v, 137 leaves : ill., (part col.) ; 26 cm. / Title page, contents and abstract only. The complete thesis in print form is available from the University Library. / Thesis (Ph.D.1974) from the Dept. of Zoology, University of Adelaide

Thesis (Ph.D. 1974) from the Dept. of Zoology, University of Adelaide.

Information on the reproductive biology and behavioural ecology of southern hairy-nosed wombats (Lasiorhinus latifrons) is limited. Field reproductive and behavioural studies have been hindered by the difficulties associated with the routine recapturing and direct observation of wombats in the wild. Additionally, due to their extremely poor breeding success within captivity and the intrinsic complications associated with the monitoring of nocturnal and semi-fossorial activity, little formal research has been conducted on captive-held individuals. The overarching objectives of this research were to gain a better understanding of (1) both male and female L. latifrons reproductive physiology/behaviour that will lead to improved captive breeding program outcomes and (2) captive L. latifrons activity rhythms and behavioural time budgets in order to identify the impacts of captivity on wombat behaviour and wellbeing. The primary experiments of this research were centred around the development of ‘non-invasive methodologies’ for the collection of biological samples (Chapter 2), direct observation of behaviour and activity (Chapter 3), assessment of male reproductive function (Chapter 4), detection of female oestrus and cyclicity (Chapter 5), characterisation of activity patterns (Chapter 6) and monitoring of stress (Chapter 7). The last two experiments also tested the efficacy of gentling (Chapter 7) and enrichment (Chapter 8) to improve the wellbeing of the captive wombats. Faecal steroid analysis is a non-invasive tool that allows the stress-free monitoring of steroid hormones and has been used on a wide range of animal species to examine their reproductive physiology and adrenal function. The usefulness of faecal hormone analysis, however, is directly related to the reliable collection and identification of individual faecal samples. In group-housed animals, the identification of faecal samples can be difficult and time consuming and is generally only accurate if a marker is incorporated into the animal’s diets. Hence, Chapter 2 examined the usefulness of non-toxic plastic glitter as a faecal marker in group-housed L. latifrons. Forty-two food treats were tested as vehicles for the oral delivery of glitter; of these, vehicle palatability (> 75% consumed) and consistency of intake (eaten > 80% of times offered) was high for six treats: (1) golden syrup with horse pellets, (2) golden syrup with weetbix, (3) pitted-dates, (4) honey with kangaroo pellets, (5) nutrigel with rolled-oats and (6) strawberry sauce with rolled-oats. Marker mean rate of passage was 2.9 ± 0.5 d, with maximal output occurring 4.2 ± 0.3 d after oral administration. A minimum marker dose of 1.6 g was necessary to achieve high labelling consistency (> 2 flecks of glitter were defaecated in > 90% of pellets) and this dosage was required every 3 days to maintain a steady and detectable state of marker output. Twelve glitter colours were tested and optimum labelling results were obtained with gold, metallic red, metallic green and metallic blue. Once established, this technique was then used to facilitate the long-term collection of faecal samples in order to characterize patterns of reproduction (Chapter 4), monitor ovarian events (Chapter 5) and to quantify stress-responses (Chapter 7) via faecal steroid analysis. The direct observation of wombats is difficult; individuals are not easily identified and the animals are often out-of-view (residing in burrows) or obscured (low-light conditions) during sampling. Published behavioural data available on L. latifrons is largely restricted to visual observations made during dawn and dusk only, whilst published activity data pertains to time spent in or out of the burrows rather than actual physical activity. Hence, Chapter 3 tested the effectiveness of two electronic monitoring systems to remotely observe wombat behaviour and physical activity. Digital video-surveillance proved to be an effective technique for the recording and monitoring of captive wombat behaviour. Animal visibility was good, behavioural events unambiguous and the system enabled the long-term, concurrent recording of behaviour with no direct human presence. Similarly, radio-telemetry proved to be an effective way of recording captive wombat physical activity. The system was reliable, removed observer error and enabled the continuous and concurrent recording of wombat activity. Once established, these two remote monitoring systems were then used to describe wombat behaviour elements and activity patterns (Chapters 3 and 6). After a year of continuous monitoring it was established that the wombats spent, on average, 69.9% sleeping, 8.8% lying resting, 5.2% feeding, 5.2% exploring, 4.3% stereotyping, 2.5% sitting-at-rest, 1.7% digging, 1.4% foraging, 0.4% being handled, 0.3% sun-basking, 0.2% grooming and 0.1% courtship/mating. Temporal patterns were bimodal for 8/12 of the wombat behaviours and unimodal for the remaining four. The mean proportion of total daily time spent active was 18.2 ± 1.8%. Daily activity patterns were characterized by a strong circadian cycle, with high nocturnal activity and low diurnal activity. Daily onset (18:19 h) and cessation (04:34 h) of activity was seasonally constant and strongly associated (P < 0.01) with sunrise/set, but not influenced by either temperature or humidity (P ≥ 0.09). At night there was an alternating rhythm of active and rest periods, with activity peaking at the beginning (19:00 h) and end (03:00 h) of each night. Activity was seasonal with annual changes in temperature, humidity and night-time length being the triggers of variation. Mean daily activity was greater during winter (19.7%) and spring (18.9%), than during summer (16.3%) and autumn (17.2%), with the degree of activity being largely governed by ambient temperature. Feeding, sleeping and stereotyping varied significantly with season. Feeding and stereotyping were negatively associated with ambient temperature and humidity, whilst being positively associated with night-time length; the inverse relationship was true for sleeping. Ambient temperature exerted its largest effect on time spent feeding; feeding times decreased by 3.1 min / 1ºC above 13ºC and compared to spring, feeding times were reduced by 41% during summer. There is, at present, very little data published on male and female wombat reproduction. The reticence of wombats to breed in captivity makes it difficult to study reproduction in captive animals and their cautious, nocturnal nature makes field studies challenging. Non-invasive techniques to monitor reproductive status/function will assist in improving the general knowledge of wombat reproduction and the development of new captive breeding management strategies, by allowing the easy monitoring of captive animals. Thus, the series of experiments in Chapters 4 and 5 explored the efficacy of a number of non-invasive methodologies to assess male reproductive function, monitor female cyclicity and predict the timing of oestrus. To address the pulsatile nature of testosterone, a GnRH agonist stimulation test was developed in the male. IM injection of 4 μg of buserelin (a GnRH agonist) resulted in an increase (P < 0.05) in plasma testosterone concentrations, with maximum secretion occurring at 90 min. Thereafter, plasma testosterone concentrations remained near maximum for 150 min. There was a strong, positive correlation (r = 0.73, P < 0.01), between pre-stimulation testosterone concentrations and the maximal concentrations achieved post-stimulation with post-stimulation concentrations varying between individuals (P < 0.01). These findings indicate that individual male wombats can show large fluctuations in plasma testosterone concentrations over time and that a GnRH agonist can be used to obtain a diagnostic index of the prevailing testosterone biosynthetic capacity of the testes. This technique was then used as part of a larger experiment (Chapter 4) to investigate seasonal changes in male wombat reproduction. To date, the effects of breeding season on captive male wombat fertility have yet to be examined and a better understanding of this phenomenon will pinpoint the most favourable times for mating. Seasonal changes in a series of male reproductive parameters were non-invasively examined over a 18 month period: (1) testosterone concentration, both plasmic and faecal, was monitored using enzyme-immunoassays (EIA), (2) testicular volume was measured manually using digital vernier callipers, (3) sperm production was evaluated by way of spermatorrhoea, whilst (4) prostate and bulbourethral gland cross-sectional areas were assessed by ultrasonography. Plasma testosterone secretion increased in early-winter, peaked late-winter and declined in early-spring (P < 0.01). No seasonal variation (P = 0.22) in faecal testosterone metabolite concentrations was apparent. Testicular volume showed no significant variation (P = 0.29) over the sampling period and spermatozoa were found in the urine throughout the year; these two observations suggest that the captive male wombats remain spermatogenically active year round. While there was no significant seasonal change (P = 0.20) in prostate size, bulbourethral gland size increased in late-autumn, peaked in mid-winter and declined in early-summer (P < 0.01). Ultimately, captive male reproductive function was influenced by seasonality, with a peak in plasma testosterone and bulbourethral gland size occurring in winter (Jun-Aug). In an attempt to characterize oestrus-specific behaviour and develop a reliable method of oestrus detection in L. latifrons, the reproductive physiology and behaviour of eight adult females was monitored for a period of 12 months (Chapter 5). The reproductive behaviour of both sexes (4♂: 8♀) was monitored using 24-h video surveillance, whilst female physical activity was remotely measured using radio-telemetry. A faecal sample was collected every three days, from each female to assess changes in faecal progesterone and oestradiol-17β metabolite concentrations. Each female also received an injection of 0.01, 0.1 or 0.2 mg/kg of oestradiol benzoate (OB) in one of two hormone trials. Video surveillance revealed that the courtship (n = 426) and mating (n = 46) ritual of L. latifrons consisted of 13 distinctive behaviours expressed over six obvious phases: investigation, attraction, chase, restraint, copulation and recovery. Reproductive behaviour was observed in five (2♂, 3♀) wombats, with female receptivity occurring at night and lasting for only 13-h. Faecal progesterone metabolite analysis proved to be a reliable method for mapping oestrous cycle activity, but was not useful for the prediction of oestrus. Six out of the eight female wombats displayed periods of elevated progesterone secretion. From these six individuals, 23 luteal phases, 12 follicular phases and 12 oestrous cycles were recorded, with a mean (± SE) length of 20.9 ± 1.1 d, 11.6 ± 0.6 d and 31.8 ± 1.1 d, respectively. In contrast, changes in the secretion of faecal oestradiol-17β metabolites provided little instructive information on oestrous cycle activity and were not associated with oestrus. Administration of OB resulted in a spike of oestradiol-17β metabolites in the faeces 3-4 d later, but was not dose dependent nor did it elicit oestrus-behaviour. Activity monitoring does not appear to be a useful method for detecting oestrus in L. latifrons, as changes were not associated with key events in the oestrous cycle. However, 24-h video surveillance proved to be a reliable method for oestrus detection in the captive L. latifrons. Threatening or aversive stimulation is experienced in wild and captive conditions alike and evokes similar physiological responses. If an animal, wild or captive, cannot cope with this stimulation it may experience stress. An uncontrollable source of stress for all captive animals lies in their interactions with their human caretakers. A high or persistent fear of people can be a source of psychological stress for animals in captivity. Positive handling is a potent method of reducing the specific fear of human beings through desensitisation. The response of animals to handling by humans has been extensively evaluated in domesticated species, but rarely assessed in wild animals. Hence, Chapter 7 examined the usefulness of a regular handling program to lower the behavioural fear and physiological stress responses of L. latifrons to human interaction. Adult L. latifrons (n = 12) were exposed to two different treatments in a replicated design: (1) daily handling: 15 min of tactile contact from a human handling 5 d/wk for 12 wk and (2) no-handling: no contact apart from that received during routine husbandry. The effect of handling was assessed using overt response, human approach, stressor and novel stimulus tests. Synthetic ACTH was used to validate a method for monitoring faecal cortisol in L. latifrons by EIA. IM injection of 250 μg of Synacthen resulted in a significant (P < 0.05) increase in plasma and faecal cortisol concentration 30 min and 2 d after administration, respectively. Handling positively affected the behavioural responses of the wombats to human approach and contact in two ways: (1) a significant reduction (P < 0.01) in the wombat’s mean flight distances to human approach and (2) a significant (P < 0.01) decline in the strength of the wombat’s behaviour-based fear reactions (i.e. fear scores) to human proximity and contact. Handling had no discernable effect on the wombat’s physiological stress responses to human contact or on their reactions to novelty. While faecal cortisol secretion increased in response to a stressor test involving human contact, it was not alleviated by regular handling (P = 0.84). Similarly, the wombat’s reactions to unfamiliar objects during novel stimulus testing were also unaffected (P = 0.17) by the handling treatment. Therefore, handling exerted its main effect on the behavioural responses of the wombats, representing fear responses to human handlers, rather than reducing their anxiety. The main difference between wild and captive environments lies in the differential ability of control, i.e. a free-living animal is able to control its amount of incoming stimulation whereas a captive animal has a limited capacity to alter the external stimulation to which it is exposed. Without natural behaviour outlets, captive animals have to rely on abnormal behaviour patterns to modify their expectations of the captive environment and exert some control over incoming stimulation. A ‘stereotypy’, defined as a repetitive movement pattern with no apparent goal or function, is a common abnormal behavioural pattern expressed by zoo animals. Stereotypies are of concern because of their association with poor welfare. A previous behavioural study revealed that common wombats (Vombatus ursinus) in captivity are very susceptible to the development of stereotypic behaviour. For that reason, the final experiment (Chapter 8) was designed to further examine stereotypic behaviour in wombats and tested whether environmental enrichment could be used to reduce its prevalence. Adult L. latifrons (n = 12) were subjected to two different treatments in a replicated design with 12 week periods: (1) Enrichment – animals received feed and olfactory enrichment, 5 d/wk and (2) No-enrichment - animals received the standard captive diet only. Wombat behaviour was remotely observed via 24-h video surveillance. Eight out of the 12 wombats displayed one of four stereotyped movements (straight-line pacing, boundary pacing, figure-8 pacing or wall-climbing), with time spent stereotyping ranging from 61-129 min/d (mean 87 ± 7 min/d); time devoted to stereotyping took up 4-9% of the daily budget. There was a significant increase in foraging (333%) and exploration (13%) in response to enrichment. Enrichment also encouraged the expression of a wider range of naturalistic foraging behaviours. Despite these positive effects, enrichment had no discernable effect on the time spent stereotyping (P = 0.87) or inactive (P = 0.13). Despite the fact that stereotypies and inactivity were not reduced by enrichment, animal welfare was still enhanced as there was a notable improvement in natural wombat-specific behavioural expression and diversity.