BEHAVIOURAL PATTERNS OF POSSUMS AND CATTLE
WHICH MAY FACILITATE THE TRANSMISSION OF
TUBERCULOSIS
A thesis presented in partial fulfilment of the requirements for the degree of Master in Veterinary Science at Massey University
Brent Maynard Paterson
1993
ABSTRACT
Behavioural patterns of a population of Australian brushtail possums with endemic tuberculosis were studied using radio telemetry, and by direct observational techniques, from November 1990 to April 1992. The study area, on the east coast of the Wairarapa, New Zealand, allowed observations of interaction between the major wildlife vector of tuberculosis in New Zealand, and cattle run as part of a commercial farming venture. The rugged 40 ha study area is mainly covered in 2-10 m high scrub, with patches of native bush and some large trees. Part of the area has been cleared of scrub and grassed. Poorer quality pasture is also available in small pockets in many other areas of the paddock.
Possums had distinct ranges that remained constant over the duration of the study period, and ranges of many possums overlapped. Males had significantly larger ranges during the breeding season, and had the largest ranges overall. The area covered by a possum in a night’s activities varied considerably between possums, and often for an individual over consecutive nights. There were no significant differences between home ranges of tuberculous and non-tuberculous possums, although there were indications that the size of nightly activity areas of the former decreased as the disease progressed. Twenty-five juvenile possums were followed over 8-10 months, 2 dispersed from the area, and 7 died within their natal home range. Most of the juveniles died from starvation and exposure.
Interactions between possums and cattle were observed in a natural setting, and also by introducing sedated possums to the same area to simulate terminally-ill tuberculous animals. Possums spent a variable amount of time feeding on pasture in the 40 sq.m observation area, ranging from a few minutes, to several hours. Normally possums appeared to avoid cattle whenever possible, and if necessary climbed trees to get away. The activities of several debilitated possums are described and their apparent indifference to external influences - in particular time of day - noted. The intense interest shown by cattle in sedated possums is described, and the possibilities for transmission of tuberculosis from possums to cattle discussed. It is concluded that transmission of tuberculosis is unlikely to occur on open pasture under normal circumstances, but that sick tuberculous possums, and later their carcases, are a source of infection.
Tuberculosis in English badgers is compared, and contrasted, with the situation in New Zealand.
Possible explanations for the absence of the disease in Australian wildlife are discussed. The differences in habitat and population density are suggested as the main reasons for the variations between, and within, countries. The importance of controlling tuberculosis in New Zealand, deficiencies in present control systems and possible areas for future research are outlined.
ACKNOWLEDGMENTS
During the past three years I have been a member of an epidemiological research unit within the Department of Veterinary Clinical Science at Massey University. Drawn from many parts of the world, and ably led by Professor Roger Morris, the group of staff, students and visiting academics have ensured that a wide ranging experience of epidemiology, veterinary medicine, camaraderie and foreign culture have come my way.
I am particularly grateful for the help and assistance from the following people. Special thanks to Professor Roger Morris, my chief supervisor, whose support, unfailing enthusiasm, and confidence that I could complete the project were invaluable. Thanks to co-supervisors Dr. Phil Cowan of Landcare Research, New Zealand, who helpfully criticised the manuscript and offered many practical suggestions, and Dr. Ed Minot from the Ecology Department, Massey University.
I am indebted to Miss Jenny Weston who helped me develop the radio triangulation system, and cheerfully spent many long and cold hours collecting data. Thanks to Mr. Bill Maunsell, owner of Waio station, for allowing access to his property, and special thanks to Mr. Ron Goile, station manager, for adjusting stock movements in line with study requirements, and help with the design and construction of hides and shelters. M.A.F. livestock officer Mr. Tony Harris provided much useful advice on possum movements in the study area, and instructed me in the basics of radio- tracking. Mr. Dick Andrews (livestock officer) and Mesdames Donna Lewis and Wendy Gabell provided technical support in the field. Messrs Dave Ward and Kevin Ley, SirTrack Electronics, gave much assistance with radio triangulation techniques and methodology. Miss Rachelle Hughes- Sparrow shared the last few weeks of radio triangulation duties with me, and processed the faecal samples. Thanks to Drs. Dirk Pfeiffer and Ron Jackson, the former for computing advice, and both for many hours of discussions and advice about my project, the longitudinal study on Waio farm, and about the universe and life in general.
Mrs Fiona Dickinson helped with advice concerning thesis layout and Mrs Robyn O’Connor assisted with solving bureaucratic and financial problems. I would also like to thank Vanessa Tilson, the post-graduate students, and various staff members of the Veterinary Faculty, who provided support in an indirect, but no less significant, manner.
The research was funded by the Animal Health Board, and through the Board, the farmers of New Zealand. I was supported by a rehabilitation grant from the Accident Compensation Corporation of New Zealand for much of the project.
The preparation of this thesis has involved input from a large number of people over the past three years and their enthusiasm and stimulating discussion over that time have opened my eyes to the diversity of man and animal on this World of ours. I owe many people a debt of gratitude, and name but a few - I thank you all.
Finally, thanks to my wife Elaine. Our courting and early months of marriage coincided with the preparation of this thesis, and the support and forbearance shown by Elaine are gratefully acknowledged.
Brent Paterson
Department of Veterinary Clinical Sciences Massey University
November 1993
TABLE OF CONTENTS
ABSTRACT ... ii
ACKNOWLEDGMENTS... iii
TABLE OF CONTENTS ... v
LIST OF TABLES ... ix
LIST OF FIGURES ... x
Chapter 1: Introduction and Project Outline ... 1
1.1. INTRODUCTION ... 2
1.1.1. PROJECT BACKGROUND ... 2
1.1.2. Tuberculosis, and Tuberculous Possums in New Zealand ... 2
1.1.3. Study Aims ... 5
1.2. STUDY AREA... 7
1.2.1. Geography and Location ... 7
1.2.2. Climate... 7
1.3. SOCIAL ORGANISATION AND BEHAVIOUR OF THE POSSUM... 10
1.3.1. Introduction ... 10
1.3.2. Activity Patterns ... 10
1.3.3. Home Range Size and Population Density... 12
1.3.3.1. Home range... 12
1.3.4. Territorial and Denning Behaviour ... 13
1.3.4.1. Possum Den Sites... 14
1.3.5. Olfactory Communication ... 15
1.3.5.1. Pouch Glands ... 15
1.3.5.2. Paracloacal Glands... 15
1.3.5.3. Urine Marking... 16
1.3.5.4. Chin Glands ... 16
1.3.5.5. Sternal Glands... 16
1.3.6. Vocalisations ... 17
1.3.7. Breeding and Reproduction ... 18
1.3.8. Dispersal ... 20
1.3.9. Summary... 21
Chapter 2: A Radio-Tracking Study of Brush-Tailed Possums and Cattle... 22
2.1. INTRODUCTION ... 23
2.2. RADIO-TRACKING - A REVIEW ... 24
2.2.1. Radio-Tracking Defined ... 24
2.2.2. History ... 24
2.2.3. Advantages and Disadvantages ... 25
2.2.4. Specific Application Of Radio-Tracking... 27
2.2.5. Radio-Tracking - System Description ... 28
2.2.5.1. Transmitters ... 29
2.2.5.1.1. Frequency... 29
2.2.5.1.2. Power supply... 30
2.2.5.1.3. Signals... 30
2.2.5.1.4. Antennae ... 31
2.2.5.1.5. Transmitter Encapsulation and Attachment... 31
2.2.5.2. Receiving System... 32
2.2.5.2.1. Receivers... 32
2.2.5.2.2. Antennae ... 32
2.2.5.2.3. Recorders ... 34
2.3. HOME RANGE ESTIMATORS - A REVIEW... 36
2.3.1. Types of Home Range Estimator... 36
2.3.1.1. Minimum area polygons ... 36
2.3.1.2. Bivariate normal estimators ... 37
2.3.1.3. Non-parametric methods... 37
2.3.2. Summary... 38
2.4. MATERIALS AND METHODS ... 39
2.4.1. Materials ... 39
2.4.1.1. Radio collars ... 39
2.4.1.2. Radio Receivers and aerials... 40
2.4.2. Methods - Radio Triangulation ... 42
2.4.2.1. Tracking procedures ... 43
2.4.2.2. Juveniles ... 44
2.4.3. Methods - Home Range Analysis ... 44
2.5. RADIO TRIANGULATION ERROR ... 45
2.5.1. Aim ... 45
2.5.2. Introduction ... 45
2.5.3. Materials and Methods ... 47
2.5.3.1. Selection of point locations and surveying... 47
2.5.3.2. Collection of Bearings for Error Data Analysis ... 48
2.5.4. Results and Discussion of Triangulation Error ... 48
2.5.5. Summary... 54
2.6. RADIO TRIANGULATION RESULTS ... 56
2.6.1. Possum Activity Areas and Home Ranges ... 56
2.6.1.1. Effect of weather on Possum movements... 63
2.6.1.2. Possums movements in relation to habitat... 63
2.6.2. Cattle Movements in the Area ... 65
2.6.2.1. Distribution of Cattle radio-locations through the site ... 67
2.6.3. Movements of Seven Frequently Tracked Possums... 69
2.6.3.1. Size of home range in relation to number of observations ... 70
2.6.3.2. Home ranges of Possums ... 71
2.6.3.3. Possum activity areas... 73
2.6.3.3.1. Changes in the size of a tuberculous possum’s activity areas ... 76
2.6.3.4. Distribution of Possum Movements. ... 76
2.6.4. Common use of Habitat by Possums and Cattle... 78
2.7. ANALYSIS OF DEN USE BY POSSUMS... 80
2.7.1. INTRODUCTION ... 80
2.7.2. RESULTS ... 81
2.7.2.1. Den use by possums in the study area ... 81
2.7.2.1.1. Multiple use of den sites ... 81
2.7.2.1.2. Use of the same den by different possums... 81
2.7.2.1.3. Classification of different den types ... 81
2.7.2.2. Analysis of den use by seven frequently tracked possums... 82
2.7.2.2.1. Distribution of dens among habitat type... 83
2.8. FATES OF RADIO-TAGGED JUVENILE POSSUMS ... 90
2.8.1. RESULTS ... 90
2.9. DISCUSSION... 94
Chapter Three: An Observational Study of Possums in a Bush- Pasture Habitat ... 98
3.1. INTRODUCTION ... 99
3.2. MATERIALS AND METHODS ... 100
3.2.1. Study Area ... 100
3.2.2. Possum Observations... 101
3.2.2.1. Observations made through the study area ... 102
3.3. RESULTS... 104
3.3.1. Observation periods and weather conditions... 104
3.3.2. Activity of possums on pasture ... 104
3.3.2.1. Movements about, and numbers of possums in, the observation area... 104
3.3.2.2. Effect of weather on possum activity ... 106
3.3.2.3. Activity patterns of possums on pasture... 107
3.3.2.4. Interactions between possums ... 108
3.3.2.4.1. Descriptions of interactions seen ... 108
3.3.2.5. Interactions between cattle and possums... 111
3.3.2.6. Possums - movements about their dens and the noises they make... 111
3.3.2.7. Miscellaneous observations ... 112
3.4. DISCUSSION... 113
Chapter Four: Interactions Between Beef Cattle and Simulated Tuberculous Possums on Pasture... 115
4.1. INTRODUCTION ... 116
4.2. MATERIALS AND METHODS ... 118
4.2.1. Behaviour Classification and Analysis... 119
4.2.1.1. Focal-animal sampling... 120
4.2.1.2. Scan-sampling... 121
4.2.2. General Description of Behaviour Patterns... 121
4.2.2.1. Possums ... 121
4.2.2.2. Cattle... 122
4.2.2.3. Possum - Cattle Interaction... 122
4.3. RESULTS... 124
4.3.1. General Description of Trials ... 124
4.3.1.1. Description of observation runs... 125
4.3.2. Focal Animal Sampling: Patterns of Interaction between Possum and Cattle... 126
4.3.2.1. Possum activity during observations ... 126
4.3.2.2. Contact between possum and cattle... 126
4.3.2.3. Interest shown by cattle towards possums... 128
4.3.3. Scan Sampling Based on Individual Cattle Beasts ... 132
4.3.3.1. Logistic regression of risk ... 135
4.3.4. Responses to Possum Carcases on Pasture... 136
4.4. DISCUSSION... 140
Chapter Five: A Preliminary Study of the Diet of Possums on Waio Station, Castlepoint, New Zealand... 144
5.1. INTRODUCTION ... 145
5.2. REVIEW OF RELEVANT LITERATURE... 146
5.2.1. The Australian Brush-tailed Possum in New Zealand... 146
5.2.2. Possum Diets ... 147
5.2.3. Dietary Analysis ... 148
5.3. MATERIALS AND METHODS ... 150
5.3.1. Sampling Method and Preparation ... 150
5.3.2. Reference Collection And Identification Of Species Eaten ... 150
5.4. RESULTS... 151
5.4.1. Seasonal Variations Of Plant Types In Faeces... 151
5.5. DISCUSSION... 158
Chapter Six: General Discussion and Conclusion... 160
6.1. DISCUSSION... 161
6.1.1. Possums as a Wildlife Reservoir of Tuberculosis ... 161
6.1.1.1. Possible Transmission Mechanisms ... 162
6.1.2. Tuberculosis in Badgers ... 162
6.1.2.1. Badger Ecology and Behaviour ... 163
6.1.2.1.1. Characteristics of Mycobacterium bovis ... 164
6.1.2.1.2. Transmission of tuberculosis between badgers and cattle... 165
6.1.2.2. Summary ... 165
6.1.3. Discussion of Study Findings ... 166
6.1.3.1. Home range size and dispersal... 166
6.1.3.2. Possum Activity on Pasture ... 166
6.1.3.3. Interactions between Possums and Cattle... 167
6.2. CONCLUSION ... 169
APPENDIX
I
... 173APPENDIX
II
... 174APPENDIX
III
... 176APPENDIX
IV
... 177APPENDIX
V
... 178APPENDIX
VI
... 179BIBLIOGRAPHY ... 188
LIST OF TABLES
Table 2.1 Mean errors from 2 tests, for each triangulation station ... 49
Table 2.2 Summary statistics of error test locations by group ... 51
Table 2.3 Error polygon areas (square metres) - means of groups ... 53
Table 2.4 Possums and Cattle radio-tracked during 1991/92 and their status at the end of August 1992... 58
Table 2.5 Measures of total home range for possums and cattle - April 1991 to April 1992 60 Table 2.6 Average size of nightly activity areas, by 4 different methods, of male and female possums. ... 61
Table 2.7 Variation in mean activity area (ha) by season ... 62
Table 2.8 Variation in mean activity areas (ha) according to status ... 62
Table 2.9 Mean night-time activity areas for cattle... 66
Table 2.10 Mean daily activity areas for cattle ... 66
Table 2.11a Variation in shape and extent of possum activity areas through the year ... 73
Table 2.11b Variation in shape and extent of possum activity areas through the year ... 74
Table 2.12 Change in size of possum activity areas over time (means, ha) ... 76
Table 2.13 Den use by 7 frequently tracked possums ... 84
Table 2.14 Dens used on more than one occasion by the possums ... 84
Table 2.15 Radio-tagged Juvenile Possums - April 1991- August 1992 ... 91
Table 3.1 Number of possums sighted during specified time intervals in observation area 105 Table 3.2 Percentage of time devoted to behaviour categories by possums ... 107
Table 4.1 Parameters of observation runs which were suitable for analysis... 124
Table 4.2 Activity of possums during observation runs - % of total time in particular activity. ... 126
Table 4.3a Percentage of total observation time that cattle were in contact with possums during Week 1* ... 127
Table 4.3b Percentage of total observation time that cattle were in contact with possums during Week 2 ... 127
Table 4.4 Interest shown by the cattle towards the possum ... 129
Table 4.5a Reaction of cattle in relation to activity status of possum - Week-1... 130
Table 4.5b Reaction of cattle in relation to activity status of possum - Week-2... 130
Table 4.6 Chi square test of week of exposure to possum versus risk to cattle ... 135
Table 4.7 Unweighted logistic regression of ‘risk’ to cattle ... 136
Table 5.1 Percentage of Plant types in possum diet each month... 153
Table A6.1 Comparison between trap-revealed and radio triangulation determined home range (compared over the same time period) for each possum. ... 182
Table A6.2 Trap revealed data for the 12 month period April 1991-April 1992, including all trapping occasions for these possums (12, 3-5 day trapping occasions) ... 183
Table A6.3 Trap revealed data for the period April 1989-May 1993, including all trapping occasions for these possums (50, 3-5 day trapping occasions) ... 184
Table A6.4 Home ranges calculated from combining trapping and den-site locations for each possum. Data for 12 month (April 1991 to April 1992), and a maximum 50 month period (April 1989-May 1993), are presented. ... 185
LIST OF FIGURES
Figure 1.1 The Australian common brush-tailed possum ... 6
Figure 1.2 Possum in a live-catch trap ... 6
Figure 1.3 Location of study site in the lower part of the North Island... 8
Figure 1.4 Digital terrain model of the Castlepoint study area... 8
Figure 1.5 Northern part of the Castlepoint study area... 9
Figure 1.6 Southern part of the Castlepoint study area... 9
Figure 2.1 Antennae used for mobile radio-tracking (from Kenward, 1987) ... 34
Figure 2.2 Radio receiver and compass rose inside hut... 41
Figure 2.3 Radio triangulation site... 41
Figure 2.4 Example of error polygon and error arcs... 46
Figure 2.5 Distribution of Bearing Errors - Initial test ... 50
Figure 2.6 Distribution of Bearing Errors - Second Test... 50
Figure 2.7 Error Polygons from Radio Triangulation Error Measurements ... 53
Figure 2.8 Graph showing time of year at which animals were tracked... 59
Figure 2.9a Possum location distribution overlaid with bush cover... 64
Figure 2.9b Common den-site areas of radio-tracked possums ... 64
Figure 2.10 Cattle range areas derived from all available triangulated locations... 67
Figure 2.11 Distribution of cattle radio locations through the area at night... 68
Figure 2.12 Distribution of cattle radio locations through the area during daylight ... 68
Figure 2.13 Number of radio locations in relation to home range size... 71
Figure 2.14a Home ranges of possums numbered 2810, 2901, 2917 and 3662 ... 72
Figure 2.14b Home ranges of possums numbered 2992, 3644 and 3724 ... 72
Figure 2.15 Shifts in centres of activity of 7 possums... 75
Figure 2.16 Distribution of radio triangulation locations for all possums tracked... 77
Figure 2.17 Distribution of radio triangulation locations of 7 frequently tracked possums... 77
Figure 2.18 Common use of habitat by Possums and Cattle (includes all possums)... 79
Figure 2.19 Common use of habitat by Possums and Cattle (7 possums only) ... 79
Figure 2.20a Surface of study site indicating home ranges of 4 possums and their den sites... 85
Figure 2.20b Surface of study site indicating home ranges of 3 possums and their den sites... 85
Figure 2.21a Enlarged view of home range and den site locations of possum 2810... 86
Figure 2.21b Enlarged view of home range and den site locations of possum 2901 ... 86
Figure 2.21c Enlarged view of home range and den site locations of possum 2917... 87
Figure 2.21d Enlarged view of home range and den site locations of possum 2992... 87
Figure 2.21e Enlarged view of home range and den site locations of possum 3644... 88
Figure 2.21f Enlarged view of home range and den site locations of possum 3662... 88
Figure 2.21g Enlarged view of home range and den site locations of possum 3724... 89
Figure 2.22 Map of Waio Station and dispersal from study site ... 92
Figure 3.1 Main observation area - from above southern boundary... 103
Figure 3.2 View of observation tower and area - looking east ... 103
Figure 4.1 Contact between cattle and possums during observation periods ... 128
Figure 4.2 Change in levels of interest shown by cattle towards possum ... 129
Figure 4.3a Week 1: Possum activity versus cattle behaviour ... 131
Figure 4.3b Week 2: Possum activity versus cattle behaviour ... 131
Figure 4.4a Contact between steers and possum - Week-1 Totals ... 133
Figure 4.4b Contact between steers and possum - Week-2 Totals ... 133
Figure 4.5a Degree of ‘risk’ to steers by observation period - Week-1... 134
Figure 4.5b Degree of ‘risk’ to steers by observation period - Week-2... 134
Figure 4.6 Group of cattle showing interest in sedated possum ... 138
Figure 4.7 Steer sniffing and nudging possum ... 138
Figure 4.8 Steer licking possum... 139
Figure 4.9 Interaction between steer and possum at night... 139
Figure 5.1 Proportions of plant types found in faeces- all samples ... 152
Figure 5.2 Proportions of plant types in faecal samples from each area ... 152
Figure 5.3 Plant type distribution between sample areas... 153
Figure 5.4a Proportion of plant types in samples by season - East area... 154
Figure 5.4b Proportion of plant types in samples by season - West area ... 154
Figure 5.5a Seasonal distribution of individual plant types - East area ... 155
Figure 5.5b Seasonal distribution of individual plant types - West area ... 155
Figure 5.6a Proportions of plant types in samples in relation to the possum breeding cycle - East area... 156
Figure 5.6b Proportions of plant types in samples in relation to the possum breeding cycle - West area... 156
Figure 5.7a Seasonal distribution of individual plant types in relation to possum breeding cycle - East area... 157
Figure 5.7b Seasonal distribution of individual plant types in relation to possum breeding cycle - West area ... 157
Figure 5.8 Damage to mingi-mingi as a result of possum bite and scratch marking ... 159
Figure A1.1 Reception patterns for various aerial configurations... 173
Figure A2.1 Illustrations of Convex and Concave home range types. From White and Garrott (1990)... 174
Figure A2.2 Illustrations of the Harmonic Mean home range ... 175
Figure A3.1 Castlepoint study area - bush regions, fencelines and trapgrid ... 176
Figure A4.1 Bush and pasture around the Observation tower ... 177
Figure A5.1 Error polygons in relation to survey locations ... 178
Figure A6.1 Radio triangulation versus trap-revealed home range comparison ... 186
Chapter 1
Introduction and
Project Outline
1.1. INTRODUCTION
1.1.1. PROJECT BACKGROUND
This work is part of a 5-year longitudinal study of the epidemiology of bovine tuberculosis in a feral possum population in contact with cattle. The study is a cooperative project between the New Zealand Ministry of Agriculture, Animal Health Board, and Massey University.
The objectives of the longitudinal study are:
(i) to describe the epidemiology and effects of bovine tuberculosis in a mixed possum and cattle population
(ii) determine the significance of each of the mechanisms of transmission of tuberculosis within possum populations, and between possums and cattle
(iii) to study the behavioural ecology of possums and cattle on the study site to determine behavioural factors in both species which facilitate transmission of infection within, and between, the two species
The studies reported in this thesis address the third objective, principally using radio triangulation techniques and direct behavioural observations.
The longitudinal project began in 1989, after confirmation that possums in the area were infected with tuberculosis (TB). A sequential capture-recapture study was developed, and for 3 nights each month possums are caught, and measurements are taken on qualifying animals in accordance with the study protocol. Each animal is uniquely identified with metal ear tags and then released. There is an estimated population of 145 possums in the study area (D. Pfeiffer, pers.comm.), almost all of which den on the northern side of the paddock. Approximately 15-20 beef cattle have been grazed in the paddock throughout the project, although were restricted to the southern half of the paddock during most of this study. Every 3 months the steers are tested for tuberculosis, using a single intradermal skin test. Additional studies are run by other members of the research group in conjunction with the routine trapping and aim to examine other aspects of tuberculosis in the area.
1.1.2. Tuberculosis, and Tuberculous Possums in New Zealand
The Australian Common Brushtail possum (Trichosurus vulpecula Kerr 1792)1, was first introduced into New Zealand about 1840 (Pracy, 1962), and has since become both abundant, and a pest.
Originally introduced for its potential as a fur producer, the possum has since spread virtually the length and breadth of the country. The possum adapted extremely well to New Zealand conditions, and can be found from sea-level through to high altitude forest. The highest concentrations of possums are found along the bush-pasture margins, where pasture species are a significant proportion of their diets (Coleman et al., 1980; Clout and Gaze, 1984; Harvie, 1973). The possum
1All future occurrences of ‘possum’ in the text refer to this species.
utilises a wide variety of food types, but particular species are often preferentially eaten. In many areas this has led to the death of native species due to defoliation, and possums compete with native birds for foods from some plant species (Fitzgerald, 1981 and 1984b).
The realisation of the damage possums cause to native bush and wildlife has led to an increased effort towards their control, and ongoing studies to quantify both the significance of the problem and effectiveness of control methods. However, the discovery in 1967 of a tuberculous possum, and subsequent evidence of links between tuberculosis in possum populations and in co-existing cattle herds, has turned the problem into one of national significance.
Tuberculosis (formerly called consumption or phthisis) has been recognised for many centuries, and probably existed in classical times. It was possibly introduced into Britain by Roman invaders, but was unlikely to have been a serious disease where few animals were kept, or they are grazed in an extensive system. As small villages began to coalesce into ever larger towns, and cattle began to be kept together in larger numbers - particularly when they were housed through much of the year - communicable diseases were likely to become a cause of significant wastage. Diseased cattle were common in dairies supplying milk to towns in 19th century Europe including Britain - in 1847 Hunting found 20% of 4,000 cattle in Durham affected with tuberculosis (Francis, 1958). It was not until the 1870s that the importance of tuberculosis in cattle was recognised, and by this time the disease was widespread.
Cattle were first introduced into New Zealand in about 1840 from England, and bovine tuberculosis arrived with one or more of the early importations. It was not formally subjected to a control program in New Zealand until 100 years later, when a voluntary ‘test and slaughter’ program for dairy herds supplying town milk was initiated. Factory supply herds were included in 1958, and the scheme became mandatory for both in 1961. Beef cattle were compulsorily tested from 1971, but not until 1977 were all cattle in New Zealand under test or surveillance (Animal Health Division of M.A.F., 1986). There were significant reductions in the incidence of infected animals as the scheme progressed, but though the number of infected herds dropped initially, it then stabilised. In several areas of the country, notably the west coast of the South Island, and in the southeast of the North Island, a significant number of herds remained infected, or were reinfected. Stricter testing and control measures in such areas failed to resolve the problem. Epidemiological studies indicated that in these areas the disease was endemic in a species other than cattle. After the initial finding of an infected possum on the west coast of the South Island in 1967, surveys of possums on farms with TB problems found local prevalences of tuberculosis greater than 10%, and these findings were repeated in other areas (Livingstone, 1988). Possum control programs began in Buller South county in 1972, and there was a rapid drop in the number of infected herds (from 95% of herds to 37%) over the next 2-3 years. Control operations, including trapping, poisoning by air and ground, and shooting, have since been the main methods used to reduce possum numbers in endemic TB areas. In conjunction with these programs, a policy of farm quarantine and movement control was introduced in 1977,
when it was realised that herds in previously non-endemic areas were becoming infected. It was thought the movement of cattle and deer from areas of endemic tuberculosis into other farming districts, was spreading the disease. In many of these new areas the possum populations, previously free of tuberculosis, were found to be infected.
There was a reduction in the number of herds on movement control through 1977 to 1981, but then began a slow, and continuing, upward trend, although the number of animals reacting to the tuberculin test has declined. It is now recognised that, in many areas, tuberculosis is maintained in the feral possum population. Furthermore, eradication of either the possums or the disease is unlikely in most of them (Animal Health Division of M.A.F., 1986). Present policy is to contain tuberculosis in endemic areas - those with a persistent TB problem - and eradicate it from others. Control in endemic areas attempts to restrict the spread of possums from foci of infection, often by the use of low possum density ‘buffer zones’ which are regularly poisoned, and surveillance testing of cattle in the area as an indicator of infection. Each endemic area has strategies developed for its particular circumstances, but these control measures have had variable success. The number of endemic areas in New Zealand has steadily increased over the past 12 years, several of them as a result of possum populations being infected by tuberculous cattle or deer brought into the area. Buffer zones have not always proved effective, and infected possum populations have been found moving down catchments at a rate of 4-8 km/year (Livingstone, 1988). In many areas possum control operations have proven successful - in the short term. Populations are reduced to relatively low numbers and cattle reactor rates decline, but, without continued pressure on the area, repopulation occurs over the next 2-3 years, and cattle reactor rates subsequently increase. Many of the techniques used in the past for controlling the spread of tuberculosis through wildlife populations are neither practical, effective, or economic, in present situations. There is also a need to consider welfare implications of the methods.
The increasing national recognition of the problem has led to an increasing research effort aimed towards an understanding of the disease.
The biology, ecology and behaviour of possums has been well researched in New Zealand and Australia. The interactions between possums, cattle, deer, Mycobacterium bovis, and other factors associated with the spread, and maintenance, of tuberculosis in New Zealand have only recently begun to be addressed.
1.1.3. Study Aims
The main aim of this work is to investigate behavioural factors in possums and cattle that may facilitate transmission of tuberculosis between and within these species. The study site is part of a commercial sheep and beef cattle farm in the Wairarapa, with a long term problem of TB in the cattle. Investigations include:
• observational studies of possum and cattle behaviour - specifically interactions between animals, particularly on the bush-pasture margin.
• investigations using radio-telemetry to determine
(i) activity areas and use of the site, by both possums and cattle and
(ii) the extent and direction of movement of any possums dispersing from the site.
The following chapters describe results from radio-triangulation studies, behavioural observations of natural activity in the study area, and an experimental study of interactions between sedated possums and cattle in a confined area. Final summaries and conclusions follow.
Figure 1.1 The Australian common brush-tailed possum
Figure 1.2 Possum in a live-catch trap Note metal identification tags in each ear
1.2. STUDY AREA
1.2.1. Geography and Location
The study was carried out on Waio Station near Castlepoint (40º51'S, 176º14'E), on the Wairarapa coastline 63 km east of Masterton (Figure 1.3). The study site covers approximately 75% of the 40 ha ‘Backdrop’ paddock, over half of which is grazed by cattle. In early 1991 a fence was constructed down the length of the paddock, restricting cattle to the southern, more developed, side. Possums den predominantly on the northern slopes, although they feed on both sides of the fence. The bowl shaped paddock varies from 60 to 270 metres above sea level (a.s.l), with an opening to the west (Figures 1.5 and 1.6). Several water courses combine to form a main creek draining the valley. The northern ridgeline drops steeply down to the valley floor, and there are several spurs off this ridge.
The southern side of the paddock, which slopes more gently, consists mainly of stands of manuka, although 30-40% of the area has been cleared and there are also large amounts of pasture under the manuka trees. The northern part of the paddock is predominantly covered with mingimingi (Leucopogon fasciculatus), manuka (Leptospermum scoparium) and gorse (Ulex europaeus), with small remnants of broadleaf forest. A line of pine trees (Pinus radiata) extends part-way along the northern ridgeline, and there are 5-8 large, individual pine trees scattered through the study area. A steep sided gully (punga gully) drops from the north ridge down to the mouth of the valley and contains the richest variety of native bush in the area. Punga gully lies on the periphery of the trapping areas, and is fenced off. There are signs indicating that a large number of possums are active in the gully, and possums denning in this area are often caught in nearby traps.
1.2.2. Climate
Coastal Wairarapa is subject to a wide variety of weather conditions, from drought to heavy rain and snow, with considerable variation along the coast. Rainfall can vary from less than 500 mm to more than 2500 mm per year, mainly falling from June to October. The region around the study area is subject to more rain, and considerably more wind, than the average along the coast. Rainfall varies between 0 and 300 mm per month in summer, to 1800 mm or more in winter months. Cold southerly storms are common in autumn and winter, though may occur at any time of the year. Strong, and persistent, north-westerly winds are the most significant climatic feature in the area. These winds are strongest between September and January and may continue for 2-3 days, with an average speed of 50-65 km/hr. Gusts of more than 100 km/hr are frequently recorded. Minimum and maximum air temperatures range between 1-15°C in winter, and 9-28°C in Summer (Meteorological data from Castlepoint lighthouse). There are occasional frosts on the valley floor and in sheltered areas.
Figure 1.3 Location of study site in the lower part of the North Island
Wellington Castlepoint Palmerston
North
Study Site
Masterton To Palmerston North
Wellington
Castlepoint
Featherston
Figure 1.4 Digital terrain model of the Castlepoint study area
Several of the figures in following chapters are based on this model. A 3-dimensional surface is obtained by laying a ‘fishnet’ surface over the terrain model and overlaying data, such as home ranges, on the surface.
Figure 1.5 Northern part of the Castlepoint study area
The study area begins at the top end of the flat grassed area (lower centre of photo)
Figure 1.6 Southern part of the Castlepoint study area
The dividing fence follows the path of the cleared, zig-zag, line down the centre of the photo.
1.3. SOCIAL ORGANISATION AND BEHAVIOUR OF THE POSSUM
1.3.1. Introduction
Possums are small, nocturnal, predominantly arboreal, herbivorous marsupials native to Australia, and introduced by Europeans to New Zealand. There are various genera within the superfamily Phalangeroidea; but of the species within the superfamily, the common brush-tail possum (Trichosurus vulpecula Kerr 1792) is the most widely spread and successful. They species has adapted to a wide variety of habitats over a substantial proportion of Australia and almost all of New Zealand. South-east Australia, Tasmania, and New Zealand have the highest population densities and presumably represent the most favourable environments (Kerle, 1984). Possums may be found from sea-level to sub-alpine woodlands. Habitats range from dense podocarp and hardwood forest, through open eucalypt forests, to scrublands, tree-lined watercourses and other protected sites in the midst of open pasturelands and semi-arid areas, and buildings in both town and country. The major proportion of their diet consists of 4-6 leaf types (mainly eucalyptus in Australia), although they eat a variety of other foods including fruits of many trees, pollen, buds and flowers from a variety of species, grasses and clovers and possibly fly larvae and moths in small amounts (Fitzgerald, 1982).
The possum has been extensively studied. Morgan and Sinclair (1983) list 808 references, from 1835 to 1982, covering all aspects of the species life history - 39 of these having ‘behaviour’ as a key word. Many more reports and papers, particularly from New Zealand, have been published since this bibliography was compiled. Many of these papers describe capture-recapture projects and radio- telemetry results from various parts of the country, and relate these findings to the prevention and control of tuberculosis in New Zealand’s possum population. However, few of the studies were of tuberculous possum populations or included direct observation of the animals. The often dense vegetation which provides much of the natural habitat of the possum in New Zealand, and its nocturnal nature, make it a difficult animal to observe in the wild. Limited numbers of captive studies have been undertaken, but these often lack relevance to the field situation (Biggins and Overstreet, 1978; Kean, 1967; Winter, 1976).
1.3.2. Activity Patterns
Possums are nocturnal creatures although sick and/or starving animals have been seen abroad in daylight (Cowan, 1990b, pg 83; Ward, 1978; T.Harris, pers.comm.). They become active in their dens 1-2 hours before dusk and generally emerge within an hour after sunset (Winter, 1976; Ward, 1978; MacLennan, 1984). Both sexes become active about the same time at dusk, although females return to their dens earlier than males, and there is more variability in the time possums return to dens at dawn (Winter, 1976). Most possums are in their daytime dens approximately an hour before dawn (Jolly, 1976b; Green and Coleman, 1986; Ward, 1978; Brockie et al., 1987). Variation in habitat type does not affect these basic patterns, which are similar for eucalyptus forests through to dense New Zealand hardwood forest. Possums spend the first 1-2 hours after emergence in the area of the dens, sitting and grooming (Ward, 1978; MacLennan, 1984) until moving away to begin
feeding. The greatest distance from the den is reached 4-6 hours after dusk. Winter's (1976) study indicated a bimodal feeding peak, either side of a 12-1 am rest period, as did Ward (1978), while MacLennan saw no clear peak. All 3 authors noted feeding preferences, some animals returning to the same tree to feed over several nights. The majority of possums begin returning to their dens 2-3 hours before dawn.
The proportion of time possums spend on various activities through the night have been studied by MacLennan (1984) and Winter (1976), although other researchers have commented on an individual possum’s particular activity at certain times of the night (Jolly, 1976b). Winter and MacLennan found possums spend more than 40% of the night immobile, generally in trees, with females stationary more often than males, and for longer periods2. Periods of immobility varied from minutes to more than 1½ hours. Possums spent 30% of their time travelling in both studies, and 16-25% of their time feeding. The remainder of the time is taken up with grooming and social interactions.
There are differences in time allocations which possums spend on activities when on the ground, compared with when they are in trees. Males and females both spent a similar amount of time (8.9%) on the ground in MacLennan's study, however the females spent significantly less time feeding but more time sitting than males. In comparison, Winter's study found males spent 18% of their time on the ground and females 9%, with both sexes feeding for about half of that time. In both studies possums spent proportionately less time sitting when on the ground than in trees. Ward (personal communication with MacLennan, 1984) found possums in a native podocarp forest in New Zealand spent 46.1% of their time inactive, 11% feeding, 22.3% grooming and 20.6% otherwise active. The latter possums spent considerably more time grooming than either of the Australian studies.
There are no reports of activity patterns of possums from a pastoral habitat in New Zealand. Pasture species are an important part of the diet in such habitat (Gilmore, 1967; Harvie, 1973). Possums are selective browsers rather than grazers and it is likely that behavioural patterns described for an arboreal environment are modified when feeding extensively on pasture - particularly the amount of time spent in the trees. A small group of trees could provide sufficient food for a possum, but if grazing the animal may need to move over a large area of pasture to obtain a similar food intake.
Pasture species are also more easily digested than leaves, and the amount of time spent immobile may decrease.
Possums spent less than 2% of their time out of the den interacting in Winter's study and less than 1% in MacLennan's. Some individuals spent considerably more time than this on certain nights, particularly when courting a female. However, few interactions were observed overall. Winter indicates the shortcomings of behavioural observations from a distance at night, and MacLennan doesn't comment on his subjects’ interactions. The majority of interactions were sexually related,
2 Winter: Males 36%, Females 44%; MacLennan: Males 50%, Females 51% of the time.
usually as a result of aggression by the female in response to a male’s advances (Winter, 1976). Jolly (1976b) reported a wider variety of aggressive encounters, but could not provide as much contextual data as Winter.
The effect of weather conditions on activity patterns has been poorly documented. Jolly (1976b), Ward (1978) and Winter (1976) found that heavy rain caused the possums to seek shelter, and delayed the emergence of possums from their dens at dusk. Strong winds did not appear to affect their activity, although heavy frosts caused them to den early (Brockie et al., 1987). In contrast, during observations carried out by the writer, it was found that few possums were out on open pasture during high winds. Jolly (1976b) cites Bamford in suggesting that maximum temperature over the previous 12 hours, and relative humidity at the time affected the number of possums seen.
1.3.3. Home Range Size and Population Density
Population densities vary widely over different habitat types. Winter (1976) determined a population density of 2.14-2.19 per ha over 3 years on a 17 ha site near Brisbane, whilst Dunnet (1956), on a 60 ha pastoral site near Canberra, found an average 0.5 possums per ha (range 0.2-2.2). Densities of 0.31/ha in 80 year old Tasmanian forest, and 0.55/ha in regenerating 3 year old bush were described by Hocking (1981). In contrast, possum densities in New Zealand are many times higher and more variable. Batchelor et al. (1967) estimated numbers in 6 broadleaf/podocarp forests and these ranged from 9.4-24.2 possums per ha. Crawley (1973) found 6-10 possums per ha using capture-recapture methods in a 14 ha area of lowland forest in the lower North Island, while Coleman et al. (1980) trapped a Westland population to extinction and calculated an average of 10.7 per ha. The Westland study was in a podocarp/mixed hardwood forest which bordered on pasture, with an altitudinal gradient of 1000 m. Densities ranged from a high of 25.4 per ha at the forest/pasture margin (250 m a.s.l), down to 1.9 possums per ha at 1200 m a.s.l (alpine grasslands). Densities of possums on developed farmland were shown by Brockie et al. (1987) to vary according to habitat. They found 0.12 possums per ha on open pasture, 6.4 per ha along tree-lined streams, and 8.4 per ha in a swamp.
Jolly (1976a and b) found similar figures for a bush/pasture habitat on Banks Peninsula with 1.1-10.5 possums per ha.
1.3.3.1. Home range
Home ranges have been calculated in most studies of possum populations. The data collection methods used vary from capture-recapture records and direct observation, to radio telemetry. The areas trapped and patrolled in these studies also vary, as does the habitat type. ‘Average’ home ranges are only meaningful if applied to similar environments - and are also complicated by varying interpretations of the term. Trends are seen, however, in that possums living in densely forested areas have smaller ranges than those from more open environments (pasture-forest margins) and males have larger ranges than females.
Green (1984) calculates overall male and female home range size for Australian possums as 5.4 ha and 2.4 ha respectively, and range length as 394 m for males and 261 m for females. However, there is a considerable variation in the observations on which the overall figures are based, varying from 3.7-7.4 ha for males, and from 1.0-4.7 ha for females3. In comparison, Green cites average home ranges for New Zealand possums as being 1.9 ha for males, and 1.3 ha for females, with range lengths of 295 m and 243 m respectively.
Studies from which the Australian values are calculated have used similar methods - mark-recapture studies and minimum area methods for calculating home ranges - and were mainly from open eucalypt forests. However, New Zealand data cited by Green has come from a variety of study types and habitats. Mark-recapture and radio-tracking data are equally represented, and environments vary from dense native forest, through to scrub/pasture mixes. Radio-tracking studies give consistently larger home ranges than do trap revealed range, and possums from habitats which have pastoral margins have larger ranges than purely forested areas. Male possums living in forest only areas had home ranges between 0.5 and 3.6 ha, females 0.3-3.8 ha. Possums with access to pasture had ranges between 2.4-65 ha (males) and 0.3-45.8 ha for females, with corresponding increases in range length.
A recent study, not included in Green’s 1984 paper, found annual home ranges for 11 adult possums on farmland varied between 2 and 105 ha (mean 30 ha) (Brockie, 1991).
Home range size will also be influenced by the period of observation (quoted figures are 6-12 months data), seasonal forays to food supplies or during the mating season, and the distribution of resources within the area (Ward, 1978; Jolly, 1976b; Green and Coleman, 1986). The effect of the latter is seen in Green and Coleman's study where possums moved up to 1.5 km down a mountainside to feed on pasture.
1.3.4. Territorial and Denning Behaviour
Home ranges within a possum population generally overlap between, and within, sexes (Green, 1984; Brockie et al., 1987). Early studies indicated ranges were defended and mutually exclusive, at least those of males (Dunnet, 1956 and 1964), but re-analysis of this data, and further work has determined that territoriality is based on mutual avoidance of co-dominants, and defence, if any, is restricted to den sites (Biggins, 1979; Green, 1984; Winter, 1976; Brockie et al., 1987). Possums in captivity form a male dominant hierarchy, although it is only the highest and lowest ranking individuals that consistently win or lose fights (Kean, 1967; Biggins and Overstreet, 1978). Middle- order possums were in a more flexible position, and the winner of encounters between these animals could not be reliably predicted. Winter (1976) suggests that dominance rights, or privileges, in wild populations are learnt through encounters between individuals rather than defence of an area.
3Individual possum home ranges vary from 0.3-20.6 ha for males, and 0.1-6.5 ha for females
In Australia Winter (1976) observed many interactions between individuals at den sites. Females dominated males over the choice of den sites on 87-100% of occasions. However, males were able to den in other locations in the same tree, whilst other females (apart from joeys) were seldom tolerated. Winter found that where dens were well spread the home ranges of established females did not overlap. Where dens were closer no animal had exclusive use of an area, except for a small region in the vicinity of the den tree. Male ranges were far less exclusive, though some such areas did exist. Extensive forays during the mating season led to both large male home ranges, and the overlaps between individuals of both sexes. None of the New Zealand studies have described territorial behaviour - that is, active defence of an area (Crawley, 1973; Ward, 1978; Brockie et al., 1987; Ward, 1986; Jolly, 1976b). These writers conclude that there is considerable overlap of home ranges of both sexes, and that most animals, once established, remain faithful to an area for most of their lives. Several workers, including Winter, suggest that in the more complex forests of New Zealand individuals cannot defend a 3-dimensional area (Green, 1984).
1.3.4.1. Possum Den Sites
Possums in Australian Eucalyptus woodland tend to use 2-5 dens per year (Winter, 1976), while New Zealand animals use 1-15, with an average of 8-10 (Brockie et al., 1987; Cowan, 1989; Green and Coleman, 1987). There is an apparent excess of den sites in New Zealand habitat compared with Australian habitat (Clout, 1977; Triggs, 1982; Green and Coleman, 1987), and populations here are more likely to be limited by food resources rather than the availability of dens (Green, 1984). The number of dens frequented by a possum increases with the period of observation (Ward, 1984).
Despite the large number of dens used in a year, many possums have been found to use a few of these locations preferentially (Cowan, 1989). Winter (1976) made similar findings in Australia.
The possum is able to adapt to a variety of den types, but prefers to den high above the ground whenever possible (Cowan, 1989). Denning habits of possums in the Orongorongo forest were similar to Australia, where most animals live in holes in the branches and stems of large trees (Cowan, 1989). Cowan found 92% of his possums denning in trees, which were generally large with significant numbers of epiphytes growing on them. Some possums denned under fallen logs, in clumps of flax, and under gorse bushes. These findings contrast with those of Green and Coleman (1987) who found 73% of their den entrances at or below ground level. In this Westland habitat large trees were uncommon, clumps of epiphytes were rare, and 40% of possum dens were less than 2 m above ground. It is unclear from the paper what proportion of actual den cavities, as distinct from den entrances, were at ground level, but it is considerably higher than at the Orongorongo site. In a bush/pasture habitat possums generally den at ground level. Common den sites include protected locations within or under clumps of flax or grass, under gorse bushes, in haybarns and ceilings, rabbit holes or in the tops of tree ferns (Jolly, 1973; Ward et al., 1986). Cowan (1989) and Ward (1978) found many of the den sites were on the periphery of the area over which the possum ranged, either the previous night, or the following night.
1.3.5. Olfactory Communication
Jones (1921, cited in Winter, 1976, pg 92) concluded that possums have a poorly developed sense of smell, “...as is not unnatural in an arboreal species...” and seems to be of little importance in obtaining food or avoiding enemies. Since this study, several sets of scent glands have been described in the possum, and the significance of olfactory behaviour in possums described. Studies have indicated that each set of glands has a particular significance (Winter, 1976; Biggins, 1979).The most obvious glands are a line down the ventral part of the chest (sternal gland) which stains the fur here a light brown. Other glands include 2 pairs of paracloacals; labial or chin glands; and pouch glands in the female. Deposition of urine on substrate may also be used as a marker. A significant difference between olfactory communication, and visual or auditory forms, is that scent deposits or marks can potentially remain for some time after the initiating animal has left the area (Biggins, 1979). Thus, information such as identity of the marker, sex, sexual status, or territory boundary markers, can be left with relatively little input, and no continuing expenditure. Similarly, other individuals moving through the area can determine the status of resident animals, and in turn, leave their ‘mark’. In solitary, nocturnal species, such as the possum, visual and tactile communication is limited and often olfactory and auditory communication is the most common social interaction (Winter, 1976; Biggins, 1979).
1.3.5.1. Pouch Glands
The pouch glands begin secreting when females are sexually active, and continue during lactation (Winter, 1976). The joey becomes stained a reddish-brown, as does the fur around the pouch opening. Biggins (1984) suggests that this may assist in individual recognition between mother and young. As the joey spends less time in the pouch the staining disappears, and auditory and visual cues may become more significant. Kean (1967) found some males followed lactating females, and suggested the pouch may be the source of the attractant. Winter only observed this after the joey had left the pouch. He suggested it corresponded to a second, spring, oestrous cycle and the male was responding to a mixture of scent cues.
1.3.5.2. Paracloacal Glands
Paracloacal glands are of 2 types: (i) a holocrine cell gland which releases entire cells and (ii) an apocrine gland producing a light coloured oily liquid. The cellular secretion is continuously being produced and is voided with the urine, or as a coating on the faeces (Biggins, 1984). There are marked differences in the number of cells produced by immature, adult male, and female possums (Winter, 1976). Kean (1967) suggests the durability of the cells, and sexual differences give the urine a persistent odour, and specific meaning. The second, apocrine, secretion is the strongest (to people) odour released by possums and is released at specific times. It has often been associated with stress or submissive responses (Thompson and Pears, 1962; cited in Winter, 1976). The secretion may also be released when the possum urinates, either dripping from the hairs around the cloaca, or smeared onto substrate (Winter, 1976). The accumulation of oil around the cloaca is probably
transferred in part when a possum sits on a branch, and would make up part of the attraction Winter observed for male possums when they closely investigated locations on branches recently vacated by females, in particular. Kean (1967) observed few instances of its deposition, and had not encountered it in the wild. He did find it on several occasions deposited at the entrance to den boxes occupied by animals known to feel “insecure” and, as with Winter, found it was often released if a possum struggled when being handled.
1.3.5.3. Urine Marking
Urine marking has been noted by Winter (1976), Biggins (1979), and Kean (1967). All writers describe possums moving their anal area along close to the ground with a sinuous motion of their tail. In some cases the urine simply dripped from the vibrissae around the cloaca, in others a distinct trail was left along a branch. Kean found it was regularly induced when transferring possums from spacious cages to restricted areas, and felt it engendered confidence. Winter’s observations were more limited, but he found urine marking associated with 2 distinct situations. The first was generally seen when males were in association with oestrous females, and urine was deposited as they followed the female in the trees. The second was secretions by females placed in front of their trailing joeys. At this stage the female was keeping the joey at a minimum distance of 1 m (except when denning), and Winter assessed the mother-joey bond as breaking down.
1.3.5.4. Chin Glands
The sebaceous chin glands are situated in small areas on the inside of the anterior part of the upper and lower lips. They are often associated with chesting (use of the sternal glands) (Winter, 1976), and this often makes it difficult to determine their use. Chinning can be distinguished from chesting when the sternum of the possum remains off the substrate being marked. When chinning the lower lips of the possum are often drawn back, and saliva may also be deposited in addition to glandular material. Winter described similar objects being marked by chinning as for chesting - tufts of grass, tree bases, ends of broken branches and den rims. However, differences were seen in that these objects tended to be smaller, and the activity less vigorous than chesting per se.
1.3.5.5. Sternal Glands
The sternal glands are most commonly used to mark objects. Chesting of the base of trees is seen most often, but branches, the ground, and rims of dens are also marked. When chesting a tree base the possum spreads the forelegs wide around the tree, pulls itself towards the tree, and then rubs the sternum forward along the trunk. The possum then lifts its chest off the tree and returns to the starting position. This action may be repeated several times (Winter, 1976; Jolly, 1976b). The possum may sniff the surface between each stroke, particularly when chesting the ground. Chinning may also be associated with the chesting movement.
In Winter's study chesting was almost always associated with the actual presence nearby, or previous presence of, a possum at that location. It was exhibited 8 times more often by males than females,
and two-thirds of the observations were in the presence of an oestrous female (Winter, 1976). Male possums also frequently mark trees as they move across the ground. There was no apparent pattern, either in number of trees marked, or extent of any marking in Winter's study, but the trees almost invariably had a recent history of possum use. Biggins (1979) noted that dominant possums marked more often. When a possum emerges from its den it may also chest the immediate vicinity, more vigorously if another possum is in the same tree. Marking by females was less vigorous and less frequently seen than that by males. The few observations noted were of chesting den trees and when other possums, usually female, were in the vicinity.
Winter (1976) suggests that the social function of chesting and chinning is to advertise the presence of the marker to a potential rival for a limited resource such as a den, or a mate. Most marking in this context occurs in locations where another male would be exposed to the scent, eg. at the base of a tree or where a female had been sitting. Biggins' 1979 study determined that male possums, exposed to scents of other males, could easily distinguish between individuals. When possums mark their den tree, or trees along the routes they follow to feeding areas, they are both distributing their own scent at focal points in the area, and informing other individuals of their presence. The scent will be freshest and most dense in frequently used locations. Scent marking of male possums may be primarily related to establishing, and maintaining, areas in which the resident male is dominant over intruders (Biggins, 1979). Self-confident resident males would have an increased chance of mating with oestrous females in that area, and also preventing establishment of other males, eg. immatures, in that vicinity.
Joeys may learn from associating scent marks deposited by their mothers during the breakdown of the mother-joey bond with possible subsequent agonistic interaction if they approach too closely (Winter, 1976; Jolly, 1976b; Biggins, 1979). Thus, scent marking reinforces, and maintains, a system of mutual avoidance between individuals as each animal assesses the freshness, density, and possibly the depositor’s identity, of markings in its immediate vicinity.
1.3.6. Vocalisations
The possum has a wide variety of vocalisations, ranging from quiet chatters and clicks, to high velocity screeches. There are a number of discrete calls, and a series of graded vocalisations which are considered to be a continuum from a low to high intensity threat response (Winter, 1976;
Biggins, 1984; Kean, 1967). Winter divided the calls into 2 groups: (i) those aimed at causing the withdrawal of another individual, generally involving some degree of threat, and (ii) calls associated with appeasement or encouraging approach from another, eg. males courting females or mother-joey interactions. Many of the group (ii) calls are quiet and not heard beyond a few metres.
The vocal behaviour of possums, and particularly their contextual associations, has not been extensively researched (Biggins, 1984). Nocturnal species are difficult to study in the wild, and captive specimens are often restricted in their behavioural repertoire - so much so that early
investigators reported an absence of vocalisations from many species (Hediger 1958, cited in Biggins, 1984). Winter observed the change in intensity of ‘screeches’ with the increase in postural threat exhibited by possums. As a possum moved from a four-legged stance, through three-legged with raised paw, to a full bipedal stance with widespread forepaws and open mouth, the screech developed from a hiss and growl, through to an ear-splitting outburst. A feature of this form of interaction is the closeness of the protagonists, whether or not a true fight ensues, and the open- mouthed hissing and screeching. ‘Grunting’ was also frequently associated with face-to-face confrontations by Winter (1976, pg 61) and, in part, resembles a short chesty cough. The grunts were given in a variety of circumstances, mostly agonistic, and were often closely associated with screeches. Winter heard grunts from possums in both offensive and defensive situations suggesting the significance of the sound is dependant on the actions of the possums giving the grunt (Winter, 1976).
The distribution of calls through the night was described by Winter for the more vocal of the calls.
These were found to peak early in the evening, and late at night (as the possums returned to their dens). This is consistent with a solitary species in which very loud calls are used in the maintenance of a minimum distance between individuals, and the graded series of calls allows for the indication of a wide variety of motivational states (Winter, 1976).
1.3.7. Breeding and Reproduction
Female possums are dioestrous, almost always monovular and ovulation is spontaneous. The 2 cm long, hairless, blind, and virtually embryonic joey is born 15-18 days after conception, and crawls unaided to the pouch. Here it attaches to one of two teats for 4-5 months before venturing outside the pouch again. The major breeding season is from March to June, with a lesser season August to September or October. The pattern is, however, variable for reasons which have not been fully documented - although the bimodal peak is most common some areas have a single Autumn peak, whilst in others possums may breed all year round (Tyndale-Biscoe, 1955; Brockie, 1991; Green, 1984). Many females are sexually mature at 1 year old (Ward et al., 1986; Winter, 1976), although in some areas maturity is delayed until the second year (Crawley, 1973; Cowan, 1990b, pg 87). Greater than 90% of sexually mature females conceive and rear a young to 4-5 months of age. Mortality of young increases rapidly from this age as the joey has reduced protection from the mother’s pouch (Tyndale-Biscoe, 1955). Joeys become increasingly independent from 8-9 months of age, though some may remain with their mothers until 12-15 months (pers. obs. R. Jackson and B. Paterson), or return at dawn to share a den with the mother (Winter, 1976).
Male possums mature later than females and sexual maturity usually occurs during the second year of life. Bodyweights continue to increase for some months after the rate of weight increase in females levels off (Cowan, 1982; Winter, 1976). However, Broom (1898), Jones (1924), and Wodzicki (1950), (all cited in Tyndale-Biscoe, 1955) state that Trichosurus vulpecula is sexually mature at the end of 1 year. Tyndale-Biscoe (1955) concurs with these writers based on analysis of
the relationship of body weight and tibial ossification to fecundity, but does not link this to a specific age of maturity. There is a significant increase in the size of the testes at puberty, and a close correlation between the length of the testes and the presence of sperm (Cowan, 1982; Tyndale- Biscoe, 1955).
Dominant animals mate with most of the females (Winter, 1976). Male show little interest in anoestrous females, but will actively follow those in oestrus (Jolly, 1976b). Winter (1976) described a consort period of up to 40 days during which the female shows an increasing degree of tolerance towards the male, and which ceased no more than 1-2 days after mating. Female possums take a passive role in courtship and do not actively encourage the male. If the male approaches beyond a certain distance, she will threaten, and even fight, with the male generally retreating. During the consort period the male persistently follows the female. Although not denning together, in Winter's study consort males were found close to females soon after dark. The male sniffs at, and often scent marks (chesting and chinning) places where the female has been sitting. Males searching for an oestrous female used scent almost exclusively to find them, although post copulatory ‘chatter’ calls by males occasionally attracted other males (Winter, 1976). The female gradually becomes more receptive, and eventually allows the attendant male to attempt copulation, after which she again threatens possums approaching too closely (Winter, 1976). The consort relationship is not necessary for a fertile mating; Winter observed several aggressive matings as did Jolly (1976b). There are few other comprehensive reports of mating behaviour in the literature. Although the female is probably receptive to mating, the males in the latter observations had not spent much time with the female, and copulations often followed fights, and/or were accompanied by loud screeching.
Whether preceded by a consort period or not, the behaviour pattern of a mating attempt is similar (Winter, 1976; Jolly, 1976b). The male approaches the female, commonly head on and, ignoring threats which would normally result in retreat, climbs over the female’s head on to the back. He turns himself around as he positions his hindquarters over the female’s rear, and attempts intromission.
The receptive female is initially aggressive, though considerably less so than when anoestrous, and may screech. As the male settles himself fully on her back, there is less and less response from her as she resigns herself to the fact. The female dislodges the male shortly after copulation.
Winter (1976) has made the most extensive observations of mother-joey interactions and notes the relative indifference of marsupial mothers to their young (Kaufman, 1974 and Russell, 1973, cited in Winter, 1976). The female possum recognises the presence, and appears to know the location of her young, as indicated below - but, out of the den, the primary responsibility for maintaining contact with the female rests with the joey. After leaving the pouch the joey may ride on the mother’s back for another 2-3 months, although attempting to retreat to the pouch on occasions (Winter, 1976).
When the joey first leaves the pouch it clings to the mother's back as best it can. Within a few weeks the joey almost invariably assumes a longitudinal position, with head at the level of the mother's shoulders, whenever travelling. As the joey gains confidence it begins to spend more time off her
