-biting behavior on horse
health remains largely anecdotal, requiring further empirical investigation and careful
documentation.

The precise etiology of crib-biting behavior has yet to be elucidated, and it is
likely that the cause is multifactorial. Several studies have been conducted to investigate
the potential biological mechanisms underlying crib-biting behavior. For example, crib-
biting has been associated with altered neuroendocrine physiology (Gillham etal., 1994;
Lebelt et al., 1998; McBride and Hemmings, 2005) and brain function (Hemmings et al.,

2007; Parker et al., 2008a). There is also some evidence to support a role of

12

gastrointestinal irritation in performance of the behavior (Mills and Macleod, 2002; Nicol
et al., 2002; Lillie et al., 2004). The findings of these studies have greatly enhanced our
understanding of the behavior, but in some cases, results have been conflicting or
insufficient, and warrant further investigation. Application of survey research
methodology to questions about crib-biting behavior has provided some insight into the
prevalence of and risk factors associated with the behavior. Specific factors found to be
associated with crib-biting behavior include time spent out of the stable, forage and
concentrate feeding, breed and sex of horse (McGreevy et al., 19950; Luescher et al.,
1998; Redbo et al., 1998; Bachmann et al., 2003b) and method of weaning (Waters et al.,
2002; Parker et al., 2008b). Some of the more recent epidemiological studies have also
attempted to assess owner awareness and perceptions regarding crib-biting behavior
(McBride and Long, 2001; Albright et al., 2009; Wickens, Chapter 3) in an effort to
determine the current level of concern with and knowledge about the behavior within the
equine community.
1.1 Aim of the review

Within the past decade, equine scientists have conducted a number of studies
designed to examine the etiology of crib-biting behavior. The purpose of this paper is to
review the existing literature on crib-biting behavior with special attention directed
toward research carried out after the publication of equine stereotypic behavior review
articles in the mid- to late- 19903 (Winskill et al., 1995; Cooper and Mason, 1998; Nicol,
1999b). Emphasis is placed on our current understanding of the role of neuroendocrine
and brain physiology and of gastrointestinal irritation in the performance of the behavior,

as well as the contribution of horse characteristics and enviromnental factors to crib-

l3

biting behavior. Recommendations concerning areas needing additional research are
made throughout.
2. Neuroendocrine physiology and brain function in crib-biting horses

The repetitive and persistent nature of stereotypic behavior has led authors in the
past to describe such behavior in horses as “obsessive compulsive disorder” or OCD
(Luescher et al., 1991; Shuster and Dodman, 1998). However, because obsessions
involve recurrent, intrusive thoughts, a capability that horses are not known to possess,
the terms “compulsive disorder” (Luescher et al., 1998), and “stereotypic behavior”
(Mills and Nankervis, 1999) are preferred. Nonetheless, implication of the serotonergic
system in compulsive disorders in both humans and horses represents a commonality
between the two Species in the underlying pathology of such repetitive or stereotyped
behavior patterns. Serotonin reuptake inhibitors have been used to treat compulsive
disorders in humans (Bandelow, 2008) and have been reported to reduce stereotypic
behavior in horses (McDonnell, 1998). However, Lebelt et a1. (1998) expressed
uncertainty regarding whether these drugs selectively affect stereotypic behavior or result
in changes in behavior by way of a general sedative effect. Lebelt et al. (1998) did find a
trend for lower basal serotonin levels in crib-biting compared to non-stereotypic horses,
suggesting that the serotonergic system of crib-biters may differ from that of non crib-
biting horses. The precise role of serotonin in the development or maintenance of the
behavior remains unclear however, and the results obtained by Lebelt et al. (1998) have
yet to be confirmed or refuted through additional experimental studies of the serotonergic

system in crib-biting horses.

14

Endogenous opioids have been suggested to facilitate and reinforce stereotypic
behavior (Dodman et al., 1987; Gillham et al., 1994; Zanella et al., 1996). In a study
conducted by Dodman et al. (1987), infusion of opioid antagonists reduced crib-biting
behavior, lending support to this hypothesis. Similarly, McBride and Cuddeford (2001)
demonstrated a reduction in crib-biting behavior by administering naloxone, but the
authors suggested that a general sedative effect of the opiate antagonist might have
influenced performance of the behavior. Measurement of plasma B-endorphin in crib-
biting horses has produced conflicting results. Gillham et a1. (1994) reported significantly
lower baseline concentrations of B-endorphin in crib-biting horses compared to non crib-
biting controls, whereas Lebelt et al. (1998) found 3 times higher basal [B-endorphin
concentrations in crib-biting horses. Pell and McGreevy (1999), however, found no
significant difference in plasma B-endorphin concentrations between crib-biting and
normal horses. Lebelt et al. (1998) and Nicol (1999) have suggested that peripheral
plasma B-endorphin concentrations may not reflect concentrations in the central nervous
system that would be responsible for producing behavioral changes. Pell and McGreevy
(1999) have proposed that a failure to detect differences in plasma B-endorphin
concentrations between crib-biting and normal horses may indicate greater sensitivity of
opioid receptors in stereotypic horses.

Crib-biting behavior also has been proposed as a means to alleviate a horse’s
stress. Heart rate and nociceptive threshold were lowered in horses during periods of crib-
biting (Lebelt et al., 1998). McBride and Cuddeford (2001) reported a significant
reduction in plasma cortisol concentration following bouts of crib-biting, providing

evidence that the act of crib-biting may reduce stress. McGreevy and Nicol (1998b)

15

found higher mean baseline concentrations of cortisol in crib-biting compared to normal
horses, but prevention of the behavior via removal of the crib-biting surface, did not
result in a rise in cortisol concentration. Subsequent studies have found no significant
differences in plasma (Pell and McGreevy, 1999; Clegg et al., 2008) or salivary (Pell and
McGreevy, 1999) cortisol between crib-biting and control horses, suggesting that levels
of arousal in stereotypic and normal horses are similar. There is some evidence
suggesting crib-biting horses react more strongly to acute stressors (Minero et al., 1999;
Bachman et al., 2003a) compared to their non crib-biting counterparts. In addition,
Minero et a1. (1999) found that heart rate and general activity of crib-biting horses
returned more quickly to basal levels following application of the stressor, providing
additional support that the behavior may serve as an adaptive response to stress.
However, interpretation of the findings obtained from these studies is difficult and
remains controversial. Results may be confounded by individual differences in
temperament, reactivity and life experiences of the crib-biting and non crib-biting horses
enrolled in such studies. An inherent limitation in many of these studies is that
measurements of cortisol have been obtained in mature horses with an established history
of performing the behavior rather than in horses just developing the behavior. Although
McGreevy and Nicol (1998b) observed higher cortisol concentrations in crib-biters, a
result that may imply a heightened stress response in the stereotypic horses, the authors
suggested that longitudinal studies would be necessary to establish whether development
of crib-biting had been successful in reducing already elevated concentrations of stress

hormones.

16

Research aimed at addressing the role of neuroendocrine physiology and brain
function in the development and continued performance of stereotypic behavior is further
complicated by the interrelationships between the hypothalamus-pituitary-adrenal (HPA)
axis and reward systems within the brain. Cabib et a1. (1998) found that stress induces
significant changes in dopamine (DA) receptor densities within the mesoaccumbens and
nigrostriatal systems in mice. In inbred strains of mice, these stress-induced changes in
dopamine neurophysiology have been associated with the development of stereotypic
behavior (Cabib et a1, 1997). More recently, McBride and Hemmings (2005) reported
significantly lower DA Dl-like receptor sub-types in the caudate nucleus (dorsomedial
striatum; DMS) and significantly higher DA D1 -like and D2-like receptor sub-types in
the nucleus accumbens (ventral striatum) of crib-biting horses. Due to the involvement of
basal ganglia and dopamine pathways in instrumental task learning, specifically goal-
directed learning and response-outcome processes, Hemmings et a1. (2007) and Parker et
a1. (2008) proposed that basal ganglia dysfunction and alterations in dopamine
physiology in crib-biters would be expressed as aberrant or impaired learning task
performance. Hemmings et a1. (2007) demonstrated that horses exhibiting crib-biting
behavior required significantly more unreinforced trials to reach extinction criterion (i.e.
stereotypic horses continued to perform button presses without receipt of the food
reward), and it was suggested that this perseverative responding might be indicative of
basal ganglia dysfunction. In the study by Parker et a1. (2008), learning performance
within a free-operant instrumental choice paradigm was compared between crib-biting
horses and non-stereotypic horses. Crib-biting horses failed to choose a more immediate

reinforcer demonstrating difficulty of the crib-biters to effectively learn the response-

17

outcome contingency. These studies are among the first to examine and provide evidence
of a behavioral correlate for neurophysiologic dysregulation in crib-biting horses.
Investigating differences in learning ability between crib-biting and non-stereotypic
horses, specifically within response-outcome paradigms offers a promising, non-invasive
approach to addressing questions pertaining to the role of brain and neuroendocrine
physiology in the performance of crib-biting behavior in horses.
3. Gastrointestinal irritation and crib-biting behavior

Free-ranging horses spend a large proportion of their time grazing and foraging.
In contrast, domesticated horses, particularly elite performance horses, are often fed high
concentrate, relatively low forage diets to meet the increased energy demands associated
with their competitive lifestyles. Concentrate and forage rations are often delivered only
2 times per day, thus subjecting horses to longer periods of feed deprivation. Feed
deprivation can result in gastric ulceration due to increased exposure of the squamous
mucosal lining to gastric acidity (Murray and Eichom, 1996). Gastrin, a peptide hormone
secreted into the blood, is a potent stimulator of gastric acid secretion (Katz, 1991), and
in horses, Smyth et a1. (1989) observed a greater and more prolonged gastrin response to
the feeding of pelleted and sweet feed diets compared to ad libitum hay feeding. Several
studies have demonstrated associations between concentrate feeding and crib-biting
behavior (Kusunose, 1992: Gillham et al., 1994; Redbo et al., 1998; Waters et al., 2002).

It has been suggested that crib-biting behavior may be an adaptive response to
gastric acidity and that the act of crib-biting may raise gastric pH as a result of increased
flow of alkaline saliva (Nicol, 1999b). Moeller et a1. (2008) demonstrated that salivation

is stimulated with crib-biting, which lends support to this theory. Crib-biting behavior has

18

recently been associated with gastric ulceration in foals, with gastric ulceration and
inflammation present in 60% of crib-biting foals compared to 20% of non crib-biting
foals (Nicol et al., 2002). In addition, crib-biting foals had greater severity of ulceration
and inflammation upon initial endoscopic examination (Nicol et al., 2002). In the same
study, the stomach condition of foals consuming a diet containing an antacid improved
and there was a trend toward reduced duration of crib-biting in supplemented foals. In
mature horses, long-term treatment with antacids has been shown to reduce the frequency
of crib-biting (Mills and Macleod, 2002), particularly in the period post-feeding. Crib-
biting horses also have been found to have lower basal and post-feeding gastric pH
compared to that of non crib-biters (Lillie et al., 2004). In the study conducted by Mills
and Macleod (2002), integrity of the gastric mucosa was not examined, thus it is unclear
whether the reduction in crib-biting frequency observed in mature horses consuming an
antacid diet was due to an increase in gastric pH alone or to an overall improvement in
stomach condition. The underlying cause of lower gastric pH in crib-biting horses
observed in the study by Lillie et a1. (2004) has not been determined. Wickens (Chapter
2) recently conducted a study to examine the integrity and function of gastric mucosa in
mature horses with a history of crib-biting behavior, but found no differences in the
number or severity of squamous mucosal lesions between crib-biting and normal horses
maintained on pasture. However, serum gastrin response to concentrate feeding was
found to be higher in crib-biting horses compared to controls, providing some additional
evidence that gastrointestinal physiology may be altered in horses exhibiting crib-biting
behavior. It was suggested that gastrin-stimulated acid secretion may be enhanced in crib-

biting horses due to greater G cell numbers or increased secretory capacity of the existing

19

G cells, but this idea can only be confirmed through further investigation involving
mucosal biopsies and molecular genetic techniques. It would be interesting to discover
whether pathways involved in gastric acid secretion are upregulated in crib-biting horses.
Additional studies employing continuous recordings of gastric pH in conjunction with
repeated blood sampling for determination of basal and post-feeding serum gastrin
concentrations in crib-biting and non crib-biting horses may also be helpful in
determining whether the gastrointestinal environment of crib-biting horses differs from
that of normal horses.

Fermentation of concentrate feeds in the cecum and large intestine is known to
reduce hindgut pH in horses (Rowe et al., 1994). In a study conducted by Johnson et a1.
(1998), increasing the amount of concentrate fed to horses resulted in the appearance of
aberrant oral behaviors, such as wood-chewing, and reductions in fecal pH.
Accumulation of lactic acid in the hindgut of horses was reduced by the addition of
virginiamycin to the diet (Rowe et al., 1994), and supplementation with virginiamycin
was shown to increase fecal pH and reduce the performance of abnormal oral behavior in
horses receiving concentrate feed (Johnson et al., 1998). However, Moeller et al. (2008)
contended that crib-biting was not one of the oral behaviors reported in the study by
Johnson et a1. (1998). A recent study conducted by Freire et a1. (2008) found no effect of
virginamycin supplementation on crib-biting behavior. Thus, the authors suggested that
established crib-biting behavior in adult horses may not be influenced by hindgut
acidosis.

Horses with gastric mucosal injury exhibit bruxism and behavioral signs of colic

(Muray 1998), thus appearing that horses are able to detect gastric acidity and mucosal

20

damage, or at least the pain it likely induces. Pain is known to bring about changes in
dopaminergic activity (Wood, 2004), and Hemmings et al. (2007) postulated that visceral
discomfort in horses may play an important role in the establishment of oral stereotypy
through alteration of basal ganglia programming. This author is aware of one report in
which a horse recovering from colic surgery had started to crib-bite. Recovery from colic
surgery usually entails periods of feed withdrawal and stall confinement, conditions
known to be ulcergenic in horses (Murray and Eichom, 1996). However, some level of
distress and general discomfort would also be associated with the procedure and recovery
process. Perhaps in this situation, crib-biting is initiated by the horse in an attempt to
reduce gastric acidity through production of alkaline saliva, but the behavior becomes
established as a result of pain-induced changes in neuroendocrine physiology. It seems
probable that a complex interrelationship between gastrointestinal and brain physiology is
involved in the etiology of crib-biting behavior and further research in this area is
warranted.
4. Findings of survey and epidemiological research on crib-biting behavior

The prevalence of crib-biting behavior reported in horses in Europe and Canada is
2.4 to 8.3% (V ecchiotti and Galanti, 1986; McGreevy et al., 1995c; Luescher et al.,
1998). Albright et al. (2009) reported an overall crib-biting prevalence of 4.4% in US.
horses, and the results of a recent behavior survey study conducted in Michigan
(Wickens, Chapter 3) provide a similar prevalence estimate of crib-biting behavior of
5.2%.

Survey studies in the UK and Canada have demonstrated an association between

various management practices and stereotypic behavior. For example, a prospective study

21

conducted by Waters et al. (2002) found that young Thoroughbred and part-
Thoroughbred horses fed concentrate feed post-weaning were 4 times more likely to
develop crib-biting behavior than foals not receiving concentrate. Weaning method also
has been associated with the performance of stereotypic behavior including crib-biting. In
a recent survey of management practices implemented on breeding farms in Europe,
North America, and Australia, natural weaning (mare allowed to wean foal) was
associated with a decrease in the chance of foals developing abnormal behavior (Parker et

al., 2008b). Post-weaning housing was also associated with the performance of abnormal

r: -. Lab- OR“ ‘1'

behavior with decreased risk of abnormal behavior in foals kept exclusively on grass
(Parker et al., 2008b). In eventing and dressage horses in the UK, increased amounts of
time spent outside the stable were associated with a decreased risk of stereotypic
behavior (McGreevy et al., 1995c). In addition, Wickens (Chapter 3) found a negative
relationship between the daily amount of tum-out and crib-biting behavior. Wickens
(Chapter 3) also found that social contact with other horses reduced the probability of
crib-biting behavior. Survey studies conducted to investigate stereotypic behavior in race
horses in Sweden (Redbo et al., 1998), and in Swiss horses of multiple breed types and
uses (Bachmann et al., 2003b) found that regular feeding of concentrates increased the
risk of performing stereotypic behavior. Specifically, Redbo et a1. (1998) demonstrated a
positive relationship between the amount of concentrate and stereotypic behavior,
including crib-biting, and a decreased risk of stereotypy with increased amount of
roughage.

An experimental study conducted by Visser et al. (2008) to investigate the effect

of two different housing conditions on the welfare of young horses exposed to stabling

Ix.)
IQ

for the first time, lends empirical support to the associations between housing, social
isolation and the performance of stereotypic behavior identified using survey
methodology. A total of 36 Dutch Warrnblood horses, 2 years of age and naive to stall
housing were enrolled in the study. Upon completion of the 12-week study, 22% of the
horses housed individually in 10.5 m2 boxes were seen exhibiting crib-biting behavior
whereas horses pair-housed in boxes (48 m2) did not begin performing stereotypic
behavior (crib-biting, weaving or box walking). Although additional studies of this nature
need to consider the welfare implications of purposely subjecting horses to management
conditions known or suspected to induce stress and behavioral disturbances, this
approach seems valuable in obtaining a better understanding of the circumstances that
elicit development of crib-biting behavior. Although the data were not shown, Visser et
a1. (2008) indicated that after 12 weeks of stabling, performance of stereotypic behavior
was still reversible in the majority of the horses exhibiting stereotypic behavior. This
finding stresses the importance of identifying behavioral problems early so that
appropriate management changes can be made before crib-biting and other stereotypic
behaviors become established.

Certain breeds of horses may be more likely to exhibit stereotypic behavior,
including crib-biting, than others. This was demonstrated by Bachmann et al. (2003b) in a
survey of stereotypic behavior in Swiss horses in which Warmbloods and Thoroughbreds
were at 1.8 and 3.1 times greater risk of performing stereotypic behavior, respectively,
compared to other breeds. Albright et a1. (2009) found that among US. horses,
Thoroughbreds were 3 times more likely to exhibit crib-biting behavior than Quarter

Horses and 5 times more likely than Arabians. Vecchiotti and Galanti (1986) suggested

23

the involvement of a genetic component in the performance of stereotypic behavior, as
evidenced by the finding that one or more relatives in 8 families of Thoroughbreds
exhibited crib-biting behavior. Luescher et a1. (1998) found a higher prevalence of crib-
biting in geldings and stallions compared to mares, and a greater risk of crib-biting
among Thoroughbred horses. The particular breed of horse may determine the primary
use of the animal, which in turn may affect the manner in which the horse is managed.
For example, many Thoroughbreds are used for competitive disciplines such as racing,
eventing, show-jumping, and dressage. Race horses in particular may be exposed during
the early part of their life to rigorous training regimens, high concentrate/ low forage
diets, very limited liberty turnout, and, depending on the track stabling, very limited
amounts of social contact. Stallions are typically managed very differently than geldings
and mares. For example they are often housed individually to prevent accidental breeding
and aggression. Stress or frustration associated with limited opportunities for tum-out or
social contact may therefore be associated with the performance of stereotypic behavior
in stallions.

Actual accounts of crib-biting behavior in other equid species are rare. For
example, this author was not able to find documented cases of crib-biting in captive
zebra. Personal communications with individuals working with large numbers of donkeys
and mules indicate that occurrence of crib-biting in this species is infrequent, with only
one donkey being reported to engage in a portion of the crib-biting sequence, i.e.
anchoring its incisor teeth on fence boards (Windsor, 2009; Taylor, 2009). Donkeys,
however, generally are not subjected to individual stall confinement and the high

concentrate feeding that many horses are exposed to. It is possible that the temperament

24

and management of donkeys is protective against development of crib-biting. Boyd
(1986) documented crib-biting in a zoo-kept female Przewalski’s horse, and in a later
publication (Boyd, 1991) referred back to this case cautioning that Przewalski’s horses
kept in small enclosures may begin crib-biting. Wood-chewing and coprophagy were
observed in these horses in the small enclosures, but not when the horses were out on
pasture (Boyd, 1986).

Interconnections between genetic and environmental factors, specifically
interactions between gender and management, or breed and management, almost
certainly play a role in the development of crib-biting behavior, and these relationships
warrant further consideration in future experimental and epidemiological studies.
Nonetheless, there is some evidence for a genetic predisposition in the display of the
behavior (Vecchiotti and Galanti, 1986; Albright et al., 2009) and identification of
specific genes responsible for crib-biting through pedigreeanalysis and association
mapping should be pursued.

A few of the more recent survey studies have included questions related to owner
and farm manager perceptions of stereotypic behavior in horses. British horse owners
have demonstrated concern regarding the performance of stereotypic behavior and the
majority of owners attempted to physically prevent horses from performing the behavior
(McBride and Long, 2001). This included the use of cribbing straps to stop horses from
crib-biting. Michigan horse owners and farm managers also expressed concern about
crib-biting behavior and 81% of owners indicated that they have tried to stop the behavior
(Wickens, Chapter 3). However, perceptions regarding the impact of the behavior on

horse performance/learning, horse health, and monetary value of the animal were

25

different between owners of non-stereotypic horses and those respondents currently
owning/managing a crib-biter. In general, respondents presently owning/managing a crib-
biting horse were less concerned about the behavior having a negative impact on
learning, health, or monetary value. Thus, it appeared that perceptions about stereotypic
behavior within the equine community at large may not coincide with those held by
individuals having first-hand experience with crib-biting horses.

Despite this discrepancy, many of the respondents currently owning/managing
crib-biting horses agreed/strongly agreed that crib-biting has a negative impact on horse
health (Wickens, Chapter 3). A small group of owners reported wear of the incisor teeth
and some problems with colic in their crib-biting horses, however a causal relationship
between the behavior and any specific health problems could not be ascertained. In the
Michigan horse behavior survey, many respondents believed that environmental variables
were largely responsible for the performance of crib-biting behavior, a finding similar to
the perceptions of US. horse owners reported by Albright et a1. (2009). These results
indicate that a large proportion of owners are aware and fairly well informed that
management practices can have an impact on the behavior of their horses.

One additional aspect of crib-biting behavior that has received some consideration
in recent epidemiological research is the question concerning whether horses learn to
copy the behavior by watching or interacting with others. Currently, there is little
evidence to support the belief that horses learn to perform stereotypic behavior by
observing others. A small percentage of crib-biting horse owners (5%) in the Michigan
study (Wickens, Chapter 3) reported that a horse had started to crib-bite after another

crib-biting horse had arrived at the farm. The management practices of those farms

26

differed from one another with respect to primary housing, social contact, and hours of
turn-out. Thus, it would be difficult to determine whether horses are in fact copying the
behavior or if the behavior is the result of exposure to common management factors,
specifically those factors previously demonstrated to be associated with an increased risk
for crib-biting behavior. Albright et al. (2009) reported that only 1% of horses surveyed
started to crib-bite after the arrival of a crib-biting horse. On the other hand, Nagy et a1.
(2008) found an increased risk of stereotypic behavior (crib-biting and weaving) in horses
exposed to stereotypic neighbors. Subsequent empirical and epidemiological
investigation is needed prior to reaching a conclusion about the ability of horses to
imitate stereotypic behavior through observation. It has been suggested that familiarity
with other horses and dominance hierarchies between horses may be important factors in
the ability of horses to learn a particular behavior by observation (Murphy and Atkins,
2007; Ninomiya, 2007) and Should therefore be examined in future studies.
Cross-sectional studies are somewhat limited in their ability to determine cause
and effect associations between management factors and crib-biting behavior, specifically
because many horses included in such studies may already be well established in the
behavior. Horses purchased or brought onto the farm already exhibiting crib-biting
behavior may have developed the behavior as a result of previous management rather
than from exposure to their present environment and management. Nonetheless, cross-
sectional studies have been extremely helpful in identifying the prevalence of, and in
generating additional hypotheses about, stereotypic behavior. Prospective
epidemiological research studies allow researchers to follow the development of behavior

in young horses exposed to various management factors, and are the favored approach to

27

identifying associations between environmental and horse-related variables and the
performance of stereotypic behavior. However, the increased cost and time commitment
associated with conducting longitudinal studies and the need for large sample sizes
represent a definite challenge to researchers. Collaboration among equine scientists and
epidemiologists from multiple institutions and regions may facilitate funding
opportunities, enhance questionnaire and experimental design, and provide increased
accessibility to farm managers/owners and available horses. Meta-analysis of existing
studies may also be of value.
5. Motivation to crib-bite and attempts to stop the behavior

Many owners attempt to physically prevent horses from performing crib-biting
behavior (McBride and Long, 2001; Wickens, Chapter 3). Specific methods used to stop
the behavior, with varying success, include removal of cribbing surfaces and application
of repellents or electric wire, cribbing straps and muzzles, aversion therapy (Baker and
Kear-Colwell, 1974) and the surgical removal of the paired omohyoideus and
stemothyrohyoideus muscles, a procedure known as modified Forssell’s technique
(Delacalle et al., 2002). The primary problem with these methods is that they fail to
address the underlying causes of crib-biting behavior and may further reduce equine
welfare (McBride and Cuddeford, 2001), particularly if the behavior serves a function in
stress reduction or alleviation of gastrointestinal discomfort. Short-term prevention of
crib-biting behavior using a cribbing strap has been shown to increase crib-biting rate
upon removal of the device (McGreevy and Nicol, 19980). It was suggested that this
post-inhibitory rebound reflected an increase in internal motivation to crib-bite during the

period when the behavior was thwarted. McGreevy and Nicol (1998c) stated that

28

behaviors that display this pattern of motivation may be considered functional to the
horse. Houpt et al. (2005) demonstrated that crib-biting horses will work to gain access to
a crib-biting surface and the results suggested that crib-biting horses valued the behavior
nearly as much as they valued food.

The use of pharmacological agents in the treatment of crib-biting behavior has to
some extent been successful in stopping or reducing the behavior (Dodman et al., 1987;
McDonnell, 1998; Rendon et al., 2001) but requires constant infiision/administration,
which would increase costs and labor inputs on farms. Moreover, the side effects and
toxicity levels of such compounds have not been adequately studied in horses. Previous
authors have recommended that management of crib-biting horses should be targeted
toward removal of the causal factors as opposed to prevention through physical means.
Increasing opportunities for horses to engage in natural foraging and social behavior is
probably the best approach in attempting to prevent the deVelopment of crib-biting
behavior and shows some promise for reducing frequency and duration of the behavior in
established crib-biters (Redbo et al., 1998; Parker et al., 2008b; Wickens, Chapter 3).
Conclusion

Experimental and survey research studies conducted within the past 12 to 15 years
have provided a wealth of knowledge regarding the potential causal factors involved in
crib-biting behavior. This information has been used extensively to help increase the
awareness within the equine community on how the routine management of horses can
affect their behavior and welfare. These studies have also given professionals engaged in
the study of equine behavior a framework from which to generate additional hypotheses

and research questions related to the development and continued performance of crib-

29

biting behavior. Some of the Specific areas meriting additional investigation include the
use of learning tasks in assessing the role of brain function and further study of the
gastrointestinal environment in crib-biting horses, application of genetic techniques to
identify specific genes involved in the behavior, the potential interactions between
genetics and management, and the role of observational learning in the performance of
crib-biting behavior. Ethologists and welfare scientists should continue to seek a
multidisciplinary approach to address questions related to crib-biting behavior, including
the use of behavioral and physiological measures, as well as application of
epidemiological research methods. Furthermore, collaboration among equine scientists is

encouraged to facilitate knowledge and resource sharing.

30

Literature Cited

Albright, J .D., Mohammed, H.O., Heleski, C.R., Wickens, C.L., Houpt, K.A., 2009. Crib-
biting in US horses: Breed predispositions and owner perceptions of aetiology.
Equine Vet. J. 41, doi: 10.2746/042516409X372584.

Archer, D.C., Freeman, D.E., Doyle, A.J., Proudman, C.J., Edwards, B., 2004.
Association between cribbing and entrapment of the small intestine in the epiploic
foramen in horses: 68 cases (1991—2002). J. Am. Vet. Med. Assoc. 224, 562-564.

Archer, D.C., Pinchbeck, G.K., French, N.P., Proudman, C.J., 2008. Risk factors for
epiploic foramen entrapment colic: an international study. Equine Vet. J. 40, 224-
230.

Bachmann, I., Audigé, L., Stauffacher, M., 2003b. Risk factors associated with
behavioural disorders of crib-biting, weaving and box-walking in Swiss horses.
Equine Vet. J. 35(2), 158-163.

Bachmann, I., Bemasconi, P., Herrmann, R., Weishaupt, M.A., Stauffacher, M., 2003a.
Behavioural and physiological responses to an acute stressor in crib-biting and
control horses. Appl. Anim. Behav. Sci. 82, 297-311.

Baker, G.J., Kear-Colwell, J., 1974. Aerophagia (windsucking) and aversion therapy in
the horse. Proc. Am. Assoc. Equine. Prac. 20, 127-130

Bandelow, B., 2008. The medical treatment of obsessive-compulsive disorder and
anxiety. CNS Spectr. 13, 37-46.

Bergeron, R., Badnell-Waters, A.J., Lambton, S., Mason, G., 2006. Stereotypic oral
behaviour in captive ungulates: Foraging, diet and gastrointestinal function. In:
Mason, G., Rushen, J. (Eds.), Stereotypic animal behavior: Fundamentals and
applications to welfare. CAB International, Wallingford, p. 20.

Boyd, L., 1986. Behavior problems of equids in zoos. In: Crowell-Davis, S.L., Houpt,
K.A., (Eds.), Behavior, The Veterinary Clinics of North America: Equine Practice
2(3), W.B. Saunders, Philadelphia, pp. 653-664.

Boyd, L., 1991. The behavior of Przewalski’s horses and its importance to their
management. Appl. Anim. Behav. Sci. 29, 301-318.

Broom, D.M., 1983. Stereotypies as animal welfare indicators. In: Indicators Relevant to
Farm Animal Welfare (Ed. by D. Schmidt), pp. 81-87. The Hague: Martinus
Nijhoff.

31

Cabib, S., Bonaventura, N., 1997. Parallel strain-dependent susceptibility to

environmentally-induced stereotypies and stress-induced behavioral sensitization in
mice. Physiol. Behav. 61, 499-506.

Cabib, S., Giardino, L., Calza, L., Zanni, M., Mele, A., Puglisi-Allegra, S., 1998. Stress
promotes major changes in dopamine receptor densities within the mesoaccumbens
and nigrostriatal systems. Neuroscience 84, 193-200.

Clegg, H.A., Buckley, P., Friend, M.A., McGreevy, PD. 2008. The ethological and
physiological characteristics of cribbing and weaving horses. Appl. Anim. Behav.
Sci. 109, 68-76.

Cooper, J .J ., Mason, G.J., 1998. The identification of abnormal behavior and behavioural
problems in stabled horses and their relationship to horse welfare: a comparative
review. Equine Vet. J. Suppl. 27, 5-9.

Cooper, J .J ., Albentosa, M.J., 2005. Behavioural adaptation in the domestic horse:
potential role of apparently abnormal responses including stereotypic behaviour.
Livest. Prod. Sci. 92, 177-182.

Delacalle, J ., Burba, D.J., Tetens, J ., Moore, R.M., 2002. NszAG laser-assisted
modified Forssell’s procedure for treatment of cribbing (crib-biting) in horses. Vet.
Surg. 31, 111-116.

Dodman, N.H., Normile, J .A., Cottam, N., Guzman, M., Shuster, L. 2005. Prevalence of
compulsive behaviors in formerly feral horses. Intern. J. Appl. Res. Vet. Med. 3(1),
20-24.

Freire, R., Clegg, H.A., Buckley, P., Friend, M.A., McGreevy, RD, 2008. Behavioural
and physiological effects of virginiamycin in the diets of horses with stereotypies.
Vet. Rec. 163, 413-417.

Gillham, S.B., Dodman, N.H., Shuster, L., Kream, R., Rand, W., 1994. The effect of diet
on cribbing behavior and plasma B-endorphin in horses. Appl. Anim. Behav. Sci.
41, 147-153.

Hemmings, A., McBride, S.D., Hale, CE. 2007. Perseverative responding and the
aetiology of equine oral stereotypy. Appl. Anim. Behav. Sci. 104, 143-150.

Houpt, K.A., McDonnell, SM. 1993. Equine stereotypies. Compendium Continuing
Education 15, 1265-1271.

Houpt, K.A., Crowley, L., Rousseliere, A., 2005. Operant learning as a means of

measuring equine motivation to crib: A vice or a need? Anthrozods, 18(3), 320-321
(Abstr.).

32

Johnson, K.G., Tyrrell, J ., Rowe, B., Pethick, D.W., 1998. Behavioural changes in
stabled horses given nontherapeutic levels of virginiamycin. Equine Vet. J. 30, 139-
143.

Katz, J ., 1991. Acid secretion and suppression. Med. Clin. North Am. 75, 877-887.
Kiley-Worthington, M. 1983. Stereotypes (sic) in horses. Equine Practice 5(1), 34-40.

Kusunose, R., 1992. Diurnal pattern of crib-biting in stabled horses. Jpn. J. Equine. Sci.
3(2), 173-176.

Lebelt, D., Zanella, A.J., Unshelm, J ., 1998. Physiological correlates associated with
cribbing behavior in horses: changes in thermal threshold, heart rate, plasma [3-
endorphin and serotonin. Equine Vet. J. Suppl. 27, 21-27.

Lillie, HQ, 2004. Comparison of gastric pH in crib-biting and non crib-biting horses.
Master’s thesis, Auburn University, Auburn.

Luescher, U.A., McKeown, D.B., Halip, J ., 1991. Reviewing the causes of obsessive-
compulsive disorders in horses. Vet. Med. 86, 527-530.

Luescher, U.A., McKeown, D.B., Dean, H., 1998. A cross-sectional study on compulsive
behaviour (stable vices) in horses. Equine Vet. J. Suppl. 27, 14-18.

Mason, G.J. 1991. Stereotypies: a critical review. Anim. Behav. 41:1015-103 7.

Mason, G.J., Latham, NR, 2004. Can’t stop, won’t stop: is stereotypy a reliable animal
welfare indicator? Anim. Wolf. 13, 857-869.

Mason, G., Rushen, J ., 2006. A decade-or-more’s progress in understanding stereotypic
behaviour. In: Mason, G., Rushen, J. (Eds.), Stereotypic animal behavior:
Fundamentals and applications to welfare. CAB International, Wallingford, p. 5-6.

McBride, S.D., Cuddeford, D., 2001. The putative welfare-reducing effects of preventing
equine stereotypic behaviour. Anim. Welf. 10, 173-189.

McBride, S.D., Long, L., 2001. Management of horses showing stereotypic behaviour,
owner perception and the implications for welfare. Vet. Rec. 148, 799-802.

McBride, S.D., Hemmings, A. 2005. Altered mesoaccumbens and nigro-striatal

dopamine physiology is associated with stereotypy development in a non-rodent
species. Behav. Brain. Res. 159, 113-118.

McDonnell, S., 1998. Pharmacological aids to behavior modification in horses. Equine
Vet. J. Suppl. 27, 50 (Abstr.).

McGreevy, P.D., Nicol, C.J., 1998a. Prevention of crib-biting: a review. Equine Vet. J.
Suppl. 27, 35-38.

33

McGreevy, P.D., Nicol, C.J., 1998b. Physiological and behavioral consequences
associated with short-term prevention of crib-biting in horses. Phys. Behav. 65, 15-
23.

McGreevy, P.D., Nicol, C.J., 1998c. The effect of short term prevention on the

subsequent rate of crib-biting in Thoroughbred horses. Equine Vet. J. Suppl. 27, 30-
34.

McGreevy, P.D., French, N.P., Nicol, C.J. 1995c. The prevalence of abnormal behaviours
in dressage, eventing and endurance horses in relation to stabling. Vet. Rec. 137,
36-3 7.

McGreevy, P., Richardson, J .D., Christine, J .N., Lane, J .G. 1995b. Radiographic and
endoscopic study of horses performing an oral based stereotypy. Equine Vet. J. 27,
92-95.

McGreevy, P., Nicol, C.J., Cripps, R, Green, L., French, N., 1995a. Management factors
associated with stereotypic and redirected behaviour in the thoroughbred horse.
Equine Vet. J. 27, 86-91.

Mills, D.S. 2002. Recent advances in the treatment of equine stereotypic behavior. [Serial
online] Animal Behavior Cognition and Welfare Group. University of Lincoln.
Available from: ,
http://research.vet.upenn.edu/HavemeyerEquineBehaviorLabHomePage/Reference
LibraryHavemeyerEquineBehaviorLab/l-IavemeyerWorkShops/HorseBehaviorand
Welfarel 3 l 6June2002/HorseBehaviorandWe1fare2/RecentAdvancesintheTreatment
oquuineStereot/tabid/ 3 1 30/Default.aspx
[Accessed 2009 February 8].

Mills, D.S., Nankervis, K.J., 1999. Equine behavior: principles and practice. Malden,
M.A., Blackwell, pp. 211-212.

Mills, D.S., Macleod, C.A., 2002. The response of crib-biting and windsucking in horses
to dietary supplementation with an antacid mixture. lppologia 13, 33-41.

Minero, M., Canali, E., Ferrante, V., Verga, M., 6dberg, F.O., 1999. Heart rate and
behavioural responses of crib-biting horses to two acute stressors. Vet. Rec. 145,
430-433.

Moeller, B.A., McCall, C.A., Silverrnan, S.J., McElhenney, W.H., 2008. Estimation of
saliva production in crib-biting and normal horses. J. Equine Vet. Sci. 28(2), 85-90.

Murphy, J ., Arkins, S., 2007. Equine learning behaviour. Behav. Processes 76, 1-13.

Murray, M.J., 1998. Gastroduodenal ulceration. In: Equine lntemal Medicine. S.M. Reed,
W.M. Bayly (Eds.), Pennsylvania, W.B. Saunders Company, pp. 615-623.

34

Murray, M.J., Eichom, E.S., 1996. Effects of intermittent feed deprivation, intermittent
feed deprivation with ranitidine administration, and stall confinement with ad
libitum access to hay on gastric ulceration in horses. Am. J. Vet. Res. 11, 1599-
1603.

Nagy, K., Schrott, A., Kabai, P. 2008. Possible influence of neighbors on stereotypic
behavior in horses. Appl. Anim. Behav. Sci. 111, 321-328.

Nicol, C.J., 1999a. Stereotypies and their relation to management. In: Harris, P.A.,
Gomarsall, G.M., Davidson, H.P.B., Green, R.E. (Eds.), Proceedings of the BEVA

Specialist Days on Behaviour and Nutrition. Newmarket, UK, Equine Vet. J ., 11-
14.

Nicol, C.J., 1999b. Understanding equine stereotypies. Equine Vet. J. Suppl. 28, 20-25.

Nicol, C.J., Davidson, H.P.D., Harris, P.A., Waters, A.J., Wilson, A.D., 2002. Study of

crib-biting and gastric inflammation and ulceration in young horses. Vet. Rec. 151,
658-662.

Ninomiya, S., 2007. Social learning and stereotypy in horses. Behav. Processes 76, 22-
23.

Odberg, F ., 1987. The influence of cage size and environmental enrichment on the
development of stereotypies in bank voles. Behav. Proc. 14, 155-173.

Owen, R.R., 1982. Crib-biting and windsucking — that equine enigma. Hill, C.S.G.,
Grunsell, F .W.G. (Eds.), The Veterinary Annual. Wright Scientific Publications,
Bristol, pp. 159-168.

Pell, S.M., McGreevy, RD, 1999. A study of cortisol and beta-endorphin levels in
stereotypic and normal Thoroughbreds. Appl. Anim. Behav. Sci. 64, 81-90.

Parker, M., Goodwin, D., Redhead, E.S., 2008b. Survey of breeders’ management of
horses in Europe, North America and Australia: Comparison of factors associated
with the development of abnormal behavior. Appl. Anim. Behav. Sci. 114, 206-215.

Parker, M., Redhead, E.S., Goodwin, D., McBride, S.D., 2008a. Impaired instrumental
choice in crib-biting horses (Equus caballus). Behav. Brain. Res. 191, 137-140.

Redbo, I., Redbo-Torstensson, P., 6dberg, F.O., Hedendahl, A., Holm, J ., 1998. Factors
affecting behavioural disturbances in race-horses. Anim. Sci. 66, 475-481.

Rendon, R.A., Shuster, L., Dodman, N.H., 2001. The effect of the NMDA receptor

blocker, dextromethorphan, on cribbing in horses. Pharm. Biochem. Behav. 68, 49-
51.

35

Rowe, J.B., Pethick, D.W., Lee, M.J., 1994. Prevention of acidosis and laminitis
associated with grain feeding in horses. J. Nutr. 124, 2742-2744.

Shuster, L., Dodman, N.H., 1998. Basic mechanisms of compulsive and self-injurious
behavior. In: Dodman, N.H., Shuster, L. (Eds.), Psychopharrnacology of animal
behavior disorders. Malden, M.A., Blackwell, pp. 185-202.

Smyth, G.B., Young, D.W., Hammond, L.S., 1989. Effects of diet and feeding on
postprandial serum gastrin and insulin concentrations in adult horses. Equine Vet. J.
Suppl. 7, 56-59.

Vecchiotti, G.G., Galanti, R. 1986. Evidence of heredity of cribbing, weaving and stall-
walking in Thoroughbred horses. Livest. Prod. Sci. 14, 91-95.

Visser, E.K., Ellis, A.D., Van Reenen, CG, 2008. The effect of two different housing
conditions on the welfare of young horses stabled for the first time. Appl. Anim.
Behav. Sci. 114, 521-533.

Waters, A.J., Nicol, C.J., French, NP, 2002. Factors influencing the development of
stereotypic and redirected behaviours in young horses: findings of a four year
prospective epidemiological study. Equine Vet. J. 34(6), 572-579.

Winskill, L., Waran, N.K., Charming, C., Young, R., 1995. Stereotypies in the stabled
horse: Causes, treatments and prevention. Current Sci. 69, 310-316.

Wood, RB, 2004. Stress and dopamine: implications for the pathophysiology of chronic
widespread pain. Med. Hypotheses 62, 420-424.

Zanella, A.J., Broom, D.M., Hunter, J .C., Mendl, M.T., 1996. Brain opioid receptors in

relation to stereotypies, inactivity, and housing in sows. Physiol. Behav. 59, 769-
775.

36

CHAPTER 2

GASTRIC ULCERATION IN MATURE HORSES WITH HISTORY OF CRIB-
BITING

Abstract

Gastrointestinal irritation has been implicated in crib-biting (CB) in
horses. Eighteen horses (9 CB and 9 NCB = non crib-biting) were used to determine
prevalence and severity of gastric mucosal damage and effect of concentrate feeding on
circulating gastrin concentrations in CB and NCB horses. Horses were maintained on
Coastal Bermudagrass (C ynodon dactylon) pasture with free access to Bermudagrass hay
and water and twice daily delivery of a pelleted (10% protein) concentrate diet. Number
of crib-bites were recorded from all CB horses in a 24 hr observation period. Endoscopic
examinations (EE) of the squamous mucosa were performed and gastric fluid sampled
after 24-28 hour feed removal. Following EE, horses were returned to pasture for 72
hours. Blood was collected at 1400 hours on pasture, following 12-hour feed removal (0
minutes), and at 60 and 120 minutes after consuming 1 kg of the pelleted diet. Mean
number of crib-bites in 24 hours was 1,558 $2 303 with CB peaking prior to and during
the afternoon feeding (1530 hours, P < 0.05). There were no differences in the number (P
> 0.05) or severity (P > 0.05) of ulcers or prevalence of hyperkeratosis (Fisher’s Exact P
> 0.05) between CB and NCB. There was no difference (P = 0.87) in gastric pH of CB
compared to NCB (3.92 vs. 3.78, respectively, SEM = 0.60). There was no effect of CB
(P = 0.56) on serum gastrin concentration (14.22 vs. 12.16 pg/ml for CB and NCB,
respectively, SEM = 2.46) with free access to hay and pasture. Concentrate feeding
increased gastrin concentration (P < 0.01). Serum gastrin concentration within CB was

greater at times 60 and 120 minutes compared to 0 minutes concentration (P < 0.01).

37

Compared to 0 minutes, serum gastrin concentration in NCB tended to differ at 60
minutes (P = 0.07) and was greater at 120 minutes (P < 0.05). Serum gastrin
concentration at 60 minutes was greater (P < 0.05) in CB compared to NCB. Compared
to NCB, there was a trend for greater serum gastrin concentration in CB horses at 120
minutes post-concentrate feeding (P = 0.06). These results suggest gastric mucosal
damage is not associated with established CB in mature horses maintained on pasture.
The greater gastrin response to concentrate feeding in CB may indicate altered
gastrointestinal function in CB, which could result in a more acidic gastric environment
following the consumption of concentrate feed. Further research into the gastrointestinal
physiology of crib-biting horses is warranted.

Keywords: Horse, Behavior, Crib-biting, Gastric ulcers, Serum gastrin

Introduction

Crib-biting is an oral stereotypic behavior unique to horses. It is characterized as
a repetitive behavioral sequence in which the horse anchors its incisor teeth on a fixed
object (e. g. fence, stall and building structures), pulls backward, contracts the neck
muscles, and draws air into the cranial esophagus emitting an audible grunt (McGreevy et
al., 1995, Dodman et al., 2005). The prevalence of CB in horses in the United States is
approximately 4.4% (Albright et al., 2009), with similar prevalence rates reported in
other countries (Nicol, 1999). Crib-biting is recognized as a welfare and management
concern. For instance, the behavior has been reported to be associated with epiploic
foramen colic (Archer et al., 2004, 2008) and has been suggested as a coping mechanism

to alleviate suffering or stress (Lebelt et al., 1998; McBride and Cuddeford, 2001). Crib-

38

biting may limit the horse’s eligibility for insurance coverage or acceptance into boarding
facilities.

Many owners perceive CB as having a negative impact on horse health and
attempt to physically prevent horses from performing the behavior (McBride and Long,
2001; Wickens, Chapter 3). Methods employed in an effort to stop the behavior, with
varying success, include the removal or alteration of crib-biting surfaces (e. g. application
of repellents, installation of electric wire), neck collars (which prevent expansion of the
throat needed to crib-bite), and even surgical removal of the muscles (paired
omohyoideus and stemothyrohyoideus) involved in the act of crib-biting, a procedure
known as the modified Forssell's technique (Delacalle et al., 2002). These approaches
may further reduce equine welfare (McBride and Cuddeford, 2001; Mills and Macleod,
2002) and fail to address the underlying causes of CB.

Although the etiology of CB remains to be elucidated, gastrointestinal irritation
has been implicated. Nicol and others (2002) demonstrated an association between CB
and gastric ulceration in foals. Upon initial endoscopic examination, CB foals had greater
severity of ulceration and inflammation of the gastric squamous mucosa compared to
normal foals. Consumption of a diet containing an antacid improved stomach condition in
the foals and tended to reduce duration of CB. In mature horses, CB has been associated
with lower basal and post-feeding gastric pH (Lillie et al., 2004), and long-term treatment
with antacids has been shown to reduce the frequency of CB (Mills and Macleod, 2002).
The underlying cause of lower gastric pH in CB horses has not been determined.
Endoscopic examinations were not performed during the Mills and Macleod study

(2002), thus it is unclear whether the reduction in CB frequency observed in mature

39

horses consuming an antacid diet was due to an increase in gastric pH alone or to an
overall improvement in the condition of the gastric mucosa. Nicol (1999) suggested that
horses may perform CB in an attempt to reduce gastric acidity through the production of
alkaline saliva. A study conducted by Moeller et a1. (2008) demonstrated that salivation is
stimulated with CB lending support to this hypothesis. Although results of previous
research have provided some evidence that gastric acidity may be involved in CB, an
association between gastric ulceration and CB in mature horses has never been reported.
Strong causal associations between gastric acidity and mucosal damage have been
demonstrated in horses (Murray and Eichom, 1996; Murray, 1999). Gastric acid secretion
is stimulated by the peptide hormone gastrin, which is released into the blood by G cells
located in the glandular portion of the stomach (Katz, 1991) in response to gastric
distension, protein, and increased luminal pH (Wolfe and 8011, 1988). Gastrin
concentration in horses has been shown to increase following feeding (Brown et al.,
1987; Young and Smyth., 1988; Wilson et al., 2007), with greater and more prolonged
gastrin secretion occurring in response to pelleted and sweet feed diets (Smyth et al.,
1989) compared to ad libitum feeding of Coastal Bermudagrass hay. Several studies have
demonstrated a positive relationship between concentrate feeding and crib-biting
behavior (Kusunose, 1992: Gillham et al., 1994; Redbo et al., 1998; Waters et al., 2002).
Specifically, Kusunose (1992) and Gillham (1994) observed an increase in crib-biting
frequency in the period following concentrate feeding. The studies conducted by Redbo
et a1. (1998) and Waters et al. (2002), found concentrate feeding to be associated with an

increased risk of crib-biting behavior.

40

The proximal portion of the equine stomach is covered by stratified squamous
mucosa. The distal portion is covered by glandular mucosa containing glands that secrete
hydrochloric acid, pepsin, bicarbonate, and mucus. The two regions are separated by a
cuticular ridge known as the margo plicatus (Figure 2-1). In mature horses, gastric ulcers
primarily occur in the squamous mucosa adjacent to the margo plicatus (Murray and
Eichom, 1996; Murray, 1997; Murray, 1998) Andrews and Nadeau, 1999, Dionne et al.,
2003). The squamous mucosa is susceptible to ulceration because it lacks intrinsic
protective factors, (mucus and bicarbonate), found in the glandular region of the stomach,
and the area adjacent to the margo plicatus is frequently exposed to high acidity (Murray
and Eichom 1996; Murray, 1997; Murray, 1999). Damage to the gastric mucosa has been
associated with bruxism and behavioral Signs of colic (Murray, 1998), thus it appears
horses are able to detect mucosal injury.

The aim of this study was to investigate the relationship between CB and the
integrity and firnction of the gastric mucosa in mature horses. We hypothesized that
horses exhibiting CB would have a higher degree of gastric mucosal damage and higher
serum gastrin response to concentrate feeding compared to non crib-biting horses. The
specific objectives were to determine: 1) prevalence and severity of gastric mucosal
damage in mature CB horses, and 2) effect of feeding on circulating gastrin
concentrations in CB and non crib-biting horses.

Materials and Methods
Animals and diet
Eighteen mature horses, 9 with history of crib-biting (CB) and 9 controls (N CB)

were used in this study. Horses were housed at the Auburn University Horse Unit in

41

Auburn, Alabama and were part of the existing university teaching and research herd.
The study was conducted in December 2007. Crib-biting and NCB horses were matched
as closely as possible based on sex, age, and breed. Characteristics of each group are
presented in Table 2-1. All horses were maintained on Coastal Bermudagrass (Cynodon
dactylon) pasture with free access to Bermudagrass hay (8% crude protein on an as-fed
basis) and water. Crib-biting and NCB horses were kept in separate pastures. Twice daily
(0730 and 1530 hours), horses received 2 kg of a 10% crude protein, 9% crude fiber

commercial pelleted diet (Nutrena® Life Design® Compete, Minneapolis, MN). The main

ingredients of the pelleted diet are presented in Table 2-2. Horses were routinely
deworrned every 8 to 12 weeks. Each horse’s body condition was assessed using a scale
of 1-9, with 1 designating poor condition (extremely emaciated) and 9 designating a
horse in extremely fat condition (Henneke et al., 1983). All experimental procedures
were carried out under Auburn University Institutional Animal Care and Use Committee
approval.
Behavioral observations

Crib-biting horses were observed on pasture at the beginning of the study for
frequency and duration of crib-biting behavior. Human observers recorded the number of
crib-bites and duration of cribbing bouts during a 24-hour observation period (0600 to
0600 hours the following day). CB was not exhibited by any of the control horses during
the study.
Endoscopic examination, blood collection, and gastric pH measurement

Forty-eight hours following behavioral observations, horses were placed in box

stalls. Feed was withheld 24 to 28 hours and water was removed 12 hours prior to

42

endoscopy to allow for adequate emptying of the stomach. Ingestion of bedding was
prevented by removing shavings from the stalls. On the morning of the endoscopic
examinations, horses were loaded onto trailers and transported 0.25 miles (402.3 meters)
to the Auburn University Large Animal Teaching Hospital. A complete blood count
(CBC) and blood gas profile were obtained from each horse prior to endoscopic
examination. Briefly, for the blood gas profile, 1 ml of blood was drawn from the jugular
vein via a heparinized syringe and placed on ice until analysis. A second blood sample (3
ml) was collected into a vaccutainer tube containing EDTA for CBC analysis. The CBC
and blood gas analyses were performed immediately following blood collection by the
Auburn University College of Veterinary Medicine Pathobiology Diagnostic Services.
Body weights were determined using a livestock scale. Horses were sedated using a
combination of butorphanol (0.02 mg/kg, IV) and detomadine (0.01 mg/kg, IV) and
confined in treatment stocks. A nose twitch was applied for further restraint. A
nasogastric tube was passed through the nasal passages into the esophagus, and a 3 meter
video endoscope (Fujinon EV-40-45-LP5-30, Wayne, NJ) was inserted through the
lumen of the nasogastric tube into the esophagus and into the stomach. Insufflation with
room air was used to facilitate visualization of the gastric mucosa. Video records
(approximately 10 minute clips) of the squamous mucosa of the saccus cecus, non-
glandular fundus, and along the margo plicatus were obtained for ulcer scoring.

During the examination, 10 to 20 ml of gastric fluid was aspirated from the
glandular fundus region of the stomach through the channel of the endoscope using a 60
ml syringe. One sample per horse was transferred to a specimen cup for determination of

pH using a digital pH meter (Omega PHH-26, Omega Engineering, Inc., Stamford, CT).

43

Endoscopic examinations were performed between 0830-1200 hours over a 2-day period.
Horses were divided into 2 groups (n = 9 in each group) for endoscopic examination. On
day 1, endoscopic examinations were performed on 5 CB and 4 NCB horses. On day 2,
endoscopic examinations were performed on 4 CB and 5 NCB horses. The order in which
CB and NCB horses underwent endoscopy was randomized on both days. Immediately
following endoscopy, each horse was returned to an outpatient stall until all horses had
been examined. Horses were transported back to the Auburn University Horse Unit and
returned to their home pasture 1 to 3 hours following endoscopy. No adverse effects as a
result of the procedure were observed.
Ulcer Scoring

Video records from each horse were viewed by 2 observers (experienced
veterinarians) blinded to the horses’ behavioral classification. Condition of the gastric
mucosa was assessed using an accepted gastric ulcer scoring system (MacAllister et al.,
1997). This system assigns both a lesion number and a lesion severity score. The non-
glandular lesion number score ranges from 0 to 4 and is defined as follows: 0 (no
lesions), 1 (1-2 localized lesions). 2 (3-5 localized lesions). 3 (6-10 lesions). and 4 [>10
lesions or diffuse (or very large) lesions]. The non-glandular lesion severity score ranges
from 0 to 5, and is defined as: 0 (no lesions). 1 (appears superficial). 2 [deeper structures
involved (lesion has greater depth than number 1)]. 3 [multiple lesions and variable
severity (l. 2. or 4)]. 4 [same as number 2 and has active (hyperemic or darkened lesion
crater] appearance, and 5 (deep structures involved. having an active appearance with
active hemorrhage or adherent blood clot). Thus. a total of4 scores were obtained for

each horse. The observers also noted the presence of hyperkeratosis; i.e. the

44

keratinization or thickening of the mucosa. indicative of exposure to high acidity
(Murray. 1992; 1997).
Serum gastrin protocol

Three days following endoscopic examination, horses were brought into
unbedded box stalls at 1345 hours following continuous access to pasture, hay and water.
At 1400 hours, 10 ml of blood was drawn from each horse via jugular venipuncture,
collected into SST vaccutainer tubes and allowed to clot on ice. Blood samples were
centrifuged (IEC Centra CL2, Therrno Electron Company, Milford, MA) at 1560 x g for
20 minutes and serum was removed and stored at -20° C until analysis. The initial sample
was collected to determine gastrin concentration in horses allowed free access to forage
(pasture and hay). Horses were fed 1.5 kg of Bermudagrass hay at 1645 hours and 1 kg of
the pelleted concentrate diet at 1830 hours. Hay and concentrate were consumed by all
horses by 1845. Water was provided in 5 gallon buckets. The next morning, three
additional blood samples (10 ml each) were collected and serum harvested using the
same protocol described above. The first sample was drawn at 0645 following 12-hour
feed removal to obtain basal metabolite values (0 minutes). Horses then were fed 1 kg of
the pelleted diet and allowed to eat for 15 minutes. All horses consumed the ration within
the allotted time period, and the second and third blood samples were collected from each
horse 60 and 120 minutes post-feeding.
Gastrin assay

Serum gastrin concentrations were measured. using a commercially available
double antibody radioimmunoassay designed for human use (Siemens Medical Solutions

Diagnostics, Los Angeles, CA). The method had been validated previously for use in

45

horses (Young and Smyth, 1988), but a validation with samples obtained from the crib-
biting and non crib-biting horses from the current study was conducted to assess the
agreement between the linear portion of the standard curve of the assay and the linear
portion of the equine serum gastrin samples. Equine serum gastrin concentrations
reported in previous studies (Brown et al., 1987; Smyth et al., 1989) have included
several values below 30 pg/ml. Therefore, the lowest standard included in the assay (25
pg/ml), was diluted using the 0 pg/ml calibrator solution to prepare a 12.5 pg/ml and a
6.25 pg/ml standard. All samples were run in duplicate and were assayed on the same
day.

Statistical analysis

Ulcer score and hyperkeratosis data were analyzed using SPSS (versionl6.0,
SPSS Inc., Chicago, IL). Ulcer number and severity scores were analyzed by Mann-
Whitney tests, and presence of hyperkeratosis was analyzed by chi-square tests of
independence. A Kendall correlation was used to evaluate agreement among scores
obtained from the 2 observers (CORR procedure of SAS, Version 9.1, SAS Institute Inc.,
Cary, NC).

Behavioral, hemoglobin, RBC count, blood gas, gastric pH, body condition and
initial serum gastrin data (samples collected following free-access to hay and pasture)
were submitted to ANOVA using the GLM procedure of SAS (Version 9.1, SAS Institute
Inc., Cary, NC). The statistical model for analysis of number of crib-bites and duration of
crib-biting included effects of horse and time. Horse was treated as a random effect. The
24-hour observation period was divided into 12, 2-hour time intervals to assess diurnal

crib-biting patterns. Multiple comparisons between time intervals were generated using

46

Tukey-Krarner adjustments. The model for hemoglobin, RBC count, blood gas, body
condition, gastric pH and initial serum gastrin concentration included the effect of
cribbing. Blood chemistry and gastric pH data also were subjected to tests for
homogeneity of variance and normality.

Serum gastrin data collected at 0, 60, and 120 minutes post-feeding, were
analyzed using the GLIMMIX procedure of SAS for repeated measures applying
Kenward-Roger's adjusted degrees of freedom to determine cribbing differences within
time (GLIMMIX 2006, SAS Institute Inc., Cary, NC). The model for serum gastrin was
cribbing, horse nested within cribbing, time, and time by cribbing. Horse within cribbing
was a random effect which was the error term used to test the effect of cribbing. Cribbing
effect within a time was tested with the slicediff option in SAS GLIMMIX.

Kendall correlations between 24-hour crib-biting frequency, ulcer number and
severity scores, presence of hyperkeratosis, gastric pH, and serum gastrin concentrations
were analyzed using the CORR procedure of SAS. Kendall correlations also were
examined between body condition score and ulcer number and severity in both groups of
horses.

Results
Behavior

Diurnal crib-biting frequency is shown in Figure 2-2. The mean number of crib-
bites in 24 hours was 1,558 i 303. Crib-biting frequency peaked prior to and during the
aftemoon feeding (1530 hours, P < 0.05). Duration of crib-biting during the 24-hour
observation period is shown in Figure 2-3. The mean duration of crib-biting in 24 hours

was 237.3 21:54.4 min. Thus, study horses spent approximately 16 to 17% of the day

47

engaged in crib-biting behavior. The duration of crib-biting behavior followed a similar
diurnal pattern as crib-biting frequency with a peak in bout length occurring during
consumption of the afternoon pelleted meal (P < 0.05). No significant correlations were
found between 24-hour crib-biting frequency, ulcer number and severity scores, presence
of hyperkeratosis, gastric pH, and serum gastrin concentrations.
Condition of gastric mucosa

Ulcer number and severity scores of CB and NCB horses reported by the 2
observers are presented in Table 2-3. Based on the scores reported by observer 1, there
were no differences in ulcer number (1.0 :t 0.50 vs. 0.7 A: 0.37, Mann-Whitney U = 35.5,
P = 0.61) or severity (0.7 :t 0.29 vs. 0.6 i 0.29, Mann-Whitney U = 37.0, P = 0.72)
between CB and NCB. There were also no differences in ulcer number (0.6 i 0.44 vs. 0.4
i 0.44, Mann-Whitney U = 36.5, P = 0.59) or severity (0.6 i 0.44 vs. 0.3 i 0.33, Marm-
Whitney U = 36.0, P = 0.54) between CB and NCB, based on the scores reported by
observer 2. Presence of hyperkeratosis in CB and NCB horses is presented in Table 2-4.
There was no difference in presence of hyperkeratosis between CB and NCB as assessed
by either observer 1 (Fisher’s Exact P = 1.00) or observer 2 (Fisher’s Exact P = 0.34).
Prevalence of ulcers observed in this study was 38.9% and 16.7% based on the
assessment of observers l and 2, respectively. Agreement among the 2 observers for
ulcer number (Kendall Correlation Coefficient = 0.63, P < 0.05) and ulcer severity score
(Kendall Correlation Coefficient = 0.64, P < 0.05) was moderate. There was less
agreement among the two observers regarding presence of hyperkeratosis (Kendall

Correlation Coefficient = 0.44, P = 0.07). Mean body condition score of CB (6.19 :t 0.22)

48

was higher (P < 0.01) compared to NCB (5.12 i 0.22). No significant correlation was
found between body condition and ulcer number or severity scores in CB or NCB horses.
Blood chemistry and gastric pH
Blood chemistry and gastric pH results for CB and NCB are presented in Table

2-5. Red blood cell count, hemoglobin, and hematocrit were higher (P < 0.05) in CB
compared to NCB. There was no difference in blood pH between CB and NCB (P =
0.52). Venous p02 and oxygen saturation tended to be higher in CB (P < 0.10, and P =
0.11, respectively) compared to NCB. Gastric pH following 24-28 hours of feed removal
was not different between CB and NCB (P = 0.87). Gastric pH data did not meet the
assumptions of normality (Shapiro-Wilk = 0.89, P < 0.05) and were log transformed. Log
transformation confirmed that gastric pH was not different between CB and NCB (P =
0.81).
Serum gastrin concentration

Results of the validation using serum samples from CB and NCB horses
confirmed that gastrin concentrations measured by the radioimmunoassay were on the
linear portion of the standard curve. When horses were allowed free access to hay and
pasture (Figure 2-4), there was no effect of CB (P = 0.56) on serum gastrin concentration.
Serum gastrin response to concentrate feeding in CB and NCB is shown in Figure 2-5.
Consumption of a pelleted concentrate meal increased gastrin concentration (P < 0.01).
The time by cribbing interaction was not found to be significant (P = 0.18). Serum gastrin
concentration within CB was greater at times 60 and 120 minutes compared to pre-
feeding (0 minutes) concentration (P < 0.01). Compared to 0 minutes, serum gastrin

concentration in NCB tended to differ at 60 minutes (P = 0.07) and was greater at 120

49

minutes (P < 0.05). Serum gastrin concentration at 60 minutes was greater (P < 0.05) in
CB compared to NCB. Compared to NCB, serum gastrin concentration tended to be
greater in CB horses at 120 minutes post-concentrate feeding (P = 0.06).
Discussion

Previous studies of crib-biting behavior in adult horses have demonstrated an
increase in crib-biting frequency shortly after the consumption of a concentrated meal
(Kusunose, 1992; Gillham et al., 1994). Thus, in the current study, an increase in crib-
biting frequency was expected following consumption of the pelleted diet during the
morning and afternoon feeding. The number of crib-bites increased during the
observation period in which horses received the afternoon concentrate meal, as did the
amount of time Spent crib-biting, however, a peak in crib-biting frequency or duration
was not observed during the morning concentrate feeding. Performance of crib-biting
may have been interrupted during the morning delivery of the pelleted diet due to a
general increase in activity associated with the arrival of the feeding crew. In addition,
crib-biting behavior in the current study was observed in horses maintained on pasture
with ad libitum access to hay. This differs from methods employed in earlier studies
(Kusunose, 1992; Gillham et al., 1994) in which horses were stabled and were not given
continuous access to forage. Kusunose (1992) showed crib-biting frequency to be low
around the time of roughage meal delivery, and Redbo et al., (1998) found the risk of
stereotypic behavior, including crib-biting, to be reduced when the amount of roughage is
increased. Deviations in the diurnal crib-biting pattern between studies also may be due
to differences in behavioral observation techniques. Behavioral data in the current study

were recorded by human observers as opposed to use of video surveillance in the study

50

conducted by Kusunose (1992). Video surveillance would have eliminated any
confounding effects of human presence on diurnal crib-biting patterns, however, project
horses were accustomed to human observers, thus human influence on crib-biting
behavior should have been minimal. Furthermore, the large pasture where crib-biting
horses were kept was not conducive to the use of video surveillance equipment, and the
authors felt it was important to assess crib-biting behavior in the environment in which
the horses were normally housed, minimizing confounding effects of stall confinement on
behavior.

The prevalence of non-glandular gastric ulcers in adult horses ranges from 11%
(Chameroy et al., 2006) to 93% (Murray et al., 1996) with prevalence and severity of
lesions found to be greatest among racing Thoroughbreds (Hammond et al., 1986;
Murray et al., 1996). Based on the evidence implicating gastrointestinal irritation as a
motivating factor for crib-biting in mature horses (Mills and Macleod, 2002; Lillie et al.,
2004; Moeller et al., 2008) and the association between gastric acidity and mucosal
damage (Murray and Eichom, 1996; Murray, 1999), prevalence and severity of gastric
ulceration was expected to be greater in horses exhibiting crib-biting behavior.
Prevalence of ulcers observed in the current study was relatively low, particularly as
assessed by observer 2. Lesion severity was mild, as only 1 horse received an ulcer
severity score greater than 3. The manner in which both crib-biting and non crib-biting
horses were routinely managed may have influenced the severity of ulceration observed
in this study. All of the horses enrolled in the study at the Auburn University Horse Unit
were maintained on pasture with free access to hay, factors shown to be protective

against gastric acidity (Murray and Schusser, 1993) and development of gastric ulcers

51

(Murray and Eichom, 1996; Pagan, 1997). Horses used in this study were not subjected
to intensive training regimens, or extended periods of stall confinement, factors that have
been considered to be ulcerogenic in horses (Hammond et al., 1986; Murray and Eichom,
1996). Agreement between observers in the number and severity of ulcers using the
scoring system of MacAllister et a1. (1997) was moderate. MacAllister et a1. (1997)
reported consistency in the way observers scored severity of non-glandular lesions, but
found significant variability between observers in the number of lesions identified in the
squamous mucosa. Thus, some level of discrepancy in scores assessed using the
MacAllister scoring system was expected. The level of agreement between observers
with regards to presence of hyperkeratosis was low, however very few studies have
assessed the presence or absence of hyperkeratosis, and therefore have not analyzed the
extent to which multiple observers’ diagnoses of hyperkeratosis coincide. The degree to
which the mucosa appears hyperkeratotic (yellow and thickened) may be highly
subjective and could vary widely between observers. Nonetheless, ulcer number and
severity scores, and presence of hyperkeratosis in the squamous mucosa as reported by
either observer were not supportive of a link between gastric ulceration and crib-biting in
mature horses, at least in this population of pastured horses.

Although a small portion of the glandular mucosa adjacent to the margo plicatus
was visible in a majority of the horses, insufficient insufflation and presence of gastric
fluid may have precluded accurate assessment of ulceration in the glandular mucosa.
Thus, the prevalence of ulcers may have been underestimated. Improvements in
endoscopic examination equipment have enhanced the ability of clinicians to visualize

the equine stomach in its entirety (Murray et al., 2001; Murray, 2002), but many authors

52

still report difficulties in viewing the glandular portion of the stomach (Andrews et al.,
2002; Charneroy et al., 2006; Dukti et al., 2006). Perhaps the challenge lies in extending
the time and effort required to allow adequate visualization of the stomach in a clinical
setting for the purposes of definitively diagnosing and treating gastric ulcers in individual
horses, to multiple horses enrolled in research studies. Murray et a1. (2001) reported
ulcer prevalences in the glandular fundus and pylorus of horses of 8% and 58%,
respectively. However, in other studies where endoscopic views of the glandular mucosa
have been obtained, occurrence of ulcers in the glandular region was less frequent
(McClure et al., 1999; Andrews et al., 2002; Dionne et al., 2003). Whereas ulcers in the
glandular mucosa are suspected to result from impairments in mucosal defense
mechanisms (Murray 1992; Murray 1997; Murray et al., 2001), ulceration in the
squamous mucosa is suggested to result primarily from exposure to high acidity (Murray,
1997; Murray et al., 2001). Compared to non crib-biting horses, gastric pH of crib-biting
horses has been shown to be lower in both the fasted and fed state (Lillie et al., 2004).
Exposure to an acidic environment more likely would affect the proximal portion of the
stomach, particularly the region adjacent to the margo plicatus, thus it was within the
scope of this study to assess and compare the condition of the squamous mucosa of crib-
biting and non-crib-biting horses.

Previous research has demonstrated an association between gastric ulceration and
poor body condition in horses (Murray et al., 1989; Dionne et al., 2003). Specifically,
prevalence and severity of ulceration were greater in horses presenting clinical signs
including poor condition and decreased appetite (Murray et al., 1989). No relationship

between body condition and ulcer number or severity was found in the current study.

53

However, ulcers in both the crib-biting and control horses were considered to be mild
(mean number and severity scores in both groups were _<_ 1), and therefore did not have a
negative impact on body condition. Although crib-biting horses received a higher mean
body condition score (BCS), a mean BCS of 5 in controls indicated the majority of non
crib-biting horses were also in good condition.

Horses with gastric ulcers have been shown to have lower RBC count and
hemoglobin compared to horses without ulcers (McClure et al., 1999). McClure and
others (1999) suggested the lower RBC count and hemoglobin may be the result of
chronic ulceration, although it was noted by the authors that mean RBC in both groups
was within the reference range (6 to 12 x 106 cells/pl). According to Murray (1998),
lesions in the squamous mucosa can be deep enough to cause bleeding, however, this
bleeding typically does not result in anemia. None of the horses in the current study
received an ulcer severity score of 5, characterizing lesions With active hemorrhage, and
mean RBC count and hemoglobin were within normal range. There is some evidence
suggesting crib-biting horses react more strongly to acute stressors (Minero et al., 1999;
Bachman et al., 2003) compared to their non crib-biting counterparts. Psychological
factors such as stress or excitement during transport can cause splenic contraction in
horses resulting in release of red blood cells into circulation producing elevated RBC
count and hemoglobin (F azio and F erlazzo, 2003). Cortisol concentration and heart rate
were not measured which prevented quantification of a heightened stress response in
crib-biting horses during study procedures. However, if crib-biting horses did experience
greater arousal and anxiety during transportation or blood sampling, it may explain the

higher RBC count and hemoglobin observed in this study. Water was withheld from all

54

horses 12 hours prior to endoscopy, thus losses in plasma volume and subsequent
elevation in RBC count and hemoglobin would have been expected to occur in both
groups. Therefore, release of RBC via splenic contraction is more likely to have
contributed to the differences observed between crib-biting horses and controls. Higher
RBC, hemoglobin, and hematocrit can increase oxygen carrying capacity, which would
explain the tendency toward greater venous p02 and oxygen saturation observed in the
crib-biting horses.

Due to the importance of mucosal blood flow in removing waste products from
the gastric mucosa (Murray, 1999), it was thought a decrease in venous blood pH might
be indicative of increased permeability of damaged gastric mucosa to hydrogen ions.
Mean venous blood pH of both crib-biting and non crib-biting horses was within normal
range. To the authors’ knowledge, no studies have found a decrease in venous blood pH
in subjects with gastric ulcers. Perhaps a relationship between gastric ulceration and
blood pH would have been demonstrated had prevalence and severity of ulcers been
greater in this sample of horses. Although it is probable that venous blood pH values that
deviate from normal may be more reflective of systemic acid-base status rather than
being useful in pinpointing a local acidotic condition within the stomach. Gastric pH of
crib-biting horses following 24 to 28 hours of feed removal (3.92) was similar to 15-hour
fasting pH (3.36) observed in the study by Lillie and others (2004). Fasting gastric pH of
non crib-biting horses in the present study (3.78) was lower compared to mean gastric pH
(5.50) of non crib-biting horses measured previously (Lillie et al., 2004). Therefore,
unlike the findings of Lillie et a1. (2004), differences in fasting gastric pH between crib-

biting and non crib-biting horses were not observed. Too much emphasis should not be

55

placed on the fasting pH measurements obtained in either study as in both research
settings, only one gastric fluid sample was obtained. The pH of gastric fluid sampled
from horses during endoscopy was highly variable, with gastric pH in crib-biting and non
crib-biting horses ranging from 1.80 to 6.24 and from 1.47 to 6.40, respectively. In
studies where pH electrodes have been used to continually monitor gastric pH in horses,
intermittent periods of alkalinization with pH readings 6.0 and greater have been
observed (Murray and Schusser, 1993; Baker and Gerring, 1993). The periods of
alkalinization became more frequent and values more variable, the longer horses were
withheld from feed. Reflux of more basic duodenal contents (pH 6.0 to 7.0) into the
stomach is reported to be a common occurrence in horses (Merritt, 1999) and is the most
probable cause of spontaneous increases in gastric pH. Therefore, repeated sampling as
was used in the study conducted by Lillie et a1. (2004) to measure post-feeding gastric pH
would be preferred because it would reduce variation and prOvide a more accurate
representation of the gastric acidity in each horse.

Basal and post-feeding serum gastrin concentrations were similar to values
obtained in other studies (Brown et al., 1987; Young et al., 1988; Smyth et al., 1989; Furr
et al., 1993). Gastrin concentrations in crib-biting and control horses following ad libitum
access to hay and pasture were also in agreement with values reported by Smyth et a1.
(1989) in horses fed Coastal Bermudagrass hay. Although gastric pH and serum gastrin
concentration data were not measured simultaneously, the author speculates that the
greater gastrin response to concentrate feeding in crib-biting horses could lead to a larger
increase in gastric acid secretion. Indeed, infusion of synthetic horse gastrin has been

shown to increase gastric acid secretion and reduce the pH of gastric fluid in horses

56

(Sandin et al., 1999). Gastrin stimulated acid secretion may be enhanced in crib-biting
horses due to greater G cell numbers or increased secretory capacity of the existing G
cells, but further studies involving mucosal biopsies and employing molecular genetic
techniques would be needed to confirm this idea. In the study conducted by Lillie et a1.
(2004), horses were fed a commercial sweet feed and given access to Bermudagrass hay
during gastric fluid sampling. Hay consumption stimulates increased saliva production
due to the longer period of mastication required to ingest forage (Meyer et al., 1985) and
bicarbonate in saliva aids in buffering the gastrointestinal tract. Indeed, Lillie and others
(2004) found that gastric pH increased with feeding in both crib-biting and non crib-
biting horses (4.21 and 5.62, respectively), but gastric pH of the crib-biting horses
remained lower than controls (P < 0.03). Thus, the response of gastric mucosa of crib-
biting horses to the concentrate portion of the diet may have been responsible for the
lower pH. Gastrin concentrations in crib-biting and non crib-biting horses consuming
pasture and hay for 6 hours were not different, indicating that gastrin response to forage
was similar in both groups. A heightened gastrin response to concentrate feeding could
result in crib-biting horses having a more acidic gastrointestinal environment, an effect
that is most likely attenuated when horses have ready access to forage. Higher gastrin
concentrations did not result in greater mucosal damage in crib-biting horses, but as
mentioned previously, management of study horses was protective against development
of gastric ulcers. Moeller et a1. (2008) suggested that the alkalinity of equine saliva, if
produced in large enough amounts, could have a significant buffering effect. It was
thought that total 24-hour crib-biting frequency would be negatively correlated with

gastric pH and positively correlated with mucosal damage and serum gastrin

57

concentration as horses attempted to alleviate the acidic gastric environment through
increased saliva production. However, no associations between 24-hour crib-biting
frequency, gastric pH, gastric ulcer number and severity, hyperkeratosis, or serum gastrin
concentrations were found.

It has been suggested by Cooper et al. (1996) that with time, stereotypies can
become emancipated from the original causal factors. Crib-biting horses used in this
study were donated to the Auburn University equine program, thus a thorough history on
each horse was not available. However, the majority of horses included in the study had
been known to crib-bite for at least 2 years. Therefore, crib-biting behavior can be
considered to be established in this group of horses. The fact that horses housed on
pasture continue to crib-bite, may be reflective of past gastrointestinal irritation rather
than being the result of an existing condition. Altered gastrointestinal function,
specifically an increased gastrin response to concentrate feeding, could be a key factor in
the development or onset of crib-biting behavior in horses. Particularly vulnerable would
be horses undergoing drastic changes in management earlier in life, such as exposure to
high concentrate:low roughage diets, increased stall confinement, and enrollment into
strenuous training regimens.

Conclusion

The results of this study suggest that gastric mucosal damage is not associated
with established crib-biting in mature horses maintained on pasture. Thus, owners of crib-
biting horses who provide their animals with adequate foraging opportunities Should not
suspect gastric ulcers as being the primary cause of the behavior. However, the greater

gastrin response to feeding in crib-biting horses is of interest as this finding may indicate

58

altered gastrointestinal function in crib-biting horses that could result in a more acidic
gastric environment following consumption of concentrated meals. An increase in gastric
acidity may be more problematic for horses receiving low forage diets in light of the
evidence that increased roughage and provision of an antacid diet are effective in
reducing crib-biting behavior even in established crib-biters. Longitudinal studies of the
gastrointestinal environment including measurement of gastrointestinal hormones in a
large population of horses exposed to various management practices are needed to further
elucidate the role of gastrointestinal irritation in crib-biting behavior.
Acknowledgements

The authors would like to acknowledge Dr. Betsy Wagner, Patty Tyler, Teresa
Fenn, Jinger Bland, and Ashley Bruce for their assistance with behavioral observations
and sample collection. Appreciation and gratitude are extended to the Auburn University
Horse Unit staff and to the Auburn University Large Animal Teaching Hospital personnel
for their cooperation and assistance during this research. The authors would also like to
thank Dr. Darrell Rankins, Jane Link, Barb Sweeney, and Runelle Phillips for their help

in preparing for and carrying out this study.

59

.omco: Ergo H 20 .wofinwseofi H m3. ._.

 

 

8035.83 .3 as...» w To use _ .acfiow m a n E macabre 82
saw Iona; .20 .me 28» one use _ @528 w a n 7c magnate
*3on ow< xom @286

 

820: .833 no 83388850 TN 033,—.

60

Table 2-2 Main ingredients of commercial pelleted diet*

 

Wheat middlings

Ground corn

Rice bran

Soybean hulls

Cane molasses

Corn oil

Glycerin

Calcium carbonate

Salt

Dehydrated alfalfa meal

Grain products

Dehulled soybean meal

Yeast culture

Vitamin A, D, and E supplement
Vitamin B 12, Riboflavin, Folic Acid, and Biotin supplement
Thiamine

L-lysine

DL—methionine

 

* Nutrena® Life Design® Compete, Minneapolis, MN. Ingredients are listed in the order

in which they appear on the feed tag.

61

Table 2-3 Ulcer number and severity scores by behavior*

 

Ulcer Number Score Ulcer Severity Score
0 1 2 3 4 0 1 2 3 4 5

 

 

 

Observer 1

Crib-biting 01:9)" 5 2 0 1 1 5 2 2 0 o 0
Non crib-biting (n=9)a 6 l l 1 0 6 1 2 0 0 0
Observer2

Crib-biting (11:9)b 7 1 0 0 1 7 1 o 0 1 0
Non crib-biting (11:9)b 8 0 0 0 1 8 o o 1 o 0
* Number of horses with ulcer number and severity scores of 0 to 4 and 0 to 5,
respectively.

a Ulcer number and severity did not differ (P = 0.61, P = 0.72, respectively) between crib-
biting and non crib-biting horses.
b Ulcer number and severity did not differ (P = 0.59, P = 0.54, respectively) between crib-
biting and non crib-biting horses.

62

Table 2-4 Presence of hyperkeratosis by behavior“

 

Hyperkeratosis Present

 

 

Yes No
Observer 1
Crib-biting (n = 9) 1a 8
Non crib-biting (n = 9) 1a 8
Observer 2
Crib-biting (n = 9) 5b 4
Non crib-biting (n = 9) 2b 7

 

* Number of horses with evidence of hyperkeratosis.
a Presence of hyperkeratosis was not different between crib-biting and non crib-biting

horses (P = 1.00).

b Presence of hyperkeratosis was not different between crib-biting and non crib-biting

horses (P = 0.34).

63

Table 2-5 Blood chemistry and gastric pH values by behavior

 

 

 

 

Behavior
CB (n=9) NCB (n=9)
Item Mean Mean SEM P-Value
RBC (x lOAé/uL)§ 9.348 7.92b 0.380 0.018
HGB (g/dL)1 16.27a 13.70b 0.671 0.016
Hematocrit (%)" 40.53a 34.61b 1.600 0.019
Venous po2 (mmHg) " 36.52a 33.98c 1.023 0.099
Venous Oxygen Saturation (%)+ 70.0721 65.34° 1.976 0.110
Venous Blood pHi 7.43a 7.44a 0.005 0.517
Gastric pH 3.92a 3.78a 0.287 0.739

 

3’1) Least squares mean estimates within a row with unlike superscripts differ at P < 0.05.

a’c Least squares mean estimates within a row with unlike superscripts tend to differ at
P S 0.1.

§ Normal range = 6 to 12 x 10A6/uL.

I Normal range = 10 to 18 g/dL.

I Normal range = 32 to 48%.

" Normal range = 17.9 to 42.2 mmHg.
I Normal range = 60 to 80%.

1‘ Normal range = 7.38 to 7.46.

64

Saccus cecus

Squamous fundus
Esophagus

Pylorus

Squamous mucosa

 
     
   
 

/

Margo plicatus

 

Glandular mucosa

Glandular fundus

Figure 2-1 Schematic representation of equine
stomach anatomy.

65

Mean number of crib-bites

450 ~

400 -
350 1
300 1
250 - be
200 - abc

150 . ab ab ab

100
50

abc abc abc
ab

    

4

l

 

 

0100 0300 0500 0700 0900 1100 1300 1500 1700 1900 2100 2300
Time Period

Figure 2-2 Mean crib-biting frequency of CB horses observed for 24 hours on
pasture. Each bar represents a 2 hour time period with the midpoint of each time
period shown along the x-axis. The pelleted diet (2 kg) was fed at 0730 and 1530

hours. Least squares mean estimates iS.E.M. are shown. Superscripts denote
significant differences at P < 0.05.

66

Mean crib-biting duration (see)

3500

00
O
O
O
l

l

2500
2000

l

1 500

I

1 000
500 ~

 

 

0100 0300 0500 0700 0900 11 00 1300 1500 1700 1 900 2100 2300
Time Period

Figure 2-3 Mean crib-biting duration of CB horses observed for 24 hours on
pasture. Each bar represents a 2 hour time period with the midpoint of each time
period shown along the x-axis. The pelleted diet (2 kg) was fed at 0730 and 1530
hours. Least squares means estimates :tS.E.M. are shown. Superscripts denote
significant differences at P < 0.05.

67

Serum gastrin (pg/ml)

18- a

 

 

16- I a
14-

l

10-

8..

6..

43

25

0

 

 

 

 

 

Crib-biting Non crib-biting

Figure 2-4 Serum gastrin concentrations in CB and NCB horses following free
access to Bermudagrass hay and pasture. Least squares mean estimates iS.E.M.

are Shown. Common superscripts indicate no effect of CB on serum gastrin
concentration (P = 0.56).

68

_‘ ~31“ 4.

 

** *

 

   

 

 

16 - b b
14 ~ I
E 12 ‘ bC
or
3 10 — --_____-——
.5 ""
a 8 ~
to
or 6
S +Crib-biting
g 4 _ a -0- Non crib-biting
2 _
0 e l l . i
0 min 60 min 120 min

Time After Feeding

Figure 2-5 Serum gastrin concentrations in CB and NCB horses following 12-
hour feed removal, 60 and 120 minutes after consuming 1 kg of the pelleted diet.
Least squares mean estimates iS.E.M. are shown. a’b Times 60 and 120 minutes
compared to 0 minutes within CB differ at P < 0.01. 8"” Within NCB, compared
to 0 minutes, time 60 minutes tended to differ at P = 0.07 and 120 minutes differs
at P < 0.05. "CB differs from NCB at time 60 minutes at P < 0.05. *CB tended to
differ from NCB at time 120 minutes at P = 0.06.

69

Literature Cited

Albright, J.D., Mohammed, H.O., Heleski, C.R., Wickens, C.L., Houpt, K.A., 2009. Crib-
biting in US horses: Breed predispositions and owner perceptions of aetiology.
Equine Vet. J. 41, doi: 10.2746/042516409X372584.

Andrews, F .M., Nadeau, J .A., 1999. Clinical syndromes of gastric ulceration in foals and
mature horses. Equine Vet. J. Suppl. 29, 30-33.

Andrews, F.M., Reinemeyer, C.R., McCracken, M.D., Blackford, J .T., Nadeau, J .A.,
Saabye, L., Sotell, M., Saxton, A., 2002. Comparison of endoscopic, necropsy and
histology scoring of equine gastric ulcers. Equine Vet. J. 34(5), 475-478.

Archer, D.C., Freeman, D.E., Doyle, A.J., Proudman, C.J., Edwards, B., 2004.
Association between cribbing and entrapment of the small intestine in the epiploic
foramen in horses: 68 cases (1991-2002). J. Am. Vet. Med. Assoc. 224, 562-564.

Archer, D.C., Pinchbeck, G.K., French, N.P., Proudman, C.J., 2008. Risk factors for
epiploic foramen entrapment colic: an international study. Equine Vet. J. 40, 224-
230.

Bachmann, I., Bemasconi, P., Herrmann, R., Weishaupt, M.A., Stauffacher, M., 2003.
Behavioural and physiological responses to an acute stressor in crib-biting and
control horses. Appl. Anim. Behav. Sci. 82, 297-311.

Brown, C.M., Sonea, I., Nachreiner, R.F., Obradovich, J .E., 1987. Serum
immunoreactive gastrin activity in horses: basal and postprandial values. Vet. Res.
Commun. 11, 497-501.

Chameroy, K.A., Nadeau, J .A., Bushmich, S.L., Dinger, J .E., Hoagland, T.A., Saxton,
A.M., 2006. Prevalence of non-glandular gastric ulcers in horses involved in a
university riding program. J. Equine Vet. Sci. 26(5), 207-211.

Cooper, J.J., 6dberg, F.O., Nicol, C.J., 1996. Limitations on the effectiveness of

environmental improvements in reducing stereotypic behavior in bank voles. Appl.
Anim. Behav. Sci. 48, 237-248.

Delacalle, J ., Burba, D.J., Tetens, J ., Moore, R.M., 2002. NszAG laser-assisted
modified F orssell’s procedure for treatment of cribbing (crib-biting) in horses. Vet.
Surg.31,111-116.

Dionne, R.M., Vrins, A., Doucet, M.Y., Pare, J ., 2003. Gastric ulcers in Standardred
racehorses: prevalence, lesion description, and risk factors. J. Vet. Intern. Med. 17,
218-222.

Dodman, N.H., Normile, J .A., Cottam, N., Guzman, M., Shuster, L., 2005. Prevalence of
compulsive behaviors in formerly feral horses. Intern. J. Appl. Res. Vet. Med. 3(1),
20-24.

70

Dukti, S.A., Perkins, S., Murphy, J., Barr, 8., Boston, R., Southwood, L.L., Bernard, W.,
2006. Prevalence of gastric squamous ulceration in horses with abdominal pain.
Equine Vet. J. 38(4), 347-349.

Fazio, E., Ferlazzo, A., 2003. Evaluation of stress during transport. Vet. Res. C ommun.
27(Suppl. 1), 519-524.

Furr, M., Taylor, L., Kronfeld, D., 1994. The effects of exercise training on serum gastrin
responses in the horse. Cornell Vet 84, 41—45.

Gillham, S.B., Dodman, N.H., Shuster, L., Kream, R., Rand, W., 1994. The effect of diet
on cribbing behavior and plasma B-endorphin in horses. Appl. Anim. Behav. Sci.
41, 147-153.

Hammond, C.J., Mason, D.K., Watkins, K.L., 1986. Gastric ulceration in mature
Thoroughbred horses. Equine Vet. J. 18(4), 284-287.

Henneke, D.B., Potter, G.D., Kreider, J .L., Yeates, BE, 1983. Relationship between
condition score, physical measurement, and body fat percentage in mares. Equine
Vet. J. 15, 371-372.

Katz, J ., 1991. Acid secretion and suppression. Med. Clin. North Am. 75, 877-887.

Kusunose, R., 1992. Diurnal pattern of crib-biting in stabled horses. Jpn. J. Equine. Sci.
3(2), 173-176.

Lebelt, D., Zanella, A.J., Unshelm, J ., 1998. Physiological correlates associated with
cribbing behavior in horses: changes in thermal threshold, heart rate, plasma [3-
endorphin and serotonin. Equine Vet. J. Suppl. 27, 21-27.

Lillie, HQ, 2004. Comparison of gastric pH in crib-biting and non crib-biting horses.
Master’s thesis, Auburn University, Auburn.

MacAllister, C.G., Andrews, F.M., Deegan, E., Ruoff, W., Olovson, S.G., 1997. A
scoring system for gastric ulcers in the horse. Equine Vet. J. 29(6), 430-433.

McBride, S.D., Cuddeford, D., 2001. The putative welfare-reducing effects of preventing
equine stereotypic behaviour. Anim. Welf. 10, 173-189.

McBride, S.D., Long, L., 2001. Management of horses showing stereotypic behaviour,
owner perception and the implications for welfare. Vet. Rec. 148, 799-802.

McClure, S.R. Glickman, L.T., Glickman, N.W., 1999. Prevalence of gastric ulcers in
show horses. J. Am. Vet. Med. Assoc. 215(8), 1130-1133.

McGreevy, P., Nicol, C.J., Cripps, R, Green, L., French, N., 1995. Management factors
associated with stereotypic and redirected behaviour in the thoroughbred horse.
Equine Vet. J. 27, 86-91.

71

Merritt, A.M., 1999. Normal equine gastroduodenal secretion and motility. Equine Vet. J.
Suppl. 29, 7-13.

Meyer, H., Coenen, M., Gurer, C., 1985. Investigations on saliva production and chewing
effects in horses fed various feeds. In: Proceedings of 9th ENPS, pp. 38-41. East
Lansing, MI.

Mills, D.S., Macleod, C.A., 2002. The response of crib-biting and windsucking in horses
to dietary supplementation with an antacid mixture. lppologia 13, 33-41.

Minero, M., Canali, E., Ferrante, V., Verga, M., 6dberg, F.O., 1999. Heart rate and
behavioural responses of crib-biting horses to two acute stressors. Vet. Rec. 145,
430-433.

Moeller, B.A., McCall, C.A., Silverman, S.J., McElhenney, W.H., 2008. Estimation of
saliva production in crib-biting and normal horses. J. Equine Vet. Sci. 28(2), 85-90.

Murray, M.J., 1992. Aetiopathogenesis and treatment of peptic ulcer in the horse: a
comparative review. Equine Vet. J. Suppl. 13, 63-74.

Murray, M.J., 1997. Overview of equine gastroduodenal ulceration. AAEP Proceedings
43, 382-387.

Murray, M.J., 1998. Gastroduodenal ulceration. In: Equine lntemal Medicine. S.M. Reed,
W.M. Bayly (Eds). Pennsylvania, W.B. Saunders Company, 615-623.

Murray, M.J., 1999. Pathophysiology of peptic disorders in foals and horses: a review.
Equine Vet. J. Suppl. 29, 14-18.

Murray, M.J., 2002. How to perform gastroduodenoscopy. AAEP Proceedings 48, 282-
286.

Murray, M.J., Schusser, G.F., 1993. Measurement of 24-h gastric pH using an indwelling
pH electrode in horses unfed, fed and treated with ranitidine. Equine Vet. J. 25(5),
417-421.

Murray, M.J., Eichom, E.S., 1996. Effects of intermittent feed deprivation, intermittent
feed deprivation with ranitidine administration, and stall confinement with ad
libitum access to hay on gastric ulceration in horses. Am. J. Vet. Res. 11, 1599-
1603.

Murray, M.J., Nout, Y.S., Ward, D.L., 2001. Endoscopic findings of the gastric antrum
and pylorus in horses: 162 cases (1996-2000). J. Vet. Intern. Med. 15, 401-406.

Murray, M.J., Schusser, G.F., Pipers, F.S., Gross, S.J., 1996. Factors associated with
gastric lesions in Thoroughbred racehorses. Equine Vet. J. 28(5), 368-3 74.

72

Murray, M.J., Grodinsky, C., Anderson, C.W., Radue, P.F., Schmidt, GR, 1989. Gastric
ulcers in horses: a comparison of endoscopic findings in horses with and without
clinical signs. Equine Vet. J. Suppl. 7, 68-72.

Nicol, C.J., 1999a. Stereotypies and their relation to management. In: Harris, P.A.,
Gomarsall, G.M., Davidson, H.P.B., Green, R.E. (Eds.), Proceedings of the BEVA
Specialist Days on Behaviour and Nutrition. Newmarket, UK, Equine Vet. J ., 11-
14.

Nicol, C.J., 1999b. Understanding equine stereotypies. Equine Vet. J. Suppl. 28, 20-25.

Nicol, C.J., Davidson, H.P.D., Harris, P.A., Waters, A.J., Wilson, A.D., 2002. Study of
crib-biting and gastric inflammation and ulceration in young horses. Vet. Rec. 151,
658-662.

Pagan, J .D., 1997. Gastric ulcers in horses: A widespread but manageable disease. World
Equine Vet. Rev. 2(4), 28-30.

Redbo, I., Redbo-Torstensson, P., 6dberg, F.O., Hedendahl, A., Holm, J ., 1998. Factors
affecting behavioural disturbances in race-horses. Anim. Sci. 66, 475-481.

Sandin, A., Andrews, F .M., Nadeau, J .A., Nilsson, G., 1999. Effects of horse gastrin on
gastric acid secretion in horses, dogs and rats. In: Sandin A, ed. Studies of gastrin
and gastric secretion in the horse. Uppsala: Acta Univ Agric Sueciae [Swedish
University of Agricultural Sciences] 1999; 69:65-78.

Smyth, G.B., Young, D.W., Hammond, L.S., 1989. Effects of diet and feeding on
postprandial serum gastrin and insulin concentrations in adult horses. Equine Vet. J.
Suppl. 7, 56-59.

Waters, A.J., Nicol, C .J ., French, NP, 2002. Factors influencing the development of
stereotypic and redirected behaviours in young horses: findings of a four year
prospective epidemiological study. Equine Vet. J. 34(6), 572-579.

Wilson, A.D., Badnell-Waters, A.J., Bice, R., Kelland, A., Harris, P.A., Nicol, C.J., 2007.
The effects of diet on blood glucose, insulin, gastrin and the serum tryptophan:
large neutral amino acid ratio in foals. Vet. J. 174, 139-146.

Wolfe, M.M., Soll, AH, 1988. The physiology of acid secretion. New Eng. J. Med. 319,
1707-1715.

Young, D.W., Smyth, GB, 1988. Validation of a radioimmunoassay for measurement of
gastrin in equine serum. Am. J. Vet. Res. 49(7), 1179-1183.

73

CHAPTER 3

INVESTIGATING WEAVING AND CRIB-BITING BEHAVIOR IN THE MICHIGAN

HORSE POPULATION VIA WEB-BASED SURVEY METHODOLOGY
Abstract

The manner in which horses are managed greatly influences their behavior and
well-being. The performance of stereotypic behavior in horses is suggested to be
indicative of either current or past sub-optimal welfare. Survey research conducted in
Europe and Canada has provided insight into the prevalence of and risk factors associated
with weaving and crib-biting behavior (WCB). Owner perceptions about WCB also have
been investigated in some of these studies. Currently, information regarding WCB in
Michigan’s horse population is unavailable. The objectives of this study were: (1)
determine whether Michigan horse owners are concerned about WCB and (2) investigate
risk factors associated with WCB. A web-based questionnaire was developed to
investigate WCB, and participants included members of 6 Michigan-based equine
associations, recipients of 2 Michigan equine periodicals, and visitors to Michigan State
University’s (MSU) Youth Equine Extension website. Complete responses from 293
individuals were received representing a total of 2,181 horses. The percentage of horses
exhibiting weaving and crib-biting was 2.7% and 5.2%, respectively. Compared to
respondents with weaving and crib-biting horses, a greater percentage of owners with
non-affected horses agreed/strongly agreed that WCB hinders learning ability (P < 0.01),
has a negative impact on horse health (P < 0.01 for weaving, and P < 0.05 for crib-
biting), and reduces the horse’s monetary value (P < 0.01). The proportion of respondents

attempting to stop crib-biting behavior (81 .0%) was greater (P < 0.01) compared to

74

weaving behavior (3 7.5%). Methods employed most frequently to stop or reduce weaving
behavior included increased tum-out (21 .4%) and provision of toys (14.3%). Cribbing
collars/straps (77.6%), alteration of crib-biting surfaces (55.3%), and increased tum-out
(54.1%) were the methods used most frequently to stop or reduce crib-biting. Many
owners used a combination of methods. Stepwise logistic regression analysis identified
associations between the probability of weaving and crib-biting and farm management
factors. A negative association was identified between hours of daily turn-out and the
probability of WCB (P < 0.01). Compared to horses allowed tum-out with conspecifics,
the risk of crib-biting was greater for horses with visual contact only (Adjusted odds ratio
(OR) = 5.61) and for horses without social contact (Adjusted OR = 6.26). In contrast to
previous studies, a larger daily amount of concentrate was associated with a decrease in
the odds of weaving (2.3 or more kg/day vs. less than 0.5 kg, P < 0.05) and crib-biting
(0.5 to 1.8 kg and 2.3 or more kg/day vs. less than 0.5 kg, P < 0.01). The results of this
study demonstrate that Michigan horse owners express concern about WCB and that
attempts are often made to control these behaviors. Additional research into the risk
factors associated with WCB is needed, as well as careful documentation of the
development of these behaviors.

Keywords: Horse, Behavior, Weaving, Crib-biting, Survey, Management factors
Introduction

Weaving and crib-biting behavior are two of the most widely recognized
stereotypic behaviors in horses. Weaving behavior is a locomotor stereotypy in which the
horse engages the forequarters, neck and head in a lateral swaying motion, repeatedly
shifting its weight from the left to right front foot (McGreevy et al., 1995a). Horses

exhibiting crib-biting behavior anchor their top incisor teeth on a fixed object, pull back,

75

arch the neck, and draw air into the cranial esophagus emitting an audible grunt
(McGreevy et al., 1995a; Dodman et al., 2005). Few experimental studies have been
conducted specifically to address weaving behavior. However, the provision of mirrors in
the stable reduced the performance of weaving (McAfee et al., 2002) suggesting that this
behavior may be performed in response to social isolation. Stable designs that increased
visual horizons, such as open stable doors providing access to views of adjacent horses
and surrounding fields, also reduced weaving behavior (Cooper et al., 2000). Ninomiya et
al. (2007) found that weaving behavior was observed mainly before feeding, and that
investigation of bedding was more likely to follow eating. These results suggest that
weaving also may be performed in an attempt to cope with frustration associated with
meal anticipation. Several studies have been conducted to investigate the potential
biological mechanisms underlying crib-biting behavior. Crib-biting has been associated
with gastrointestinal irritation (Mills and Macleod, 2002; NiCol et al., 2002), and with
altered brain function (Hemmings et al., 2007; Parker et al., 2008) and neuroendocrine
physiology (Gillham et al., 1994; Lebelt et al., 1998; McBride and Hemmings, 2005).
However, the etiology of these behaviors has yet to be completely elucidated.
Application of epidemiological research methods to questions about equine
stereotypic behavior has provided some insight into the prevalence of and risk factors
associated with weaving and crib-biting behavior. A summary of 5 cross-sectional
surveys conducted in Europe and Canada demonstrated average prevalence of weaving
and crib-biting behavior of 3.3% and 4.1%, respectively (Nicol, 1999a). A similar
prevalence of crib-biting behavior (4.4%) has recently been reported in horses in the

United States (Albright et al., 2009). Some of the management factors found to be

76

associated with weaving and crib-biting behavior include time spent out of the stable,
extent of social contact, forage and concentrate feeding, and breed and sex of horse
(McGreevy et al., 1995a,b; Luescher et al., 1998; Redbo et al., 1998; Bachmann et al.,
2003). Based on previous studies, Thoroughbreds (Luescher et al., 1998; Redbo et al.,
1998; Albright et al., 2009) and Warmbloods (Bachmann et al., 2003) have been
identified as the breeds at greatest risk of displaying stereotypic behaviors, and there is
some evidence to suggest a genetic predisposition for stereotypic behavior (Vecchiotti
and Galanti, 1986).

It is widely reported in the literature that weaving and crib-biting behavior may
endanger horse health and that these behaviors are viewed by owners as being
problematic and undesirable (Kiley-Worthington, 1983; Houpt and McDonnell, 1993;
Nicol, 1999; Mills, 2002). Owner perceptions of stereotypic behavior have been

investigated to some extent using survey research methodology. McBride and Long

(2001) reported that the majority of horse owners in the United Kingdom (UK) attempted

to stop horses from performing stereotypic behavior and that a large proportion of owners

were concerned about the impact of the behavior on horse health and monetary value.

Many owners believe that horses learn to perform stereotypic behavior by observing

others, i.e. that these behaviors are copied (McGreevy et al., 1995b; Albright et al., 2009).

The recent survey study conducted by Albright et a]. (2009) also provided information on

what horse owners in the United States (US) perceive to be the predominant
contributing factor to crib-biting behavior. The majority of respondents indicated the

horse’s environment was the primary cause.

77

Survey research studies have been conducted in the state of Michigan to
determine the economic impact of the equine industry, providing valuable information
about the number of horses, breed inventories, geographic distribution of horses, as well
as equine operation demographics (Michigan Equine Survey, 2007). However,
stereotypic behavior has never been investigated in the Michigan horse population. The
objectives of this study were to assess the perceptions of weaving and crib-biting
behavior held by Michigan horse owners and to investigate risk factors associated with
these behaviors in the Michigan horse population. It was hypothesized that Michigan
horse owners would express concern regarding weaving and crib-biting behavior and that
the majority of owners attempt to stop or reduce the performance of these behaviors.
Additionally, the authors hypothesized that certain housing and feeding strategies, for
example those which limit a horse’s time out of the stable, limit contact with other
horses, or limit foraging opportunities, would result in an increased probability of horses
exhibiting weaving and/or crib-biting behavior. Some of the specific predictions were
that increased turn-out would reduce the probability of a horse being a weaver or crib-
biter, and that the probability of being a weaver or crib-biter would be increased in horses
with no visual or tactile contact with other horses. Furthermore, that those horses fed
larger amounts of concentrate would be more likely to weave or crib-bite.

Materials and Methods
Horse behavior questionnaire

In collaboration with Michigan State University’s (MSU) Office for Survey
Research (OSR), a comprehensive online survey instrument was constructed and
administered to Michigan horse owners to collect information about weaving and crib-

biting behavior. The questionnaire consisted of a combination of multiple choice, Likert

78

scale, and open-ended question types (Table 3-1). Questions were focused on stable and
management information (type of operation, number of horses, housing, feeding,
exercise), and horse behavior (familiarity and level of concern with weaving and crib-
biting behavior, type of behavior observed on each farm and number of horses exhibiting
the behavior, demographics on weaving and crib-biting horses, and methods used in an
attempt to reduce or stop the behavior). Previous survey research studies conducted in
Canada and the United Kingdom to investigate stereotypic equine behavior served as a
starting point for question generation (McGreevy et al., 19953,b; Luescher et al., 1998;
McBride and Long, 2001; Waters et al., 2002). Prior to the beginning of the study, the
survey was pilot tested on a small group (n = 12) of equine owning faculty, staff, and
graduate students within MSU’S Department of Animal Science for the purpose of
assessing questionnaire content, readability, and functionality of the online format.
Comments obtained from the pilot group were used to modify the questionnaire. A copy
of the final questionnaire is included in Appendix A.
Sampling frame and subject recruitment

The sampling frame for this study consisted of self-selected recipients of the MSU
Equine Newsletter and Saddle-Up magazine in addition to web site recruitment of
owners, managers, and barn employees belonging to Michigan-based equine
organizations. Participating organizations included the Michigan Horse Council, Arabian
Horse Association of Michigan, Michigan Quarter Horse Association, Central Michigan
Horsemen’s Association, Wolverine Morgan Horse Association, and the Michigan Horse
Show Association. The sampling frame was further supplemented by providing a link to

the survey on the MSU Youth Equine Extension Site. The sampling frame was

79

constructed in this manner in an effort to collect information across a wide variety of
horse breeds/types and riding disciplines and was created in consultation with OSR staff.
Subscribers to the MSU Equine Newsletter and Saddle-Up magazine were
informed of the survey study through a brief project description posted in both
publications explaining the purpose of the study and directing readers to a web link from
which they could access the electronic survey. The contact information for the project
leader also was provided in case interested persons without computer access wanted to
request a paper copy of the questionnaire. The various horse organizations were
contacted in July and August 2006 by phone or email and asked to provide a link to the
online questionnaire on their organization's home page and/or to post a project
announcement in their fall newsletters. Subjects were encouraged to participate even if
they did not currently own or manage a weaving or crib-biting horse. This allowed
assessment of potential risk factors based on management information collected for non-
stereotypic as well as stereotypic horses. Data collection was initiated in October 2006
and concluded at the end of February, 2007. In March 2008, a subset of respondents from
the online survey, who had indicated their willingness to participate in future research,
was contacted either by post or email to request a follow-up telephone interview. The
purpose of conducting a follow-up interview was to collect horse characteristics (breed,
sex, and age) of the non-stereotypic horses on the respondents’ farms as well as
additional information about the history of their weaving and crib-biting horse(s). All
survey research procedures were approved by MSU’s Committee for Research Involving

Human Subjects.

80

Data collection

Data were collected and stored electronically at MSU through the OSR server
using 128 bit encryption software. Data from the server were downloaded to one specific
login/password protected personal computer. The project manager at OSR was
responsible for uploading and downloading data and was the only person with access to
the sample files and the software used for collecting data. The software program allowed
data to be uploaded into a database compatible with SPSS (Statistical Package for the
Social Sciences, SPSS Inc., Chicago, IL) statistical software to facilitate later data
analysis. The online format allowed for inclusion of a detailed battery of questions while
minimizing the time commitment required of respondents. To the author’s knowledge,
this is one of the first studies to utilize a web-based survey instrument to collect
information about stereotypic behavior in horses.
Statistical Analysis

Farm management, owner perception, and horse behavior data were analyzed

using SPSS (Version 16.0; SPSS Inc., Chicago, IL). Data are presented as frequency
counts. Where Bernoulli proportions were conducted (Lindgren, 1976), question
responses were first ranked from highest percentage of respondents to lowest, and then
the proportion for each situation was compared with that of the next highest proportion
on the list using pairwise Z-test statistics. In addition, univariate and multivariate logistic
regression models were fitted separately to the probability of a horse being a weaver and
to the probability of a horse being a crib-biter using SAS statistical software (Version 9.1 ,
SAS Institute Inc., Cary, NC). In each case, explanatory variables of interest considered

for model inclusion were the categorical variables: primary housing, type of turn out,

81

social contact, kilograms of concentrate fed per day, amount of controlled exercise per
week, and number of times per day horses had access to forage (Table 3-2), as well as the
continuous variables: hours of turn out and number of times per day horses were fed
concentrate. Selection of these variables as potential risk factors for weaving and crib-
biting behavior was based on their known or suspected associations presented in the
literature (McGreevy et al., 1995a,b; Luescher et al., 1998; Waters et al., 2002). A
stepwise model selection procedure was implemented for each response variable to
determine a final model of best fit, which was then used for analysis. Criteria for
inclusion and stay of factors/covariates in the model were p-values of 0. l 0 and 0.15,
respectively. Results are presented as estimated odds ratios (OR) and 95% confidence
intervals. The odds of an event occurring is defined as the probability that the event will
occur to the probability that it will not. The odds ratio is the ratio of the odds of
stereotypic behavior (e. g. odds of a horse being a weaver) in the exposed group (e.g. no
visual or tactile contact with other horses) to the odds of stereotypic behavior in the
unexposed group (e. g. turned out with other horses). An odds ratio greater than 1.0
indicates an increase in risk while an odds ratio less than 1.0 indicates decreased risk.
Questions related to the daily amount of forage fed to horses as well as type of
forage offered, type of stall bedding used, and type of concentrate feed offered were
included in the online questionnaire. However, these variables were excluded from the
statistical analysis due to extreme category problems (i.e. under representation of each
behavior type across levels of the categorical variable) or because of difficulties in
collecting detailed information. For example, when asked to report the amount of forage

fed to horses on average each day, some owners provided the actual amount in pounds

82

per day while many others indicated amount in terms of flakes of hay per day. There was
also overlap among the types of grain fed to horses in approximately 50% of the cases,
e. g. several owners reported feeding more than one type of grain.
Comparison of sample data with 2007 Michigan Equine Survey results

The 2007 Michigan Equine Survey (MES) project constituted a state-wide census
of horse operations based on a comprehensive list sampling frame and enumeration.
Comparisons were made between sample demographics from the current study and
results of the MES (where possible based on question wording and category choices) to
assess the degree to which participating farms/owners and their horses were
representative of the larger Michigan equine population.
Results

A total of 293 useable responses were received, representing a total of 2,181
horses. This included 290 responses to the online questionnaire and 3 paper copies of the
survey instrument which were included in the analysis. The percentages of horses
exhibiting weaving and crib-biting behavior were 2.7% (n=58) and 5.2% (n=113),
respectively. The median number of horses per facility was 4 (range 1-100 horses/farm).
There were 191 farms without any stereotypic behavior observed in their horses, 17 farms
with both non stereotypic horses and weavers, 60 farms with both non-stereotypic horses
and crib-biters, and 25 farms indicating that non-stereotypic, weaving and crib-biting
horses were owned/managed. Respondents were not asked detailed questions as to
whether a particular horse performed both weaving and crib-biting behavior, thus it was

difficult to determine with certainty if horses exhibited more than one stereotypic

83

behavior. However, careful examination of individual horse information suggested that a
small subset of horses (n=9) probably exhibited both behaviors.

A comparison of farm/respondent demographics between this study and the 2007
Michigan Equine Survey (MES) is presented in Tables 3-3 and 3-4. The percentage of
operations with 3-9 head of horses was similar between the two studies as was the
distribution of operations by primary housing. Wording of our questionnaire was
developed prior to the development of the MES. Thus, responses to the two survey
instruments are not completely parallel. Based on the recent MES data, 76% of farms
surveyed were private residence/backyard operations with the majority of horses (37.1%)
being used for recreation/pleasure. In the current study, the majority of farms (37.9%)
also indicated pleasure/backyard as type of operation.

Owner perceptions about weaving and crib-biting behavior

Of those respondents owning or managing weaving horses on their farm (n=42
respondents; 58 horses represented), 57.1% believed environmental factors were the
primary cause of the behavior, 2.4% considered genetics to be the primary cause, and
40.5% believed that a combination of genetic and environmental factors contributed to
the behavior. A similar trend was found within respondents owning or managing crib-
biting horses (n=84 respondents; 112 horses represented) with 54.8%, 2.4%, and 42.9%
believing environment, genetics, or a combination of genetics and environment,
respectively, predominantly contribute to the behavior. The level of concern with
stereotypic behavior, specifically perceptions about the impact of weaving and crib-biting
behavior on the horse’s learning ability, health, and monetary value, differed depending

on whether the respondent currently owned or managed an affected horse (Figures 3-1

84

and 3-2). Compared to respondents with weaving and crib-biting horses, a greater
percentage of responders with non-affected horses agreed/strongly agreed that weaving
and crib-biting behavior hinder learning ability (P < 0.01), have a negative impact on
horse health (P < 0.01 for weaving, and P < 0.05 for crib-biting), and reduce the horse’s
monetary value (P < 0.01). The apparent triggers to weaving and crib-biting behavior
based on the observations of the respondents are shown in Figure 3-3. The factors
reported to initiate bouts of weaving and crib-biting behavior appear to differ. Horses are
observed to begin weaving before feeding (54.8%), before tum-out (35.7%), or when the
horse is separated from other horses (38.1%). The majority of farms reporting on crib-
biting behavior indicated that crib-biting starts after feeding (45.9%), but that the
behavior also occurs before feeding (38.8%), and as a result of separation from other
horses (20.0%). Of the respondents owning or managing weaving and crib-biting horses,
a small percentage indicated that a horse had started to weave (2.5%) or crib-bite (5.0%)
after another weaving or crib-biting horse had arrived at their farm.

Respondents indicated that they receive equine behavior information from a
variety of sources including the Internet, equine magazines, equine association and/or
extension newsletters, equine seminars/workshops, their veterinarian, and personal
contacts (e. g. fellow horse owner). The most frequently marked resources included
equine magazines (70.3%), personal contacts (62.8%), veterinarian (49.8%) and the
intemet (44.4%). Equine newsletters and seminars/workshops were utilized less

frequently (28.7% and 25.6%, respectively).

85

Attempts at stopping or reducing weaving and crib-biting behavior

Although many respondents attempt to stop horses from performing weaving and
crib-biting behavior (Figure 3-4), the proportion of respondents attempting to stop crib-
biting behavior (81.0%) was greater (P < 0.01) compared to weaving behavior (37.5%).
The methods employed by owners/farm managers in an attempt to stop or reduce
weaving and crib-biting behavior are shown in Figures 3-5 and 3-6, respectively.
Methods most frequently used to stop or reduce weaving behavior included increased
tum-out (21.4%) and provision of toys (14.3%). The methods used most frequently to
stop or reduce crib-biting behavior included fitting horses with cribbing collars/straps
(77.6%), alteration of crib-biting surfaces by, for example, electrifying fences and/or
applying distasteful substances to barn and pasture fixings (55.3%), followed by
increased turn-out (54.1%). Many farms utilized a combination of methods. The
percentage of respondents reporting that attempts to stop weaving and crib-biting
behavior were successful was only 26.7% and 27.9%, respectively. Attempts to reduce
weaving and crib-biting behavior were more successful (66.7% and 49.3%, respectively)
than attempts to completely extinguish the behavior. Consistent use of a cribbing collar
was reported as the only means of effectively stopping a horse from performing crib-
biting behavior (n=11). The method reported as being successful in stopping weaving
behavior was keeping horses turned-out rather than confined to a stall (n=3). Increased
turn-out, often in combination with increased social contact (n=8) and providing horses
with a consistent routine (n=2) were the methods respondents reported as being effective

in reducing weaving behavior. Methods reported by respondents as being effective in

86

reducing crib-biting behavior included increased turn-out (n=14), increased forage (n=5),
and removal/alteration of crib-biting surfaces (n=10).
Characteristics of weaving and crib-biting horses

Sex distribution of weaving and crib-biting horses is presented in Table 3-5.
lnforrnation about sex of horse was reported for 56 weaving and 106 crib-biting horses.
Of the 56 weaving horses, 25 (44.6%) were geldings, and 30 (53.6%) were mares. One
weaving horse was a stallion (1.8%). Of the 106 crib-biting horses, 73 (68.9%) were
geldings and 33 (31.1%) were mares. Breed distribution of weaving and crib-biting
horses is presented in Table 3-6. Breed data were reported for 56 weaving and 107 crib-
biting horses. Large percentages of weaving horses were Quarter Horses (35.7%) and
Thoroughbreds (21.4%). Many of the horses exhibiting crib-biting behavior were also
Quarter Horses (47.7%) and Thoroughbreds (14.0%). No draft horses were reported as
exhibiting either behavior. The majority of both weaving (71.9%) and crib-biting (60.7%)
horses were aged 6-17 years. Data related to age distribution of horses were not included
in the MES (2007) survey, but data collected on weaving and crib-biting horses were in
good agreement with age demographics reported in the 2005 US. Department of
Agriculture Equine Health and Management survey (56.7% of horses were 5-19 years of
age). The sex, breed, and age of non-stereotypic horses were not queried in the web-
based survey.
Associations between farm management factors and weaving behavior

At the univariate level (Table 3-7), an association was identified between primary
method of housing and the probability of weaving (P = 0.041). Horses housed equally

between stalls and pasture were 2 times more likely to display weaving behavior (OR =

87

2.10) compared to horses housed completely on pasture. An association also was found
between hours of turn out and the probability of weaving (Maximum Likelihood Estimate
= - 0.0319, P = 0.053). An increase in 1 hour of tum-out was associated with a 0.97
change (i.e. with a 3% decrease) in the odds of a horse being a weaver. Factors remaining
in the final multivariate logistic regression model (Table 3-8) included kilograms of
concentrate fed per day (P = 0.043) and hours of turn-out (Maximum Likelihood Estimate
= -0.0619, P = 0.003). The odds of being a weaver decreased by 75% in horses that
received 2.3 or more kilograms of concentrate per day (Adjusted OR = 0.25) compared to
horses that received less than 0.5 kilogram of concentrate per day. An increase in 1 hour
of turn-out was associated with a 0.94 change (i.e. with a 6% decrease) in the odds of a
horse being a weaver.
Associations between farm management factors and crib-biting behavior

At the univariate level (Table 3-7), associations were identified between primary
housing (P = 0.066), type of turn-out (P = 0.037), social contact (P < 0.001) and the
probability of crib-biting. Horses housed primarily in stalls were 1.8 times more likely to
display crib-biting behavior (OR = 1.83) compared to those housed on pasture. Horses
turned out in an indoor arena were 7.2 times more likely to display crib-biting behavior
(OR = 7.24) than horses receiving pasture tum-out. Compared to being turned out with
other horses, horses allowed visual contact only and horses with no visual or tactile
contact with other horses were 5.5 (OR = 5.48) and 6.6 (OR = 6.64) times more likely to
display crib-biting behavior, respectively. An association was also found between hours
of turn out and the probability of crib-biting (Maximum Likelihood Estimate = - 0.0310,

P = 0.009). An increase in 1 hour of tum-out was associated with a 0.97 change (i.e. with

88

a 3% decrease) in the odds ofa horse being a crib-biter. Factors remaining in the final
multivariate logistic regression model (Table 3-8) included social contact (P < 0.001),
kilograms of concentrate fed per day (P = 0.001) and hours of turn-out (Maximum
Likelihood Estimate = -0.0425, P = 0.006). Compared to being turned out with other
horses, horses allowed visual contact only and horses with no visual or tactile contact
with other horses were 5.6 (Adjusted OR = 5.61) and 6.3 (Adjusted OR = 6.26) times
more likely to display crib-biting behavior, respectively. The odds of a horse being a crib-
biter were decreased by 44% in horses that received 0.5-1.8 kilograms of concentrate per
day (Adjusted OR = 0.56) compared to horses that received less than 0.5 kilogram of
concentrate per day. The odds of being a crib-biter were decreased by 82% in horses that
received 2.3 or more kilograms of concentrate per day (Adjusted OR = 0.18) compared to
horses that received less than 0.5 kilogram of concentrate per day. An increase in 1 hour
of turn-out was associated with a 0.96 change (i.e. with a 4% decrease) in the odds of a
horse being a crib-biter.
F allow-up interviews

Requests for a follow-up telephone interview were sent via post or email to 124
respondents. These participants had indicated in their responses to the web-based
questionnaire that they had an interest in participating in additional research. A total of 38
individuals provided their telephone number (30.6% response rate). However, repeated
attempts to contact 12 farms were unsuccessful, thus interviews were conducted with
only 26 respondents. Therefore, the actual response rate was 21.8% representing a total of
149 horses, 4 of which were weavers and 13 of which were crib-biters. A time lapse of

approximately 13 months passed between completion of data collection using the online

89

 

questionnaire and the follow-up interviews. Follow-up interviews were conducted for the
purpose of collecting sex, breed, and age of stereotypic and non-stereotypic horses on
respondents’ farms simultaneously and to ask about the history/early life of weaving and
crib-biting horses. Unfortunately, due to the small sample size, associations between
stereotypic behavior and individual horse characteristics could not be examined. One
owner did report that a newly acquired yearling Percheron drafi horse stallion had started
to crib-bite. The horse suffered a leg injury shortly after weaning and had been confined
to a stall with limited exercise in the form of hand-walking and round pen tum-out.
Qualitative assessment of responses to questions pertaining to the history/early life of
weaving and crib-biting horses and behavior management strategies revealed several
interesting themes and are presented in Table 3-9.
Discussion

This study is the first to investigate stereotypic behaviors in Michigan horses. The
percentages of horses in Michigan exhibiting weaving and crib-biting behavior were
similar to the prevalence reported previously in other countries (Nicol, 1999).
Specifically, the percentages observed were almost identical to the approximately 2.6%
and 5.3% of Canadian horses reported to exhibit weaving and crib-biting behavior. The
findings are also in good agreement with the 4.4% prevalence of crib-biting recently
reported for US. horses (Albright et al., 2009). Although the sampling frame for this
study consisted of self-selected respondents, it would appear based on comparisons with
demographic data from the 2007 MEMS survey, that the sample was fairly representative
of the Michigan equine population with respect to number of horses per operation, type

of operation and primary use of horses. and manner in which horses were primarily

90

housed. Calculating a precise response rate for the web-based survey was not possible
due to the passive style of recruitment used to invite subjects to participate (Vate-U-Lan,
2006). Passive invitation methods, e. g. posting a link to the questionnaire on an
association web page, do not allow the researcher to determine the contact rate. Knowing
the total membership or readership of the various participating Michigan-based equine
organizations and publications would not necessarily provide an accurate assessment of
the number of people who viewed the survey advertisement, and as a result would greatly
underestimate the response rate.

In the study conducted by McBride and Long (2001), British horse owners
demonstrated concern regarding the performance of stereotypic behavior and attempts
often were made to physically prevent horses from performing the behavior. This
included the use of anti-weave bars and cribbing straps to stop horses from weaving and
crib-biting, respectively. Michigan horse owners and farm managers also expressed
concern about weaving and crib-biting behavior, but perceptions regarding the impact of
these behaviors on horse performance/learning, horse health, and monetary value of the
animal were different between owners of non-stereotypic horses and those respondents
currently owning/managing a weaver or crib-biter. Overall, respondents presently
owning/managing weaving and crib-biting horses were less concerned about the
behaviors having a negative impact on learning, health, or monetary value. Thus,
perceptions about stereotypic behavior within the equine community at large may not
coincide with those held by individuals having first-hand experience with weaving and
crib-biting horses, particularly if a horse exhibiting stereotypic behavior has been a

beloved companion, performs well, and/or has not encountered any major health

91

problems. For example, a relatively low percentage of respondents with weaving and
crib-biting horses indicated that they are concerned the behavior hinders learning or
performance, thus it would not appear that these behaviors substantially interfere with
training or with the overall usefulness of the horse. However, in light of recent research
providing some evidence that learning ability is impaired in stereotypic horses
(Hausberger et al., 2007; Parker et al., 2008), this is an area that warrants further
investigation. Additional inquiry into the experiences and observations of horse owners
may provide valuable insight into how weavers and crib-biters respond to various training
and handling procedures compared to their non-stereotypic counterparts.

Concern with weaving and crib-biting behavior having a negative impact on horse
health was lower among respondents with these behaviors on their farms compared to
respondents with only non-stereotypic horses, but it Should be noted that the respondents
currently owning/managing crib-biting horses still expressed a high level of concern with
the behavior from a horse health standpoint. Indeed, qualitative assessment of responses
obtained from owners of crib-biting horses revealed that crib-biting had resulted in at
least some wear of the incisor teeth and that some crib-biting horses had experienced
repeated bouts of colic. It is not possible to ascertain from the qualitative interview data,
given the small sample size and lack of information about colic in non-stereotypic horses,
whether a true link between crib-biting and colic exists, but an association between crib-
biting behavior and epiploic foramen entrapment, a specific type of colic, has been
demonstrated (Archer et al., 2004, 2008). Crib-biting behavior also has been associated
with gastric ulceration in foals (Nicol et al., 2002), and with lower basal and post-feeding

gastric pH in mature crib-biting horses (Lillie et al., 2004). In light of the work conducted

92

 

by Mills and Macleod (2002) demonstrating a reduction in crib-biting frequency in adult
horses receiving an antacid product, it was interesting to learn that some Michigan horse
owners had administered ulcer treatments to their crib-biting horses. The success of these
products in reducing crib-biting frequency, as reported by owners, varied. In fact, only
one owner indicated that the ulcer treatment appeared to result in a decrease in the
horse’s crib-biting behavior, and this effect would need to be validated through an on-
farm visit. The relatively high percentage of respondents perceiving some ill effect of

crib-biting behavior on horse health may partially explain why more respondents with

 

crib-biting horses attempt to stop the behavior compared to those respondents
owning/managing horses exhibiting weaving behavior. The possibility also exists that the
behavioral sequence of crib-biting represents a greater nuisance to horse owners/barn
managers (i.e. it is more aesthetically displeasing) or that there are simply more available
devices for physically preventing crib-biting behavior (e. g. collars, muzzles, electric wire,
and distasteful paints) compared to weaving behavior. The two groups compared using
Bernoulli proportions (owners with weaving horses and owners with crib-biting horses)
are neither independent nor entirely dependent. There would have been some degree of
overlap between the two groups as 25 respondents indicated having both weaving and
crib-biting horses on their farms and a very small percentage of horses exhibited both
behaviors. In general, the majority of respondents believed that environmental variables
were largely responsible for the performance of stereotypic behavior, similar to the
perceptions of US. horse owners reported by Albright et a1. (2009). This finding

indicates that Michigan owners and farm managers are aware and fairly well informed

93

that management practices do have a direct impact on the behavior of the horses under
their care.

The majority of weaving and crib-biting horses in Michigan were aged 6-17
years, therefore the stereotypic behavior exhibited by these horses was likely well
established. This would explain why the efforts of respondents to completely stop the
behaviors were largely unsuccessful. The use of cribbing collars and the removal or
alteration of cribbing surfaces by owners and managers in the current study could be
classified as a reactive solution to the performance of stereotypic behavior, while turnout
represents a more proactive management strategy aimed at minimizing or eliminating
causal factors. Increasing the amount of turn-out, social contact, and forage were reported
as being effective in reducing performance of the behaviors.

Situations or daily events reported as triggering bouts of stereotypic behavior
seemed to agree with factors identified in experimental behavior studies. In studies
conducted by Ninomiya et al. (2007) and Clegg et a1. (2008), weaving behavior was
mainly observed prior to feed delivery. Clegg et a1. (2008) also observed an increase in
the frequency of weaving in the hour preceding turn-out. Overall it appears from these
studies that weaving behavior is closely related to periods of high activity and frustration
associated with anticipation of an event, specifically feeding. Bouts of crib-biting were
reported as beginning after feeding which also coincides with results from previous
studies demonstrating increased frequency of crib-biting following consumption of
concentrate feed (Kusunose, 1992; Gillham et al., 1994; McGreevy et al., 1995c; Clegg et

al., 2008).

94

 

To date, there is very little evidence to support the belief that horses learn to
perform stereotypic behavior by observing others. The percentage of Michigan
respondents indicating that a horse had learned to weave or crib-bite after another
weaving or crib-biting horse had arrived at the farm was low, and on these farms,
management practices differed with regards to primary housing, social contact, and hours
of tum-out. Thus, it is difficult to distinguish whether horses are in fact copying the
behaviors or if the behavior is the result of exposure to common management factors,
specifically those factors previously demonstrated to be associated with an increased risk
for stereotypic behavior. In the study of crib-biting behavior in US. horses, only 1% of
horses started to crib-bite after the arrival of a crib-biting horse (Albright et al., 2009). In
contrast however, Nagy et a1. (2008) recently demonstrated an increased risk of
stereotypic behavior in horses exposed to stereotypic neighbors. Again, this is an area

that requires further empirical and epidemiological investigation before coming to a

conclusion about the ability of horses to learn a stereotypic behavior through observation.

Without information about the sex and breed of non-stereotypic horses, statistical
analysis of associations between horse characteristics and weaving and crib-biting
behavior were not possible. However, in looking at the percentages of weaving and crib-
biting horses that fall into each sex and breed category in relation to the percentages
reported in the MES, it appears that there is a relationship between horse characteristics
and the performance of stereotypic behavior. For example, based on these preliminary
findings related to weaving and crib-biting behavior in Michigan horses, it would appear
there is a gender-based factor involved. Specifically, mares appear to be more at risk for

locomotor stereotypy, e. g. weaving, while geldings appear to be at greater risk for oral

95

 

stereotypy, e. g. crib-biting. Breed of horse also appeared to be an important factor.
Thoroughbreds and Quarter Horses appeared to be at higher risk for stereotypic behavior
in general compared to other breeds. It is not known whether these apparent differences
reflect a genetic predisposition to stereotypic behavior or a gender by management or
breed by management interaction of some kind, and additional investigation would be
required in order to validate these findings. Compared to geldings, Luescher et a1. (1998)
found a higher prevalence of weaving in mares, and a greater risk of both weaving and
crib-biting among Thoroughbred horses. The study conducted by Luescher et a1. (1998)
also demonstrated a higher prevalence of weaving and crib-biting in stallions when
compared with the other sexes. The management of stallions is typically quite different
from geldings and mares. Stallions often are housed individually to prevent accidental
breeding and aggression, thus stallions may be more prone to developing stereotypic
behavior as a result of stress or frustration associated with limited social contact and/or
turn-out. None of the crib-biting horses and only 1.8% of weaving horses in the current
study were stallions. Approximately 5% of respondents to the Michigan horse behavior
survey reported breeding as the type of their operation. This is similar to the 4.6% of
operations in the MES survey reporting breeding as the primary activity of their
operation. It is likely that current stallion numbers in Michigan in general are much lower
than the number of geldings and mares and that very few of the farms participating in the
behavior survey presently owned or managed a stallion. The particular breed of horse
also may determine the primary use of the animal, which in turn may ultimately affect the
manner in which the horse is managed. For example, many Thoroughbreds and Quarter

Horses are used for competitive Sports including racing, eventing. or physically

96

 

demanding western events such as barrel racing and cutting. Race horses in particular
may be exposed during the early part of their life to rigorous training regimens, high
concentrate/comparatively low forage diets, very limited liberty turnout, and, depending
on the track stabling, very limited amounts of social contact. Thus interconnections
almost certainly exist between genetic and environmental factors and these relationships
warrant further consideration in future experimental and epidemiological studies.
Several studies have demonstrated associations between management factors and i

stereotypic behavior in horses. The risk of performing stereotypic behavior was decreased

Ayn.

 

in Thoroughbred horses fed 6.8 kilograms or greater of forage per day and offered forage , .
more than 3 times daily (McGreevy et al., 1995a). Stall designs that minimized social
contact between horses and use of bedding other than straw increased the risk of
stereotypic behavior (McGreevy et al., 1995a). It should be noted however, that in the
study conducted by McGreevy et al. (1995a), it appeared thatresults were reported only
for weaving, woodchewing, or abnormal behaviors as a whole. The authors indicated that
crib-biting behavior was included in the study, but the number of horses exhibiting crib-
biting behavior in the yards sampled was not clearly stated, and specific relationships
between management factors and crib-biting were not identified. However, a general
note included in the authors’ discussion indicated that yards offering less forage per day
had a higher prevalence of oral-based stereotypies (i.e. crib-biting). Wood-chewing
behavior is often included in epidemiological studies, although this behavior may be
more appropriately classified as an aberrant or redirected oral behavior as opposed to a
true stereotypic behavior pattern as it is more variable in form and more nearly resembles

normal ingestive behavior. Nonetheless, there is some evidence to suggest that wood-

97

chewing may be a precursor to crib-biting. Waters et al. (2002) documented that in foals
which developed crib-biting behavior, 74% had been wood-chewers. Therefore the
associations between various management factors and wood-chewing behavior are of
interest to investigators. In eventing and dressage horses in the UK, a greater amount of
time spent out of the stable was associated with a decreased risk of stereotypic behavior
(McGreevy et al., 1995b). Survey studies conducted to investigate stereotypic behavior in

race horses in Sweden (Redbo et al., 1998), and in Swiss horses of multiple breed types i

i.

and uses (Bachmann et al., 2003) found that regular feeding of concentrates increased the

 

risk of performing stereotypic behavior. Specifically, Redbo et al. (1998) reported a
positive relationship between the amount of concentrate and stereotypic behavior. In the
current study, the negative relationship identified between weaving and crib-biting and
hours of tum-out in both the univariate and multivariate regression models is in good
agreement with the previous finding that increased amount of time spent outside the
stable reduces the risk of stereotypic behavior (McGreevy et al., 1995b). In light of
previous studies demonstrating a role of social isolation in the performance of weaving
behavior (McGreevy et al., 1995a; Cooper et al., 2000; McAfee et al., 2002), it was
expected that the probability of a horse being a weaver would decrease in horses allowed
social contact with others, but no such association was demonstrated in the present study.
However, because visual contact with other horses alone has been shown to reduce the
frequency of weaving (Cooper et al., 2000), and all of the respondents reporting on
weaving horses indicated that horses were provided with at least visual contact with
others, the social contact factor did not contribute to the risk of weaving. It is interesting

that the risk of weaving was greater for horses housed equally between stalls and pasture

98

compared to those housed primarily on pasture. Perhaps horses housed in this manner
experience greater arousal or frustration in anticipation of scheduled turn-out times or are
more sensitive to stall confinement after having been left out for a longer period of time
during the day, as opposed to horses accustomed to living primarily in stalls. Although
increased tum-out in general appears to be protective against both forms of stereotypic
behavior, a minimum threshold level of turn-out required to prevent weaving and crib-
biting behavior from developing in horses remains to be determined.

Other individual level factors associated with crib-biting behavior included
primary housing, type of tum-out, and social contact. Horses primarily housed in stalls
and those receiving turn-out in an indoor arena only, were at greater risk of performing
crib-biting behavior. This finding is not surprising since these practices would greatly
limit the horse’s opportunity to engage in natural foraging behavior and social contact.
The extent or quality of social contact appeared to be an important factor as those horses
without social contact or visual contact only were at an increased risk for crib-biting
compared to horses being turned out with other horses or allowed both visual and tactile
contact. The magnitude and direction of the association between social contact and crib-
biting behavior was similar in the final model after adjusting for potential confounders.

The direction of the relationship between the amount of concentrate and
stereotypic behavior was surprising to these authors, especially for the risk of crib-biting
behavior. The authors had expected to see an increased risk of crib-biting behavior in
horses fed larger amounts of concentrate per day rather than a reduced risk. Associations
between increased concentrate feeding and increased crib-biting behavior have been

documented in several studies (Kusunose, 1992; Gillham et al., 1994; Waters et al.,

99

 

2002). Gillham et al. (1994) and Kusunose (1992) observed an increase in crib-biting
frequency shortly after the consumption of concentrate meals, and in the prospective
study conducted by Waters et al. (2002) foals receiving concentrate feed after weaning
were 4 times more likely to develop crib-biting behavior than those receiving only forage.
Data on the amount of concentrate fed to horses in the current study was not collected on
a continuous scale. Limiting owners to the selection of only a few categories may have

hindered the ability to detect a similar association between this variable and stereotypic

'1' WV!
.1

behavior found in previous studies. For most horses, it only takes a short amount of time

 

to complete a small grain meal. Thus, it is possible that horses receiving less than 0.5 ~'- -_
kilogram of concentrate per day experience some frustration associated with feeding and

as a result, redirect their thwarted feeding motivation toward aberrant oral or locomotor

behavior. The type of concentrate offered also may affect the association between grain

feeding and stereotypic behavior. For example, consumption 0f highly palatable sweet

feed was shown to increase crib-biting frequency (Gillham et al., 1994). Many of the

Michigan horse owners and barn managers reported feeding a variety of ingredients and

the protein, carbohydrate and fiber content of the diets was not queried. Therefore, in this

study, it is unknown how the type of concentrate, and perhaps more specifically, how the

nutrient composition of the diet may influence the risk of stereotypic behavior.

In light of the unexpected nature of the results regarding the relationship between
greater amounts of concentrate fed and apparently lower rate of stereotypic behaviors in
horses, a post-hoe analysis was used to investigate this association further. The author
looked at the effect of an interaction between daily amount of concentrate and the number

of times horses were fed grain per day on the predicted probability of stereotypic

100

behaviors. This analysis confirmed the original finding that greater daily amounts of
concentrate were associated with an increase in the probability of a horse being a weaver
or crib-biter. More specific questions about the diet of horses should be included in
epidemiologic research studies in an effort to further elucidate the role of concentrate
feeding in the development and continued performance of stereotypic behavior.

Finally, an association between the number of times per day horses had access to
forage and the risk of stereotypic behavior was also suspected in this study. It would
seem that offering forage more frequently throughout the day would potentially increase

the total amount of time horses spend eating, resulting in less time engaged in stereotypic

 

behavior. However, it is likely in light of the findings of McGreevey et al. (1995a), that
the total daily amount of forage is the more important factor in minimizing the risk of
stereotypic behavior. The role of gut fill and satiety mechanisms in the development and
continued performance of stereotypic behavior, specifically the oral stereotypy of crib-
biting, remains to be determined. Clustering (i.e. the non-independence of horses housed
within the same farm) was not accounted for in this study. Thus it is possible that by not
fitting a random effect of farm, some of the associations identified between the
probability of weaving and crib-biting and the various management factors may have
been influenced by differences in management of horses between farms. However, even
if clustering of horses is considered in the statistical model, it may still fail to account for
“sub-clusters” within farms. For example, within a given farm, horses exhibiting
stereotypic behavior may be managed differently from non-stereotypic horses, just as
management strategies implemented for horses of a particular sex or age group may vary

within a farm.

101

Conclusion

The results of the current study demonstrate that weaving and crib-biting occur in
Michigan horses with the same frequency reported in other countries, and that owners are
concerned about these behaviors. Although some of the associations identified between
management factors and weaving and crib-biting behavior were in good agreement with
those reported previously, some of the relationships investigated produced conflicting

results and warrant further investigation. Reports by owners/barn managers that increased

' .".

tum-out and social contact and/or increased amounts of forage are effective in reducing

 

weaving and crib-biting behavior concur with previous studies in which housing and
feeding strategies have greatly influenced the performance of stereotypic behavior.
Management strategies that provide horses with increased opportunities to engage in
natural behaviors should be employed as the first means of moderating stereotypic
behavior, as they are aimed at removing the causal factors of the stereotypic behavior
rather than at physical prevention that may result in reduced welfare. These practices
would be especially crucial to implement on farms raising young horses in order to
prevent the development of stereotypic behavior or for those owners with horses just
starting to engage in weaving or crib-biting behavior so as to prevent the behaviors from
becoming established. Many owners in Michigan look to veterinarians for information
about stereotypic behavior, highlighting the importance of providing veterinary
professionals with current, science-based information about performance of stereotypic
behavior in horses. Continued research into the etiology of weaving and crib-biting
behavior is needed. Acquiring a better understanding of the interactions between horse

characteristics and management factors in the development of stereotypic behavior, and

102

the impact of these behaviors on the learning ability and health of affected horses will be
essential in making sound recommendations to horse owners and farm managers on how
best to care for their horses.
Acknowledgments

Sincere gratitude is extended to all who participated in the survey, for their
interest in this research and their dedication in filling out the questionnaire. The authors
would like to thank all of the participating Michigan equine associations/organizations I
for their help in providing their members with access to the web-based questionnaire.

Thank you to Karen Waite and Dr. Christine Skelly for their assistance in further

 

developing the sampling frame through the MSU Equine Newsletter and Youth Equine
Extension web page. We would also like to thank Kim Dobson for her help with follow-
up interview request mailings. Finally, we wish to thank Dr. Adroaldo Zanella for his

assistance in the initiation of this project.

103

Table 3-1 Examples of question types used

 

Question type Question asked

Response choices

 

How are horses
primarily housed
during the majority of
the year (2 9 months)?

Multiple choice

I am concerned with
cribbing behavior

Likert because it negatively
impacts the health of
the horse

What techniques or
management changes
have been successful
in reducing crib-biting
behavior?

Open ended

I.) Mainly in stalls 2.) Mainly
on pasture 3.) Nearly equal
time between stalls and
pasture

Strongly Agree, Agree,
Neither Agree nor Disagree,
Disagree, Strongly Disagree

 

Survey participants were
provided a text box in which
to type their responses

 

104

Table 3-2 Levels of the explanatory variables (housing and management factors)

 

 

 

Variable Description
Primary Stalls Pasture Equal
housing between
stalls and
pasture
Type of turn Pasture Grass Dry lot Indoor
out paddock arena
Social contact Turned Visual Visual and No visual
out with contact tactile or tactile

other only contact* contact

horses
Kilograms Less than 0.5-1.8 2.3 or more
concentrate 0.5 kg kg kg
(kg/day)
Controlled Less than 1-3 times 4-6 times 7 times
exercise? 1 time
(times/week)
Access to 1 time 2 times 3 times 4 times Free
forage choice
(times/day)

 

* Horses can touch over fence line, stall door, between bars on stall.
I Horse is ridden, driven, lunged.

105

Table 3-3 Percentage of operations by number of horses

 

 

Number of Horses Sample (n=293) MES 2007* (n=35,000)
1-2 head 33.1 44.0
3-9 head 47.7 47.4
10-29 head 13.4 7.6
30+ head 5.3 1.0

 

I Michigan Equine Survey (MES).

106

 

Table 3-4 Percentage of operations by housing category

 

 

Housing category Sample (n=293) MES 2007 (n=35,000)
Primarily stalls 11.9 9.7
Primarily pasture 42.0 38.7
Equally between stalls and pasture* 46.1 51.6

 

I Michigan Equine Survey (MES) labeled this category as Partially Stalled. Equine
operations in this category were described as providing horses with 4 or more hours of
turnout daily.

107

 

I'i".‘s'r I.
3

Table 3-5 Sex distribution of weaving and crib-biting horses (percentage of horses in
each category) in the present study

 

 

Sex Weaving (n=56) Crib-biting (n=106)
Geldings 44.6 68.9
Mares 53.6 3 1 .1
Stallions l .8 0.0

 

 

108

Table 3-6 Breed distribution of weaving and crib-biting horses compared to general
Michigan horse population (number of horses in each breed category)*

 

 

Breed Weaving Crib-biting MES 2007+

Arabian 8 (14.3) 4 (3.7) 12,500 (8.3)

Half-Arabian 2 (3.6) 0 (0.0) 3,400 (2.3)

Draft 0 (0.0) 0 (0.0) 13,900 (9.3)

Morgan 1 (1.8) 8 (7.5) 4,000 (2.7)

Pony 2 (3.6) 3 (2.8) 9,300 (6.2)

Quarter Horse 20 (35.7) 51 (47.7) 41,000 (27.4)

Saddlebred l (1.8) l (0.9) 1,800 (1.2)

Standardbred 2 (3.6) 2 (1.9) 12,000 (8.0)

Thoroughbred 12 (21 .4) 1 5 (14.0) 7,100 (4.7)

Warmblood 0 (0.0) 2 (1.9) 3,800 (2.5) 7"
Other 8 (14.3) 15 (14.0) 37,300 (24.9)

Grade 0 (0.0) 6 (5.6) 3,700 (2.5) ,2
Total 56 107 149,800 E --

 

 

* Percentage of horses Shown in parentheses.
I Michigan Equine Survey (MES).

109

 

 

8.33 m2 333 who sous: 52: son

53 and oases 8:868
oedema as Said So be} use a 3
2.32 as 2.32 8.2 seas. w. 72
2: of an} no 55 a3

0o — .O m 34.0 38800200 mguwomvm
Seuss Se - - .289 a new: oz
E. Zone and 5.345 one sees one as?»
Samoa men 8.3% as 33.5
. . mumpos
OO ~ 00 fl .5503» SCIESP

Sod v See noseoo _soom
$.53; e3 - - .aeoa 8er
m _ sees as: Smog am: so. an
3. Zoe 8.. on: no one sooeeea new
of of ocean

Red need so Eeco 8?
3.2% Ea 3.3;; SA .95
oo._ oo. fl 28mg
e n .33 a: 32:3 a: 2.8m

Sod zoos mason BEE

5 some one See use; 6 some one see use; ozsa>

Seance

magma?

*co_mm2w8 ouflwo— co comma 338:8 we
532.09 wcufibto one 9:283 @258qu no em: on“ was moose.“ EoEomewE ES.“ coosaon 83323—8 Butafig o5. brm «Zak.

110

 

 

nix—£8 2: Bot 32:88 was» one momcon war/83 o: cog—2: bowoueu e
.o_.o v m as EeéEwR Segovia—8 £82m _o>o_ 126365 ...
2.785 a: $3 8.723 m: 33 ran can no.2 nose 1
36.36 nod oood oo. 3.06 nod mmod >23 So EBB 33$ 1
52.33 mud wo._-mm.o mmd 823: 3..
m _ .mémd $6 3.2:; o: .3338: an
em. 33 e2 $.32 _2 Senna: m
2: a: sewage
3.2-86 mm; moémdmd L _.m 338:0
owed mi .o awash 8 mmooo<
ofiTnmd cod 3. Tmmd mmd Seance: A
we. Toad mad Sdémd we; x8388: 0%
co; co; 0.8.386: m;

32:50 5. ozne

 

 

.Amocfl $000 08380 0.3 880$ 0050 000 083.60 mote £000 *

 

 

00.0-3.0 00d coed wodéod v0.0 mood \4:00 0:0 E0000 2:0:
$6-30 3o 333.0 mg 53 use 8 mm D.
$.32 03 :. 30o NS 53 m. 72 1
co; co; 339— md :05 $04
50.0 90.0 0000000000 mfifiwozz
S. _mbw; 0N0 - - 0:008 00 55> 02
omdéod n _ A - - 0:008 28 13% >
mm; Tow.m $6 - - .035
. - - 8&0:
cc _ 0050}, 30-53.
Sod v .3. 008000 300m
5 $3 0:8 300 02?-0 _0 $3 0:8 £000 020>i 03093

fix. u 5 $05.35

QR u 5 @333

23883.. ouflwB 00 0300 003838 m0 0030000
@5505 new @0333 wEESfoQ m0 xmt 05 0:0 30000.0 EquwmcmE Sufi 0003000 3330028 00053036 05. ”4” «Ssh.

 

 

 

' LIP .. 0

0E 0 0:00 .00 00000 0000000900 00; 00002

0 _00E

\0 m E00w 00000 00 mats—0 000:. 00300000 000000000;

0 0030000 05 0000000000

E N 300300 .0000 “0003 0300.00 000 0000— 2

$5000

m m 30:0 E 000000 00 ,000000 0.053000 03000::

N m wEE000 0000 E 000000 00 ,0000: 0000-xm

000 00 0000» N :05

0 _ 000000w 00 00300000 $080803 :0w00 0000:

0 0w0 00 0000» m

E m :05 000_ 00 0030000 wEE00000Q :0w00 00005

u08000.0 02000000000

0 m 00 0030000 wEE000000 =0w00 0000:

u..00000005000000000

m w :05» 0030000 wEE000000 300:0 003 00003
$00002 0000:
010 0:333 s 7.5 0:03-000 80 25.:

 

00mw000000 0:0E0w00000 0:0 30000000000200
3000 .w.0v E0000: 0000: 00 0000—00 0000000030 00 0000000000 00030 ,00 0008000000 030002030 05 E000 00805 E00000 0-0:. 030,—.

113

 

.000 00 000000 N 0005 000_

00 8000 0000830 000 00 wE0E-E00 :0w00 008 0000: 0080000 000 000000000 0003 9003-000 00000—0 003 005 0000; 0 00: 000030 0:0 *

 

e:

00:

s:

s:

3:

0030000
08
E 0000000000 0 300 000 80800000 0008 000%

0030500 05 E 000000000
0 030000 00: 000 :5 080800000 000—: 000001

005.000

0000; 00— 000.0 30: 0:0 00:00 000000-000 00th
”00000000000 8080w0c02

£000 00000E 00 0003 0800 00: 0000:

0000: :0 Ew00>> 00000 00 0003

503 0.003 00 000 “00008008 0:8 0_ 0000:

00:00:00 0-0 02.;

114

 

80 * *

 

 

 

 

 

 

Percentage of respondents who
strongly agree/agree

Weaving hinders Weaving Weaving reduces
learning negatively impacts monetary value
heahh

I Farms without stereotypic horses I Farms with weavers

Figure 3-1 The percentage of respondents agreeing with three separate statements
about weaving behavior. The percentage of strongly agree/agree responses for each
respondent category were ranked from highest to lowest and treated as Bernoulli
proportions (Lindgren, 1976). The Bernoulli proportion was compared for each
situation to that of the next highest proportion using pairwise Z—test statistics.

* Denotes a significant difference between farms without weaving horses and farms
with weaving horses in each of the three cases at P < 0.01.

115

 

Percentage of respondents who
strongly agree/agree

Crib-biting hinders Crib-biting Crib-biting
learning negatively impacts reduces monetary
health value

 

I Farms without stereotypic horses I Farms with crib-biters

Figure 3-2 The percentage of respondents agreeing with three separate statements
about crib-biting behavior. The percentage of strongly agree/agree responses for each
respondent category were ranked from highest to lowest and treated as Bernoulli
proportions (Lindgren, 1976). The Bernoulli proportion was compared for each
situation to that of the next highest proportion using pairwise Z-test statistics.

* Denotes a significant difference between farms without crib-biting horses and
farms with crib-biting horses in each of the three cases at P < 0.05, ** 1? < 0.01.

116

so fi———i — a

 

 

 

 

 

 

 

Percentage of respondents

Before After Before During show When

feeding feeding turn-out or event separated
from other
horses

 

I Weaving I Crib-biting

Figure 3-3 Apparent triggers to weaving and crib-biting as reported by owners.
Several respondents indicated more than one trigger.

117

"5.
a,

 

I Weaving
I Crib-biting

Percentage of respondents

 

Yes, attempt to stop behavior

Figure 3-4 Percentage of respondents attempting to stop horses from performing
weaving and crib-biting behavior. The percentage of yes responses for each
respondent category were ranked from highest to lowest and treated as Bernoulli
proportions (Lindgren, 1976). The Bernoulli proportion was compared for each
situation to that of the next highest proportion using pairwise Z-test statistics.

* Denotes a significant difference at P < 0.01.

118

 

 

More forage/less grain

More social contact fi

Provison of toys

 

 

 

 

 

More turn-out

 

l l I r

5 1O 15 20 25

Percentage of respondents

O

I Methods employed in an attempt to stop or reduce weaving behavior

 

Figure 3-5 Methods employed by farms in an attempt to stop or reduce weaving
behavior. Many respondents used a combination of methods.

 

119

More social contact
More forage/less concentrate

Provison of toys

 

More turn-out

Alteration of crib-biting surfaces

 

Cribbing collars/straps

 

 

 

 

O 20 40 60 80 100
Percentage of respondents

I Methods employed in an attempt to stop or reduce crib-biting behavior

Figure 3-6 Methods employed by farms in an attempt to stop or reduce crib-biting
behavior. Many respondents used a combination of methods.

 

 

Literature Cited

Albright, J .D., Mohammed, H.O., Heleski, C.R., Wickens, C.L., Houpt, K.A., 2009. Crib-
biting in US horses: Breed predispositions and owner perceptions of aetiology.
Equine Vet. J. 41, doi: 10.2746/042516409X372584.

Archer, D.C., Freeman, D.E., Doyle, A.J., Proudman, C.J., Edwards, 3., 2004.
Association between cribbing and entrapment of the small intestine in the epiploic
foramen in horses: 68 cases (1991-2002). J. Am. Vet. Med. Assoc. 224, 562-564.

Archer, D.C., Pinchbeck, G.K., French, N.P., Proudman, C.J., 2008. Risk factors for
epiploic foramen entrapment colic: an international study. Equine Vet. J. 40, 224-
230.

Bachmann, 1., Audigé, L., Stauffacher, M., 2003. Risk factors associated with
behavioural disorders of crib-biting, weaving and box-walking in Swiss horses.
Equine Vet. J. 35(2), 158-163.

 

Clegg, H.A., Buckley, P., Friend, M.A., McGreevy, RD, 2008. The ethological and
physiological characteristics of cribbing and weaving horses. Appl. Anim. Behav.
Sci. 109, 68-76.

Cooper, J .J ., McDonald, L., Mills, D.S., 2000. The effect of increasing visual horizons on
stereotypic weaving: implications for the social housing of stabled horses. Appl.
Anim. Behav. Sci. 69, 67-83.

Dodman, N.H., Normile, J .A., Cottam, N., Guzman, M., Shuster, L., 2005. Prevalence of
compulsive behaviors in formerly feral horses. Intern. J. Appl. Res. Vet. Med. 3(1),
20-24.

Gillham, S.B., Dodman, N.H., Shuster, L., Kream, R., Rand, W., 1994. The effect of diet
on cribbing behavior and plasma B-endorphin in horses. Appl. Anim. Behav. Sci.
41, 147-153.

Hausberger, M., Gautier, E., Miiller, C., Jego, P., 2007. Lower learning abilities in
stereotypic horses. Appl. Anim. Behav. Sci. 107, 299-306.

Hemmings, A., McBride, S.D., Hale, CE, 2007. Perseverative responding and the
aetiology of equine oral stereotypy. Appl. Anim. Behav. Sci. 104, 143-150.

Houpt, K.A., McDonnell, S.M., 1993. Equine stereotypies. Compendium Continuing
Education 15, 1265-1271.

Kiley-Worthington, M., 1983. Stereotypes (sic) in horses. Equine Practice 5(1), 34-40.

121

 

Kusunose, R., 1992. Diurnal pattern of crib-biting in stabled horses. Jpn. J. Equine. Sci.
3(2), 173-176.

Lebelt, D., Zanella, A.J., Unshelm, J ., 1998. Physiological correlates associated with
cribbing behavior in horses: changes in thermal threshold, heart rate, plasma [3-
endorphin and serotonin. Equine Vet. J. Suppl. 27, 21-27.

Lillie, H.C., 2004. Comparison of gastric pH in crib-biting and non crib-biting horses.
Master’s thesis, Auburn University, Auburn.

Lindgren, B.W., 1976. Statistical Theory. 3rd Edition. New York: Macmillan.

Luescher, U.A., McKeown, D.B., Dean, H., 1998. A cross-sectional study on compulsive
behaviour (stable vices) in horses. Equine Vet. J. Suppl. 27, 14-18.

McAfee, L.M., Mills, D.S., Cooper, J .J ., 2002. The use of mirrors for the control of
stereotypic weaving behaviour in the stabled horse. Appl. Anim. Behav. Sci. 78,
159-173.

McBride, S.D., Long, L., 2001. Management of horses showing stereotypic behaviour,
owner perception and the implications for welfare. Vet. Rec. 148, 799-802.

McBride, S.D., Hemmings, A., 2005. Altered mesoaccumbens and nigro-striatal
dopamine physiology is associated with stereotypy development in a non-rodent
species. Behav. Brain. Res. 159, 113-118.

McGreevy, P.D., French, N.P., Nicol, C.J., 1995b. The prevalence of abnormal
behaviours in dressage, eventing and endurance horses in relation to stabling. Vet.
Rec. 137, 36-37.

McGreevy, P., Richardson, J .D., Christine, J .N., Lane, J .G., 1995c. Radiographic and
endoscopic study of horses performing an oral based stereotypy. Equine Vet. J. 27,
92-95.

McGreevy, P., Nicol, C.J., Cripps, R, Green, L., French, N., 1995a. Management factors
associated with stereotypic and redirected behaviour in the thoroughbred horse.
Equine Vet. J. 27, 86-91.

Michigan Equine Survey. 2007. United States Department of Agriculture. National
Agricultural Statistics Service. Available from:
http://www.nass.usda.gov/Statistics_by_State/1\/Iichigan/Publications/Michigan_Rot
ational_Surveys/equine07/equine.pdf
[Accessed 2009 February 8].

122

 

 

Mills, D.S., 2002. Recent advances in the treatment of equine stereotypic behavior.
[Serial online] Animal Behavior Cognition and Welfare Group. University of
Lincoln. Available from:
http://research.vet.upenn.edu/HavemeyerEquineBehaviorLabHomePage/Reference
LibraryHavemeyerEquineBehaviorLab/HavemeyerWorkshops/HorseBehaviorand
Welfare 1 3 1 6June2002/HorseBehaviorandWelfare2/RecentAdvancesintheTreatment
oquuineStereot/tabid/ 3 l 30/Default.aspx
[Accessed 2009 February 8].

Mills, D.S., Macleod, C.A., 2002. The response of crib-biting and windsucking in horses
to dietary supplementation with an antacid mixture. lppologia 13, 33-41.

Nagy, K., Schrott, A., Kabai, P., 2008. Possible influence of neighbors on stereotypic
behavior in horses. Appl. Anim. Behav. Sci. 111, 321-328.

Nicol, C.J., 1999. Stereotypies and their relation to management. In: Harris, P.A.,
Gomarsall, G.M., Davidson, H.P.B., Green, R.E. (Eds.), Proceedings of the BEVA
Specialist Days on Behaviour and Nutrition. Newmarket, UK, Equine Vet. J ., 11-
14.

Nicol, C.J., Davidson, H.P.D., Harris, P.A., Waters, A.J., Wilson, A.D., 2002. Study of
crib-biting and gastric inflammation and ulceration in young horses. Vet. Rec. 151,
658-662.

Ninomiya, S., Sato, S., Sugawara, K., 2007. Weaving in stabled horses and its
relationship to other behavioural traits. Appl. Anim. Behav. Sci. 106, 134-143.

Parker, M., Redhead, E.S., Goodwin, D., McBride, SD, 2008. Impaired instrumental
choice in crib-biting horses (Equus caballus). Behav. Brain. Res. 191, 137-140.

Redbo, 1., Redbo-Torstensson, P., tidberg, F.O., Hedendahl, A., Holm, J ., 1998. Factors
affecting behavioural disturbances in race-horses. Anim. Sci. 66, 475-481.

USDA. 2006. Equine 2005, Part 1: Baseline Reference of Equine Health and
Management. USDAzAPHIS:VS, CEAH. Fort Collins, CO #N451-1006, p. 15.

Vecchiotti, G.G., Galanti, R., 1986. Evidence of heredity of cribbing, weaving and stall-
walking in Thoroughbred horses. Livest. Prod. Sci. 14, 91-95.

Vate-U-Lan, P. 2006. An invitation method enhance intemet-based survey response rates.
Special Issue of the Journal of the Computer, Internet and Management. 14, 34.1-
34.6. Available from: http://www.ijcim.th.org/v14nSP1/pdf/p34.1-6-fin—07-upd-
ver.pdf [Accessed 2009 March 9].

 

 

 

Waters, A.J., Nicol, C.J., French, NP, 2002. Factors influencing the development of
stereotypic and redirected behaviours in young horses: findings of a four year
prospective epidemiological study. Equine Vet. J. 34(6), 572-579.

124

 

 

CHAPTER 4

SUMMARY AND CONCLUSIONS

The overall goal of this dissertation was to provide further insight into the factors
associated with the performance of specific stereotypic behaviors in horses. The aim of
the first study was to investigate the relationship between crib-biting and the integrity and
function of the gastric mucosa in mature horses. We hypothesized that horses exhibiting
crib-biting would have a higher degree of gastric mucosal damage and greater serum
gastrin response to concentrate feeding compared to non crib-biting horses. The number
and severity of gastric ulcers did not differ between crib-biting and non crib-biting
horses. However, serum gastrin response to concentrate feeding was greater in crib-biting
compared to non crib-biting horses. The results of this study suggest that gastric mucosal
damage is not associated with established crib-biting in mature horses maintained on
pasture. Owners of crib-biting horses who provide their animals with adequate foraging
opportunities should not suspect gastric ulcers as being the primary cause of the behavior.
However, the increased gastrin response to concentrate feeding in crib-biting horses is of
interest as this finding may indicate altered gastrointestinal function in crib-biting horses
that could result in a more acidic gastric environment following consumption of
concentrated meals. An increase in gastric acidity may be particularly problematic for
horses receiving low forage diets in light of the evidence that increased roughage and
provision of an antacid diet are effective in reducing crib-biting behavior even in
established crib-biters. Longitudinal studies of the gastrointestinal environment including

measurement of gastrointestinal hormones in a large population of horses exposed to

125

 

 

various management practices are needed to further elucidate the role of gastrointestinal
irritation in crib-biting behavior.

The aim of the second study was to investigate weaving and crib-biting behavior
in the Michigan horse population. It was hypothesized that Michigan horse owners would
express concern regarding weaving and crib-biting behavior and that the majority of
owners attempt to stop or reduce the performance of these behaviors. Additionally, the
author hypothesized that certain housing and feeding strategies, for example those which
limit a horse’s time out of the stable, limit contact with other horses, or limit foraging
opportunities, would result in an increased probability of horses exhibiting weaving
and/or crib-biting behavior. Some of the specific predictions were that increased tum-out
would reduce the probability of a horse being a weaver or crib-biter, and that the
probability of being a weaver or crib-biter would be increased in horses with no visual or
tactile contact with other horses. Furthermore, that those horses fed larger amounts of
concentrate would be more likely to weave or crib-bite.

This study was the first to examine stereotypic behavior in the Michigan horse
population and also among the first to utilize an online questionnaire. The percentages of
horses exhibiting weaving and crib-biting behavior were 2.7% and 5.2%, respectively.
These results were similar to prevalences reported previously in other countries, and
more recently reported for US. horses. Michigan horse owners and farm managers
expressed concern about weaving and crib-biting behavior, but perceptions regarding the
impact of these behaviors on horse performance/learning, horse health, and monetary
value of the animal were different between owners of non-stereotypic horses and those

respondents currently owning/managing a weaver or crib-biter. Overall, respondents

126

 

presently owning/managing weaving and crib-biting horses were less concerned about the
behaviors having a negative impact on learning, health, or monetary value. Thus,
perceptions about stereotypic behavior within the equine community at large may not
coincide with those held by individuals having first-hand experience with weaving and
crib-biting horses. A relatively low percentage of respondents with weaving and crib-
biting horses indicated that they are concerned the behavior hinders learning or
performance, thus it would not appear that these behaviors substantially interfere with
training or with the overall usefulness of the horse. However, additional inquiry into the
experiences and observations of horse owners may provide valuable insight into how
weavers and crib-biters respond to various training and handling procedures compared to
their non-stereotypic counterparts.

Information regarding the sex and breed of non-stereotypic horses was not
collected in the online questionnaire, thus statistical analysis of associations between
horse characteristics and weaving and crib-biting behavior were not conducted. However,
in comparing weaving and crib-biting horse demographics with sex and breed
distribution data reported in the 2007 Michigan Equine Survey, it appeared that there was
a relationship between horse characteristics and the performance of stereotypic behavior.
In general, based on the sample data, a gender-based factor appeared to be involved.
Mares appeared to be more at risk for locomotor stereotypy, e. g. weaving, while geldings
appeared to be at greater risk for oral stereotypy, e. g. crib-biting. Breed of horse also
appeared to be an important factor. Thoroughbreds and Quarter Horses appeared to be at
higher risk for stereotypic behavior in general compared to other breeds. It is not known

whether these apparent differences reflect a genetic predisposition to stereotypic behavior

127

 

 

or a gender by management or breed by management interaction of some kind, and
additional experimental and epidemiological research is needed in order to validate and
further elucidate these potential relationships.

A negative relationship between weaving and crib-biting and hours of daily tum-
out was identified in this study. Specifically, a one hour increase in tum-out was
associated with a decrease in the probability of a horse being a weaver or crib-biter. An
association was also identified between crib-biting behavior and extent of social contact.

Horses without social contact or those allowed visual contact only were at an increased

 

risk for crib-biting compared to horses being turned out with other horses.

In contrast to previous findings, larger daily amounts of concentrate were
associated with a decrease in the probability of a horse being a weaver or crib-biter. Data
regarding the amount of concentrate fed to horses was not collected on a continuous
scale. Limiting owners to the selection of only a few categories may have hindered our
ability to detect a similar association between this variable and stereotypic behavior
found in previous studies. The type of concentrate offered also may affect the association
between grain feeding and stereotypic behavior. Many of the Michigan horse owners and
barn managers reported feeding a variety of ingredients and the protein, carbohydrate and
fiber content of the diets was not queried. Therefore, in this study, it is unknown how the
type of concentrate, and perhaps more specifically, how the nutrient composition of the
diet may influence the risk of stereotypic behavior. More specific questions about the diet
of horses should be included in epidemiologic research studies in an effort to further
elucidate the role of concentrate feeding in the development and continued performance

of stereotypic behavior.

128

 

163

be:

 

 

The majority of respondents believed that environmental variables were largely
responsible for the performance of stereotypic behavior, indicating that Michigan owners
and farm managers are aware that farm management practices have a direct impact on the
behavior of the horses under their care. Methods reported by owners/barn managers as
being effective in reducing weaving and crib-biting behavior included increased tum-out
and social contact and/or increased amounts of forage. Management strategies that
provide horses with increased opportunities to engage in natural behaviors should be
employed as the first means of moderating stereotypic behavior, as they are aimed at
removing the causal factors of the stereotypic behavior rather than at physical prevention
that may result in reduced welfare. These practices would be especially crucial to
implement on farms raising young horses in order to prevent the development of
stereotypic behavior or for those owners with horses just starting to engage in weaving or
crib-biting behavior so as to prevent the behaviors from becoming established. Many
owners in Michigan look to veterinarians for information about stereotypic behavior,
highlighting the importance of providing veterinary professionals with current, science-
based information about performance of stereotypic behavior in horses.

Acquiring a better understanding of the interactions between equine physiology
and management factors in the development of stereotypic behavior, and the impact of
these behaviors on the health of affected horses, will be essential in making sound
recommendations to horse owners and farm managers on how best to care for their
horses. Thus, continued research into the etiology of weaving and crib-biting behavior is

warranted.

129

 

 

Equ

The
crit
pre
cau
onl
an:
W

Th
ab
lihi

Appendix A

Michigan Horse Behavior Questionnaire in Microsoft Word Format
Equine Behaviors: A Study of the Michigan Horse Population

The purpose of this research is to learn more about certain behaviors, weaving and
cribbing, that may occur in the Michigan horse population. Little is known about the
prevalence of these behaviors and there is still much to learn about the underlying
causes of these behaviors and about how best to treat and/or prevent them. The
only way to do this is to gather information from horse owners, managers,
and trainers whose animals do not exhibit these behaviors as well as those
whose animals do exhibit weaving and cribbing.

The results of this research will be used to educate owners, managers, and trainers
about these behaviors and provide them with information to assist them in managing
these stereotypic behaviors in the Michigan horse population.

This research is completely voluntary and anonymous. It should take 15 minutes or
less to complete depending on your answers. You can skip any question that you do
not want to answer. You can also end your participation at any time.

If you have any questions about this research, you need clarification or assistance in
answering the questions, or would like a paper copy of the questionnaire, please
contact Carissa Wickens at 517.353.2260 or by email at W. If you
have technical difficulties with this survey, please contact Karen Clark at the Office
for Survey Research at MSU at 517.353.1764 or by email at glarkk@msu.egu .

 

I indicate my voluntary consent to participate in this research study.

I: Yes 1:] No

Part 1: Stable and Management Information

1.Which of the following best describes the type of stable or farm you own, manage,
or keep your horse(s) at?

E] Breeding

[:1 Training

E1 Boarding and/or riding lessons
E] Race (e.g., flat, harness)

El Performance (e.g., dressage, eventing, endurance, competitive trail riding,
western pleasure, English pleasure)

130

[3 Pl:
3 01

2.1

the
ma

Kl

Lil]

b

[.3

C7

[‘1

[I1

Cl Pleasure, "backyard" (leisure, trail riding, etc)
D Other

 

2. What is the total number of horses that you own or manage?

horses

 

3. For the majority of the year, that is, for at least nine months, is your horse/are
the horses primarily housed . . . If horses are primarily housed on pasture, you
may skip questions 4-7.

[3 In stalls
[:1 In a pasture
Cl Equal between stalls and pasture

4. What type of bedding do you use most often?

D Wood shavings
Ci Sawdust

Cl Straw

El Paper

El Pellet

D Other

If you selected "other", please specify:

5. How many hours of turn-out is given on a typical day?

hours per day (please answer in whole numbers)

Sa. Does the amount of turn-out vary from season to season, for example, do you do
this more in the summer than the winter?

[:1 Yes D No

b. Please describe the changes in turn-out from season to season.

131

 

 

11

[J

l l

6. In general, what type of turn out is provided most often?

El Pasture

El Small grass lot or paddock
El Dry lot

El Indoor arena

7. On average, how many hours per day is your horse/are the horses allowed to
graze, that is, given access to grass during the grazing season?

hours per day (please answer in whole numbers)

 

8. How often does your horse/do the horses receive controlled exercise (riding,
driving, lunging)?

[1 Less than once a week

13 1 -3 times per week

El 4 - 6 times per week

[:1 Daily

9. Which of the following statements best describes your horse's/the horses' social
contact?

El Horse or horses are turned out with another horse or group of horses

[:1 Allowed visual contact with other horses (e.g., over stall door, through windows in
stall)

[:1 Allowed visual and tactile contact with other horses (e.g., horses can touch over
fence line or stall )

El Horses do not have social contact with others (e.g., horses are separated by stalls
with solid walls, or are housed separately with no visual or tactile contact)

10. How many times per day is your horse/are the horses fed concentrate (grain)?

times per day (please answer in whole numbers)

11. Does the amount of grain fed vary from season to season, for example, do you
feed your horse(s) more grain in the winter than the summer?

El Yes [:1 No

12. Please describe the feeding changes you make from season to season.

132

 

 

13. What type of concentrate or grain do you feed? (Please select all that apply)

ClOats
UCorn
[:JSoybean meal

1:]Corn and oat mixture
DCommercial sweet feed
DCommercial pelleted feed
[:lOther

 

14. On a typical day, how much concentrate does your horse/a horse at your barn or
stable receive?

DLess than one pound

|:11 - 2 pounds
1:13 - 4 pounds
1:15 - 6 pounds

[:17 or more pounds

15. On a typical day, about how many times a day is your horse/are the horses given
access to forage, such as hay?

[:lZero

1:10ne

[:lTwo

I:|Three

DFour

CIAd libitum (free access)

16. What type of forage is usually fed?

DGrass hay

DLegume hay (e.g, alfalfa, clover)
[Mix of grass and legume hay
[:IHaylage

DHay cubes

[:IPasture

DOther

If you selected "other", please specify:

17. Please describe how often and how much forage is fed to your horse/horses on a
typical day.

133

 

Part II: Equine Behaviors

This next set of questions focuses on two stereotypic behaviors, weaving
and cribbing.

Weaving is defined as swaying back and forth in the stall or paddock and
shifting weight between left and right front feet. The neck and head may
also be heavily engaged in the swaying motion.

Cribbing or crib-biting is the grasping of a fixed object with the incisor
teeth, arching the neck, pulling back, and emitting a grunting sound.
Cribbing is not wood chewing. When a horse exhibits cribbing or crib-biting,
wood is not eaten and wood does not disappear.

1. How familiar are you with weaving behavior in general?
DVery Familiar

DSomewhat Familiar

DNot Familiar At All

2. Do you think that weaving is . . .

DPredominantly a genetic disorder
1:1Predominately environmental (horse's surroundings and management)
L—JCombination of genetics and environment v

3. What do you believe to be the underlying cause or causes of why horses develop
weaving behavior?

4. Please indicate to what extent you agree or disagree with each of the following
statements about weaving behavior.

Strongly
Agree

Strongly Don't

Agree Disagree Disagree know

Weaving hinders a horse’s ability
to learn, train, or perform
effectively at tasks or during
events.

134

 

 

We
iml

Ha
thl

r-nr’ll’fi

Weaving behavior negatively
impacts the health‘ of a horse

Having‘ weaving behavior reduces v!"
the monetary value of the horse. . ,

5. How familiar are you with cribbing behavior in general?

DVery Familiar
1:]Somewhat Familiar
DNot Familiar At All

6. Do you think that cribbing is . . .

1:1Predominantly a genetic disorder
[ZIPredominately environmental (horse's surroundings and management)
UCombination of genetics and environment

7. What do you believe to be the underlying cause or causes of why horses develop
cribbing behavior?

8. Please indicate to what extent you agree or disagree with each of the following
statements about cribbing behaviors.

Strongly . -'~':. Strongly Doni'tz '
Agree Agree Disagree Disagree know

Cribbing behavior negatively
impacts the health of a horse.

Having cribbing behavior reduces
the monetary value of the horse.

Cribbing hinders a horse’s ability
to learn, train, or perform
effectively at tasks or during
events.

135

cril

al

9. Does your horse(s) or any of the horses you are familiar with exhibit weaving,
cribbing, or both of these behaviors?

Weaving is defined as swaying back and forth in the stall or paddock,
shifting weight between left and right front feet. The neck and head may
also be heavily engaged in the swaying motion.

Cribbing or crib-biting is the grasping of a fixed object with the incisor teeth,

arching the neck, pulling back, and emitting a grunting sound. Cribbing is not

wood chewing. When a horse exhibits cribbing or crib-biting, wood is not eaten
and wood does not disappear.

DWeaving only

DCribbing Only

DBoth weaving and cribbing (e.g, either one horse that does both behaviors or
horse(s) that cribs and another that weaves)

|:|No, neither behavior

 

If you selected “No, neither behavior”, you may skip to Part III: Source of
Information.

This section focuses on weaving behavior.
W1. How many of your horses exhibit weaving behavior?

horses

In order to better understand weaving behavior, we would like to gather
some background information on your horse or each of the horses (up to
five) that you own or manage that exhibit this behavior.

lst Horse or Only Horse (If reporting on multiple horses, please answer questions
1-10 for each horse, up to five horses, on a separate sheet of paper. Be sure to label
as horse 2, 3, etc.)

 

1. What is your relationship to the horse (e.g. owner/co-owner, leaser, trainer)?

2. What is the breed of the horse?

136

 

3 [Fl—E

3. Is the horse a. . .
DGelding

DMare

1:15tallion

4. What is the age of the horse?

5. What is the primary discipline this horse is used for? (e.g. Dressage, Racing,
Leisure/trail riding)

5a. On average, in how many events per year does this horse participate (e.g,
shows, competitions)?

Number of events

6. Are you familiar with this horse's first year of life? If you selected “No" you
may skip questions 7-10.

DYes
DNo

7. Which, if any, of the following relatives of this horse exhibit weaving? (Please
select all that apply)

Yes No Not sure
Asibling
Sire
Dam

8. At what age was the horse weaned?

DWeaned at less than 4 months of age
|:]Between 4 and 6 months of age
1:]Weaned at more than 6 months of age
CIDon't know

137

 

9. Was the horse. . .

EIAbruptly weaned, foal moved to own stall

DAbruptly weaned, the foal left with companions (other foals), mare removed
DGradually weaned, time mare and foal spent apart was gradually increased
DDon’t know

EIOther

10. To the best of your knowledge, did the horse experience any of the following?

Don't
Yes No . 3..
fiffiow
A major illness as a foal? “i“
1' , v . \ .. 'I 1
Had an injury or experienceda traumatic eventas a foal? .3 fly .
Was exposed to regular human contact (e.g., imprinted to a
human handler ., " ‘
at birth, groomed and worked with on a regular basis)? A" has: ‘3
.- ' ',< . “c. ' V ' s

U»

i ~ . ‘ "2 » lg.
Received a high concentrate diet/prepped for showing or sale?

0 ','0.t

Received mainly pasture or lots of turn out? '»

If you have more than one horse that exhibits weaving behavior, when
answering the following questions, please think about the horse that
displays this behavior most often or the horse that you are most familiar
with who displays this behavior.

Please indicate which horse (the lst, 2nd, 3rd, etc) that you will be referring to for
the following set of questions.

138

W2. Did the horse start to weave after another horse that weaves arrived at your
farm or stable?

DYes
DNo

W3. Have any horses brought into direct contact with the weaving horse started to
weave?

DYes
EINo

W4. On average, how many hours per day do you spend directly watching or
interacting with the horse?

[:lLess than one hour

1:11 - 2 hours
1:13 - 4 hours
[:15 - 6 hours
[:17 - 8 hours
C19 - 10 hours

[More than 11 hours

W5. When you are observing the horse, how often does the horse exhibit weaving
behavior when the horse is not being handled or restrained?

DAII of most of the time
USome of the time
[:lRarely

W6. When does the horse begin weaving or what appears to trigger the weaving
behavior in the horse? (Please select a// that apply)

DBefore feeding [:IAfter riding or DAt a show or event
[:lAfter feeding exercise [When the horse is
1:]Before riding or DBefore turnout separated

exercise DDuring stall cleaning 1:]Other

If you selected "other", please specify:

139

 

W7. Have you ever tried to stop the weaving behavior?

L—JYes
EINo

W8. Which of the following techniques have you tried? (Please select all that apply)

DAnti-weave bars

[:lElectric wire in stable or pasture
L—JSurgery (e.g., myectomy, neurectomy)
1:]Other

If you selected "other", please specify:

W9. Which of the following management changes have you made? (Please select all
that apply)

[:lIncreased turn out or grazing opportunities
Dlncreased exercise

DIncreased social contact

DProvided toys

DIncreased forage in diet and decreased concentrate
[:lOther

If you selected "other”, please specify:

W10. Have any of these techniques or management changes been successful in
stopping the weaving behavior?

DYes
[:lNo
DNot Sure

W11. What techniques or management changes have been successful in stopping the
weaving behavior?

W12. Have any of these techniques or management changes been successful in
reducing the weaving behavior?

[:lYes
[No
[Not Sure

140

 

W13. What techniques or management changes have been successful in reducing
weaving behavior?

W14. Have your efforts to prevent or reduce weaving behavior led to the horse
performing other noticeable behavior(s)?

1:]Yes
[No
[Not Sure

W15. What types of behaviors have you noticed?

W16. On average, how much money do you spend each year on training aids,
devices, or informational resources related to weaving behavior? These costs could
be for items such as anti-weave bars, horse toys, etc.

$.—

This section focuses on cribbing behavior.

C1. How many of your horses exhibit cribbing behavior?

horses

In order to better understand cribbing behavior, we would like to gather
some background information on your horse or each of the horses (up to
five) that you own or manage that exhibit this behavior.

Ist Horse or Only Horse (If reporting on multiple horses, please answer questions
1-10 for each horse, up to five horses, on a separate sheet of paper. Be sure to label

as horse 2, 3, etc.)

 

1. What is your relationship to the horse (e.g. owner/co-owner, leasor, trainer)?

2. What is the breed of the horse?

141

a

1:.

r -.-s__0

 

3. Is the horse a . . .
[:lGelding

EIMare

DStallion

4. What is the age of the horse?

5. What is the primary discipline this horse is used for? (e.g. Dressage, Racing,
Leisure/trail riding)

5a. On average, in how many events per year does this horse participate (e.g,
shows, competitions)?
Number of events

6. Are you familiar with this horse's first year of life? If you selected “No" you
may skip questions 7-10.

DYes
[No

7. Which, if any, of the following relatives of this horse exhibit cribbing? (Please
select all that apply)

Yes No Not sure
A sibling
Sire

Dam

8. At what age was the horse weaned?

[:1Weaned at less than 4 months of age
[Between 4 and 6 months of age
1:]Weaned at more than 6 months of age
DDon't know

142

 

9. Was the horse. . .

EIAbruptly weaned, foal moved to own stall

1:]Abruptly weaned, the foal left with companions (other foals), mare removed
DGradually weaned, time mare and foal spent apart was gradually increased
I:1Don’t know

1:]Other

10. To the best of your knowledge, did the horse experience any of the following?

063??“

Yes No
r * . know

A major illness as a foal? 7 L ,g . TIE
.-' . “1'12 . 1 “45‘s '

Had an injury or experienceda traumatic eii'ent as a foal? z
». fififi -“ '1 :- aT'iaa . 1%

Was exposed to regular human contact (e. g., imprinted to a

human handler r 0,,

at b,lrth groomed and worked with on a regular basis)? ’ Wee .4395;

Received a high concentrate diet/prepped for showing «fate?

.19?

Received mainly pasture or lots' of turn out? ‘ * 1 “P...

If you have more than one horse that exhibits cribbing behavior, when
answering the following questions, please think about the horse that
displays this behavior most often or the horse that you are most familiar
with who displays this behavior.

Please indicate which horse (the lst, 2nd, 3rd, etc) that you will be referring to for
the following set of questions.

143

C2. Did the horse start to crib after another horse that cribs arrived at your farm or
stable?

DYes
DNo

C3. Have any horses brought into direct contact with the cribbing horse started to
crib?

DYes
DNo

C4. On average, how many hours per day do you spend directly watching or
interacting with the horse?

DLess than one hour

D1 - 2 hours
D3 - 4 hours
D5 - 6 hours
D7 - 8 hours
D9 - 10 hours

DMore than 11 hours

C5. When you are observing the horse, how often does the horse exhibit cribbing
behavior when the horse is not wearing a device designed to stop or prevent the
behavior?

DAll of most of the time
DSome of the time
DRarely

C6. When does the horse begin cribbing or what appears to trigger the cribbing
behavior in the horse? (Please select all that apply)

DBefore feeding DAfter riding or DAt a show or event
DAfter feeding exercise DWhen the horse is
DBefore riding or DBefore turnout separated

exercise DDuring stall cleaning DOther

If you selected "other", please specify:

144

C7. Have you ever tried to stop the cribbing behavior?

DYes
DNo

C8. Which of the following techniques have you tried? (Please select all that apply)

DCribbing collars/straps

DElectric wire in stable or pasture

DSurgery (e.g., myectomy, neurectomy)

DRemoval of cribbing surfaces or use of distasteful paints/wood coatings
DOther

If you selected "other", please specify:

C9. Which of the following management changes have you made? (Please select all
that apply)

DIncreased turn out or grazing opportunities
DIncreased exercise

DIncreased social contact

DProvided toys

DIncreased forage in diet and decreased concentrate
DOther

If you selected ”other", please specify:

C10. Have any of these techniques or management changes been successful in
stopping the cribbing behavior?

DYes
DNo
DNot Sure

C11. What techniques or management changes have been successful in stopping the
cribbing behavior?

145

C12. Have any of these techniques or management changes been successful in
reducing the cribbing behavior?

DYes
DNo
DNot Sure

C13. What techniques or management changes have been successful in reducing
cribbing behavior?

C14. Have your efforts to prevent or reduce cribbing behavior led to the horse
performing other noticeable behavior(s)?

DYes
DNo
DNot Sure

C15. What types of behaviors have you noticed?

C16. On average, how much money do you spend each year on training aids,
devices, or informational resources related to cribbing behavior? These costs could
be for items such as cribbing collars, removing cribbing surfaces, horse toys, etc.

5

Part III: Source of Information

1. From which of the following do you get information about horse behavior in
general?

DComputer via Worldwide Web sites and the internet

DHorse magazines (e.g. Equus, Horse Illustrated, etc.)

DHorse newsletters (e.g. horse association and/or horse extension newsletters)
DOral presentations /programs (e.g. workshops, seminars, conferences)
DRadio or television programs

DVeterinarian

DOther horse owners/family/friends

DOther

146

If you selected "other", please specify:

2. From which of the following do you get information about stereotypic horse
behaviors such as weaving and cribbing?

DComputer via Worldwide Web sites and the internet

DHorse magazines (e.g. Equus, Horse Illustrated, etc.)

DHorse newsletters (e.g. horse association and/or horse extension newsletters)
DOral presentations /programs (e.g. workshops, seminars, conferences)
DRadio or television programs

DVeterinarian

DOther horse owners/family/friends
DOther

If you selected "other", please specify:

3. Please feel free to make additional comments about these or other stereotypic
behaviors.

4. If you would like results of this study, please provide the following information:
Name:

Stable or Farm Name:

Address:

City:

State:

Zip Code:

5. If you would be interested in participating in further research studies involving
stereotypic horse behavior (such as additional surveys or on-site visits to observe
your horse(s), please check the box below.

[:1 Yes, I am interested in participating in further research studies involving stereotypic horse
behavior.

Thank you for your assistance!

147

   

M'llllllllllljljlfillljijlllllljlllllllEs