|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 AN ANALYSIS OF HUMAN INTERACTION AS ENVIR ONMENTAL ENRICHMENT FOR CAPTIVE WOLVES A ND WOLF DOG HYBRIDS By LINDSAY RENEE MEHRKAM A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2012
2 2012 Lindsay Renee Mehrkam
3 To my Mom and Dad, for their enduring support, to the sanctuaries, for their enduring trust, and to Kobe, for his enduring companion ship
4 ACKNOWLEDGMENTS I wish to thank the co founders of Big Oak Wolf Sanctuary, John and Debra Knight, for their continuous support, research hospitality, and access to their residents for this study. I am also grateful to the many volunteers at the s anctuary for their time, assistance, technical support, and suggestions dur ing data collection procedures. Finally, I thank my committee, Dr. Clive Wynne, Dr. Jane Brockmann, and Dr. Jesse Dallery, for helpful comments and suggestions on a previous version of this thesis.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 6 LIST OF FIGURES ................................ ................................ ................................ .......... 7 ABSTRACT ................................ ................................ ................................ ..................... 8 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 10 Human Interaction as Env ironmental Enrichment for Captive Animals ................... 10 Comparing Training Versus Human Interaction Alone ................................ ..... 11 Limitations of Current Researc h ................................ ................................ ....... 12 Sanctuary Animals: An Ideal Population for Human Interaction Research ............. 12 Wolf Dog Hybrids as Scientific Subjects ................................ ........................... 13 Study Aims ................................ ................................ ................................ ....... 14 2 METHODS ................................ ................................ ................................ .............. 15 Subjects ................................ ................................ ................................ .................. 15 Procedure ................................ ................................ ................................ ............... 17 Ethogram ................................ ................................ ................................ ................ 19 Video and Statistical Analyses ................................ ................................ ................ 20 3 RESULTS ................................ ................................ ................................ ............... 25 Species Typical Affiliative Behavior ................................ ................................ ........ 25 Proximity Measures ................................ ................................ ................................ 26 Activity Levels ................................ ................................ ................................ ......... 28 4 D ISCUSSION ................................ ................................ ................................ ......... 34 Is Human Interaction Enriching for Wolves and Wolf Dog Hybrids? ....................... 34 Variables Influencing Enrichment Efficacy ................................ .............................. 37 Understanding Play and Welfare Is Play Spontaneous? ................................ ...... 39 Behavior Analysis as a Tool for Identifying Functional Relationships Between Play and the Environment ................................ ................................ .................... 41 LIST OF REFERENCES ................................ ................................ ............................... 44 BIOGRAPHICAL SKETCH ................................ ................................ ............................ 47
6 LIST OF TABLES Table page 2 1 Demographics of subjects at Big Oak Wolf Sanct uary ................................ ....... 22 2 2 Ethogram of species typical and abnormal behaviors for wolves and wolf dog hybrids ................................ ................................ ................................ ................ 23 3 1 Percent of social play obs erved in human interaction conditions for wolves and wolf dog hybrids ................................ ................................ .......................... 33
7 LIST OF FIGURES Figure page 3 1. Duration of human and conspecific affilia tion across replications for subjects in Pairs 1 & 2. Changes in conditions and replications are represented by single vertical lines and double vertical lines, respectively. ................................ 29 3 2. Duration of human and conspecific affiliation across replications for subjects in Pairs 3 & 4. Changes in conditions and replications are represented by single vertical lines and double vertical lines, respectively. ................................ 30 3 3. Proportions of scans per condition that pair housed conspecifics were observed in close proximity to one another during baseline and human animal interactions (HAI) across replications. ................................ ..................... 31 3 4. Activity levels for all subjects. Series represent mean percentage of baseline and human animal interaction (HAI) conditions across replications that subjects were recorded as active. ................................ ................................ ...... 31 3 5. Activity and pacing levels for Peter in Replication 1. Peter displayed higher rates of inactivity during human interaction sessions; however, much of his activity during baselines was pacing, which decreased when a volunteer wa s present in the enclosure. ................................ ................................ .................... 32 3 6. Duration of cage chewing behavior emitted by Ava during Replication 2. Baseline (5 min) and intervention (10 min) conditions are separated by dashed lines. ................................ ................................ ................................ ...... 32
8 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science AN ANALYSIS OF HUMAN INTERAC TION AS ENVIRONMENTAL ENRICHMENT FOR CAPTIVE WOLVES AND WOLF DOG HYBRIDS By Lindsay Renee Mehrkam May 2012 Chair: Clive D.L. Wynne Major: Psychology The benefits of environmental enrichment for captive mammals are well documented in the scientific liter ature. However, many facilities may have limited resources to implement traditional enrichment strategies. One possible solution is to provide socialized animals with positive interaction sessions with experienced volunteers. Although prior research has qu estioned whether h uman interaction alone may serve as enrichment, this topic has received little empirical attention to date. The present study aimed to evaluate whether human interaction could be enriching for socialized, pair housed wolves and wolf dog h ybrids. Behavioral observations of each subject were conducted in a reversal design to measure species typical affiliation, activity levels, and aberrant behaviors when volunteers were both present and absent. Results demonstrate high levels of human direc ted affiliation and significantly higher levels of conspecific directed affiliation when human interaction was made available to subjects relative to baseline conditions. Significantly higher activity levels were observed for two thirds of replications dur ing human interaction sessions compared to baselines. Individual data across conditions revealed reduced levels of pacing during
9 human interaction sessions for some subjects. One wolf dog hybrid engaged in abnormal cage chewing in some baseline conditions following human interaction sessions. Social play also increased when caregivers were present, suggesting that environment. The potential for human interaction to be est ablished as a scientifically validated, cost effective enrichment strategy for captive animals is discussed, as well as the potential for play as an indicator of good welfare.
10 CHAPTER 1 INTRODUCTION Human Interaction as Environmental Enrichment for Capt ive Animals The benefits of environmental enrichment for zoo and laboratory mammals are widely documented in the empirical literature. Three acknowledged aims of environmental enrichment are to 1) promote opportunities for the occurrence of species typical behaviors, 2) to promote novel sensory stimulation, and 3) to concurrently reduce the occurrence of aberrant behaviors that may be indicative of poor welfare (Shepherdson, 1998). Despite these benefits, enrichment strategies often cost time and resources to design, implement, and maintain. This suggests a need to establish cost efficient enrichment methods that are equally effective in producing desired behavioral changes in captive animals. In addition, the evaluation of enrichment strategies is of great importance when determining if a particular strategy is satisfying these goals. Enrichment practices that incorporate non stressful stimuli and target the dominant sense for the species of interest are likely to result in the greatest benefits for animal w should not always be assumed to be meaningful to a captive individual of that species. In some cases, stimuli that do not occur naturally in the wild but are readily available in the captive environment may be more cost effective in promoting overall welfare (Wells, 2009). Given that captive animals are provisioned with basic survival needs by human caretakers, the potential for human animal interactions to be enriching has been p roposed for captive animals that have positive relationships with humans (Claxton, 2011).
11 Comparing Training Versus Human Interaction Alone Most empirical attention on human animal interactions in this context has focused on the use of positive reinforce ment training (PRT) to promote desirable behaviors while simultaneously providing an opportunity for positive interactions between animals and their caregivers (Laule, Bloomsmith, and Schapiro, 2003; Baker et al., 2010; Claxton, 2011;). Although PRT is cle arly relevant to developing target behaviors that make aversive husbandry techniques obsolete (Laule et al., 2003), such methods generally cannot assess changes in behavioral welfare (e.g. species typical behavior, activity levels) through observations (Ba ker, 2004). An additional limitation of PRT is that human animal interaction during training sessions is often confounded with food reinforcement, thus obscuring the effects of human interaction alone on the behavioral welfare of captive animals. What has perhaps been more seriously overlooked is the provision of free contact interactions with captive animals and their human caregivers. Few studies have examined the benefits of positive human interaction for captive animals, mostly in non human primates. Fo r example, solitary and socially housed laboratory chimpanzees (Waitt et al., 2002; Baker, 2004; Bloomsmith et al., 2004) showed reduced stereotypic behavior; macaques and socially housed common marmosets exhibited increased grooming when positive interact ions with caretakers were available (Baker, 2003; Waitt et al., 2002; Manciocco et al., 2009). In addition, human interaction has been evaluated for shelter dogs (Coppola, Grandin, and Enns, 2006; Normando et al., 2009; Bergamasco et al., 2010). However, the literature concerning the impacts of human interaction alone for captive animals is limited by restricted species generality, inconsistencies in the
12 literature regarding the efficacy of such interactions, even within related species, and a lack a syste matic evaluation of individual data. For example, Wood (1998) observed lower rates of positive species typical behaviors in zoo housed chimpanzees in the presence of larger crowds of visitors. Male rhesus macaques showed higher rates of abnormal behaviors during human interaction sessions (Maloney et al., 2007). Limitations of Current Research In addition, none of the prior studies were designed in a manner that would permit a direct analysis of cumulative behavior change, as a result of alternating condit ions of caregiver presence and absence. Baker (2004) only examined changes of the restrictions on contact with subjects, her study could not examine the effect of free cont act interactions. This is problematic because important changes in behaviors of interest may occur not only when enrichment is available, but following its removal as well. Other limitations have been staff availability to provide interactions because of l arge group housings (Manciocco et al. 2009). Furthermore, replications were generally not conducted in these studies, which would allow for the assessment and reliability of enrichment effectiveness. Finally, because behavioral welfare is a property of ind ividual animals, group designs and aggregate data are of only limited usefulness when evaluating enrichment efficacy. Sanctuar y Animals: An Ideal Population for Human Interaction Research One population that may be ideal to implement and evaluate human in teraction is that of privately run sanctuaries for exotic animals. Many exotic animals that arrive at these sanctuaries have a unique history of extended contact with humans, as a result of being purchased by breeders from a young age, and subsequently han d reared as
13 companion animals in pet homes among humans. Yet this population of exotic animals still has direct application to enrichment evaluation for non domesticated zoo animals. In addition, many such facilities are non profit organizations that are run entirely on volunteer assistance and private donations, often making traditional environmental enrichment strategies both physically and financially difficult to implement. In this case, human interaction may serve as a financially feasible and readily accessible strategy to promote welfare of captive animals. Wolf Dog Hybrids as Scientific Subjects A large proportion of the animals that exist in such facilities are wolves and wolf dog hybrids. A wolf dog hybrid is a cross between a wolf ( Canis lupus ) a nd a dog ( Canis familiaris ) or between animals where one or both parents contain recent wolf genes (Cusdin, 2000). At present, there are estimated to be anywhere from 300,000 to 1.5 million wolf dog hybrids in captivity in the United States (Busch, 2007). In many cases, wolves and wolf dog hybrids are relinquished voluntarily by their owners or confiscated by wildlife officials from situations in which the animals were illegally obtained and/or possessed or improperly cared for (e.g. in cases of abuse or ne glect). At present, however, no peer reviewed publication presents data on the behavioral management of captive wolf dog hybrids. Though objective information on this population is greatly needed, there are numerous limitations to the scientific literatur e on wolf dog hybrids. First, the identification of wolf dog hybrids is often a subject of primary concern and can be difficult when the hybridizing species are closely related and do not possess diagnostic genetic markers (Vil et al. 2003). Second, legal ownership of wolf dog hybrids varies by state, and even county, in the United States, contributing to substantial controversy. In addition, because wolf dog hybrids are neither
14 domestic nor exotic wildlife, studies proposing them as subjects of scientific inquiry are hard pressed to find appropriate sources of funding to carry out objective behavioral evaluations. Study Aim s The study of wolf dog hybrids in particular has applications to both the management of exotic and domesticated species. Wolf dog hyb rids are often provided life long care in sanctuaries, and in many cases, are not adopted out into the community. Thus, the need for effective environmental enrichment is central to the welfare of this population. Given the limitations of implementing trad itional enrichment practices in these settings, human interaction seems an ideal strategy to evaluate with wolves and wolf dog hybrids a population that has not received empirical attention in this area of inquiry. Therefore, the objective of this study was to assess if human interaction could serve as enrichment for pair housed wolves and wolf dog hybrids at a sanctuary. It was predicted that contact with familiar caregivers would increase the occurrence of species typical affiliative behavior, promote n ovel sensory stimulation (and thus, increase activity levels), and decrease the occurrence of any aberrant behaviors that were observed when human interaction was not provided.
15 CHAPTER 2 METHODS Subjects The subjects were two socialized wolves and six wo lf dog hybrids ( Table 2 1). Wolf wolf content was determined by recording physical similarities to a wolf phenotype (Cusdin, 2000). All wolf dog hybrids in this study were i dentified as such by state officials prior to the onset of the study (Knight, 2011, personal communication), and were categorized either as high content wolf dogs (80 99% wolf phenotype), mid content wolf dogs (40 79% wolf phenotype), and low content wolf dogs (5 39% wolf phenotype) (Florida Lupine Association, 2011, personal communication). All subjects were housed at Big Oak Wolf Sanctuary (Green Cove Springs, Florida, United States) at the time of this study. The sanctuary is a privately owned, non prof it organization that provides long term residence and care to neglected and confiscated wolves and wolf dog hybrids. The facility was not open to the general public, but all animals were accustomed to interaction sessions several times per week with sanct uary volunteers. All subjects were housed in similar enclosures (ranging from 2,286 to 3,048 m 2 ) surrounded by chain link fencing (approximately 3.1 m in height) around the perimeter. All enclosures included a 2.6 kL pool, a wooden platform (3.1 m in heigh t and 3.7 m by 3.7 m) above the pool, and an underground den (approximately 4.9 m by 2.4 m). All subjects were adults that had been housed at the sanctuary for a minimum of 2 years at the time of the study. Each was housed with a conspecific of the opposit e sex (e.g., wolf dog hybrids were paired with another wolf dog hybrid and wolves were paired
16 with wolves). All pairs were housed together for a minimum of 1 continuous year prior to the study. All subjects remained in their original pairs throughout the s tudy. One pair o f wolf dogs Peter and Sara was relocated to a different enclosure between the first and second replications for management reasons. Subjects were chosen for this study based on how well socialized they were to familiar people and their ability to interact safely with a single person in the enclosure. All subjects were spayed or neutered prior to the onset of the study. Sampson and Spirit were sibling hand reared wolves purchased from a private breeder in Texas. Both had resided together at two other sanctuaries for approximately 5 months when they were under 1 year of age. They had resided at this sanctuary for 4 years prior to this study. Sabbath and Ava were sibling, reportedly low content, wolf dog hybrids previously kept privately b y separate owners. Reasons for relinquishment are unknown. Both Sabbath and Ava had been at the sanctuary for approximately 2 years. Peace and Lea were reportedly high content and low content wolf dog hybrids respectively, purchased from breeders for priva te ownership. Lea was reportedly multiple owners before being relinquished. Peace and Lea had resided at the sanctuary for 2 and 3 years, respectively. Peter and Sara were rep ortedly low content wolf dog hybrids. Sara was purchased from a breeder at 12 weeks of age and brought to the sanctuary 3 weeks later, after a neighbor reported physical abuse to Sara and another wolf dog hybrid. Peter was adopted from an animal shelter, b ut subsequently surrendered when
17 authorities informed owners that he was a wolf dog hybrid. Peter had resided at the sanctuary for 3 years. Procedure An A B A B A reversal design was used to evaluate the effects of unstructured human interaction for pair h oused wolves and wolf dog hybrids. A reversal design entails repeatedly alternating no treatment (baseline: A) conditions and treatment (intervention: B) conditions within each period of observation. Each pair was observed for 35 min., which consisted of a n initial 5 min. baseline (A) condition, followed by a 10 min. treatment (B) condition, a subsequent 5 min. return to baseline (A) condition, a second 10 min. treatment (B) condition, and a final 5 min. baseline (A) condition, respectively. Three replicati ons of observations were conducted for each pair, with 2 weeks between each replication. Replications were conducted in order to assess the reliability of the data obtained (Johnston and Pennypacker, 1993) for each pair. Baseline conditions occurred in the absence of a person in the enclosure, but still in the presence of a conspecific, with which the subjects were always paired. However, because it was not always possible to ensure that volunteers were not visible or audible to the subjects during baseline conditions, volunteers were specifically instructed to avoid providing direct attention to the subjects during baseline conditions. During intervention conditions, a familiar female volunteer (with at least 3 months of experience of providing food and soc ial interaction to subjects) was either instructed either by the lead author or signaled by a pre set timer to enter the enclosure and interact with the subjects if they approached her. Social interactions included providing both tactile petting and ve rbal praise to subjects if the subjects engaged with or were in close
18 encourage play with the subjects (e.g. providing toys, grabbing onto any part of the dy) or solicit interactions (e.g. following subjects around the enclosure, intervention condition. If interactions between conspecifics occurred during intervention c onditions, volunteers were permitted to monitor the interaction, but were instructed to avoid responding in any way that could potentially interfere with the interaction (e.g., ible nonverbal cue). Tangible items (e.g., edible treats, novel items, toys, training materials) were never present during either baseline or intervention conditions. All observations were subsequently viewed from videotape as a treatment integrity procedu re. All observations were collected between the hours of 17:00 h and 19:00 h between June and September. Observations were never conducted within 30 min. of a feeding or within 6 hours of volunteer interactions. Two female volunteers participated in each separate intervention condition per replication. The order of the volunteers within each session was held constant across all replications for all subjects. Volunteers had sufficient free contact experience with all subjects, were knowledgeable about the t emperament of the subjects and trained how to interact safely with each animal. In subjects, make quick or sudden movements, pet quickly or forcefully, stare at the sub jects directly in the eyes for prolonged periods of time, or give reprimands. All observations were video recorded using a high definition camera (Kodak Zi6 HD Pocket Video Camera, Eastman Kodak Company, Rochester, NY, USA) that was positioned several met ers from the front of the enclosure, or on an adjacent platform
19 Ethogram The following behaviors were measured: affiliation, proximity to human, proximity to conspecific, and inactivity (Table 2 2). All durations of conspecific, human, and other affiliation were recorded as measures of species typical behaviors. Affiliation was defined as any of the following behaviors: tail wagging, licking volun teer or conspecific, pawing, muzzle grabbing, sitting in lap, or social play (Bekoff, 1974; Cools, Van Hout, & Nelissen, 2008). Proximity measures were scored as either in contact, proximate (within from conspecific), or distant conspecific) and were recorded instantaneously every 10 s in each condition. The percentage of time spent inactive for each condition was also calculated for each subject. Not visible was coded if subject was not visible for 30 s or more or for any duration in which the previous and subsequent behavior coded were not identical. All durations of aggression, scent rolling or rubbing, pacing, aber rant behavior (e.g. cage chewing), and vocalizations were also recorded. Other behaviors were defined as any other behavior not explicitly defined in the ethogram, and thus was mutually exclusive from all other behaviors in the ethogram. The requirement fo r recording a bout of behavior was that the behavior had to be observed for at least 3 s; similarly, the end of a bout required observation of at least 3 s in the absence of that behavior (with the exception of the not visible condition).
20 Video and Stati stical Analyses All observations were coded from digital video on computer for interobserver reliability (minimum of 90% agreement in all conditions). All behavior measures were recorded using Microsoft Excel v.11.5.0. Durations of inactivity, affiliative behavior, pacing, and aberrant behaviors were calculated across all conditions and replications and subsequently converted to proportions for each condition to account for the differences in lengths of baseline and intervention conditions. Human affiliati on and social play were recorded as durations. The duration of social play observed in each condition was recorded, and then divided by the total time elapsed in each condition and multiplied by 100% to obtain the percentage of each condition in which soci al play was observed. Human p roximity and conspecific proximity measures were calculated as percentages per condition across all conditions and replication s. Data for all behavioral measures was subsequently transformed to generate arcsine transformations of proportions of behavior observed for each condition within subjects. For each replication, a t test was used to compare proximity measures between intervention and baseline conditions, as well as between intervention conditions only, to determine if t here was an effect of caregiver. As all subjects were pair housed, each pair (not subject) was considered as an individual case for the purposes of statistical analyses on transformed data for all measures of social behavior. These measures included conspe cific proximity, total affiliation, and conspecific affiliation. The mean raw durations of both total affiliation and conspecific affiliation between two pair members for conspe cific proximity was identical; as such, the actual raw duration data for each condition was transformed. Each subject was considered an independent sample for
21 activity level data. A repeated measures Analysis of Variance (ANOVA) was subsequently used to co mpare arcsine transformations of proportions of behavior observed for each condition within subjects. Planned unpaired t tests were also conducted to generate pair wise comparisons between conditions for each replication. All statistical analyses were con ducted using Microsoft Excel v.11.5.0 or SPSS v18 (SPSS, Inc., Chicago, IL, USA).
22 Table 2 1. Demographics of subjects at Big Oak Wolf Sanctuary Pair/Names Gender Age (years) Breed Pair 1 Sampson Male 5 Alaskan gray wolf Spirit Female 5 Alaskan gray wolf Pair 2 Sabbath Male 3 Wolf dog hybrid Ava Female 3 Wolf dog hybrid Pair 3 Peace Male 3 Wolf dog hybrid Lea Female 4 Wolf dog hybrid Pair 4 Peter Male 4 Wolf dog hybrid Sara Female 5 Wolf dog hybrid
23 Table 2 2 Ethogram of species typical and abnormal behaviors for wolves and wolf dog hybrids Behavior Definition Conspecific affiliation Tail wagging, licking (us ually face, but may be any part of body); or submission (may include crouching and licking while greeting, rolling on back or side and exposing underbelly) while oriented toward conspecific. Human affiliation Tail wagging, licking (usually face, but may be any part of body); or submission (may include crouching and licking while greeting, rolling on back or side and exposing underbelly) and sniffing while oriented toward volunteer Other affiliation Tail wagging, ears back, but not oriented toward any ide ntifiable target Aggression Any aggressive response directed toward another animal or person; includes all non contact aggression (e.g. growling, snarling, raised hackles, teeth baring) and contact aggression (e.g. inhibited biting, injurious biting, sho ving, knocking over) Inactive sleeping, resting). Head may be erect or moving (e.g. looking around, panting). Not visible Any occurrence in which you cannot clearly see more than s body and/or such that you cannot identify second period in which the subject meets the above criteria and in which the previously recorded and subsequently recorded beha viors are not identical. Other Subject is alert (eyes open). Record other anytime the subject is moving and not stationary (except while standing) (e.g. walking, grooming, digging) Pacing Subject repeatedly (>3 times, with no more than 3 seconds pause a t any time) walks in a fixed route in enclosure (adapted from Hubrecht, Serpell & Poole, 1992).
24 Table 2 2. Continued Behavior Definition Scent rolling Subject presses and subsequently rubs his or her body and/or head onto volunteer (can also be ont o another object); may include rolling on back from side to side. Social play Any affiliative, non aggressive interaction between two pair housed animals occurring for at least 5 seconds; may include wrestling, relaxed body posture, reciprocal. Vocaliza tion Any audible sound emitted from subject. May commonly include barking, howling, whining, or yelping. Undesirable May include cage chewing, tail chasing, concentrated self biting and/or self licking that is compulsive (lasts for more than 15 seconds) and/or results in hair removal, irritated skin, bleeding, etc. Note: Definitions of all occurrences behaviors were recorded in durations (s). A minimum of 3s constituted the onset and offset of each behavior.
25 CHAPTER 3 R ESULTS Species Typical Affiliative Behavior Transformed proportions of conspecific directed affiliation were significantly higher in intervention condition s relative to baselines for Replication 2 (F = 4.97, P < 0.05), and Replication 3 (F = 74.0, P < 0.0 01, where d.f. = 4, 12 for all F tests) (Fig. 3 1a & b). No significant difference in conspecific affiliation between conditions was observed for Replication 1 (F = 2.89, P = 0.07). For Replication 2 planned pair wise comparisons revealed significant diff erences in conspecific affiliation between Baseline 1 and Intervention 2 (t = 3.29, P < 0.05) and Baseline 2 and Intervention 2 (t = 3.29, P < 0.05; d.f. = 3 for all t tests). In Replication 3 significant differences were found between Baseline 1 and I ntervention 1 (t = 18.0, P < 0.001), Baseline 1 and Intervention 2 (t = 3.86, P < 0.05), Intervention 1 and Baseline 2 (t = 30.5, P < 0.001), Intervention 1 and Baseline 3 (t = 9.51, P < 0.05), and Intervention 2 and Baseline 3 (t = 3.84, P < 0.05, d.f. = 3 for all t tests). Transformed proportions of human directed affiliation were also higher in intervention conditions relative to baselines in all replications, as human directed affiliation was never observed during baselines (Fig. 3 1a & b). Transforme d proportions of total affiliation observed were significantly higher in intervention conditions than in baseline conditions for Replication 1 (F = 24.8, P < 0.001), Replication 2 (F = 30.8, P < 0.001), and Replication 3 (F = 92.2, P < 0.001, where d.f. = 4 12 for all F tests). Planned pair wise comparisons revealed significant differences in the first replication between Baseline 1 and Intervention 1 (t = 10.8, P < 0.01), Baseline 1 and Intervention 2 (t = 4.04, P < 0.05), Intervention 1 and Baseline 2 ( t = 8.34, P < 0.01), Intervention 1 and Baseline 3 (t = 6.30; P < 0.01), Baseline 2 and
26 Intervention 2 (t = 7.57, P < 0.01), and Intervention 2 and Baseline 3 (t = 7.46, P < 0.01: d.f. = 3 for all t tests). In Replication 2 significant differences were f ound between Baseline 1 and Intervention 1 (t = 10.8, P < 0.01), Baseline 1 and Intervention 2 (t = 11.6, P < 0.01), Intervention 1 and Baseline 2 (t = 7.50, P < 0.01), Intervention 1 and Baseline 3 (t = 4.80; P < 0.05), Baseline 2 and Intervention 2 (t = 10.7, P < 0.01), and Intervention 2 and Baseline 3 (t = 3.91, P < 0.05: d.f. = 3 for all t tests). In Replication 3 significant differences were found between Baseline 1 and Intervention 1 (t = 10.4, P < 0.01), Baseline 1 and Intervention 2 (t = 21.7 P < 0.001), Intervention 1 and Baseline 2 (t = 12.2, P < 0.01), Intervention 1 and Baseline 3 (t = 10.4, P < 0.01), Baseline 2 and Intervention 2 (t = 12.0, P < 0.01), and Intervention 2 and Baseline 3 (t = 21.7, P < 0.001, where d.f. = 3 for all t test s). There was no significant difference in total or conspecific affiliation betw een any baseline conditions across replications (t = 1.8, d.f. = 3, P = 0.11). In Replication 1 and Replication 2, no significant difference between intervention conditions was detected in total or conspecific affiliation. In Replication 3 a significant difference in conspecific affiliation (t= 10.7, d.f.= 3, P < 0.01) was found between the Intervention 1 and Intervention 2 conditions. Durations and percentages of social play observed for each intervention condition for all pairs are reported in Table 3 2. Social play was never observed in baseline conditions. Proximity Measures The transformed proportion of proximity scans in which the subject was observed in contact, proxima te to, or distant from both the other conspecific and human was calculated for all conditions. Proximity to conspecific is shown in Figure 3 2. No
27 significant difference was found in the proportion of time conspecifics spent near one another when baseline and intervention conditions were compared in Replication 2 and Replication 3 A significant difference was found in the Replication 1 (F = 3.8, d.f. = 4, 12, P < 0.05); however, pair wise comparisons did not reveal any significant differences in any compar ison of baseline and intervention conditions. For three of the four pairs, higher proportions of time per condition that conspecifics spent in close proximity to one another were observed in baseline conditions following human interaction relative to the i nitial baseline (Fig. 3 2) On average, subjects spent 66.03% of intervention conditions in close proximity the volunteer during interaction sessions (90.44% on average ) and Spirit spent the least amount of time near the volunteer (53.55% on average). In addition, there was a significant difference in the proportions of time subjects spent near th e volunteer between intervention conditions for Replication 2 (t = 3.56, d. f. = 7, P < 0.001) and Replication 3 (t = 2.4, d.f. = 7, P < 0.05). This effect was not observed for Replication 1 Interactions were observed between human proximity data and conspecific proximity data for all individual subjects during intervention cond itions. Mean percentages of scans in which subjects were observed near (defined as either in contact or proximate to) both the volunteer and conspecific during intervention conditions were as follows: 67.9% for Ava (Range: 21.4% to 94.8%), 76.4% for Sabbat h (Range: 50.9% to 100%), 82.3% for Peace (Range: 37.2% to 100%), 74.0% for Lea (Range: 32.6% to 88.6%), 62.4% for Sampson (Range: 42.3% to 75.0%), 73.2% for
28 Spirit (Range: 19.2% to 89.3%), 91.7% for Peter (Range: 85% to 100%), and 77.0% for Sara (Range: 5 5.0% to 94.2%). Activity Levels Figure 3 3 shows the percentage of time (calculated from transformed proportions) observed in activity for all subjects on average in baseline and during human interaction. There was a significant difference in activity lev els between baseline and intervention conditions in Replication 2 (F = 7.45, d.f. = 4, 28, P < 0.001) and Replication 3 (F = 13.5, d.f. = 4, 28, P < 0.001), such that duration of inactivity was significantly lower in baseline conditions than in interventio n conditions. Planned pair wise comparisons in Replication 2 further revealed an effect for Baseline 1 compared to Intervention 1 (t = 3.25, d.f. = 7, P < 0.02), Baseline 2 Intervention 2 (t = 3.2, d.f.= 7, 28, P < 0.01), and Intervention 2 Baseline 3 (t = 2.95, d.f. = 7, 28, P < 0.02) and in Replication 3 for Baseline 1 Intervention 1 (t = 8.74, d.f. =7, 28, P < 0.001), Baseline 1 Intervention 2 (t = 8.40, d.f = 7, 28, P < 0.01), and Intervention 2 Baseline 3 (t = 4.2, d.f. = 7, 28, P < 0.01). This effect was not observed for Replication 1. In Replication 1 three subjects exhibited much higher rates of activity in intervention conditions relative to baselines, while the remaining five subjects did not (Fig. 3 3 ). Durations of pacing observed per condition were also analyzed for all subjects that emitted this behavior. Three subjects exhibited significantly higher rates of pacing (P < 0.05) during baseline conditions relative to human interaction conditions in some replications (Fig. 3 4). One subject, Ava, exhibited increased rates of cage chewing during baseline conditions following huma n interaction conditions for Replication 2 (Fig. 3 5)
29 Figure 3 1. Duration of human and conspecific affiliation a cross replications for subjects in Pairs 1 & 2. Changes in conditions and replications are represented by single vertical lines and double vertical lines, respectively. Sampson Replication 1 Replica tion 2 Replication 3 Duration of behavior observed per minute (s) Time elapsed per condition (min) Spirit Peace Lea
30 Time elapsed per condition (min) Figure 3 2 Duration of human and conspe cific affiliation across replications for subjects in Pairs 3 & 4. Changes in conditions and replications are represented by single vertical lines and double vertical lines, respectively. Sara Duration of behavior observed per minute (s) Sabbath Ava Peter Replication 1 Replication 2 Replication 3
31 Figure 3 3. Proportions of scans per condition that pair hous ed conspecifics were observed in close proximity to one another during baseline and human animal interactions (HAI) across replications. Figure 3 4. Activity levels for all subjects. Series represent mean percentage of baseline and human animal interac tion (HAI) conditions across replications that subjects were recorded as active. Replication 1 Repl ication 2 Replicatio n 3
32 Figure 3 5. Activity and pacing levels for Peter in Replication 1 Peter displayed higher rates of inactivity during human interaction sessions; however, much of his activit y during baselines was pacing, which decreased when a volunteer was present in the enclosure. Figure 3 6. Duration of cage chewing behavior em itted by Ava during Replication 2 Baseline (5 min) and intervention (10 min) conditions are se parated by dashed lines.
33 Table 3 1. Percent of social play observed in human interaction conditions for wolves and wolf dog hybrids Pair Replication 1 Replication 2 Replication 3 H 1 H 2 H1 H2 H1 H2 Sampson & Spirit 0% 0% 0% 3.7% 6.0% 0% Peace & Lea 4.7% 0% 16.7% 9.3% 3.7% 4.2% Sabb ath & Ava 0% 21.5% 33% 0% 10.5% 0% Peter & Sara 7.0% 0% 0% 4.3% 4.0% 0% Note: First and second human interaction conditions are labeled H1 and H2, respectively across replications.
34 CHAPTER 4 DISCUSSION Is Human Interaction Enriching for Wolves and Wolf Dog Hybrids? we observe its effects on specific behaviors (Hoy et al., 2009); from these effects, being. Our results suggest that all subjects showed noticeably higher rates of positive species typical affiliatio n toward both humans and conspecifics during human interaction sessions relative to baseline conditions. The presence of a familiar volunteer not only prompted greeting behavior and affiliative responses towards the human, but also promoted social inter actions among each pair. This increase in social interaction among conspecifics is chimpanzees. Conspecifics also spent a significantly greater proportion of time in close prox imity to one another when a volunteer was present in the enclosure in two thirds of replications Thus, free contact interactions with familiar human caregivers appear to be an effective strategy for increasing species typical social affiliation and enhanc ing conspecific interactions in socialized, pair housed wolves and wolf dog hybrids. In addition, the results of this study provide an informative evaluation of a specific stimulus change to serve as environmental enrichment. Given the different historie s of the subjects used in this study, the presentation of individual data allows one to examine the consistency and reliability of behavior patterns observed throughout intervention and baseline conditions across subjects. Individualized data presented cum ulatively over time and replicated repeatedly also allow for the detection of unique changes in behavior within each individual subject that would otherwise be masked by
35 group designs, averaging across conditions, and other forms of aggregate data treatmen t. Welfare is a property of the individual; thus, it is especially appropriate, instructive, and perhaps ideal, to study changes in behavior using a single subject design when evaluating enrichment efficacy. However, several limitations should be noted. F irst, the effects for conspecific proximity should be interpreted cautiously. In this study, enhanced conspecific proximity for all pairs appears to be related to human proximity for some intervention conditions. That is, our results suggest that in some i nterventions, the majority of scans in which conspecifics were scored as near one another, pair members were also scored as conspecific proximity was enhanced simply because both animals were engaging with the volunteer, and not necessarily with one another. Although our results suggest that conspecific proximity is maintained for at least a brief time following human interaction, the dramatic increase observed in af filiative behaviors appears to be temporary. Most subjects, with the exception of Peace and Lea, exhibited affiliative responses only during the intervention conditions. This suggests that these high rates of increased social interaction may not be long la sting and possibly require the continued attention of a familiar volunteer to be maintained. This finding may be of interest both to theoretical questions about the function of play behavior (discussed below ), as well as practical questions in applied sett ings. Many facilities that house captive animals seek ways to continuously promote positive, naturalistic social interactions among their animals, including at times that staff are not available to interact with the animals. Therefore, different schedules of social
36 reinforcement should be explored for their ability to promote maintenance of positive intraspecific behavior in order to maximize welfare in captive canid populations. Trends in activity levels were somewhat inconsistent across replications. Si gnificant increases in activity levels were observed during intervention conditions in Replication 2 and Replication 3 ; however, no such effect was found in Replication 1 Upon closer examination of the data, it appears that activity levels were confounded with high rates of stereotypic behavior for two subjects in Replication 1 Five subjects exhibited significantly longer durations of inactivity during baseline conditions than in treatment conditions during Replication 1 The remaining three subjects Sa ra, Peter, and Peace showed higher levels of inactivity during treatment sessions relative to baselines. However, Peter and Peace exhibited a high frequency of pacing behavior in the initial baseline conditions, followed by a dramatic elimination of paci ng in the first subsequent intervention condition. Ava was the only other subject that exhibited an aberrant behavior (cage chewing occ repl urred only during baseline conditions following an intervent ion condition in Replication 2 Thus, increases in activity levels do not necessarily reflect an improvement in welfare. Future studies that evaluate enrichment in captive species should take steps to be sure that the type of activity emitted is being asse ssed, in addition to the general activity durations. In chewing which occurred in 22.2% of return to baseline conditions further illustrates the importance of evaluating the effects of enrichment on an individual basis replicating observations, and monitoring behavior following the removal of enrichment, and not solely while enrichment is being provided.
37 A recent study that surveyed over 200 personnel from 13 zoological institutions reported that less than half of all respondents reported collecting formal data on enrichment efficacy (Hoy et al., 2009). Finally, one concern was that the presence of a single person in the enclosure would create an opportunity for increased aggression between the pairs. However, the low levels of aggression observed in all baseline conditions for all subjects were maintained throughout all human interactions sessions. Only one bout of aggression (3 s in duration) was recorded for one pair during human interaction. Thus, these data suggest that human interaction alone has the potential to be a highly enriching strategy for socialized wolves and wolf dog hybrids who are pair housed for an extended period of time. This is supported by higher rates of species typical affiliation toward humans and conspecifics, higher activity levels for the majority of subjects, low levels of aggression, and decreased pacing for all subjects that exhibited this behavior in initial baselines. Variables Influencing Enrichment Efficacy A supplementary measure of their cage mate (during both baselines and human interaction conditions) and to the volunteer (during human interaction conditions only). No difference was found in the proportion of time spent in close pr oximity to a conspecific across all conditions during Replication 1 ; however, there was an increase in the proportion of time conspecifics spent near one another from the initial baseline to the first intervention condition, but not between subsequent base lines and interventions. This is likely because pair housed animals simply remained in close proximity to one another after human interaction was no longer available. Three out of four pairs remained in close proximity to one another in
38 the final baseline condition (Fig. 3). In other words, the effects of human interaction on proximities appeared to be maintained through subsequent conditions. All subjects spent, on average, more than half of each human interaction cond ition near the volunteer. In Replicat ion 2 and Replication 3 there was a significant difference between the proportion of time spent in close proximity to the volunteer between the first and second intervention conditions, but this was not the case for Replication 1 There are several possib le explanations for this. One possibility is that all subjects may have a preference for one volunteer over another. The volunteer that was present in all first interaction sessions was the co owner of the facility, and thus more familiar to the subjects t han the second volunteer, who had been with the sanctuary for 3 months. A second possibility is that the subjects satiate somewhat to human interaction at 10 min. session durations. It would thus be worthwhile to examine the effect of a longer inter trial interval and return to baseline condition if there are limited volunteers or time to allocate to human interaction sessions. Another variable that may contribute to the efficacy of human interaction in evoking species typical affiliation is the length o f time since last feeding. An inverse relationship between length of time since last feeding and enrichment effectiveness has been suggested by Tarou and Bashaw (2007) and has also been reported with boomer balls in captive maned wolves ( Chrysocyon brachyu rus ) (Cummings et al., 2007). Future studies should examine potential motivating operations including proximity to feedings and length of time between caregiver interactions that could influence the efficacy of human interactions.
39 Finally, the results in activity levels and affiliative behaviors presented here may be conservative estimates due to the high temperatures and humidity throughout the study. Although we only conducted observations during perhaps the coolest daylight hours, it may have still been to energetically expensive to exert much activity under such conditions for these species. Thus, future studies should examine the efficacy of human interactions for this population across seasonal variation. Understanding Play and Welfare Is Play Spontaneous? The increases observed in conspecific affiliation during human interaction include d multiple occurrences of social play across pairs. The presence of a familiar caregiver was also associated with increased levels of play among conspecifics. F urthermore, these increases in play were associated with concurrent decreases in stereotypic behavior during human interaction for subjects that exhibited pacing. Though the absence of social play in baseline conditions did not permit statistical analysis, it appears that a systematic trend emerged with respect to this behavior. All pair housed subjects engaged in social play during human interaction sessions in every replication; within replications, however, social play repeatedly occurred only in one hum an interaction session for two pairs Sabbath and Ava and Peter and Sara. Furthermore, this effect was not repeatedly observed solely in the presence of the same caregiver (although social play did occur more often when the first, more familiar caregiver was present). One possibility is that social play was too energetically expensive to engage in, given the seasonal time of the study, with a return to baseline interval of only five minutes in length. Peace and Lea, however, engaged in social play in both human interaction sessions in both Replication 2 and Replication 3 It may be the case that these subjects, being the youngest of all the subjects, were also the
40 relative fittest and could thus physically engage in such behavior more often. These questions suggest that, in addition to examining seasonal variation, the schedules of human interaction should also be investigated; perhaps a longer interval between human interaction sessions would result in maximizing the efficacy of this strategy. This has also been pr oposed by Tarou and Bashaw (2007 ). These findings support the notion play behavior is associated with the environmental context in which animals find themselves and thus may be a potential indicator of animal welfare (Held and Spinka, 2010). Previ ous research has claimed that animals will only play if they are well fed, healthy, and not subjected to stressful conditions (Burghardt, 2005; Fagen, 1981; Martin and Caro, 1985). This is also consistent with observational findings that social play in cap tive wolves occurs when conflict among group members is absent and stress is thus low or minimal (Cordoni, 2009). However, the benefits of play behavior remain to be fully understood, as the task of defining welfare indicators experimentally is difficult. Furthermore, play has not been given much empirical attention in large part because of the premise that the behavior has no readily observable function; this implies that play does not contribute to the immediate survival of the organism (Burghardt, 2005) It may be argued, however, that failure to identify a function for a behavior may be due to a shortcoming of the observer rather than a true lack of function to the behavior. In this study, conspecific social play occurred only during human interaction conditions. As such, social play does not appear to be spontaneous nor independent of external stimulation, but rather, at least partially related to the presence of an interspecific social stimulus in this population.
41 An interesting observation is that so cial play did not occur in baselines even if a volunteer was visible to the animals during that time. Prior to each observation, the lead author instructed all volunteers to ignore the subjects during baselines if the volunteer could not avoid being visi ble to the observed pair. It is possible that due to their extensive history of receiving reinforcement from human caregivers, subjects were sensitive to whether or not volunteers were attending to them. In contrast, volunteers were continuously providing attention to the subjects during human interaction conditions. It is thus possible that interspecific attention may be an environmental variable maintaining positive social interactions, including conspecific play. Dyadic play between domestic dogs was fou nd to be directly correlated with conspecific attention, (Horowitz, 2009); however, the possibility that interspecific social attention could maintain social play among conspecifics has not previously been proposed. Perhaps in captive environments and hous eholds, where animals are dependent on human caretakers for survival essentials (e.g. food, shelter), play may function as an attention maintained behavior. Behavior Analysis as a Tool for Identifying Functional Relationships Between Play and the Environm ent In applied behavior analysis, functional analysis is a highly effective tool in identifying variables that maintain self injurious and other problematic behaviors in humans (e.g., Iwata et al., 1994) and non human primates (Dorey et al., 2009; Martin a nd Bloomsmith, 2011). It has not yet been demonstrated whether functional analysis could also effectively identify environmental variables that maintain positive behaviors, including social play. Such findings would be useful in addressing both basic quest ions surrounding the theories of play, as well as addressing captive management issues in
42 applied settings. Future studies may also examine ways to objectively test play as an indicator of good welfare and work toward determining a satisfactory and useful dependent measure of play for canids, and other species. The functions and environmental variables affecting play behavior are not fully understood, and play itself does not have a universal definition. Because the focus of this study was to evaluate an extant husbandry strategy at a single facility, we had limited ability to manipulate variables that could have contributed to a more functional analysis of play behavior in this population. This is a valuable direction for future research. Our results here allow us to state that access to a familiar interspecific social stimulus evokes immediate, significant increases in positive species typical behaviors, including social play, in socialized wolves and wolf dog hybrids. Thus, it is possible that social pla y in this population is at least partially related to external stimulation in the captive environment. A Note on Generality While the results of this study suggest that human interaction can be a potent enrichment strategy for the present subjects, this is likely not the case for all wolves and wolf dog hybrids housed in captivity. First, only socialized individuals were used in this study to ensure the safety of the human volunteers. However, many exotic animals arrive at sanctuaries through confiscation, often because they have been abused or neglected by their previous owners. Consequently, it is plausible that a substantial proportion of animals living in sanctuaries are not well socialized, but show fear or aggression to humans, as a result of prior exp eriences that make human presence aversive. Such animals are unlikely to respond as positively to human entrance into
43 their enclosures, even if such a strategy were considered safe for the human volunteers involved. Aggression can be treated as an operant behavior, and at least one unpublished study has successfully used distance from an aversive stimulus to reduce aggressive responses and shape desirable alternative behaviors in domestic dogs (Snider, unpublished, 2007). It would be interesting to determin e whether such a negative reinforcement shaping procedure can effectively reduce fearful behaviors that occur in response to human volunteers (e.g., pacing, prolonged staring, tail tucking, and crouching). To make human interaction an effective form of en richment for more individuals, future research should explore such interventions that can reduce both fear and aggression, as well as promote affiliative behaviors in relatively non socialized animals. This would be hugely beneficial for management purpose s in any facility providing long term residence to captive animals.
44 LIST OF REFERENCES Baker, K.C., 2004. Benefits of positive human interaction for socially housed chimpanzees. Anim. Welf. 13, 239 245. Baker, K.C., Bloomsmith, M.A. 2003. Se lf injurious behavior and response to human interaction as enrichment in rhesus macaques. Am. J. Primatol. 60, 94 95. Baker, K.C., Bloomsmith, M.A., Neu, K., Griffis, C., Maloney, M.A., 2010. Positive reinforcement training as enrichment for singly housed rhesus macaques (Macaca mulatta ). Anim. Welf. 19, 307 313. Bekoff, M., 1974. Social play and play soliciting by infant canids. Am. Zool. 14, 323 340. Bergamasco, L., Osella, M.C., Savarino, P., Larosa, G., Ozella, L., Manassero, M., Badino, P., Odore, R., Barbero, R., Re, G., 2010. Heart rate variability and saliva cortisol assessment in shelter dog: Human animal interaction effects. Appl. Anim. Behav. Sci. 125, 56 68. Bloomsmith, M.A., Laule, G.E., Alford, P.L., Thurston, R.H., 1994. Using training t o moderate chimpanzee aggression during feeding. Zoo Biol. 13, 557 566. Burghardt, G.M., 2005. The Genesis of Animal Play: Testing the Limits. Cambridge, Massaschusetts: MIT Press. Busch, R.H., 2007. The Wolf Almanac: A Celebration of Wolves and Their Wo rld. Guilford, Connecticut: The Lyons Press. Carlstead, K., 2009. A comparative approach to the study of keeper animal relationships in the zoo. Zoo Biol. 28, 589 608. Claxton, A.M., 2011. The potential of the human animal relationship as an environmen tal enrichment for the welfare of zoo housed animals. Appl. Anim. Behav. Sci. 133, 1 10. Cools, A.K.A., Van Hout, A.J. M., Nelissen, M.H.J., 2008. Canine reconciliation and third party initiated postconflict affiliation: Do peacemaking social mechanisms i n dogs rival those of higher primates? Ethology 114, 53 63. Coppola, C.L., Grandin, T., Enns, R.M., 2006. Human interaction and cortisol: Can human contact reduce stress for shelter dogs? Appl. Anim. Behav. Sci. 87, 537 541. Cordoni, G. 2009. Social play in captive wolves ( Canis lupus ): not only an immature affair. Behaviour 146, 1363 1385.
45 Cummings, D., Brown, J.L., Rodden, M.D., Songsasen, N., 2007. Behavioral and physiologic responses to environmental enrichment in the maned wolf ( Chrysocyan brachyur us ). Zoo Biol. 26, 331 343. Cusdin, P.A., Greenwood, M.A., 2000. The keeping of wolf hybrids in Great Britain. Royal Society for the Prevention of Cruelty to Animals. Dorey, N.R., Rosales Ruiz, J., Smith, R., Lovelace, B., 2009. Functional analysis and t reatment of self injury in a captive olive baboon. J. Appl. Behav. Anal. 42, 785 794. Fagen, R. 1981. Animal Play Behavior. Oxford: Oxford University Press. Held, S.D.E., Spinka, M., 2010. Animal play and animal welfare. Anim. Behav. 81, 891 899. Horo witz, A., 2009. Attention to attention in domestic dog ( Canis familiaris ) dyadic play. Anim. Cognition 12, 107 118. Hoy, J.M., Murray, P.J., & Tribe, A., 2009. Thirty years later: Enrichment practices for captive mammals. Zoo Biol. 29, 303 316. Iwata, B. A., Dorsey, M.F., Slifer, K.J., Bauman, K.E., Richman, G.S., 1994. Toward a functional analysis of self injury. J. Appl. Behav. Anal. 27, 197 209. Johnston, J.M., Pennypacker, H.S., 1993. Strategies and Tactics of Behavioral Research. Hillsdale: Lawrence Erlbaum Associates, Inc. Maloney, M.A., Baker, K.C., Griffis, C., Neu, K., Bloomsmith, M.A., Martinez, M., 2007. Behavioral responses by singly housed rhesus macaques ( Macaca mulatta ) during positive reinforcement training and human interaction. Am.J.Pri matol. 69, 119 120. Manciocco, A., Chiarotti, F., Vitale, A., 2009. Effects of positive interaction with caretakers on the behaviour of socially housed common marmosets. Appl. Anim. Behav. Sci. 120, 100 107. Martin, A.L., Bloomsmith, M.A., 2011. Functio nal analysis and treatment of human directed undesirable behavior exhibited by a captive chimpanzee. J. Appl. Behav. Anal. 44, 139 143. Martin, P., Caro, T.M., 1985. On the function of play and its role in behavioral development. Advances in the Study of Behavior. 15, 59 103.
46 Normando, S., Corain, L., Salvadoretti, M., Meers, L., Valsecchi, P., 2009. Effects of an using a novel nonparametric test. Appl. Anim. Behav Sci. 116, 211 2 19. Shepherdson, D., 1998. Tracing the path of environmental enrichment in zoos. In: Shepherdson, D.J., Mellen, J.D., Hutchins, M. (Eds.), Second Nature: Environmental Enrichment for Captive Animals. Smithsonian Institution Press, Washington, DC. Snider, K.S., (Unpublished results). A constructional canine aggression treatment: using a negative reinforcement shaping procedure with dogs in home and community settings. University of North Texas, Denton, TX, USA. Tarou, L.R., Bashaw, M.J., 2007. Maximizing the effectiveness of environmental enrichment: Suggestions from the experimental analysis of behavior. Appl. Anim. Behav. Sci. 102, 189 204. Waitt, C., Buchanan Smith, H.M., Morris, K., 2002. The effects of care taker primate relationships on primates in the laboratory. J. Appl. Anim. Welf. Sci. 5, 309 319. Wells. D.L., 2009. Sensory stimulation as environmental enrichment for captive animals: a review. Appl. Anim. Behav. Sci. 118, 1 11. Wood, W., 1998. Interactions among environmental enrichment, viewin g crowds, and zoo chimpanzees ( Pan troglodytes ). Zoo Biol. 17, 211 230.
47 BIOGRAPHICAL SKETCH Lindsay Renee Mehrkam was born in 1987 in Allentown, Pennsylvania. She graduated summa cum laude from William Allen High School in 2005. She s ubsequently earned he r Bachelor of Arts in animal b ehavior from Franklin and Marshall College in 2009, where she was also awarded the Roger and Elizabeth Thompson Award in Animal Behavior for her research on captive wolf dog hybrids. Her study was conducted in collaboration wi th Dr. Roger Thompson, and was hosted at both the Wolf Sanctuary of Pennsylvania and the California Wolf Center. Upon graduating in 2009, Lindsay was assigned a research internship at Animal Kingdom Theme Park at Walt Disney World, in Bay Lake, Florida. T his position provided her with the opportunity to collect behavioral data on a wide range of non human primates and avian species housed in captivity, to later be used for management and conservation purposes. In addition, Lindsay assisted with a project o n Key Largo woodrats, a critically endangered species endemic to Florida. This field position allowed her to work temporarily with the U.S. Fish and Wildlife Service and the Florida Fish and Wildlife Commission, in conjunction with the Walt Disney Company Psychology allows her the opportunity to study human animal interactions with captive canids. In addition, Lindsay collaborates with the Santa Fe College Teaching Zoo on an e valuation study of environmental enrichment for Galapagos tortoises. Upo n the completion of her Master of Science in psychology Lindsay will conti nue in the doctoral program in p sychology at the University of Florida, specializing in welfare research on c aptive wolf dog hybrids and other exotic animals.