Lincoln University Digital Thesis Copyright Statement The digital copy of this thesis is protected by the Copyright Act 1994 (New Zealand). This thesis may be consulted by you, provided you comply with the provisions of the Act and the following conditions of use: 


you will use the copy only for the purposes of research or private study you will recognise the author's right to be identified as the author of the thesis and due acknowledgement will be made to the author where appropriate you will obtain the author's permission before publishing any material from the thesis. Frontispiece. Thumbnail images of 3341 bitemark cast photographs from 467 possum-bitten
WaxTags in field trial used to make 504 identifications of 20 individuals.
Forensic Approaches to Monitoring and Individually Identifying
New Zealand Vertebrate Pests
A thesis
submitted in partial fulfilment
of the requirements for the Degree of
Doctor of Philosophy
at
Lincoln University
by
Keisuke Sakata
Lincoln University
2011
Abstract of a thesis submitted in partial fulfilment of the
requirements for the Degree of Doctor of Philosophy.
Forensic Approaches to Monitoring and Individually Identifying
New Zealand Vertebrate Pests
by
Keisuke Sakata
The brushtail possum (Trichosurus vulpecula) is New Zealand‘s number-one vertebrate pest
in terms of its ecological damage to native flora and fauna. One of the basic knowledges
needed to plan, perform and evaluate wildlife management inputs and outcomes is population
abundance. Although the most common method used in New Zealand for possum abundance
monitoring is the trap-catch method, the bait interference technique has advantageous
properties, including its minimal risk to non-target fauna, low operational cost, and ease of
use. Similar to the trap-catch method, animal abundance is inferred by an index of proportion
of baits interfered with. However, non-toxic bait can be interfered with by many individual
animals, and likewise a single individual can interfere with multiple baits, which is an
intrinsic drawback to the method, confounding the results due to an unknown probability of
multiple interference of bait.
This study assessed the issue of visits to a single bait by multiple species and individuals and
visits to multiple baits by a single individual by using forensic science approaches in
developing a novel method to identify the responsible species and individual animals from the
bitemarks that animals leave as interference.
The issue of identifying multiple species biting on bait was assessed by applying the bait to
captive animals of known species and defining the species from its bitemark‘s characteristics.
Possum and rodent bitemarks were defined by measuring single tooth-mark widths where
possums were 2.87 ± 0.51 mm, as opposed to 0.94 ± 0.28 mm for combined rat species and
0.46 ± 0.07 mm for mouse. Paired t-tests showed significant differences (p<0.05) between
ii
possum and rodent tooth mark widths. Qualitative description of striation- and impressiontype bitemarks also identified characteristic features of possums and rodents.
In order to assess the prevalence of multiple visits, a novel method to identify individual
animals was developed by applying the forensic firearm and toolmark examination principle
to microscopic features observed on the bitemarks. Trials using known simulated bitemarks,
blind test of random bitemarks, known ‗live‘ bitemarks from captive animals, and finally, a
random bitemark selection from known individuals which reached 89% precision and 99%
accuracy. This demonstrated that bitemarks can be used under controlled laboratory
conditions to reliably identify individual animals.
Field application of the above methods were conducted for 10 nights in paired blocks of 10 m
interval 12 × 12 grids, separated by 100 m. One block was treated with a food lure to follow
the most recent pest monitoring protocol. The number of interfered WaxTags® was
significantly higher (p<0.05) with food lure up to the third night, and showed no difference
afterward. By applying the novel method to the field, previously unknown information was
obtained: The mean and range of nightly activities recorded by the 20 identified individuals
were 2.5 tags (range 1–33) interfered with, 128.1 m (0–311) travelled distance and 0.11 ha (0–
2.8) in covered area. Up to three individuals bit one tag in one night, and WaxTags® were
repeatedly bitten for up to six nights at the same location. This study‘s novel individual
identification method has added another tool in wildlife management techniques.
Keywords: wildlife management, population monitoring, bait interference method, WaxTag®,
brushtail possum, Trichosurus vulpecula, individual identification, forensic science, bitemark,
toolmark
iii
Acknowledgements
I would like to express my huge gratitude to all of you great guys that allowed me to do this
study. Dr Shaun C. Ogilvie, my supervisor, supported me to get through some never-before
experienced difficulties, which was way beyond his duty as a supervisor. Dr Adrian M.
Paterson was also always encouraging, looking at the positive sides of everything. The two
supervisors were more than excellent in providing appropriate feedback in guiding my
research. Such exciting results would never have been produced if it weren‘t for you both.
Malcolm Thomas, Director of Pest Control Research, was THE best source of information on
WaxTag® and its use, besides that he assisted with the captive animal facility and provided
me with the WaxTags® used in the study. Dr. James G. Ross was essential in providing
statistical guidance in handling the huge amount of data.
The Postgraduate administrator Ms Bernadette Mani was always sympathetic towards
students‘ difficulties, and kept on supporting me to get through my research. Bruce Chapman,
who initially saw the start of my research as the Divisional Research Committee Chair and
carried on in the Academic Administration Committee, was exceptionally understanding on
what many students go through during the process and was ever so encouraging.
I can never fully thank my wife Akemi for her caring support while tolerating the hardships
that were part of this long journey. My parents were supportive from the first day I had this
crazy idea of studying again. My other family members Pommie, Chibi and Pula, all on fourwheel-drive and with wagging tails, were of great help by just being themselves.
James, Hamish, Doreen, Jeanie, Fiona, Chris, Robin, Blair, Denise et al., the staff at Student
Health and Support are awesome! I could have never made it this far without your support!.
My thanks also go to Anna Belle, Emma, Benoit, Craig, and Akemi, who participated in the
repeatability trials, and all the volunteers who helped with the simulated bitemark study. Your
input provided valuable data that formed part of this thesis.
Thanks all!
iv
Table of Contents
Abstract ..................................................................................................................................... ii
Acknowledgements .................................................................................................................. iv
Table of Contents ...................................................................................................................... v
List of Tables .......................................................................................................................... viii
List of Equations ...................................................................................................................... ix
List of Figures ........................................................................................................................... x
Chapter 1 Introduction and Background............................................................................... 1
1.1 Wildlife management ........................................................................................................ 1
1.1.1 Animal abundance ................................................................................................. 3
1.1.2 Population monitoring ........................................................................................... 5
1.1.3 Mammals in New Zealand .................................................................................... 6
1.1.4 Monitoring animals ............................................................................................... 8
1.1.5 Bait Interference techniques .................................................................................. 9
1.1.6 Contagion ............................................................................................................ 11
1.1.7 Bait presentation and contagion .......................................................................... 14
1.1.8 Lure attractability and contagion......................................................................... 15
1.1.9 Behaviour and contagion ..................................................................................... 17
1.2 Forensic science .............................................................................................................. 18
1.2.1 Individual identification ...................................................................................... 18
1.2.2 Forensic odontology ............................................................................................ 22
1.2.2.1 Dental variation ..................................................................................... 22
1.2.2.2 Individual identification from bitemarks .............................................. 25
1.2.2.3 NAS report and forensic odontology .................................................... 30
1.2.3 Firearm and toolmark examination ..................................................................... 32
1.2.3.1 Toolmark variation ................................................................................ 32
1.2.3.2 Toolmark generation ............................................................................. 35
1.3 Identification in wildlife .................................................................................................. 36
1.3.1 Wildlife forensics ................................................................................................ 36
1.3.2 Species identification .......................................................................................... 37
1.3.2.1 Footprints .............................................................................................. 38
1.3.2.2 Sound .................................................................................................... 38
1.3.2.3 Hair........................................................................................................ 39
1.3.2.4 Genetic material .................................................................................... 39
1.3.2.5 Bitemarks .............................................................................................. 40
1.3.3 Individual identification ...................................................................................... 43
1.3.3.1 Markings ............................................................................................... 43
1.3.3.2 Biological material ................................................................................ 44
1.3.3.3 Traces and signs .................................................................................... 45
1.3.3.4 Bitemarks .............................................................................................. 45
1.4 Objectives ........................................................................................................................ 46
1.5 Structure of thesis ............................................................................................................ 47
Chapter 2 Species Identification From Bitemarks on WaxTags®...................................... 48
2.1 Introduction ..................................................................................................................... 48
2.2 Methods ........................................................................................................................... 49
2.2.1 Captive animals of known species ...................................................................... 50
v
2.3
2.4
2.2.2 Bitemark detection device: WaxTag® ................................................................. 50
2.2.3 Presentation of WaxTags® .................................................................................. 52
2.2.4 Photographic data capture ................................................................................... 53
2.2.4.1 Setup...................................................................................................... 53
2.2.4.2 Measurements ....................................................................................... 53
Results ............................................................................................................................. 57
2.3.1 Bitemark type recognition and description ......................................................... 57
2.3.2 Bitemark width measurements ............................................................................ 61
2.3.2.1 Toolmark identification ......................................................................... 64
Discussion ....................................................................................................................... 64
2.4.1 Bitemark description ........................................................................................... 64
2.4.2 Bitemark measurement ........................................................................................ 65
2.4.3 Further study ....................................................................................................... 65
Chapter 3 Identification of Individual Possums From Bitemarks on Wax ...................... 66
3.1 Introduction ..................................................................................................................... 66
3.2 Methods ........................................................................................................................... 68
3.2.1 Simulated bitemarks from known individual‘s dentition .................................... 68
3.2.1.1 Individual animals ................................................................................. 69
3.2.1.2 Wax medium ......................................................................................... 70
3.2.1.3 Bitemark simulation .............................................................................. 70
3.2.1.4 Observation ........................................................................................... 72
3.2.1.5 Comparison schedule ............................................................................ 75
3.2.2 Blind comparison of simulated bitemarks ........................................................... 76
3.2.2.1 Cast preparation and observation .......................................................... 76
3.2.3 Bitemarks from captive known individual possums ........................................... 78
3.2.3.1 Individual animals, WaxTag® presentation and bitemark observation . 78
3.2.4 Repeatability and performance............................................................................ 78
3.3 Results ............................................................................................................................. 82
3.3.1 Bitemark features for identification .................................................................... 82
3.3.2 Image capture and enhancement ......................................................................... 83
3.3.3 Individual match trials ......................................................................................... 84
3.3.4 Repeatability and performance............................................................................ 87
3.3.4.1 Uncontrolled observation ...................................................................... 87
3.3.4.2 Controlled observation .......................................................................... 87
3.4 Discussion ....................................................................................................................... 91
3.4.1 Observation ......................................................................................................... 91
3.4.2 Identification ....................................................................................................... 92
3.4.3 Comparison of bitemark casts ............................................................................. 93
3.4.4 Repeatability........................................................................................................ 93
3.4.5 Further study ....................................................................................................... 94
Chapter 4 Individual Identification of Wild Possums From Bitemarks ........................... 95
4.1 Introduction ..................................................................................................................... 95
4.2 Methods ........................................................................................................................... 98
4.2.1 Study area ............................................................................................................ 98
4.2.2 Site selection within the study area ..................................................................... 99
4.2.3 WaxTag® layout design ..................................................................................... 100
4.2.4 WaxTag® field deployment ............................................................................... 101
4.2.5 WaxTag® retrieval and recording ...................................................................... 105
4.2.6 Data collation .................................................................................................... 106
4.2.7 NPCA protocol .................................................................................................. 107
vi
4.3
4.4
Results ........................................................................................................................... 107
4.3.1 Retrieved WaxTags® ......................................................................................... 107
4.3.2 Influence of food lure on WaxTags® interference ............................................ 108
4.3.3 Repeated visits to tags ....................................................................................... 112
4.3.4 Identification of individual possums ................................................................. 114
4.3.5 Identified individuals......................................................................................... 116
4.3.6 Activity patterns of individual possums ............................................................ 117
4.3.7 BMI emulation .................................................................................................. 151
Discussion ..................................................................................................................... 152
4.4.1 Individual identification .................................................................................... 152
4.4.2 Multiple visits.................................................................................................... 153
4.4.3 Food reward....................................................................................................... 154
4.4.4 Further ecological information .......................................................................... 154
4.4.5 NPCA protocol .................................................................................................. 155
4.4.6 Further directions .............................................................................................. 156
Chapter 5 General Summary and Conclusions ................................................................. 158
5.1 Introduction ................................................................................................................... 158
5.2 General discussions ....................................................................................................... 159
5.2.1 Forensics............................................................................................................ 159
5.2.2 Bitemark stability .............................................................................................. 159
5.2.3 Wildlife management ........................................................................................ 160
5.3 Study components ......................................................................................................... 160
5.3.1 Species identification ........................................................................................ 161
5.3.2 Individual identification — laboratory .............................................................. 161
5.3.3 Individual identification — field ....................................................................... 162
5.4 Conclusions ................................................................................................................... 162
5.5 Future directions............................................................................................................ 163
5.5.1 Quantified identification ................................................................................... 163
5.5.2 Objective identification criteria ......................................................................... 164
5.5.3 Computer assisted identification ....................................................................... 164
5.5.4 Other applications of WaxTags® ....................................................................... 166
5.6 Contribution to knowledge ............................................................................................ 168
References.............................................................................................................................. 169
Appendix A Wildlife abundance assessment methods ...................................................... 196
Appendix B Forensic odontology terminology (excerpts) ................................................. 197
B.1 Bite mark vs. bitemark .................................................................................................. 197
B.2 A characteristic (as it pertains to bitemarks) ................................................................. 197
B.3 Bitemark definition ....................................................................................................... 198
Appendix C Forensic toolmark examination glossary (excerpts) .................................... 199
C.1 Toolmark examination .................................................................................................. 199
C.2 Firearm terminology...................................................................................................... 201
C.3 Firearm anatomy and sources of toolmarks .................................................................. 201
Appendix D Site selection (random plots) .......................................................................... 204
Appendix E Individuals identified on WaxTags® from each night .................................. 206
Appendix F Nightly activity by individual possums. ......................................................... 213
Appendix G Striation quantification trial of bitemark cast image .................................. 215
vii
List of Tables
Table 2-1. Mean single-tooth mark width measurments from possums and rodents. .............. 62
Table 3-1. Confusion matrix to evaluate identification performance. ...................................... 79
Table 4-1. Number of WaxTags® retaining bitemarks by possums and rodents. ................... 108
Table 4-2. Individual identification made from possum bitemarks on WaxTags® ................ 116
Table 4-3. Cumlative bitten WaxTags® and grid locations up to 3, 7 and 10 nights. ............ 127
Table 4-4. Summary WaxTag® biting activity of individual possums over 10 nights. .......... 128
Table 4-5. BMI emulation from WaxTag® grids.................................................................... 152
viii
List of Equations
(3.1) Accuracy .......................................................................................................................... 80
(3.2) Precision ........................................................................................................................... 80
(3.3) Sensitivity ........................................................................................................................ 80
(3.4) Specificity ........................................................................................................................ 80
(3.5) Geometric mean ............................................................................................................... 81
(3.6) Matthews correlation coefficient ..................................................................................... 81
(3.7) F-measure......................................................................................................................... 81
(3.8) Youden's index ................................................................................................................. 82
ix
List of Figures
Frontispiece. Thumbnail images of 3341 bitemark cast photographs from 467 possumbitten WaxTags in field trial used to make 504 identifications of 20 individuals.
Figure 1-1. Principles of toolmark creation. ............................................................................. 36
Figure 2-1. WaxTag® structure and dimensions. ...................................................................... 51
Figure 2-2. Origins of inter-dental ridges on WaxTag® bitemark made by possum upper
incisors. Numbers describe FDI nomenclature of individual teeth. ...................... 55
Figure 2-3. Tooth mark width measurement scheme for striation bitemarks by (a) possums
and (b) rodents as cross-section profiles in wax medium and corresponding
dentition creating respective marks. Observation and image capture was done
by orientating the bitemark with the optic axis perpendicular to the tangent at
the median ridge (corresponding to the median inter-incisor gap). Tooth mark
width was measured from the median ridge to the border of the distal ridge
immediately distal to the median ridge, or to the outer edge of the tooth mark
when recognised as the outer edge of the median incisor, i.e. full tooth mark
width. Numbers describe FDI two-digit nomenclature of individual teeth. .......... 56
Figure 2-4. Bitemark types by possums and rats observed from retrieved WaxTags® with
photograph insets of sample bitemarks: (a) possum impression bitemark, (b) rat
impression bitemark, (c) possum striation bitemark, and (d) rat striation
bitemark. ................................................................................................................ 58
Figure 2-5. Schematic illustration of striation type bitemarks of (a) rodent and (b) possum
with cut-away section views to show profiles. ...................................................... 59
Figure 2-6. Morphological characteristics of possum lower incisor impression bitemark
cast observed from dorsal view of occlusal facet of the incisal arch..................... 60
Figure 2-7. Morphological characteristics of upper incisal (occlusal) facet observed on
ventral view of incisal arch impression casts. Full-depth (a) and partial-depth
(b) full arch impression with teeth outline overlay, illustrating direction and
degree of tooth facet inclination among incisor teeth, and (c) schematic
topographical illustration of upper incisal facet with sharp, medium and
moderate concave inclination, and direction of inclination. .................................. 61
Figure 2-8. Single-tooth width measurement distribution of bitemarks by possums and
rodents.................................................................................................................... 63
Figure 2-9. Box plot of tooth width measurements of bitemarks by possums and rodents. .... 63
Figure 3-1. Bitemark simulation setup: (a) wax block setup on stand in relation to possum
cranium and its movement to simulate bite; (b) simulation process with teeth
scraping on wax surface; (c) wax block dimensions and block number label
position in relation to simulated bitemark, with azimuth designations in relation
to bite direction where bite started at azimuth 000 and ended at azimuth 180...... 71
x
Figure 3-2. Possum bitemark microscopic characteristics observation setup. (a, b)
simulated bitemark photograph capture setup with digital camera, illumination
and wax block with bitemark; (c) bitemark cast observation setup with stereo
microscope, illumination and silicone cast of bitemark. ....................................... 73
Figure 3-3. Image cropping to 200 × 800 pixels. ..................................................................... 75
Figure 3-4. Silicone cast preparation on bitemark: (a) silicone polymer application onto the
bitemark and (b) bitemark cast removal after polymer curing. The mirror
reversal of left (L) and right (R) between actual bitemark and cast is
demonstrated. ......................................................................................................... 77
Figure 3-5. Sample bitemark images cropped (a, d), enhanced (b, c) and aligned (b–c). A
schematic sample generated from greyscale intensity values (e) demonstrate
positive identification possibility from a pair of unequal sized images. ............... 85
Figure 3-6. Example of bitemark characteristics on casts used to identify individual
animals. .................................................................................................................. 86
Figure 3-7. Reference match trial identification performance by three observers — percent
correct identification. ............................................................................................. 88
Figure 3-8. Reference match trial identification performance by three observers —
sensitivity, specificity, precision and accuracy...................................................... 89
Figure 3-9. Reference match trial identification performance indices by three observers —
g-mean, Matthews correlation coefficient, F-measure and Youden's index.......... 90
Figure 3-10. Reference match trial identification performance indices by three observers —
correct identification rate fluctuation with trial repeats. ........................................ 90
Figure 3-11. Blind trial identification performance at differing azimuths (observer KS) ........ 91
Figure 4-1. Study area: Bottle Lake Forest Reserve, Christchurch, New Zealand................. 102
Figure 4-2. Sampling grid arrangement. Grids were composed of 12 × 12 nodes at 10 m
intervals, resulting in two 110 m × 110 m square plots. Minimum distance
between closest nodes in grid (a) and (b) was 100 m. Node location notation
was presented as a row + column combination, e.g., C10 for 3rd row 10th
column node. Node A01 was designated as grid start reference. ........................ 103
Figure 4-3. Components of hip-chain and its functional principle. Free end of thread is
fixed to a reference location, e.g. tree, while the mobile unit (yellow box) is
carried by the researcher along the pre-determined transect. Distance travelled
from the reference location is measured by length of thread pulled out from
mobile unit recorded by the trip meter. ............................................................... 104
Figure 4-4. WaxTag® field deployment. WaxTags® and glo-tags were attached together to
tree trunk by a wood screw at 30 cm above ground. In one grid plot, food lure
(flour/sugar mixture) was applied to the tree below the WaxTag®. Photograph
taken after one night. ........................................................................................... 105
Figure 4-5. Cumulative number of WaxTags® bitten per line without replacement. Both
numbers of bitten WaxTags® with and without food lure reach asymptote after
xi
five nights. Error bar = SEM. Significantly more tags with food lure were
bitten throughout the 10 nights. ........................................................................... 109
Figure 4-6. Cumulative number of WaxTags® bitten per line with replacement. Both
numbers of bitten WaxTags® with and without food lure do not reach
asymptote and continuously increase. Error bar = SEM. Significantly more tags
with food lure were bitten throughout the 10 nights. .......................................... 110
Figure 4-7. Number of WaxTags® bitten per line each night. Error bar = SEM.
Significantly more tags with food lure were bitten up to the third night. No
significant difference between lure type was observed after the fourth night.
(p<0.05) ............................................................................................................... 111
Figure 4-8. Locations where WaxTags® were repeatedly bitten during 10-night period.
Grid (a) with no food lure and grid (b) with food lure. Actual distance between
grids was 100 m. Result shows that no saturation occurred after 10 nights. ....... 113
Figure 4-9. Grid locations and microscopic images of sample bitemark casts on WaxTags®
0077 and 4986 recovered from the first night, and WaxTag® 6118 recovered
from the second night. Identification was made on the bitemarks from the three
tags, interpreted as the same individual possum biting across a spatial span of
grid locations C02, J10 and L11, and across a temporal span of one night......... 115
Figure 4-10. Box-and-whisker plots of nightly (a) number of tags bitten, (b) distance
between furthest bitten tags and (c) minimum convex polygon area, of
individual possums. Individuals are sorted to the order of least to most number
of tags bitten in all three charts for comparison. ................................................. 118
Figure 4-11. Number of individual possums identified per line. Error bars = SEM. ............. 119
Figure 4-12. Distribution of WaxTags® bitten by individual possum 0077 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative
number of visits and nightly shift of centroid (greyscale plots). Legend for
colour plots is shown in ―Legend 1‖. Legend for greyscale plot is shown in
―Legend 2‖. .......................................................................................................... 130
Figure 4-13. Distribution of WaxTags® bitten by individual possum 0375 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative
number of visits and nightly shift of centroid (greyscale plots). Legend for
colour plots is shown in ―Legend 1‖. Legend for greyscale plot is shown in
―Legend 2‖. .......................................................................................................... 131
Figure 4-14. Distribution of WaxTags® bitten by individual possum 2353 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative
number of visits and nightly shift of centroid (greyscale plots). Legend for
colour plots is shown in ―Legend 1‖. Legend for greyscale plot is shown in
―Legend 2‖. .......................................................................................................... 132
Figure 4-15. Distribution of WaxTags® bitten by individual possum 0866 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative
number of visits and nightly shift of centroid (greyscale plots). Legend for
colour plots is shown in ―Legend 1‖. Legend for greyscale plot is shown in
―Legend 2‖. .......................................................................................................... 133
xii
Figure 4-16. Distribution of WaxTags® bitten by individual possum 1885 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative
number of visits and nightly shift of centroid (greyscale plots). Legend for
colour plots is shown in ―Legend 1‖. Legend for greyscale plot is shown in
―Legend 2‖. .......................................................................................................... 134
Figure 4-17. Distribution of WaxTags® bitten by individual possum 0139 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative
number of visits and nightly shift of centroid (greyscale plots). Legend for
colour plots is shown in ―Legend 1‖. Legend for greyscale plot is shown in
―Legend 2‖. .......................................................................................................... 135
Figure 4-18. Distribution of WaxTags® bitten by individual possum 2272 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative
number of visits and nightly shift of centroid (greyscale plots). Legend for
colour plots is shown in ―Legend 1‖. Legend for greyscale plot is shown in
―Legend 2‖. .......................................................................................................... 136
Figure 4-19. Distribution of WaxTags® bitten by individual possum 6990 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative
number of visits and nightly shift of centroid (greyscale plots). Legend for
colour plots is shown in ―Legend 1‖. Legend for greyscale plot is shown in
―Legend 2‖. .......................................................................................................... 137
Figure 4-20. Distribution of WaxTags® bitten by individual possum 3443 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative
number of visits and nightly shift of centroid (greyscale plots). Legend for
colour plots is shown in ―Legend 1‖. Legend for greyscale plot is shown in
―Legend 2‖. .......................................................................................................... 138
Figure 4-21. Distribution of WaxTags® bitten by individual possum 0131 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative
number of visits and nightly shift of centroid (greyscale plots). Legend for
colour plots is shown in ―Legend 1‖. Legend for greyscale plot is shown in
―Legend 2‖. .......................................................................................................... 139
Figure 4-22. Distribution of WaxTags® bitten by individual possum 0560 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative
number of visits and nightly shift of centroid (greyscale plots). Legend for
colour plots is shown in ―Legend 1‖. Legend for greyscale plot is shown in
―Legend 2‖. .......................................................................................................... 140
Figure 4-23. Distribution of WaxTags® bitten by individual possum 0719 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative
number of visits and nightly shift of centroid (greyscale plots). Legend for
colour plots is shown in ―Legend 1‖. Legend for greyscale plot is shown in
―Legend 2‖. .......................................................................................................... 141
Figure 4-24. Distribution of WaxTags® bitten by individual possum 2382 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative
number of visits and nightly shift of centroid (greyscale plots). Legend for
colour plots is shown in ―Legend 1‖. Legend for greyscale plot is shown in
―Legend 2‖. .......................................................................................................... 142
xiii
Figure 4-25. Distribution of WaxTags® bitten by individual possum 3082 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative
number of visits and nightly shift of centroid (greyscale plots). Legend for
colour plots is shown in ―Legend 1‖. Legend for greyscale plot is shown in
―Legend 2‖. .......................................................................................................... 143
Figure 4-26. Distribution of WaxTags® bitten by individual possum 8355 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative
number of visits and nightly shift of centroid (greyscale plots). Legend for
colour plots is shown in ―Legend 1‖. Legend for greyscale plot is shown in
―Legend 2‖. .......................................................................................................... 144
Figure 4-27. Distribution of WaxTags® bitten by individual possum 4242 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative
number of visits and nightly shift of centroid (greyscale plots). Legend for
colour plots is shown in ―Legend 1‖. Legend for greyscale plot is shown in
―Legend 2‖. .......................................................................................................... 145
Figure 4-28. Distribution of WaxTags® bitten by individual possum 8349 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative
number of visits and nightly shift of centroid (greyscale plots). Legend for
colour plots is shown in ―Legend 1‖. Legend for greyscale plot is shown in
―Legend 2‖. .......................................................................................................... 146
Figure 4-29. Distribution of WaxTags® bitten by individual possum 1833 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative
number of visits and nightly shift of centroid (greyscale plots). Legend for
colour plots is shown in ―Legend 1‖. Legend for greyscale plot is shown in
―Legend 2‖. .......................................................................................................... 147
Figure 4-30. Distribution of WaxTags® bitten by individual possum 0919 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative
number of visits and nightly shift of centroid (greyscale plots). Legend for
colour plots is shown in ―Legend 1‖. Legend for greyscale plot is shown in
―Legend 2‖. .......................................................................................................... 148
Figure 4-31. Distribution of WaxTags® bitten by individual possum 4458 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative
number of visits and nightly shift of centroid (greyscale plots). Legend for
colour plots is shown in ―Legend 1‖. Legend for greyscale plot is shown in
―Legend 2‖. .......................................................................................................... 149
Figure 4-32. Overlay of all 20 identified individual possums‘ MCP plots in Figure 4-12 to
Figure 4-31, summarising individuals‘ nightly records over 10 nights. From left
to right: (a) all individuals in greyscale, possums that visited WaxTags® (b)
with no food lure, and (c) with food lure on first night, respectively, and
overlap of two lure-type initiated groups (d). Area colour intensity levels
correspond to degree of polygon overlap. ........................................................... 151
Figure 4-33. Relationship of emulated BMI with observed possum abundance. ................... 152
Figure 5-1. Relationship between study components, and the key outputs. ........................... 167
xiv
Chapter 1
Introduction and Background
In the realm of scientific enquiry, investigating the presence or non-presence of an object of
interest is one of the fundamental pursuits. Accordingly, quantitatively measuring the
abundance of the object is one of the first jobs a researcher will look into to assess the matter.
This thesis lies in the premises of such basic knowledge pursuits that most if not all
researchers aspire to achieve.
1.1 Wildlife management
Wildlife management practices are heavily dependent on having reliable population size
information to decide what management action to take and to assess the achievement of such
management activities, such as in protecting a threatened species or controlling a pest species
(Caughley and Sinclair 1994). In New Zealand, where the brushtail possum (Trichosurus
vulpecula Kerr) is a major conservation and agriculture pest, the trap-catch technique has
been widely used to obtain reasonably reliable information on its abundance, and the success
of population control operations, though it is not well suited to meet recent needs to monitor
possums at low to very low densities (Fraser et al. 2002). It also has the risk of trapping native
ground-dwelling birds, and alternative methods are being developed. The bait interference
method measures animal abundance from the frequency of non-toxic bait stations that are
interfered (Bamford 1970), and the use of wax-block baits to detect animals has the
advantages of being (1) lightweight, low-cost, and easy to set and maintain in the field,
allowing larger sample sizes to improve precision, (2) highly sensitive so it can detect
possums at low densities where trap-catch methods cannot, and (3) not threatening to nontarget species as it poses no risk of being caught in traps (Thomas et al. 2004). Bait
interference methods, on the other hand, have a potential problem that individuals may learn
1
the presence of attractive bait and actively search for more, a phenomenon known as
contagion (Bamford 1970). When contagion occurs, bait stations are not independent
sampling units, which is a major drawback in adopting the method as a robust monitoring tool
(Warburton 2000), and previous studies have attempted to prevent or contain contagion (e.g.,
Bamford 1970, Jane 1979, Bearman 2002, Thomas et al. 2004). However, it is not clearly
established what influences the occurrence of contagion, and how it occurs.
As bitemarks on wax blocks enable species differentiation between those that marked the
device, the potential of individual differentiation from the unique bitemarks was suggested
(Kavermann 2004). Individual identification methods allow investigation of the extent and
degree of behavioural traits that affect the occurrence of contagion, and possibly permit the
use of mark-recapture population estimation method that can generate accurate and precise
population abundance information.
Knowledge of contagion effects on bait-interference derived indices will increase
understanding of biases and possibly allow indices to be related to actual abundance, thus
improving accuracy. It shall also provide means to modify the technique to reduce contagion
effects to negligible levels, approximating the animal‘s random encounter with the bait, and
allow direct estimation of population density using frequency measurements of encounters
(Caughley 1977a). The outcome of this study shall provide insight into the wider issue of
contagion recognised internationally in various methods to measure population abundance
such as scent-station techniques and track-station techniques.
2
In any management profession, the manager needs to know the present status of the managed
resource as baseline information, and equally needs to know the relative information of the
resource to evaluate performances or design inputs to meet specified goals. In wildlife
management, obtaining reliable population size information for species is an essential
component. Apart from using this information to decide what action is needed for a given
population, such information is often essential for assessing the achievement of management
activities. Such achievement goals are frequently expressed as increasing, decreasing, or
maintaining the population at a specific level, as exemplified in managing threatened species,
pest species, and harvested species respectively (Caughley and Sinclair 1994, Lancia et al.
1994).
1.1.1 Animal abundance
Abundance of the population can be presented as population size or density, and is measured
as absolute abundance and relative abundance, with direct and indirect methods to obtain the
measurement. An index is another measurement of abundance that does not itself describe the
population size or density but which has a proportional relationship to it. Each resulting
abundance measure should be accompanied by a measure expressing the degree of the
measure‘s reliability, which can be statistically calculated and compared.
Absolute abundance of the population is the exact or estimated number of individuals in a
population at a given time point (Davis and Winstead 1980, Caughley and Sinclair 1994,
Lancia et al. 1994). For example, in exterminating a pest species from an island sanctuary the
absolute abundance needs to become nil. Absolute abundance is obtained from total counts
such as total-mapping of territorial birds (Verner 1985), estimating from sample counts (Jolly
1969), or from indirect methods such as mark-recapture or change-in-ratio by marking or
3
manipulating the population and calculating the original population size from the response of
population to the manipulation (Otis et al. 1978, Seber 1982).
Relative abundance is often used in wildlife management to compare abundance over
spatially or temporally different data sets. Provided the data sets are comparable, relative
abundance is sufficient to make many management decisions (Caughley and Sinclair 1994),
which may significantly reduce operational costs compared to obtaining absolute abundance.
Comparison of animal abundance at different time intervals may also provide information on
the population trend, i.e., whether the population is stable, decreasing or increasing. Relative
abundance is often useful to evaluate a management input that results in the change of animal
abundance, for example, percentage reduction of rats after a control operation.
When it is not possible or necessary to obtain the direct count of the animals‘ abundance, and
when an attribute changes in a predictable manner with changes in the absolute abundance, it
can be used to index the population abundance (Davis and Winstead 1980, Caughley and
Sinclair 1994, Lancia et al. 1994). Such attributes can be observable phenomena or traces of
activities (field signs) that can indicate the degree of presence at the location. Indices are
mostly used to make comparisons between populations from the same location at different
times (trend) or between populations from different locations at the same time, indicating
relative differences in abundance (Lancia et al. 1994). When indices are directly correlated
with absolute abundance or when the bias is known or measurable and thus can be corrected
for, it can also be used to calculate the absolute population size estimate. Different indices,
however, are derived from different principles, and are not directly comparable between
different methods (Caughley and Sinclair 1994).
4
1.1.2 Population monitoring
Monitoring a wildlife population provides information on population change over time and is
designed to answer specific questions. Monitoring involves using the measures of abundance
associated with the measures of reliability. Numerous methods to monitor animal populations
have been developed to meet various end-user needs with specific levels of reliability, as well
as to accommodate operational difficulties, either by reducing costs at the same reliability
level, or to increase outcome reliability with a given cost (Davis and Winstead 1980,
Caughley and Sinclair 1994, Lancia et al. 1994). Each method has its intrinsic and operational
advantages and disadvantages, and the choice of monitoring method largely depends on
whether the outcome is useful to the end-user to answer a specific question.
The reliability of animal abundance measurements to be used for monitoring, expressed as
accuracy and precision, can be statistically measured and should always accompany the result.
Accuracy is a measure of how close a result is to the true value. Monitoring methods can
result in giving a sufficiently accurate measure but having very wide confidence intervals due
to high variation among samples. Precision, on the other hand, is the probability of obtaining
the same results on the same measurement on the same object. For example, when a study
aims to determine whether an animal population has declined or increased, and not how large
it is, the accuracy of the measure is secondary to its precision (Caughley and Sinclair 1994,
Lancia et al. 1994).
Sensitivity of a monitoring technique can be categorised as the ability to detect animal
presence at a low density and as the ability to detect small changes in abundance. A sensitive
detection method that is capable of functioning at very low animal densities may face
problems of saturation at higher densities (Kavermann 2004, Thomas et al. 2004).
5
1.1.3 Mammals in New Zealand
The terrestrial mammal fauna of New Zealand is, with the exception of bats, entirely made of
introduced species that have established successfully mostly with the assistance of humans.
While some were vigorously protected during initial introduction periods, all are now
classified as statutory pests (King 1990).
Three of the naturalised rodent species are widespread in New Zealand: ship rat (Rattus rattus
Linnaeus), Norway rat (Rattus norvegicus Berkenhout), and house mouse (Mus musculus
Linnaeus) (King 1990). Besides competing with native birds for food sources such as
invertebrates, fruit and seeds, they are serious pests especially because of predation of native
vertebrate and invertebrate species (Innes 1990, Moors 1990, Murphy and Pickard 1990).
They also provide important food sources for predators such as stoats (Mustela ermine
Linnaeus) that are likewise predatory pest species that threaten the native fauna (Murphy and
Dowding 1995). Rodents are controlled by trapping and poisoning with toxicants such as
brodifacoum and cholecalciferol. The rodent population is monitored by track-stations and
trap-catch index from a trap-grid, though uncertainty remains in applying track station
visitation rates as a measure of population abundance (Brown et al. 1996, Blackwell et al.
2002).
The brushtail possum was introduced into New Zealand from mid-19th century to establish a
fur industry (Cowan 1990, Clout and Ericksen 2000). Lack of natural predators and its
opportunistic approach in feeding strategy (Nugent et al. 2000) allowed it to become a
widespread and major pest of agricultural and conservation values of New Zealand. Possums
6
are considered pests because they cause damage to crops and native forests, (Norton 2000,
Nugent et al. 2000, Payton 2000), are known to predate on native bird eggs and chicks (Hay
1981, Nugent et al. 2000), and are vectors and reservoirs of the economically important
bovine tuberculosis (Tb) (Ekdahl et al. 1970, Coleman and Caley 2000). Possums are
controlled by poison including sodium monofluoroacetate (1080) delivered in forms of cereal
pellet, paste and carrot baits, cyanide in paste or capsule forms, brodifacoum and
cholecalciferol both in cereal pellet forms (Morgan and Hickling 2000), and non-toxic
techniques, including hunting and trapping (Montague and Warburton 2000). Alternative
biological control methods, such as using parasites and diseases that cause high mortality, and
using biological agents that interfere with physiological processes involved in the production
of young, are being investigated (Cowan 2000). Possum populations are monitored by
spotlight counts, bait interference, bait-take from stations, faecal pellet counts, trap catch, and
mortality-sensing radio-transmitters (Warburton 2000, Fraser et al. 2002), to obtain estimates
of abundance and to evaluate control operation outcomes.
Currently the most widely used method to monitor possums is the trap-catch method
(Batcheler et al. 1967) with an operational protocol developed to ensure the method is used
consistently under standardised conditions (National Possum Control Agencies = NPCA
2004). The method is a basic form of catch-per-unit-effort model based on the assumption that
the number of possums caught is proportional to the effort put into catching them, and it has
been used to produce sufficiently reliable estimates of percentage kill in possum control
operations and to provide measures of relative abundance (Residue Trap Catch Index = RTCI)
(Warburton 2000). Recently there has been an increasing need to measure very low residual
trap-catch rates to meet more stringent residue possum density targets for controlling Tb, with
an acceptable degree of statistical robustness. Furthermore, vector managers are becoming
7
less interested in overall average possum abundance of the managed area, and more interested
in the spatial distribution of the residual possum population from the standpoint of controlling
the transmission of Tb (Fraser et al. 2002). To further achieve these requirements, refinement
of the trap-catch method and development of alternative techniques are being pursued
(Thomas 1998, Brown and Thomas 2000, Thomas et al. 2003, Brown et al. 2004). Recent
developments in alternative methods to meet current monitoring requirements include bait
interference methods which detect the presence of possums from bitemarks made on wax
blocks (Thomas et al. 1999, McGlinchy and Warburton 2000, Warburton 2000).
1.1.4 Monitoring animals
Much of wildlife management practices are based on, and evaluated by monitoring population
abundances, using various methods chosen to meet end-user needs of the information, and its
practicality (Bookhout 1994, Caughley and Sinclair 1994). Some methods may only provide
statistically imprecise and/or inaccurate estimates, and yet can answer a broad range of
ecological and management questions to an acceptable level of approximation. For example,
aerial survey is considered the only practical means of estimating the abundance of large
animals inhabiting a large area while providing population estimates of acceptable reliability,
and has been used successfully by wildlife management authorities in Africa (NortonGriffiths 1975, Caughley 1977b).
In New Zealand continuous efforts to control brushtail possums have necessitated the need to
generate abundance information through monitoring, currently done by using trap-catch
method as a principal tool under a set protocol (NPCA 2004, 2008b). While the trap-catch
index has so far shown to be practical for detecting large changes in possum density, the
method is inaccurate and imprecise at the very low possum densities that are now commonly
8
targeted (Fraser et al. 2002). The trap-catch technique involves setting heavy leg-hold traps at
prescribed distances along a trap line, checking each trap daily within 12 hours of sunrise as
required by the Animal Welfare Act (1999), releasing or killing animals caught, and resetting
sprung traps, for three nights. The high degree of required labour input restricts its use for low
to very low densities (Brown and Thomas 2000) as larger sample sizes to acquire a reasonable
level of reliability may significantly increase monitoring costs. Moreover, traps pose a risk of
catching non-target species, such as weka (Gallirallus australis) and kiwi (Apteryx spp) (Reid
1983). By-catch is currently reduced by raising traps off the ground (NPCA 2008b) although
this may reduce its detection ability (Thomas and Brown 2001). Alternative methods are
under development with a bait interference technique using wax blocks to detect interference
by target species as one of potential monitoring tools (Thomas et al. 1999, McGlinchy and
Warburton 2000, Warburton 2000, Fraser et al. 2002).
1.1.5 Bait Interference techniques
Monitoring animal abundance using interference methods operates on the principle that
animals randomly encounter detection devices, are positively attracted to the devices to leave
identifiable marks, from which frequency or proportion of marked detection devices is
collated, providing an index that is presumably proportional to the true animal abundance.
Scent-station or scent-post methods rely on animals attracted to an olfactory lure and mark
footprints on a patch of smoothed sand or soil, and the frequency of target species visitation is
recorded during a specified time frame (Linhart and Knowlton 1975, Travaini et al. 1996).
Track-plates or track-tunnels are footprint detection plates, and the method assumes that the
frequency of detection is proportional to the number of animals. Animals are lured to the
detection plates and pass a plate where its footprints are recorded by ink, soot or powder
9
(Brown et al. 1996, Zielinski and Stauffer 1996, Drennan et al. 1998, Thomas 1998,
Blackwell et al. 2002). The bait interference method detects the animal‘s visitation by luring
the animals to bait and recording any interference to the bait, such as partial or whole
consumption, disappearance, or scratch / bitemarks, using palatable bait such as flour paste
(Bamford 1970), flour and soy oil mixture (Spurr 1995), apple (Byers 1981, Arulchelvam and
Brown 1995), orange (Thomas et al. 2003), or unpalatable baits such as wax blocks (Thomas
and Meenken 1995, Thomas et al. 2003).
General advantages of these methods include detection of inconspicuous animals, such as
nocturnal species, non-destructive detection permitting monitoring of threatened species or
contamination by other species at minimum risk, and relative cost-effectiveness in that lowcost devices can be used in large quantities to improve detection sensitivity and precision
(Thomas 1998, McGlinchy and Warburton 2000, Fraser et al. 2002, Thomas et al. 2004).
Disadvantages include the contamination of multiple species that can interfere with the device
(Arulchelvam and Brown 1995), which is relatively easy to solve by identifying target species
from the marks, or prevent by choice of bait and presentation method. Another major issue
with using interference techniques to measure abundance is that the probability of individuals
marking multiple devices and individual devices being marked by multiple animals is
unknown (Linhart and Knowlton 1975, Roughton and Sweeny 1982, Zielinski and Stauffer
1996, Warburton 2000, Warrick and Harris 2001, Bearman 2002, Fraser et al. 2002, Sargeant
et al. 2003). Warburton et al. (2004) compared the bait interference method using bitemarks
on wax blocks and the abovementioned trap-catch index with mortality of radio-collared
possums to assess the indices‘ accuracy in detecting population change, and concluded that
further studies to assess the extent of potential bias in the bait interference method is
necessary to reliably use it to monitor possum abundance.
10
1.1.6 Contagion
As interference detection techniques rely on lures, there is a possibility of animals acquiring
learned attraction to these devices and to actively seek other detection devices over temporal
and spatial scales. Such behaviour can increase bait visitation frequencies and inflate
detection rates disproportional to its abundance. This phenomenon is known as contagion, or
multiple visits, creating an unknown probability of more than one individual animal being
recorded on one detection device, one individual being recorded on more than one device, or
the same animal marking one detection device repeatedly over time (Bamford 1970, Spurr
1994, Zielinski and Stauffer 1996, Sargeant et al. 1998, McGlinchy and Warburton 2000,
Warburton 2000, Bearman 2002, Sargeant et al. 2003, Thomas et al. 2003, Thomas et al.
2004). While detecting a species presence by these devices is straightforward and can be
readily accepted, application of the information to measure population abundance has been
varied due to the lack of evidence as to the frequency and form of contagion. The likelihood
that individual detection devices are not independent sampling units is one of the major
drawbacks in this method being adopted as a robust monitoring tool (Warburton 2000, Fraser
et al. 2002).
Contagion can occur in several forms that don‘t directly reflect the true animal abundance,
with differing degrees of effects to the method. When one individual marks more than one
device from a change in behaviour that it actively seeks more devices, it negates the
assumption of random encounter, and directly inflates the frequency of visitation. In the case
of multiple individuals marking one device, it will return only one positive result for an
unknown number of animals, confounding the presumed positive relationship between
devices interfered with and true animal abundance. When one individual marks a device
11
repeatedly, one device will return a positive result for one animal and will not influence the
overall estimate of abundance if each individual animal is consistently visiting the same
device, though there has been no method to provide evidence of such occurrence, or to
differentiate from multiple individuals‘ visitation.
Conversely, a phenomenon opposite to contagion may occur when animals acquire learned
aversion after its first encounter with the bait and actively avoid other detection devices it
encounters in the future, commonly known as repulsion. While contagion seriously affects the
reliability of the bait interference method by confounding the detection devices‘ positive
results, repulsion may be beneficial by preventing the same individual from marking multiple
devices, when the method is used as a one-off operation for a particular area. On the other
hand, when multiple trials are scheduled for the same animal population over a specific time
interval, such as before and after control operations, repulsion can be problematic as the
probability of detecting the animal has changed between the sampling operations, precluding
any comparisons of the results until the learned behaviour wanes from the animals‘ memory.
Previous studies assessed the occurrence of contagion from statistical probabilities of
diverging from random encounter with baits (Bamford 1970, Spurr 1995), from evidence of
increased bait interference with time (Thomas et al. 2003), or from the patchiness or
clustering of bait interference (Thomas et al. 2003, Hunter 2004).
Bamford (1970) found that as the level of bait interference increased nightly, for the first 3-5
days of baiting it arose from a summation of random encounters rather than active bait
seeking. In comparing bait lines with various bait spacings after 10 days of baiting, those of
12
over 40 yards (approx. 36 m) were considered free of contagion where the overall level of
interference had not increased significantly from the first day. Spurr (1995) assessed the
probability of contagion by whether the sequences of bait interferences along the lines of bait
stations were random. He found that the spatial distribution of bait interference was random at
least for the first night, and usually for at least five nights (non-randomness resulting from too
many consecutive interferences or non-interferences, which may indicate contagion, occurred
only after five nights), supporting earlier studies (Bamford 1970, Jane 1981) that contagion is
unlikely when bait stations are 40 m or more apart. However, he also found that individual
baits may be interfered with by more than one possum and that individual possums were
interfering with more than one bait, on bait lines with 40 m bait spacings, and suggested that it
could occur even at 200 m distances. Thomas et al. (2003) measured contagion by
determining whether significant increases occurred between the three baited nights, and found
that when spacing the bait at 20 m intervals, contagion occurred in orange baits, which
showed significant increase in average proportion of baits interfered with from night 1 to
night 3, but not in wax blocks. Hunter (2004) concluded that from the lack of patchiness of
possum interference in her line of cruciform tag stations spaced 100 m apart, that the distance
was effective to prevent contagion.
The prevalence of contagion, however, has not been extensively studied from the viewpoint of
individual animals. It is not clearly established what influences the occurrence of contagion,
and how it occurs. Identification of individual animals that visit the detection devices will
enable further studies into the phenomenon, evaluate its prevalence, and search for ways to
reduce, contain, or measure its occurrence. Such information is vital for developing new
methods of using the detection frequencies to index animal abundance.
13
1.1.7 Bait presentation and contagion
Thomas and Meenken (1995) used unpalatable wax blocks spaced at 20 m intervals along
sample lines, a method following closely the trap-catch protocol (NPCA 2004) and repeated
as similar layouts in later studies (Thomas and Hickling 1990, Ogilvie et al. 1996, Thomas
1998, Thomas et al. 1999, Ogilvie et al. 2000, Thomas et al. 2003). In attempts to reduce
effects of contagion that can occur in sample line layout pattern, clusters of multiple detection
devices were treated as single sampling units, which were then arranged at intervals presumed
sufficiently distant to treat contagion as negligible (Sargeant et al. 1998, 2003). A cruciform
layout with four wax blocks placed at 10 m distances around a centre wax block was designed
to accept contagion within the cluster and thus reduce the unknown bias and variability that
originates from contagion (Bearman 2002). Later studies have followed the cruciform design
(Hunter 2004, Kavermann 2004, Thomas et al. 2004, Thomas and Maddigan 2004) though
none had specifically provided evidence that the inter-station distances prevented contagion.
While Kavermann (2004) believed that any change in design was unwarranted, possible
alterations in the number of wax blocks per station and arrangement were suggested by
Hunter (2004) to maximize efficiency and reduce costs, but no study has determined the
optimal presentation of detection devices in terms of reducing contagion and optimising costefficiency.
Similarly in scent-station surveys, where the detection devices usually are arrayed
systematically in lines to reduce travel time and expedite data collection, scent-stations within
sample lines generally produce spatially dependent data including that of individuals visiting
multiple stations (Sargeant et al. 1998). The proportion of lines visited by carnivores,
compared to the proportion of individual stations visited was used to avoid the problem of
possible contagion (Sargeant et al. 1998). Sargeant et al. (2003) compared clustered data from
14
scent-station lines and data from each station, and argued that whilst the clustering method
reduced the undue influence of individuals visiting multiple stations it reduced sample size
with the given numbers of stations that can be practically set up, concomitantly increasing
sampling variability. Tracking tunnels are also prone to contagion, and in monitoring fishers
(Martes pennanti) and American martens (Martes americana) Zielinski and Stauffer (1996)
used small grids of tracking plates as sample units that are spaced almost twice the size of the
species home range diameter to reduce possibilities of spatial dependency.
1.1.8 Lure attractability and contagion
Studies have used the percentage wax block interfered or percentage wax removed to compare
relative attractiveness of various lures to wax blocks, as it affects sensitivity of the device
(Abrams et al. 2004, Hunter 2004, Kavermann 2004, Thomas and Maddigan 2004). In the bait
interference method, however, increased attraction to a detection device may also raise the
probability of animals changing behaviour upon initial or repeated contact with the lure and
actively seeking baits, therefore creating contagion.
Spurr (1995), and McGlinchy and Warburton (2000) suggested the high palatability of baits
as probable cause for contagion. McGlinchy and Warburton (2000) believed that solely visual
attractant should be used to avoid possums actively seeking other baits, and similarly Allen et
al. (1996) suggested that some attractants may bias the scent-station method by artificially
increasing dingo activity and influence the index over time. For example, compared to using
orange-quarter bait that showed evidence of contagion, unpalatable wax bait was suggested to
have prevented possums from actively seeking other baits (Thomas et al. 2003). Kavermann
(2004) suggested the lack of potential food sources (other than WaxTag® bait) in pine
plantations may have contributed to high levels of interference recorded in his study. Low
15
palatability of wax blocks, on the other hand, may induce learned aversion discouraging
animals from biting other devices (McGlinchy and Warburton 2000), which in effect should
prevent contagion.
Visual attractants to lure possums may be effective (Carey et al. 1997), and luminescent strip
lures have been proved more efficient compared to white, coloured and UV enhanced lures
(Hunter 2004, Thomas and Maddigan 2004, Ogilvie et al. 2006a, Ogilvie et al. 2006b). In an
extension study, it was concluded, however, that flour blaze (flour and icing sugar 5:1
mixture) attractants were significantly preferred by possums over purely visual (luminescent)
lures (Abrams et al. 2004, Ogilvie et al. 2006a). These studies suggest that highly attractive
lures may cause animals to actively seek devices thus increasing the method‘s sensitivity
when bait is used to detect possum presence, though no previous research studied the effect of
increased attraction on probable contagion, which is problematic when the method is used to
assess animal abundance. Similarly no study tested the effect of palatable lures compared to
visual lures as a probable cause of contagion (McGlinchy and Warburton 2000, Thomas et al.
2003).
The mechanism of how lure-attracted individuals actually interfere with detection devices is
unknown for interference devices. In scent station techniques, animals are lured to the
olfactory source and, upon approach, unintentionally leave footprints as a behaviour which is
independent to the attraction cue, though the occurrence of individuals that are lured but do
not approach the source close enough to pass the detection area, possibly influenced by wind
or individual wariness, is unknown. Track-tunnels are similar in that leaving footprints
derives from a behaviour independent to the olfactory and gustatory (taste) cues that lured the
animals to the device. In contrast, bait-interference using gustatory lures have direct
16
connections between the lure and the detection device which needs to be bitten or consumed
to detect animal visitation. When using visual lures, it is commonly believed that the animal‘s
curiosity attracts them to the detection device (Carey et al. 1997, Hunter 2004, Thomas and
Maddigan 2004), though the mechanism of switching from visual attraction to a device
attraction that encourage animals to mark detection devices is unknown. While studies
suggest that high palatability of bait may be a probable cause of contagion (Spurr 1995,
McGlinchy and Warburton 2000, Thomas et al. 2003), there is no evidence that directly
connects attraction behaviour and device marking behaviour to the occurrence of contagion.
Whether the attractants, using olfactory, gustatory, or visual cues are necessarily causing
contagion is unknown, though it is known to affect sensitivity (Warburton and Yockney 2000,
Hunter 2004, Thomas and Maddigan 2004). It is possible that although high numbers of
animals may be attracted by one or a number of combined cues, the detection devices may be
marked independent of the initial attraction, resulting in an increased device sensitivity while
keeping the contagion probability unchanged.
1.1.9 Behaviour and contagion
As contagion is essentially the animal‘s behavioural response to encountering (and
subsequently re-encountering) a novel object, it is difficult to objectively observe or measure.
Earlier observation trials using video recordings at wax block stations showed that at some
stations none of the wax blocks were interfered with for the first two nights although
numerous possums were recorded approaching the devices and licking the flour lure but
leaving no mark of presence, whereas other stations recorded nearly every possum
investigating the wax blocks (Bearman 2002). The study, however, could not make conclusive
explanations on how its observations were related to contagion, as it was outside its scope.
17
Although possums are solitary animals, social interactions occur upon encounter (Day et al.
2000). Such interactions are oriented mostly towards forming reproductive relations or
maintaining dominance relationships (Green 1984). Possums defend high-quality foodsources but will not continue to defend it after they cease feeding (Day et al. 2000), and
although feeding behaviour may influence the prevalence of contagion, most of the previous
behavioural studies have focused on its significance in possum control applications.
1.2 Forensic science
Forensic science is a discipline that endeavours to establish and re-construct the nature and
sequence of events of legal significance from traces, signs, or symptoms, left behind by those
who were present and by activities that occurred at the time and/or space. One core premise
centres on the domain of individual identification: the conclusion of identification can only be
reached where the source of marks possess a recognisable set of features that is unique to the
particular individual, and when the unique features are transferred as marks with sufficient
precision to discern the feature's characteristics (Kiely 2005).
1.2.1 Individual identification
Eyewitness testimony (Nelson et al. 2009) may be considered a historical form of individual
identification in the legal arena. Human facial recognition is an everyday task done
effortlessly for the human eye and brain (Kevin Zhou et al. 2009) although such human
behaviour has the risk of being subjective and non-consistent.
Biometric identification is an approach to objectively quantify various biological properties of
individuals, which has its roots in anthropometry (Bertillon 1893 quoted in Dessimoz and
18
Champod 2008). Recent Bayesian approach to the use of biometric identification systems was
introduced by Gonzalez-Rodriguez et al. (2006). Current use of normally unalterable
biological features that can be used in biometric identification cover a wide spectrum as
exemplified in the following:
Fingerprint, or more precisely, friction ridge examination compares its pattern which is
assumed to be unique to individual (Stoney 2001, Langenburg 2009, 2011). The identification
process can be automated using objective criteria in the form of an automated fingerprint
identification system (AFIS) (Meuwly 2009). Su and Srihari (2009) evaluated the probability
of random correspondence in fingerprint matches by mathematical models of fingerprint
features. Footprint analysis similarly applies the individuality of footprints for identification
purposes (Brown and Rutty 2005). Earprints (Alberink and Ruifrok 2007, Meijerman et al.
2009) and lips (Coward 2007) are also found useful in recognizing individuals. Compared to
the subjective eyewitness, objective facial recognition performed automatically by machines
is derived from research areas in computer vision, pattern recognition and image
understanding (Cattaneo and De Angelis 2009, Kevin Zhou et al. 2009). Iris and retinal
characteristics offer an identification source in humans (Yoon et al. 2005, Daugman 2009, AlDiri et al. 2010), as well as cattle (Allen et al. 2008). Individual recognition by comparing
radiographic images of the frontal sinus is likewise a feasible method (Besana and Rogers
2010). Dynamic biological characteristics such as gait can also identify individuals in humans
(Cunado et al. 2003, Lynnerup and Vedel 2005, Bashir et al. 2010, Simsik et al. 2010,
Bouchrika et al. 2011, Chen et al. 2011), and in animals (Sharma et al. 2005). Individual
variability in dental characteristics also offer identification opportunities (American Board of
Forensic Odontology = ABFO 1997, Glass 2005), which is reviewed in further detail in the
following section.
19
Genetic identification is a highly accurate method (Graham 2009) which was the basis of a
series of overturned convictions which were based on other forms of forensic evidence
(Neufeld and Scheck 2010). On the other hand, traditional forms of evidence such as
osteological and dental evidence, still provide forensic evidential value (Steadman et al. 2006).
Non-biological physical evidence that assist in reconstruction of an event include ballistics,
glass, and toolmarks. Ballistic examination of an firearm projectile‘s flight path could
determine the firearm‘s location when discharged (Jackson 2009, Maiden 2009). Glass
fragment examination could classify its refraction properties and identify its origin, as well as
reconstructing the impact point from the fragment‘s morphological characteristics (Curran
and Hicks 2009, Lovelock 2009).
Toolmarks are used to identify a suspect tool to a toolmark collected at the crime scene.
Firearm identification is a specialised branch of toolmark identification where the harder
material components of the firearm, usually steel, leaves toolmarks on the softer material
components such as lead bullets and brass cartridge cases. (Association of Firearm and Tool
Mark Examiners = AFTE 1998, Cork et al. 2008). Shoe and tyre prints can also take the form
of toolmarks that leave morphological features of the shoe sole or tyre impressed on soft
substances such as snow or sand (Bodziak 2005a, b). Similarly, bitemark analysis in forensic
odontology, a discipline based on pattern matching of bitemarks with the suspect‘s dentition‘s
characteristics, can be also considered a specialised branch of toolmark identification (Vale
2005).
20
Identity is established when the examiner is satisfied that one or more characteristics of the
examined objects sufficiently coincide beyond pure chance. In toolmark / bitemark analysis
the examiner determines whether the characteristics are from the same source e.g., a tool, a
firearm, or a set of teeth. Conversely, a conclusion of negative identification, i.e., exclusion or
elimination can be reached when the examiner determines that observed characteristics
sufficiently disagree between each analysed sample to the extent that the same source cannot
have made the marks (AFTE 1992, ABFO 1997).
The subjective nature of the human decision to declare a ‗match‘ is one of the main issues by
forensic identification science critiques (Saks 1994, Committee on Identifying the Needs of
the Forensic Science Community and National Academy of Sciences = NAS 2009). A typical
example can be the eyewitness testimony in identifying a suspect (Nelson et al. 2009) where
there is nothing more subjective than the witness‘s own interpretation and memory. Although
the testimony itself is not a scientific method performed by a forensic expert, its evidential
value is increased though various operational methods derived from sound cognitive and
psychological science (Peterson and Loftus 2009, Ruifrok 2009).
Cognitive bias has been recognised as having influence in decisions (Dror 2005, Dror et al.
2005). This may be problematic in forensic science disciplines that rely on decisions to come
to a conclusion (Saks et al. 2003, Dror and Charlton 2006, Dror et al. 2006, Dror and Cole
2010), as emotional and contextual biases are shown to have influences in such decisions
(Hodge 1988, Hall and Player 2008, Dror et al. 2011, Lange et al. 2011). Further research into
understanding human cognition was suggested to reduce cognitive bias (Dror 2009, Dror et al.
2011). Random auditing with a penalty for blind proficiency tests in an experimental setup
demonstrated it can significantly reduce the error rates (Cowan and Koppl 2011). Another
21
approach is to automate identification using objective criteria (e.g., Blackwell and Framan
1980, Alexandre 1996, Banno 2004, Banno et al. 2004, 2005). However, the human eye and
brain combination in recognising selecting, inputting, analysing, comparing, and interpreting
patterns is highly complex and is yet to be surpassed by computers (Dailey 2005, Kevin Zhou
et al. 2009). This can be evident in CAPTCHAs (Completely Automated Public Turing test to
tell Computers and Humans Apart), commonly used in security systems, and as a supplement
to computerised character recognition system for recognising characters that computers have
given up (von Ahn et al. 2003, Chew and Baird 2003, von Ahn et al. 2008). Ongoing studies
in human cognition, machine learning and pattern recognition demonstrate that ‗objective‘
machines are yet to fully take over what the ‗subjective‘ human eye and brain are
accomplishing (e.g., Fageth et al. 1996, Dailey and Cottrell 1999, Lewin and Herlitz 2002,
Hayward et al. 2008, Hoffman and Logothetis 2009, Deng et al. 2010, Woolley et al. 2010).
1.2.2 Forensic odontology
Because teeth have a low biological turnover rate and undergo mineralization which render
them the strongest structure in the human body, they contain more physically and biologically
stable information about the individual than any other biological structure apart from DNA
(Whittaker 1995). The discipline of forensic odontology or forensic dentistry deals with
mainly two aspects of identification from dental features, i.e., body identification by antemortem and post-mortem comparison of dental records, and bitemark analysis by comparing
bitemarks and suspect dentition (ABFO 1997, Glass 2005).
1.2.2.1
Dental variation
Variation in dental features can result from any of the biological (inherited) and
environmental (acquired) factors during the formation and use of teeth (Bowers and Bell
1997). Most mammals are diphyodont, i.e., have deciduous and permanent teeth during its
22
lifetime which changes with age (Bishara et al. 1989, Bishara et al. 1995, Harari et al. 2005).
Most mammals are also heterodont, i.e., they possess different functional types of teeth,
mostly in the forms of incisors, canines, premolars, and molars, which differs by location in
the jaw. Each tooth type is morphologically different from others by size, shape and root
numbers.
Compared to the taxon‘s standard number of teeth, e.g., 32 in Homo sapiens, the number may
be excessive (hyperdontia) or lacking (hypodontia) (Kindelan et al. 1998). Individual tooth
size, dental arch dimension, tooth location and arrangement, relative tooth position in the
dental arch or with neighbouring teeth, and tooth orientation e.g. rotation in relation to
‗normal‘ dental arch orientation, may vary in different degrees due to the individual‘s normal
and pathological history. Intraspecific dental variations in primate populations, including man,
have been studied extensively by Kieser and Groeneveld (e.g., 1987a, b, 1988a, b, 1989a, b).
One of the prominent variations that may occur with the use of teeth is surface wear (e.g.,
Western et al. 1983, Kieser et al. 1985, Xiaohu et al. 1992, Solheim 1993, Zheng et al. 2003,
Chattah and Smith 2006, Zheng and Zhou 2006, Yun et al. 2007). Tooth wear can be part of a
normal process (attrition) or can be the result of pathological or traumatic condition (erosion,
abrasion, fracture), either which can add to the variability of individual dental pattern
(Litonjua et al. 2003). For example, occlusion abnormalities including overbite (overshot) and
undershot are known to be associated with abnormal wear (Casanova-Rosado et al. 2005,
Janson et al. 2010, Oltramari-Navarro et al. 2010).
23
Where the degree of tooth wear with time is generally uniform across a population, its extent
can be associated with an estimated age class (e.g., Kim et al. 2000, González-Colmenares et
al. 2007). Age estimation from tooth wear can be further combined with tooth development,
i.e., eruption and calcification (Bowers 1997, Franklin 2010). Association of tooth
development and wear with age has also been a widely used method in wildlife age estimation
(e.g., Laws 1966, Schroeder and Robb 2005).
Another important source of dental variation is particularly anthropogenic, which
encompasses all forms of artificial manipulation to the dental components and structures
including restorations, crowns, bridges and implants (e.g., ABFO 1997, Bowers and Bell
1997, Glass 2002). On the other hand, orthodontics, by definition, mainly contribute to
lessening dental variation of individuals to bring it closer to the perceived population
‗standard‘, although its procedures, such as application of braces or surgical re-construction,
can provide individual variation to the overall picture.
Given that the abovementioned variations can occur singularly or in combination in an
individual, such dental feature(s) can assist in determining the identity of the individual‘s
dental pattern. Studies on genetically identical monozygotic twins confirm that they do not
necessarily possess the same dental features (Sognnaes et al. 1982, Kindelan et al. 1998).
Adams (2003) examined the civilian and military dental databases of the United States, and
found that from the combination of dental variations as outlined above, a very high degree of
diversity was found, comparable to the level known in mtDNA, making it an ―excellent
source for forensic identification.‖.
24
The subjective nature of visual dental comparison in establishing identity can result in
examiner disagreement (Pretty et al. 2003) or misidentification (Kieser et al. 2001), where
proper training of the observer is essential. Computer assisted identification can be a valuable
tool in confirming an identification or narrowing down possible candidate identities (e.g.,
Dailey 1997).
Apart from traditional morphological analysis to compare and match dental characteristics,
teeth are increasingly being recognised as a valuable source of genetic material useful for
individual identification. This is particularly important when other body parts have been
destroyed, as teeth are one of the hardest structures of the human body which can withstand
adverse conditions, e.g., extreme heat and decomposition. Viable DNA can be extracted from
the tooth root and pulp to make a positive identification in cases where most other methods
are impossible, e.g., in mass disaster victims (Manjunath et al. 2011).
1.2.2.2
Individual identification from bitemarks
A bitemark can be defined as a pattern manifested as physical and/or biological response,
including damage, to the forceful contact of dental structures on a substrate, e.g., skin, food,
or any other substance softer than teeth (Bernstein 2005, Kieser et al. 2005). Provided the
patterned mark retains a reasonable degree of detail, it can make available a near-identical
imprint of physical characteristics of an individual‘s tooth pattern and teeth positioning.
Bitemark analysis procedure uses the dental imprints in a combination of metric analysis and
pattern association that link the marks to the suspects‘ dental models from observable features
(Sweet 1997). Bitemarks subject to analysis and comparison are not restricted to human
sourced marks, and may involve those made by non-human animals (Kieser et al. 2005,
Souviron 2005, Tonn et al. 2005, Kannikeswaran and Kamat 2009).
25
According to the ABFO bitemark analysis scoring guidelines, features including (1) gross
characteristics, e.g., presence / absence of individual teeth; (2) tooth position, e.g. alignment
and rotation; and (3) intradental features, e.g., fractures, are scored to evaluate its match
(ABFO 1997). Dailey (2005) outlined the comparison procedure using currently available
methods noting that there currently exists no single all-encompassing method to identify
individuals from bitemarks, and that forensic odontologists use two methods as a standard in
the comparison process.
Rawson et al. (1984) statistically demonstrated that human anterior dental arrangements are
uniquely individual from a general sample of bitemarks on wax wafers (n = 1200). Though
the study‘s weakness lies in the assumption that each individual tooth variation is independent
of each other (MacFarlane et al. 1974, Pretty and Sweet 2001a), the authors correctly
recognised that dental uniqueness does not directly translate to bitemark uniqueness. Bernitz
et al. (2006) also used bitemarks on wax wafers for a metric analysis method of anterior tooth
rotation, tested in a limited sample population of 300 volunteers, and concluded that a
population frequency database can be built by this method. Kieser et al. (2007) likewise
demonstrated the uniqueness in human anterior dentition from a geometric morphometric
analysis of two-dimensionally scanned dental casts, while also cautioning that such
uniqueness would not necessarily transfer as bitemarks.
The degree of detailed dental characteristics that can be accurately transferred and retained
largely varies with the bitten substrate. Bitemarks on foodstuff generally provide a good
representation of the biter‘s dental features if the perishable food is recovered before
desiccation or disintegration, and an impression cast made to retain its features (Stoddart
26
1973). Identification of bitemarks was possible from cheese (Layton 1966, Solheim and
Leidal 1975, Webster and MacDonald 1979, Webster 1982, Aksu and Gobetti 1996, Bernitz
and Kloppers 2002), sandwich (Simon et al. 1974), cakes (Aboshi et al. 1994), and chocolate
(McKenna et al. 2000). Non-animate, non-food substances such as soap (Corbett and Spence
1984) and wax (Whittaker et al. 1975) are known to produce a high-quality replica of the
dental characteristics, where the latter is commonly used to produce golden standard sample
bites (exemplars) (Sweet 1997, Sweet and Bowers 1998, Dailey 2005, Johnson 2005, Bernitz
et al. 2006).
Contrastingly, bitemarks on elastic substances, such as skin, pose an entirely different issue.
Pertaining to the difficulty in analysing bitemarks on human skin, Taylor (1963) predicted
that bitemarks on skin may be ―..unlikely to establish convincing proof in most cases.‖ Ström
(1963) recognised the various sources of possible distortion in analysing bitemarks on human
skin compared to those on foodstuff, and commented that it would be easier to establish a
non-match with an apparent inconsistency between the dentition and bitemark than finding
consistent characteristics and decide on a ‗positive match‘. DeVore (1971) demonstrated the
possible extent of human skin distortion that can occur when the body position is changed,
and strongly cautioned against a careless application of the transparency overlay method if
done with disregard to such possible distortion. By comparing bitemarks on ―good recording
properties‖ e.g., wax, and those on porcine skin, Whittaker et al. (Whittaker et al. 1975) found
that identification accuracy significantly differed between the bitten substance, with wax
reaching 99% if good impressions or photographs can be obtained, and skin marks falling as
low as 9% if photographs were taken after a time lapse from the actual bite. They also found
that variability between observers were greatest in skin bitemarks, suggesting that a more
subjective assessment criteria had to be applied.
27
While most studies took a cautionary approach in interpreting distorted bitemarks, Harvey et
al (1976) concluded that identification from distorted bitemark images is possible — provided
that unique characteristics are transferred as high-quality evidence. They illustrated ―[t]he
importance of characteristics which can be clearly recognized throughout changes of posture‖
where ―[c]haracteristics are clearly identifiable in spite of distortion from surface spread,
flexion or rotation.‖ Such strong claim without scientifically sound evidence may not be
readily acceptable in the current legal climate of admitting only scientifically proven forensic
methods (NAS 2009).
Computer generated digital overlays from dentition models by using commercially available
image processing software made possible the objective production of the acetate overlay to
compare with the bitemarks (Naru and Dykes 1996, Pretty and Sweet 2001b, Martin-de las
Heras et al. 2007). Mathematical comparison of random pairs of computer-generated overlays
from experimental bitemarks and stone casts of volunteers concluded that the ―2D polyline‖
and ―painting‖ methods were applicable and reliable although the latter required visual
assessment of the data (Al-Talabani et al. 2006). Naru and Dykes (1997) attempted
computerised comparison from cross-correlation scores, where the conclusion was conditional
that ―[the] method may have value when applied to certain mark types‖.
Assessment of bitemark identification performances by examiners with different levels of
training and experience showed that the effect of training was important on the examiners‘
performance (Whittaker et al. 1997). Where the difference between examiners are large
enough to reach different conclusions from the same evidence material, bitemark examiner
disagreement may occur (Bowers and Pretty 2009).
28
Although rectification of photographic distortion of two-dimensional bitemark images were
studied (e.g., Rawson et al. 1986, Bowers and Johansen 2002), rectification of bitemark
distortion caused by the dermal elasticity as an intrinsic biological characteristic has been
limited. Harvey et al. (1976) suggested that pre-existing tension lines, i.e., Langer‘s lines, in
human skin can be one of the factors that determine skin distortion, but did not propose a
rectification method. Bush et al. (2010a) commented that skin distortion concerns variables in
skin too complex to allow for predicting a single distortion factor, while affine
transformations of distorted bitemarks (Stols and Bernitz 2010), are some examples of recent
attempts to objectively rectify bitemark distortion.
Recognition of limitations in two-dimensional analysis from the fact that bitemarks are threedimensional, combined with the advancement in 3D data capture technology brought forth
studies that capture and analyse bitemarks in 3D (Thali et al. 2003, Blackwell et al. 2007,
Lasser et al. 2009, Martin-de las Heras and Tafur 2009).
Whereas forensic bitemark analysis historically involved examining the physical pattern of
the bitemark and the biter‘s dentition, recent advances in molecular biology techniques made
possible the collection of genetic material from bitemarks, including saliva (Sweet and
Hildebrand 1999, Sweet and Shutler 1999, da Silva et al. 2006, Kenna et al. 2011) and
bacterial flora from saliva (Borgula et al. 2003, Rahimi et al. 2005, Tompkins 2005, Kennedy
2011), that are of evidential value to identify the biter.
29
1.2.2.3
NAS report and forensic odontology
The recent NAS report (2009) recognised that lack of scientifically established proof in (1)
the uniqueness of human dentition; (2) the ability of the dentition to transfer a unique pattern
to human skin and the ability of skin to retain that uniqueness; and (3) the standard of type,
quality and number of individual characteristics to reach evidentiary value threshold; are
some of the basic problems inherent in bitemark analysis and interpretation. While the report
has been an impetus in a number of legal reviews (e.g., Garrett and Neufeld 2009, Neufeld
and Scheck 2010), the scientific community has been aware of such issues and has been
continuously attempting to address them (Pretty 2005a).
Reliability issues in analysing bitemarks on skin were recognised from early stages of the
method, and warnings were repeated in subsequent studies (e.g., DeVore 1971, Whittaker et
al. 1975, Pretty and Turnbull 2001, Pretty 2005b, Saks and Koehler 2005, Bowers 2006,
2008). The complicated process in the creation and retention of bitemarks on elastic, threedimensional skin, where its manifest on a living victim is further subject to changing
dynamics of vital response to the injured tissue, and the limitations in current methods to
capture, record, analyse, compare and evaluate the marks, are among the issues that bitemark
identification experts and researchers acknowledge (Bernstein 2005). Positive elimination of a
suspect from discrepancies in dental features are far more easier and reliable than to assess a
case where matching features are present (Ström 1963). ABFO terminology guidelines (1997)
specifically recognises ―that a statement of absolute certainty such as "indeed, without a
doubt", is unprovable and reckless‖ in declaring a match. Pretty and Turnbull (2001) reported
a case where the dental features in a bitemark did not sufficiently represent the dental
uniqueness between two suspects, demonstrating that dental uniqueness does not
automatically translate to bitemark uniqueness. Miller et al. (Miller et al. 2009) also
demonstrated that individuals with similarly aligned dentitions were difficult to distinguish
30
the individual biter from bitemarks, with some cases where known non-matches appeared
better correlated to the bite. Ante-mortem and post-mortem bites were simulated on porcine
skin to evaluate the error rates by examiners with different levels of bitemark analysis
experiences (Avon et al. 2010).
A series of studies by Bush and colleagues (Bush et al. 2010b, Bush et al. 2011a, b)
demonstrated the difficulty in reproducing similar bitemarks on human skin even when the
dentition was the same, and concluded that ―statements of dental uniqueness with respect to
bitemark analysis in an open population are unsupportable.‖.
While uniqueness was generally accepted as an essential premise in establishing individuality,
some recent forensic scholars are questioning the dogma. Interestingly, a law review paper
criticised the notion of individualisation as non-existing (Saks and Koehler 2008), while Page
et al. (2011) countered by not only accepting that reaching definite proof of uniqueness is
considered impossible by modern scientific methods, but to suggest that uniqueness is not
even the essential requirement for forensic conclusions of identity. They argue that it is the
forensic expert‘s role ―to provide the trier of fact with additional probative information,
within the bounds of their expertise, that may strengthen or weaken the likelihood of guilt‖,
not to individualise or identify. Cole (2009) also applied the Bayesian model to fingerprint
identification arguing that forensic trace element disciplines ―can live without these concepts‖
of individualisation and uniqueness. The use of likelihood ratio under the Bayes‘ Theorem
was likewise demonstrated in bitemarks to assess its evidential value in court (Kittelson et al.
2002, Kieser 2005).
31
1.2.3 Firearm and toolmark examination
Firearm and toolmark examination is a forensic science discipline associated with the analysis,
comparison and evaluation of a mark to determine its source tool (Davis 1958, AFTE 1998,
Dutton 2009, Katterwe 2009, Hueske 2011). By broader definition, tools include a much
larger entity such as tyre marks and bitemarks where these are likewise interpreted as specific
cases of tools, i.e., tyres or teeth, leaving marks on the softer substance, i.e., snow or cheese.
Sweet (1997) noted the similarity of bitemark analysis with other forensic methods that
compare the characteristics of physical evidence. Firearm and toolmark analysis is one of
such approaches.
For the purposes of this study bitemarks were considered a specialised form of toolmark,
where the harder substance tool (tooth) produces a toolmark (bitemark) as a physical
deformation on the softer substance (wax) as the result of forceful contact and movement
(bite).
1.2.3.1
Toolmark variation
Toolmark features that are of the forensic examiner‘s primary interest to establish it source
are categorised into class, sub-class and individual characteristics (AFTE Training and
Standardization Committee 2001, Hueske 2011).
A class characteristic is a feature indicative of a restricted group source. It can originate from
the tool type, size, or make, which could vary widely from a hammer to hacksaw. The
toolmarks‘ features specific to the group of similar tools form the basis of class characteristics
A sub-class characteristic is an incidentally produced feature during tool manufacture
common in a subset of a class group, which can originate from the manufacturing tool, e.g.,
32
die, press-forge, press-cutting. An individual characteristic is a feature sourced from applied
or acquired surface irregularity as a result of incidental or random process during manufacture.
Individual characteristics can also originate from incidental irregularities as the result of tool
use, wear or modification. Incidental damage, normal wear and tear, intentional modification
can all leave characteristic toolmarks unique to the individual tool at that point in time.
In firearms, there are many sources of variation that constitute characteristics used in
identifying the source (Hamby and Thorpe 1999, Hueske 2011). Calibre, model and make of
the firearm are the primary sources of class characteristics that define the firearm‘s
specifications, and are results of design prior to the manufacture. Rifling marks on bullet
surfaces result from the bullet forced though the spiral grooves and lands (rifling) on the inner
surface of the barrel, collecting the rifling features as striation marks. The number and widths
of grooves/lands, the twist ratio and its direction, are class characteristics of the rifling
imprinted on the bullet surface. Depending on the manufacture process rifling marks can also
differ, i.e., cut, broach, button and swage rifling. Incidental imperfections during the
manufacturing process are the source of individual characteristics in rifling marks. When the
firearm is discharged, the force of expanding gasses not only propels the bullet forward
through the barrel, it is applied in all directions from the cartridge case internal cavity towards
the chamber wall and breech face, resulting in impression toolmarks on the cartridge case wall
and head. Class characteristics can be recognised as the firing pin/hole shape and type, while
incidental machining marks in the breech face or firing pin surface can be a source of
individual characteristics. Individual marks can be imprinted on the cartridge case by the
chamber wall as a result of any pre/post discharge processes, i.e., chambering, expansion
during firing, and extraction. Additionally in multiple-round automatic, semi-automatic and
bolt/lever action repeating firearms, components such as the extractor, ejector, and magazine
33
lips which come into contact with the cartridge case during the firing cycle can likewise be
sources of class or individual characteristics. A historic overview of methods applied and
attempted to compare class and individual characteristics is summarised in Biasotti (1989),
from visual microscopic comparisons, photography, hologram, replica, to 3D analysis.
In order to determine a match or non-match in toolmark examination, the examiner must be
satisfied that within-source variation is sufficiently smaller than between-source variation.
The criteria to conclude an identification is defined by the AFTE (1992). Biasotti (1959)
successfully made an attempt to statistically individualise the striation marks by defining a
match from by counting the number of consecutively matching striations (CMS). The AFTE
criteria and the seemingly objective CMS, however, are recognised as having an subjective
component ―even when based on specific guidelines‖ and is therefore prone to errors (Biasotti
and Murdock 1984, Miller and McLean 1998).
The recognition of such issues in the discipline has resulted in a large body of research that
have studied the individuality and repeatability of toolmarks, and efforts to develop an
objective identification criteria (e.g., Skolrood 1975, Freeman 1978, Deinet 1981, Murdock
1981, Hall 1983, Singh and Aggarwal 1984, Matty 1985, Uchiyama 1986, 1988, Biasotti
1989, Hall 1992, Uchiyama 1992, Brown and Bryant 1995, Vanderkolk 1995, Masson 1997,
Brundage 1998, De Kinder et al. 1998, Miller 2001, Grzybowski et al. 2003, Faden et al.
2007, Howitt et al. 2008, Hamby et al. 2009, Bachrach et al. 2010, Hamby 2010). The large
number of research papers illustrates that efforts to strengthen the scientific basis of the
firearm and toolmark examination discipline is up and alive, regardless of the
recommendations of the NAS report (2009)
34
1.2.3.2
Toolmark generation
The principles in toolmark generation and its identification are summarised below from Davis
(1958):
A toolmark is defined as two materials coming in forced contact with each other, particularly
where one of the two materials is harder than the other. The forced contact typically results in
physical alteration, e.g., disfiguration, of the softer material. During the contact process,
alteration could also occur on the harder material, albeit to a lesser extent. The harder of the
two materials is referred to as the tool, while the patterned disfiguration of the softer material
is referred to as the toolmark.
Depending on the direction of movement between the two materials, toolmarks can be either
an impression type, or a striation type toolmark. Firstly an impression type toolmark occurs
when the tool moves against and away from the marked object in a direction which does not
involve lateral movements of either the tool or the material. This action results in the tool‘s
physical characteristics transferred as a negative image of the characteristics onto the marked
object (Figure 1-1,(b)) A striation type toolmark occurs when lateral movement is combined
while the tool is in contact with the substance. This results in the tool‘s contact points being
continuously marked as the lateral movement progresses, thus transferring the drag marks of
the tool‘s physical characteristics onto the substance. In striation type toolmarks, the physical
characteristics of the tool which made last contact with the marked material will be resulting
in the final striation pattern after the tool has passed (Figure 1-1,(a, c)). Combinations of the
two toolmark types can occur depending on the dynamic relationship between the tool and the
marked substance while in contact (Figure 1-1). It is therefore possible that different striation
35
toolmarks can be generated from the same tool depending on the exact contact point of the
tool with the marked substance.
(a)
(c)
(b)
2
1
Striation
Impression
Striation
Tool
Material being marked
Tool movement/pressure
Point(s) of contact with tool
Toolmark
Area of contact with tool
Figure 1-1. Principles of toolmark creation.
1.3 Identification in wildlife
1.3.1 Wildlife forensics
By definition, forensics has its mandate to serve for legal purposes in the human domain.
Where wildlife management concerns enforcing legal protection of wildlife resources or
investigating a nuisance source in pest control, wildlife forensics provide information to
relevant authorities acting on the issues. Varied fields in veterinary medicine were identified
to aid in this pursuit, which included forensic pathology, gunshot and arrow wounds,
36
poisoning, and identification techniques such as species, sex, origin of tissue, hair, and carcass
part matching (Stroud and Adrian 1996, Cooper and Cooper 1998, Stroud 1998).
In contrast to human forensics, wildlife forensic examination is often complicated by (1) the
lack of species-specific definitions for evidence which frequently takes the form of body part
or product, and (2) that a dead animal may mean either a legal or illegal killing depending on
circumstances (Goddard 2011). Where cases involve only a part of the body to be traced back
to the species level, molecular techniques provide reliable identification methods (Verma and
Singh 2003). For example, illegal trade was investigated in elephant (Loxodonta africana)
tusks and rhino (Ceratotherium simum, Rhinocerus unicornis and Diceros bicornis) horn
(Bollongino et al. 2003), tiger (Panthera tigris) bone (Wetton et al. 2004), and whale meat
(Baker et al. 1996, Lukoschek et al. 2009, Baker et al. 2010).
1.3.2 Species identification
Differentiation between species forms the basis of traditional taxonomy, although it is not
always easy to identify the species from its morphological characteristics. Furthermore,
genetic investigation methods have changed the landscape of taxonomy by providing a highly
precise tool to differentiate the population, e.g., Hector dolphin (Cephalorhynchus hectori)
subspecies (Baker et al. 2002), or disputing morphological classification of lichen species
Usnea florida and U. subfloridana (Articus et al. 2002).
In field situations, species are commonly identified from its outer morphological
characteristics, such as body configuration and colour, when it is observable (Salmon 1968,
Haltenorth and Diller 1980, Robertson and Heather 1999). However, direct observation is not
37
always feasible and identification from its trace or field signs is necessary, especially when
they are rare or elusive species (e.g., Murie 1954, MacKenzie et al. 2004).
1.3.2.1
Footprints
Footprints (tracks, spoor) are one of the commonly used traces that indicate the presence of
animals, which has been used in many species including mountain lions (Puma concolor)
(Van Dyke et al. 1986, Beier and Cunningham 1996) gray wolves (Canis lupus), coyotes (C.
latrans), red foxes (Vulpes vulpes), skunks (Mephitis mephitis and Spilogale putorius),
raccoons (Procyon lotor), and bobcats (Felis rufus), (Sargeant et al. 1998). Ratz (1997)
described New Zealand small mammal species that can be identified from their inked
footprints in track tunnels (King and Edgar 1977), while computational differentiation of the
three introduced rat species, i.e., ship rat, Norway rat and Pacific rat (= kiore, Rattus exulans),
were done from their footprints (Yuan et al. 2005). Palma and Gurgel-Gonçalves (2007)
found that multivariate analysis can identify Brazilian small mammal footprints, and further
access intraspecific variation, e.g., sex and geographic, from morphometrical analyses on
footprints. In Asia, black bears (Ursus thibetanus) and sun bears (Helarctos malayanus) were
distinguished from claw mark measurements (Steinmetz and Garshelis 2008). African
elephant footprint lengths were found be significantly correlated to the individual‘s shoulder
height and was used to assess the population‘s age structure (Western et al. 1983).
1.3.2.2
Sound
Acoustic identification from audio signals unique to the animal can be used to determine the
presence of a species, particularly when visual contact is difficult. Traditionally this method
relied heavily on the observer‘s (listener‘s) experience and/or knowledge of the species‘ call
or sound. Recent incorporation of computing technology has greatly enhanced the recognition
38
capability, including the use of sound levels that are normally outside the normal human
hearing range. For example, species were correctly identified from vocal emissions by nine
frog species (Eleutherodactylus spp)and three bird species (Patagioenas squamosa, Loxigilla
portoricensis and Coereba flaveola) of Puerto Rico (Acevedo et al. 2009) and sounds made
by four British grasshopper (Orthoptera) species (Chesmore and Ohya 2004). In mammals,
bat calls from 20 species were successfully classified using multiple discriminant function
analysis, classification and regression trees, and artificial neural networks, with varied levels
of ease depending on the bat family in the study, i.e., Vespertilionidae and Rhinolophidae
(Preatoni et al. 2005). Seismic recordings of animal footfalls were also used to accurately
(~82%) differentiate between species (Wood et al. 2005).
1.3.2.3
Hair
Microscopic examination of hair morphology, e.g., configuration, medulla/cortex ratio,
cuticle, pigment granules, and ovoid bodies, can be used to identify the animal origin
including humans (Deedrick and Koch 2004). Identification of pine marten (Martes martes)
from hair traps were used to monitor its presence and abundance (Lynch et al. 2006).
1.3.2.4
Genetic material
Genetic species identification can be performed on trace material left by the animals. With the
technological advances allowing for less stringent sample material requirements, genetic
analysis is becoming a common method to identify species. Mitochondrial DNA from faecal
matter was used to identify three sympatric lagomorphs species in New England (Kovach et
al. 2003), an elusive animal visiting a zoo was found to be a leopard from its faeces (Verma et
al. 2003), the distribution of the endangered Chinese tiger (Panthera tigris amoyensis) was
confirmed from faeces (Wan et al. 2003), and differentiation between sympatric Mexican
39
gray wolves (Canis lupus baileyi) and coyotes, which can be mistaken for each other because
of their similar appearances, was demonstrated by faecal DNA analysis (Reed et al. 2004).
Saliva retrieved from domestic sheep carcasses identified coyote to be the perpetrator in 95%
of the 18 cases (Williams et al. 2003).
1.3.2.5
Bitemarks
Determination of the biting animal species from its bitemark on human skin has important
forensic implications (Souviron 2005). Intercanine widths, (Tonn et al. 2004), and
additionally the size and shape of the dental arch (Tonn et al. 2005) were studied by
comparing visual patterns and measurements from bitemark records and animal skulls.
Murmann et al. (2006) studied 486 museum specimens of 12 carnivore species to compare the
jaw shapes and bitemark patterns of the species. The study further examined simulated
bitemarks made by the skulls and developed a modified intercanine measurement method that
incorporates the variation of canine puncture wound width depending on the depth of
penetration.
In a case report, carnivore bitemarks on a human victim were identified to be from a mountain
lion, and its sex and approximate age were likewise profiled. The bitemark analysis performed
in this case was later confirmed by matching the hunted animal‘s dentition to the victim‘s
wound and detecting her DNA in the animal‘s claw (Rollins and Spencer 1995).
Bitemarks are likewise useful in palaeontology, where pathological / traumatic marks in
fossilised remains are examined to reconstruct the event and understand ecological aspects, by
identifying the extinct species from its dental configuration such as tooth type, arrangement,
40
number and size. Haynes (1983) examined biting damage to 326 femora and tibiae of 7 extant
herbivore species by 10 extant species of carnivores to apply the findings for understanding
what damage marks on fossil bones can imply. Identified species in fossil remains were used
to interpret events such as Tyrannosaurus predation on Triceratops and Edmontosaurus
(Erickson and Olson 1996), mosasaur predation on nautiloids, ammonoids (Tsujita and
Westermann 2001, Kauffman 2009) or another mosasaur (Everhart 2009), shark predation on
an elasmosaur (Everhart 2005) or a dolphin (Bianucci et al. 2010).
Similar to human forensic odontology, class characteristics of a bitemark can assist in
determining the origin species responsible for the bite. Bitemarks left in prey animal and plant
species, are useful to assess the biology, ecology and behaviour of the biter, which provide
important information to make management decisions. Murie (1954) included bitemarks as
part of field signs to identify North American animals, and provided tooth-mark width
measurements to identify selected species. Ratz et al. (1999) found from necropsies of 56
New Zealand birds that 4 yellow-eyed penguin (Megadyptes antipodes) chicks, 3 royal
albatross (Diomedea epomophora) chicks and one little penguin (Eudyptula minor) adult had
predator tooth puncture holes on their skin or bills. They determined the likelihood of stoats,
ferrets (Mustela furo), and cats (Felis catus) being the biter by comparing the measured
bitemark intercanine distances to intercanine distance distribution in a sample pool of
museum specimens. Lyver (2000) similarly overlaid distances between bitemark puncture
holes on sooty shearwater (Puffinus griseus) carcasses with intercanine distances of probable
predator species caught in southern New Zealand, and found that direct examination of
bitemarks on bone underlying the bitten soft tissue can increase statistical reliability of the
method. Examination of intercanine distances of bitemarks on Hutton‘s shearwater (Puffinus
41
huttoni) carcasses with a similar method showed that stoats were the main predator of this
species (Cuthbert 2003).
Species contamination in trace detection methods is avoided by the ability to clearly
distinguish the interference marks between species (spoor, bitemarks) (Linhart and Knowlton
1975, Conner et al. 1983, Thomas et al. 2003), or by eliminating multiple species issues by a
species specific device design with specific placement or type of bait (Bamford 1970). False
identification of the target species may severely affect the output.
The usefulness of bitemarks on bait to determine the presence of an animal depends on the
ability of the bait material to retain bitemark information to allow species identification from
the marks. Commercially available bait-interference blocks make use of this ability to enable
differentiation between species from bitemarks (e.g., WaxTag®, Census™). Bait types to
detect rodent presence from its bitemarks were found to be preferred by the biter in the order
of, chocolate, cheese, soap, wax and oiled wood (Ji et al. 1999), while the least favoured wax
and oiled wood were found the most durable in terms of bait longevity and environmental
decay.
Identification of New Zealand mammal species, i.e., rodents, possums, European hedgehogs
(Erinaceus europaeus) and rabbits (Oryctolagus cuniculus) from bitemarks is possible on
wax-block bait (Thomas 1998, Thomas et al. 1999, Thomas et al. 2002, 2003) but not on
flour-paste, apple bait, or orange quarters. The actual species differentiation, however, has
never been documented in detail, and there exists no accessible bitemark type reference
material for using wax blocks to identify the presence of possums or rats. A new type of bait42
interference multi-species detection device called the chew-track-card was recently
developed, which combines (1) a food-lured corrugated plastic board as the material to be
bitten, with (2) an ink-based footprint detection device. The combination increases the
probability to detect and identify the interfering animal from its bitemark pattern and/or
footprint (Sweetapple and Nugent 2011).
1.3.3 Individual identification
The concept of identifying individuals is applied in wildlife management to obtain population
parameters such as social behaviour (Katona et al. 1980, Moss 1983, Frank 1986, Moss 1996)
and population abundance (Allen et al. 1996, McGregor and Peake 1998, Kelly 2001, Wilson
and Delahay 2001, Lukacs and Burnham 2005, Sharma et al. 2005, Reisser et al. 2008).
Differentiation between individual animals rely on detectable variation of characteristics
within the population. Such characteristics could be: (1) outer appearance of the animals if
they are observable, e.g., size, pattern and markings; (2) biological material, e.g., faeces, hair
and saliva, if samples could be collected; or (3) traces of activity such as footprints, calls or
bitemarks.
1.3.3.1
Markings
Spot patterns in a variety of vertebrate and invertebrate species could be used (Speed et al.
2007), including whale shark (Rhincodon typus) pattern (Arzoumanian et al. 2005, Meekan et
al. 2006), cheetah (Acinonyx jubatus) spot pattern (Kelly 2001), and polar bear (Ursus
maritimus) whisker spots (Anderson et al. 2007). Palatine maculation in the European badger
(Meles meles) could be used to identify individuals although the pattern was found to be
43
influenced by parasitic infection as cubs (Nouvellet et al. 2011). Advances in computed face
detection technology demonstrated that individual lions (Panthera leo) in video recordings
could be recognised and tracked (Burghardt et al. 2004a).
Apart from naturally occurring variations, incidental scar marks or nicks, e.g., in whale sharks
(Speed et al. 2008), cetacean fin and fluke edges (Katona et al. 1980, Würsig and Jefferson
1990, Black et al. 1997, Gowans and Whitehead 2001, Auger-Méthé and Whitehead 2007),
and African elephant ears (Moss 1996) could be used to identify individual animals. When it
is necessary to identify individuals that have few distinguishable characteristics, artificial
variation could be applied by visually, chemically or electronically marking individuals, using
invasive marking methods, e.g., tagging , and ear notching, or non-invasive marking methods,
e.g., collaring and painting (Silvy et al. 2005). Although artificial marking is a useful method,
it may necessitate extra consideration to animal welfare (Beausoleil et al. 2004).
1.3.3.2
Biological material
Biological excretions such as faeces, urine and saliva frequently include epithelium cells of
the host, which could offer genetic material to analyse its individuality. Hair, shed skin and
feathers not only could morphologically determine the host species, but its associated cells
could similarly yield genetic material to identify individual animals. Genetic analysis is a
powerful tool to study the population‘s demography and genetic composition (DeYoung and
Honeycutt 2005, Waits and Paetkau 2005, De Barba et al. 2010), while individualisation
offers information commonly used for population abundance estimation studies such as in
martens (Mowat and Paetkau 2002) and black bears (Boersen et al. 2003).
44
1.3.3.3
Traces and signs
When direct observation of the animal or collection of biological material is not feasible,
variations in traces or signs of the animal‘s activity could be used to identify individuals. For
example, footprints of mountain lions (Smallwood and Fitzhugh 1993, Grigione et al. 1999,
Lewison et al. 2001), tigers (Panthera tigris) (Sharma et al. 2005), black rhinoceros (D.
bicornis) (Jewell et al. 2001), and fishers (Herzog et al. 2007) were used to identify
individuals. Variation between individual acoustic characteristics from monkeys (Ceugniet
and Izumi 2004) bats (Siemers and Kerth 2006), and birds (Cheng et al. 2010) also allow
identification.
1.3.3.4
Bitemarks
Variation in dentition patterns could also yield information useful to identify individuals
among a population, as demonstrated in human forensic odontology. Individual variation in
intercanine and inter incisal distances positively identified three out of five stray dogs in a dog
attack case (Santoro et al. 2011). In wild-living individuals, considerable variation was
observed in the dentition of 350 skulls from Banks Peninsula possums (Gilmore 1966),
mostly involving the absence of vestigial upper first premolars and lower second incisors,
with only 48.6% of specimens having the typical dentition formula (3/2, 1/0, 2/1, 4/4). The
possum‘s use of teeth in feeding is described by Gilmore (1966) as food held in the anterior
teeth and torn by a sudden quick pulling of the head inward to the body. The biting action of a
possum leaves a bitemark on wax blocks distinguishable from rodents, hedgehogs and rabbits
(Thomas and Meenken 1995, Thomas et al. 1998, Thomas et al. 1999, Hunter 2004,
Kavermann 2004, Thomas et al. 2004), though no study has documented detailed dentition
patterns on the blocks. Rodents‘ gnawing behaviour leaves small multiple bitemarks scratched
on the bitten surface. Rodents‘ and lagomorphs‘ first incisors continuously grow and wear
down at the same turnover rate to ensure a sharp cutting edge at all times. Such feature may
45
possibly make the edge inconsistent over extended time spans resulting in different dentition
pattern of the same individual animal, though the degree of change or comparability is
unknown.
Positively identifying individual animals from dentition patterns imprinted on wax as
bitemarks has not been hitherto attempted. This study shall allow evaluation of multiple visit
occurrence and its effect on the bait interference technique, and has the potential to expand
into a wide range of applications in wildlife research.
For the purpose of this study, physical characteristics of bitemarks made by animals on wax
medium were classed as a specialised form of toolmarks. Enamel and dentine of the animals‘
teeth which come to contact with the wax during biting process are considerably harder than
the wax substrate, and therefore the resulting deformations in wax were interpreted as a
toolmark made by the tool, i.e., teeth.
1.4 Objectives
The abovementioned issues were identified in monitoring New Zealand mammalian pests
with the current use of bait-interference methods — WaxTags® in particular. The species
identification component of the method provided insight to forensic science approaches,
which in turn presented the possibility to identify individual animals from its bitemarks. This
research was an attempt to develop a means to investigate ecological questions with a forensic
science approach.
46
The objectives of this PhD study were to:
1. Develop a methodology using WaxTags® that allows the species identification of
possums, ship rats, and Norway rats, with definitions of the bitemark characteristics
for each of these three species (Chapter 2).
2. Develop, using laboratory and captive trial techniques, a methodology for using
WaxTags® to identify individual possums from their bitemarks (Chapter 3).
3. Undertake field trials to develop a WaxTag® methodology that allows the
identification of individual possums in the wild (Chapter 4).
4. Determine the prevalence and extent of multiple visits by possums to WaxTags®, and
the influence this has on the reliability of the WaxTag® as a monitoring tool
(Chapter 4).
5. Investigate the utility of WaxTags® to elucidate new information about the home
range, distribution, and activity patterns of individual possums (Chapter 4).
1.5 Structure of thesis
In order to fulfil the objectives, this thesis is arranged as follows: a general introduction and a
literature review (Chapter 1), the defining characteristics of bitemarks on WaxTags®
(Chapter 2), the development of a novel method to identify individual animals (Chapter 3),
the field application of the individual identification technique (Chapter 4), and finally general
discussion and conclusions (Chapter 5).
47
Chapter 2
Species Identification From Bitemarks on WaxTags®
2.1 Introduction
The bait interference method originated by Bamford (1970) was designed to infer the target
animal population abundance from the number of non-toxic bait interfered with by the study
species. In order to ensure interference by the target species, the brushtail possum, Bamford
limited accessibility to the flour-paste bait by raising and shielding the bait to exclude rats and
mice. Later development of the method used wax as the non-toxic medium to be interfered
with, where ―species-specific bite-marks‖ were identified from wax detection devices and
thus eliminating need for elaborate device design to restrict accessibility to the bait (Thomas
et al. 1999, Thomas et al. 2003). Thomas et al. (1999) noted that from bitemarks by captive
animals and from simulated bitemarks using the skulls of the animals ―marks could be easily
identified‖ when made by hedgehogs, possums, rats and rabbits. However, these studies did
not detail the bitemark characteristics that were specific to the species.
Current bait-interference method for monitoring possum abundance which uses the WaxTag®
as the detection device is standardised by a protocol published by the National Possum
Control Agencies (NPCA) — a collective of government authorities, researchers and contract
operators — (NPCA 2008a). According to the protocol species identification of the bitemark
is to be made by field operators who are trained and accredited through NPCA training
courses. The minimum identification achievement level required for a qualified operator is
stipulated in the protocol (NPCA 2008a) as follows: ―Of 40 bitten tags of known (caged
animals) origin, comprising approximately 50% possum bitten tags, the examinee may record
no more than 4 as unknown, and must identify at least 90% of the possum bitten tags
correctly, and may not identify more than one non-possum bitten tag as a possum.‖.
48
Whilst high accuracy (99%) was achieved in correctly identifying possums from simulated
bitemarks created by the skulls (Thomas et al. 2007), qualitative or quantitative criteria to
distinguish the species responsible for bitemarks is limited.
Although biological definitions of the four species can characterise its dentition‘s anatomical
arrangement, number and function, the WaxTag® method requires the species to be identified
from its resultant bitemark characteristics as the method can observe only the bitemarks, not
the animals. In contrast with static dental casts commonly used by dentists for faithful
reproduction of physical characteristics, bitemarks do not necessarily reflect all teeth
characteristics of the biter as it introduces additional variables including movement of the
teeth, movement of the bitten medium, ability of the medium to retain the mark, and
observability of the mark. Bitemarks made by captive target animal species on WaxTags®
were examined to assess and define bitemark characteristics of each target species.
The objective of this chapter was therefore:

To develop a methodology, using WaxTags®, that allows the identification of
possums, and rodents, by defining bitemark characteristics for each species.
2.2 Methods
In order to correctly observe characteristics of bitemarks made by a specific animal species,
animals of known species were exposed with WaxTags® to enable them to interfere, i.e., leave
bitemarks.
49
2.2.1 Captive animals of known species
A total of 36 brushtail possums, 17 ship rats and one Norway rat were live-trapped from
Banks Peninsula and mid-Canterbury of the South Island, New Zealand between April and
November 2005. An additional 21 mice that were live-trapped from the same areas between
October and November 2007, and 15 Norway rats (Wistar laboratory rats) obtained from a
biological supplier were also included in the study. Field specimens‘ species were identified
by its external morphological characteristics as described in King and Barrett (2005).
Animals were held in individual wire cages sized 100 × 40 × 52 cm for possums and 57 × 36
× 30 cm for rodents, with a minimum acclimatization period of two weeks. Each animal was
provided with a nesting chamber, food and water ad libitum. Maintenance feed for both
species were cereal based Opossum Pellets (Weston Animal Nutrition, 23 Southbrook Rd.
Rangiora, New Zealand). All animals were kept and handled under approved conditions by
the Lincoln University Animal Ethics Committee (Animal Welfare Act 1999).
2.2.2 Bitemark detection device: WaxTag®
The detection device used in this study was a commercially available product, the WaxTag®
(N.Z. patent 516900, Pest Control Research, Christchurch, New Zealand), designed to detect
the presence of animals that bite and leave marks on a wax block. The device consisted of a
rounded pyramid-shape 12 cc block of microcrystalline wax measuring 2 × 2.5 cm at its base
and 2.5 cm high moulded onto the acute end of a plastic isosceles triangle tag 6 cm wide and
12 cm tall (Figure 2-1).
50
6 cm
Plastic tag
Wax block
2.5 cm
12 cm
2.5 cm
2 cm
Figure 2-1. WaxTag® structure and dimensions.
Its application in New Zealand by the possum control industry is standardized the NPCA to
index the possum activity from the proportion of WaxTags® bitten per tag line each with 10 or
20 tags placed at 10m intervals in lines positioned at minimum 100m nominal distances,
retrieved after 3 or 7 days (NPCA 2005, 2008a).
The protocol requires tags to be used in the field by nailing or stapling the tag onto a tree or
log at 30 cm above the ground with the wax-block end of the tag pointing downwards. Lines
of 10 to 20 WaxTags® spaced at 10 m intervals are used as sampling units to calculate the
percentage of tags bitten.
Apart from the visual attraction by the plastic tag, additional attraction is provided by a visual
and/or gustatory lure such as a 5:1 mixture of flour and icing sugar rubbed onto the tree bark
above the tag in pre-protocol studies (e.g., Thomas et al. 2003, Thomas et al. 2004).
51
The WaxTag® operational principles are that an animal is visually attracted to the white
coloured tag (Hunter 2004, Kavermann 2004). In field application, the wax block portion of
the device collects interference data by the animal‘s biting behaviour leaving bitemarks on the
wax. From 2005, the WaxTag® producer modified the plastic tag colour from white to blaze
orange to assist in field tag relocation and recovery.
2.2.3 Presentation of WaxTags®
WaxTags® were individually marked with unique 4-digit identification codes from a random
number table to avoid recognition of the tag or animal to which it was presented, thereby
ensuring a double-blind trial upon bitemark observation. The tags were presented to each
individual cage 30 minutes before sunset and tag identification codes were recorded with
corresponding cage number and date. WaxTag® presentation to captive animals simulated
field conditions. Tags were attached vertically onto the cage wall by means of tie-down wire
through a hole opened on the wide base of the tag where it would instead be attached to trees
by a nail or staple in the field. An additional wire was tied at the base of the wax block
stabilizing the tag onto cages. WaxTags® for possums were not lured, and those for rodents
were presented with additional lure of approximately 2 ml peanut butter (half-peanut volume)
onto the wax block surface. Some individual animals were observed to start biting on the wax
immediately after presentation which necessitated checks for the first two 60-minute intervals
post-application for retrieval to avoid total stripping of the wax. Remaining tags were
collected the following morning. The process was repeated for up to five consecutive nights
from June 14, 2005 and October 6, 2005 to attempt retrieval of a minimum of three samples
bearing macroscopically observable bitemarks.
52
2.2.4 Photographic data capture
All collected WaxTags® were brought to the laboratory for further investigation. For the
purpose of measuring a sizable number of specimens in quick succession, the study took full
advantage of the recent affordability of digital photographic data collection and recording.
This non-contact process reduced the risk of damage of the original bitemarks on soft wax by
measurement apparatus, e.g., slide callipers. Additionally, Whittaker et al. (1975) noted that
tooth measurements from photographs are most accurate.
2.2.4.1
Setup
Collected specimens were observed with a stereoscopic microscope (Olympus SZX12) and
magnified images were captured with a digital camera (Olympus DP12) fitted onto the
microscope. Each specimen was observed at × 7 and × 10 magnifications for visible signs of
bitemarks, from which a minimum of one clearly demarcated bitemark was selected per bitten
tag, for taking measurements. Illumination was applied to the specimens by a cold light
source (Schott KL1500).
2.2.4.2
Measurements
In order to standardise the measurements the WaxTag® was attached onto a ball-jointed
mount allowing for universal angle adjustments. The wax block portion was set at an angle so
that the bitemark faced the microscope optic axis at 90 degrees. This was done by placing the
bitemark floor at the farthest distance to the microscope at its median ridge. The resulting
bitemark was positioned to the microscope‘s optical axis perpendicular to the tangent of the
dental arch at the median ridge formed by the inter-dental conjunction between the mesial
incisors (FDI: 11–21, 31–41) (Figure 2-2). Object illumination was supplied with a flexible
goose-neck output adjusted at approximately 45 degrees to the wax, providing a high-contrast
53
image to highlight the bite-mark delineations. WaxTag® and lighting position settings were
re-adjusted for each observation and subsequent image capture to enable a clear view of each
observed bitemark. Each tooth of the dental arch recorded on the wax was defined by the
aforementioned inter-dental ridge. Individual tooth width measurements were defined from
the optical axis of 90 degrees to the bitemark, i.e., from the biting animal‘s viewpoint (Figure
2-3).
In Figure 2-3 tooth mark width measurement scheme is illustrated for striation bitemarks by
(a) possums and (b) rodents as cross-section profiles in wax medium and corresponding
dentition creating respective marks. Observation and image capture was done by orientating
the bitemark with the optic axis perpendicular to the tangent at the median ridge
(corresponding to the median inter-incisor gap). Tooth mark width was measured from the
median ridge to the border of the distal ridge immediately distal to the median ridge, or to the
outer edge of the tooth mark when recognised as the outer edge of the median incisor, i.e. full
tooth mark width. Numbers describe FDI two-digit nomenclature of individual teeth.
54
12 11
21 22
1 cm
Median ridge
®
WaxTag
Figure 2-2. Origins of inter-dental ridges on WaxTag® bitemark made by possum upper
incisors. Numbers describe FDI nomenclature of individual teeth.
55
13
Optic axis
Optic axis
(a)
23
41
12
31
22
Edge
11
Ridge
21
Median
ridge
Median
ridge
Tangent
Tangent
Width
11
Optic axis
Optic axis
(b)
41
21
Edge
1 mm
Width
31
Edge
Median
ridge
Median
ridge
Tangent
Width
Tangent
Width
1 mm
Wax medium and surface
Possum incisor profiles: upper and lower incisors
Rodent incisor profiles: upper and lower incisors
Figure 2-3. Tooth mark width measurement scheme for striation bitemarks by (a) possums
and (b) rodents as cross-section profiles in wax medium and corresponding dentition
creating respective marks. Observation and image capture was done by orientating the
bitemark with the optic axis perpendicular to the tangent at the median ridge
(corresponding to the median inter-incisor gap). Tooth mark width was measured from
the median ridge to the border of the distal ridge immediately distal to the median ridge,
or to the outer edge of the tooth mark when recognised as the outer edge of the median
incisor, i.e. full tooth mark width. Numbers describe FDI two-digit nomenclature of
individual teeth.
56
Image acquisition and record maintenance was done by an image database and analysis
software analySIS® (Soft Imaging System GmbH, Munster, Germany). The camera was set at
auto-exposure measured at the image centre. Images were captured at a nominal × 7 optical
magnification while each bitemark image was accompanied by an identical frame with a scale
added into the field to calibrate image magnification. The captured images were entered into
the image database component of the analySIS® software at 2048 × 1536 pixel resolution
TIFF format. Bitemark widths were determined from an average of five on-screen
measurements taken from the stored images.
2.3 Results
2.3.1 Bitemark type recognition and description
Brushtail possum bitemarks typically display a continuous arc-shaped biting edge of the
combined incisors with differing tooth participation in upper and lower jaws. The upper
incisors form a near-half-circle with all 6 teeth (I1 – I3) participating to form the arc. By
contrast, the first incisors, I1, are the only teeth that constitute the half-circle arc in the lower
jaw. These incisor arrangement patterns can readily be observed as arc-shaped indentation
bitemarks, or as deep U-shaped engravements in bite-through bitemarks.
In rodents, a typical bitemark had two parallel grooves with a curved or flat-bottomed ‗w‘
shaped section with a ridge between representing the inter-incisor gap. Unlike possums, both
upper and lower rodent incisors comprise one pair each of I1 (I1) characteristically leaving
paired bitemarks.
57
1 mm
1 mm
1 cm
(a)
1 cm
(b)
1 mm
1 mm
1 cm
(c)
1 cm
(d)
Figure 2-4. Bitemark types by possums and rats observed from retrieved WaxTags® with
photograph insets of sample bitemarks: (a) possum impression bitemark, (b) rat
impression bitemark, (c) possum striation bitemark, and (d) rat striation bitemark.
58
(a) Rodent bitemark
(b) Possum bitemark
(a) Rodent upper teeth profile
(b) Possum upper teeth profile
(a) Rodent lower teeth profile
(b) Possum lower teeth profile
Figure 2-5. Schematic illustration of striation type bitemarks of (a) rodent and (b) possum
with cut-away section views to show profiles.
59
41
31
Enamel
Dentine
Pulp
Figure 2-6. Morphological characteristics of possum lower incisor impression bitemark cast
observed from dorsal view of occlusal facet of the incisal arch.
60
(a)
(b)
(c)
11
12
21
Enamel layer
22
Sharp inclination
13
Medium inclination
Moderate inclination
23
Direction of inclination
Figure 2-7. Morphological characteristics of upper incisal (occlusal) facet observed on ventral
view of incisal arch impression casts. Full-depth (a) and partial-depth (b) full arch
impression with teeth outline overlay, illustrating direction and degree of tooth facet
inclination among incisor teeth, and (c) schematic topographical illustration of upper
incisal facet with sharp, medium and moderate concave inclination, and direction of
inclination.
2.3.2 Bitemark width measurements
Width measurements were made of 511 single-tooth bitemarks made by the incisors of
possums, ship rats and Norway rats. Mean bitemark widths of single teeth of each species and
averages of each individual animal are summarised in Table 2-1.
61
Paired t-tests showed that bitemark widths differed significantly (p<0.05) between the possum
and each rodent species. Mean measurements for individuals also showed significant
differences (p<0.05) between the possum and rodents. Histograms of independently measured
bitemarks and those averaged for each individual animal demonstrate distinct differences
between possum and rodent bitemarks (Figure 2-8, Figure 2-9).
Table 2-1. Mean single-tooth mark width measurments from possums and rodents.
Tooth marks (mm)
Species
mean ± SD
T. vulpecula (combined)
2.87 ± 0.51
Upper incisors
Lower incisors
(sample size)
Individual’s mean (mm)
mean ± SD
(sample size)
(n=386)
2.88 ± 0.28
(n=30)
2.66 ± 0.40
(n=256)
2.68 ± 0.29
(n=29)
3.27 ± 0.44
(n=130)
3.34 ± 0.27
(n=25)
R. rattus
0.83 ± 0.19
(n=281)
0.84 ± 0.08
(n=11)
R. norvegicus
1.25 ± 0.25
(n=106)
1.24 ± 0.21
(n=7)
Wild
0.85 ± 0.16
(n=11)
0.85 ± [n/a]
(n=1)
Wistar (laboratory)
1.30 ± 0.22
(n=95)
1.31 ± 0.13
(n=6)
- Rattus spp (combined)
0.94 ± 0.28
(n=387)
1.00 ± 0.25
(n=18)
M. musculus
0.46 ± 0.07
(n=73)
0.45 ± 0.21
(n=3)
- Rodentia (combined)
0.87 ± 0.31
(n=460)
0.92 ± 0.30
(n=21)
62
M. musculus (0.46 ± 0.07 mm)
M. musculus (n=73)
R. rattus (n=281)
R. norvegicus (wild) (n=11)
R. norvegicus (Wistar) (n=95)
T. vulpecula - upper incisor (n=256)
T. vulpecula - lower incisor (n=130)
Normal distribution: species (mean ± SD)
Number of samples
R. rattus (0.83 ± 0.19 mm)
50
R. norvegicus (wild, 0.85 ± 0.16 mm)
R. norvegicus (Wistar, 1.30 ± 0.22 mm)
T. vulpecula - upper incisors (2.66 ± 0.40 mm)
T. vulpecula - lower incisors (3.27 ± 0.44 mm)
T. vulpecula (2.87 ± 0.51 mm)
0
0
1
2
3
Tooth width measurement (mm)
4
5
Figure 2-8. Single-tooth width measurement distribution of bitemarks by possums and rodents
Species
n = toothmarks (individuals)
n =73 (3)
M. musculus
R. rattus
n =281 (11)
R. norvegicus
n =11 (1)
(wild)
n =95 (6)
R. norvegicus
(Wistar)
n =106 (7)
R. norvegicus
(all)
n =256 (29)
T. vulpecula
(upper incisor)
n =130 (25)
T. vulpecula
(lower incisor)
n =386 (30)
T. vulpecula
(all incisors)
1
2
3
4
Tooth width measurement (mm)
Tooth width measurement (mm)
Figure 2-9. Box plot of tooth width measurements of bitemarks by possums and rodents.
63
5
2.3.2.1
Toolmark identification
Identification of toolmarks is the process of comparing and determining whether a toolmark in
question originates from a particular source. This procedure includes the matching of a
toolmark against another toolmark, or a toolmark against a tool. In order for such comparison
to be made possible, underlying precepts are two-fold: (1) observable or measurable unique
characteristics are present on tools to sufficiently differentiate one tool from another, and (2)
the unique characteristics are transferable and are reproduced on the toolmark.
The species biting WaxTags® produced a clear measureable feature of single-tooth width
which quantitatively defined the species. Class characteristics of bitemarks on WaxTags®
were described and found effective in identifying the species which interfered.
2.4 Discussion
2.4.1 Bitemark description
This Chapter demonstrated, by qualitatively describing bitemarks made by the brushtail
possum and rodent species, that it was possible to descriptively define the species from its
bitemarks. To my knowledge, this is the first study to descriptively define the species‘
bitemarks in detail, beyond what is ―obvious‖ or ―easily identified‖ as a ―V-shape‖ or ―chisel
shape‖ (e.g., Thomas et al. 1999).
Such class characteristics included the number of teeth involved in creating the bitemark, the
incisal arch shape and configuration in terms of participation ratio of each tooth to create the
semi-circular arch in the case of possums, or a more flattened arrangement in rodents.
64
2.4.2 Bitemark measurement
Width measurements of each medial incisor (FDI: 11, 21, 31, 41) were taken from bitemarks
of four species, i.e., brushtail possums, Norway rats, ship rats and mice, and its measurement
distribution was plotted in a histogram (Figure 2-8) and box plot (Figure 2-9). The results
show clearly that possum teeth measurements from bitemarks on WaxTags® were
significantly larger than any of the rodents.
2.4.3 Further study
The class characteristics of bitemarks on WaxTags® were qualitatively described and
quantified by measuring the single-tooth widths made by the species. This should allow for a
more objective verifiable identification of the species which interfered with the WaxTags®. It
should also reinforce the advantages of the WaxTag® method to provide for a technique to
monitor possum population presence and abundance as stipulated by the NPCA (2008a)
protocol.
Furthermore, investigation into subclass characteristics, e.g., sex, age-class, feeding habits or
animal‘s original habitat, that may have the potential to influence bitemark patterns, may
provide an easy and inexpensive method to detect further detailed population parameters, e.g.,
demographics, and behaviour.
By further observing the bitemarks in detail, it was assumed that individual characteristics can
be distinguished and provide a novel tool to individualise interferences in wax, which has the
potential to offer a variety of new ecological studies. This is attempted in Chapter 3.
65
Chapter 3
Identification of Individual Possums From
Bitemarks on Wax
3.1 Introduction
Following on investigation into bitemarks‘ characteristics on wax made by known species in
Chapter 2 to define and differentiate the species, this chapter investigates further detailed
characteristics of bitemarks by one of the species, the brushtail possum Trichosurus vulpecula
to develop and establish a novel method to identify individual animals.
On microscopic examination of possum bitemarks on WaxTags®, parallel scratch lines in the
direction of tooth travel were frequently observed on the bitemark floor. Such marks were
hypothesised to be caused by the biting animal‘s tooth edge scraping on the wax surface. By
observing possum dentition and considering the possum‘s biting behaviour (Gilmore 1966),
the responsible tooth type was likely assumed to be the incisors from the anterior dental arch.
As the edges of possums teeth did not appear to be identical, most possibly because of natural
or pathological wear, the features created as incidental (random) features seemed analogous to
individual characteristics in firearms and toolmark examination (AFTE Training and
Standardization Committee 2001). If the observed incidental imperfections on the incisors
were transferred to the wax in the form of striation toolmarks, it may offer means to identify
individual animals. Unlike human forensics, there is no need or possibility at this point to
identify the very individual animal who bit the wax because operational data that can be
collected from WaxTags® are limited to bitemarks, not the animals. However, if it was
possible to determine whether the same individual animal bit more than one tag, or whether
more than one individual bit a single tag, the most confounding issue in the bait interference
66
method may be addressed, i.e., two bitten tags do not necessarily mean two possums
(Bearman 2002). The core hypothesis to be tested in this chapter was whether bitemarks on
WaxTags® can be used to identify individual possums. As the bitemark on the wax medium is
essentially a toolmark created by the harder material (tooth) on the softer substance (wax),
and the marks were of the striation type, where the biting process made the teeth move in a
lateral direction while being pressed on the wax portion, a forensic toolmark identification
approach to match striation type toolmarks was attempted (Hueske 2011).
By establishing an identification method from bitemarks on WaxTags®, the first attempt made
for any animal species, it provides the potential means to study the multiple visit (contagion)
issue and strengthen the NPCA protocol (2008b). Furthermore, it can be used in a vast range
of research approaches to address biological, ecological, and epidemiological questions, e.g.,
behaviour and population abundance, where identifying the individual animal is necessary.
This may be particularly important for New Zealand possums as they are known to carry and
transmit bovine tuberculosis to livestock, necessitating to control its population density to a
level low enough to prevent transmission: such low density population is extremely difficult
to monitor. Obtaining presence/absence and location information of sparsely distributed
individual animals with an identification tool therefore has the potential to greatly enhance the
ability to monitor and control very small numbers of individual possums.
The objective of this chapter was therefore:

To develop, using laboratory and captive trial techniques, a methodology for using
WaxTags® to identify individual possums from their bitemarks.
67
3.2 Methods
In order to develop and verify a new method to identify individual animals from its bitemarks,
a three-stage process was designed. The first stage was to test the hypothesis that individual
characteristics that occur from animal teeth scraping on wax, and therefore leaving striation
type marks, can be used to identify individual animals. This was done by simulating
bitemarks on wax by scraping possum teeth from skulls, on wax surfaces, and investigating its
utility to identify the animal that produced the mark. Simulated bitemark samples were
compared to other samples of known individuals and classified to each individual.
The second stage involved a blind test of unknown combinations of individual possum skulls
to create simulated bitemarks. This was achieved by asking volunteers to produce simulated
bitemarks from possum skulls of their choice.
The third stage further developed the identification method so as to be applicable to live
samples done by known captive animals. This was further blind-tested from random samples
of bitemarks.
3.2.1 Simulated bitemarks from known individual’s dentition
Simulated bitemarks made in wax from skulls of known individual possums were used to
determine if microscopic examination of bitemarks could identify and differentiate individual
animals.
68
3.2.1.1
Individual animals
Individual possum carcasses were obtained from Landcare Research Limited — Manaaki
Whenua (PO Box 69 Lincoln, New Zealand, hereinafter Landcare). The animals were livecaptured from Canterbury Region, undergone acclimatisation and later used for reproductive
studies in captivity. Ten male and ten female post-experiment carcasses minus its
reproductive organs were frozen (-18°C) at Landcare and donated to this study May to June
2005.
All animals were labelled and kept frozen in Lincoln University freeze storage, until one to
two individuals each day were taken out to thaw overnight at room temperature. Heads were
separated from the body at the atlanto-occipital joint. Skin and soft tissue were removed as
much as possible but carefully enough not to damage the animal‘s teeth with the dissecting
instruments and its remaining body was discarded. Each individual was processed separately
and its head was refrigerated in individually labelled containers.
Each of the 20 individuals‘ heads were placed in separate compartments of fine-meshed nylon
net (recycled women‘s panty-hose) together with a waterproof paper label. Individually
contained heads with its mandibles were soaked in warm water (approximately 30°C) for one
hour, brought to boil and allowed to simmer for three hours before cooling at room
temperature. Individual mesh-net compartments were opened, its contents sieved, and
remaining soft tissue was removed from the bones and teeth using running water, brush,
spatula, needle, and syringe. Each individual was processed separately to avoid mixture of
body parts, i.e. cranium, mandible and teeth. Three days of air-drying at room temperature
was followed by 24 hours drying in a 40°C incubator.
69
All teeth of each individual were identified by its morphological characteristics and confirmed
by its fit into its respective tooth socket. Teeth were re-inserted into each socket, and
cemented by hot-melt adhesive (hot-glue). Mandible halves were aligned by temporarily
articulating the temporomandibular joint using putty-like adhesive, e.g. Blu-Tack®, at
maximum intercuspation, and later joined by hot-glue at the mandibular symphysis. The reconstructed cranium and mandible of each individual were labelled with unique four-digit
identification codes from a random number table, and recorded on a reference table to later
refer to its individuality and origin.
3.2.1.2
Wax medium
The wax medium on which bitemarks were to be simulated was prepared in 10 × 30 × 40 mm
block forms from microcrystalline wax, identical to that used for WaxTags®. Each blank
block were given a unique four-digit identification code, and labelled with a paper strip
melted into one long side of the block close to a corner. The label position was kept constant
to later assist in orientating the block and bite simulation directionality.
3.2.1.3
Bitemark simulation
The cranium was applied by hand to the wax block resting on a bench set at an approximately
45° angle, then pulled backwards while directing downward pressure to simulate the possum
biting a WaxTag® (Figure 3-1a). The degree of applied pressure was guided by the width of
resulting bitemark produced. In order to obtain a sufficient strip of teeth included in the
bitemark, the incisal edges of a minimum number of four teeth from the centre (I1 to I2 FDI:
12, 11, 21, 22) were to be in contact with the wax surface. This procedure ensured that the full
widths of I1 (FDI: 11, 21) made contact and their bitemarks were constantly produced. The
process was repeated until five simulated bitemarks were made for each individual cranium.
70
Cranium movement
(a)
Possum cranium
Bite direction
Wax block
(working surface)
Block number label
θ = 45°
(b)
1 cm
(c)
3 cm
000
Azimuth
090
Start
Working surface
270
4 cm
180
Bite direction
End
Block number label
Figure 3-1. Bitemark simulation setup: (a) wax block setup on stand in relation to possum
cranium and its movement to simulate bite; (b) simulation process with teeth scraping on
wax surface; (c) wax block dimensions and block number label position in relation to
simulated bitemark, with azimuth designations in relation to bite direction where bite
started at azimuth 000 and ended at azimuth 180.
71
Additionally, 10 volunteers were each asked to prepare six simulated bitemarks from skulls of
their choice. The volunteer‘s choice could be anything between one bitemark from each of six
individuals, to six bitemarks of one individual. Visual demonstrations and oral guidance of the
procedures were provided to the volunteer, although no instructions were given except the
total number of required simulations. The bitemark set produced by volunteers had added
variability to the number of individuals as well as the simulated biting process, with the
intention to emulate more closely the reality of factors influencing the bitemarks by actual
possums.
All wax block and cranium identification numbers were recorded to reference the matching
individuals after blind observation trials. A total of 160 bitemarks were produced, i.e., 100 by
all 20 possums, and 60 made by a masked number of individuals.
3.2.1.4
Observation
Stereomicroscopic observations were performed on the bitemarks and images were recorded
by digital camera (Konica-Minolta DiMAGE X31, 4.7–14.1 mm focal length, 320M pixel
capability, JPEG 2048 × 1536 pixel image quality). Image captures were made on an
adjustable platform where the wax block surface distance to the camera lens was adjusted to
4.5 cm to fix its magnification in macro mode (Figure 3-2). Each bitemark image frame was
accompanied with an identical frame including a scale to confirm its magnification size.
Frame numbers were recorded with the wax block identification numbers.
72
Camera
(a)
(b)
Light
source
Wax block with bitemark
Stereo
microscope
Light source
Specimen:
silicone cast
(c)
Figure 3-2. Possum bitemark microscopic characteristics observation setup. (a, b) simulated
bitemark photograph capture setup with digital camera, illumination and wax block with
bitemark; (c) bitemark cast observation setup with stereo microscope, illumination and
silicone cast of bitemark.
Illumination azimuths of 0, 90, 180 and 270 degrees in relation to the simulated bitemarks‘
start–end direction were trialled in various attack angles to the wax block surface, to explore
73
optimum data capture conditions (Figure 3-1). Aforementioned capture of digital images was
done according to each azimuth, resulting in a set of 400 known individual files, each
accompanied with a scale frame. Volunteer-produced bitemarks of masked individuals were
captured as 240 files each accompanied with a scale frame. The 640 captured image files were
further re-assigned 4-digit random-number filename to mask the image‘s origin during
observation trials, and the accompanied scale files were given a suffix ‖-scale‖ to the
corresponding image to allow for magnification standardisation, totalling to 1280 image files.
After image magnification adjustments were made to compensate for the camera‘s auto-focus
function, a 200 × 800 pixel portion of each bitemark image was cropped with the 400 pixel
image centre aligned to the wax ridge that corresponds to the medial inter-tooth gap between
I1 (FDI: 11, 21) to standardise bitemark images for side-to-side comparison (Figure 3-3).
74
Figure 3-3. Image cropping to 200 × 800 pixels.
Cropped images were digitally enhanced with algorithms built in image processing software
(Corel® PHOTO-PAINT versions 9 & 12) in order to develop an effect that allowed rapid and
accurate visual comparison of features for matching individuals. Enhanced images were
compared against each other to assess the accuracy of the matching method (Figure 3-5).
3.2.1.5
Comparison schedule
Cropped and enhanced bitemark images were grouped according to illumination azimuth and
each group was compared independently. For each azimuth group, the first stage comparison
involved randomly selecting a reference image from the five images of each individual
possum, obtaining a reference group of 20 images. The remaining 80 images were pooled as a
comparison group. All 80 images were visually compared side-to-side to the 20 reference
75
images and decided whether a match was recognised from the image characteristics (Figure
3-5). Only positive identification was recorded as a match, and the number of comparison
images matching any reference image was masked during the trial to blind the number that ―is
supposed to match‖ a particular reference. The theoretical number of comparisons made in
this trial is 20 × 80 = 1600 comparisons. This matching process was repeated 10 times for
each of the four azimuths, giving a total number of comparisons of 56,000.
3.2.2 Blind comparison of simulated bitemarks
The second stage of developing the comparison method involved comparing randomly
selected bitemark images to further simulate field conditions where no ‗reference‘ possum
exists to compare other bitemarks with.
At this stage it was found that observing silicone convex casts of concave bitemarks gave
more flexibility in observing and digitally capturing bitemark characteristics for comparison.
3.2.2.1
Cast preparation and observation
Trials at this stage revealed that lighting inside a concave surface at an acute angle to enhance
microscopic characteristics posed a significant challenge, especially as the clear or whitish
colour of wax gave limited diffuse reflection to recognise the surface topography. Convex
casts of the bitemarks were made from an elastic silicone polymer impression material, vinyl
polysiloxane, or polyvinyl siloxane, commonly used in dental and forensic applications
(Morgano et al. 1995, Du Pasquier et al. 1996, Sweet 1997, Wolff et al. 1998). The product
used in this study was Forensic Sil (Armor Forensics, Jacksonville, Florida, USA) due to its
neutral gray colour which creates higher tonal contrast in black-and-white images, and ease of
76
application by a self-mixing applicator. A sample image of bitemark cast production is
illustrated in Figure 3-4.
WaxTag
WaxTag
21 11
11 21
R
L
L
41 31
Silicone polymer
R
31 41
Bitemark cast
(b)
(a)
)
Figure 3-4. Silicone cast preparation on bitemark: (a) silicone polymer application onto the
bitemark and (b) bitemark cast removal after polymer curing. The mirror reversal of left
(L) and right (R) between actual bitemark and cast is demonstrated.
Observation and image capture of casts was done under a stereo microscope fitted with a
× 0.7–4.0 zoom objective lens and × 10 ocular lenses and using an elongate light source from
a fluorescent tube, and from a fixed azimuth of 90 degrees in relation to the bitemark start-toend direction. Micrographic images of the casts were captured on an Olympus DM12 digital
camera fitted to an Olympus stereoscope SZX12 at × 16 objective zoom and × 10 ocular
lenses.
77
Captured image files followed the convention of assigning 4-digit random-number
identification codes to mask its origin. After the initial trial using direct capture images of
bitemarks, it was found that image cropping was not necessary as long as the median interdental gap was recognisable to serve as a reference point in aligning the compared images
against each other. Direct side-to-side comparison of image files were done to determine the
feasibility of using casts to identify individual possums from the observed characteristics.
3.2.3 Bitemarks from captive known individual possums
The third stage used live possums to bite on WaxTags® to simulate field conditions. Thirtysix captive possums in individual cages were exposed nightly to WaxTags®, the bitten tags
collected, and the bitemarks were compared from direct images of the bitemarks, and silicone
polymer casts. This section determines whether actual bitemarks made by live animals on
WaxTags® can be used to identify individuals.
3.2.3.1
Individual animals, WaxTag® presentation and bitemark observation
The same source of 36 captive brushtail possums used in species identification (Chapter 2)
was used to collect 107 bitten WaxTags®. All 107 tags were examined for bitemarks and
silicone polymer casts were made for all bitten surfaces. Each individual was caged
independently and therefore it was possible to attribute the bitemark to the biting individual
animal.
3.2.4 Repeatability and performance
An unsupervised trial by a novice observer was made with an open-method approach. The
observer was instructed to classify 100 simulated bitemark images from 20 individual possum
78
skulls using an image enhancement algorithm of choice to the observer as most suited for the
purpose.
Supervised trials were made by two novice observers. Each received a minimum of one day in
training the matching procedure, which continued up to a time when the observer was
confident to start the comparison trial. The first stage comparison with simulated bitemarks
was repeated by the observers and compared to the results to evaluate the repeatability of the
method.
The performance evaluation and inter-observer comparison of results were done using
specific performance measures, i.e., accuracy, precision, specificity and sensitivity. Each
measure was derived from counts of true positives, false positives, true negatives and false
negatives scored by the observer as shown in Table 3-1.
Table 3-1. Confusion matrix to evaluate identification performance.
Actual value
Match
Observed value
Non-Match
Total
Match
True positive
(TP)
False negative
(FN)
Actual matches
Non-match
False positive
(FP)
True negative
(TN)
Actual
non-matches
Total
Observed
matches
Observed
non-matches
Where: true positive (TP) is equivalent with ―hit‖; true negative (TN) is equivalent with
―correct rejection‖; false positive (FP) is equivalent with ―false alarm‖ or ―Type I error‖; and
false negative (FN) is equivalent with ―miss‖ or ―Type II error‖.
79
Calculations to derive performance measures are described in the following equations (3.1) to
(3.4). Additionally, single-measure statistics, i.e., g-mean, Matthews correlation coefficient,
F-measure and Youden's index, were calculated to evaluate the performance quality of
identifications according to equations (3.5) to (3.8).
(3.1) Accuracy
Accuracy, or correctness, is the proportion of correct observations among all observations:
Accuracy 
TP  TN
TP  FN  FP  TN
(3.1)
(3.2) Precision
Precision, equivalent to the positive predictive value, is the proportion of correct positive
identifications (TP) with all observed positive identifications, i.e., the correctness of positive
identifications:
Precision 
TP
TP  FP
(3.2)
(3.3) Sensitivity
Sensitivity, also known as the true positive rate, hit rate or recall, is the proportion of correct
positive identifications (TP) against the actual positive matches:
Sensitivit y 
TP
TP  FN
(3.3)
(3.4) Specificity
Specificity, or true negative rate, is the proportion of correct negative identifications, i.e.,
exclusions (TN), against actual negative matches:
Specificit y 
80
TN
FP  TN
(3.4)
(3.5) Geometric mean
Geometric mean (g-mean) is a single-measure statistic where the resulting value between 0
and 1 indicates poor to perfect match performance:
g  mean  Sensitivit y  Precision
(3.5)
TP
TP

TP  FN TP  FP

(3.6) Matthews correlation coefficient
The Matthews correlation coefficient (MCC) (Baldi et al. 2000), also referred to as the phi
coefficient (φ) or mean square contingency coefficient, takes into account any classification
size differences and is particularly useful when the two classes are of very different sizes. The
calculated coefficient value is between -1 and 1, with 0 as a random match performance, 1 as
a perfect match performance, and -1 as a perfect inverse match performance:
MCC 
TP  TN  FP  FN
TP  FP  TP  FN  TN  FP  TN  FN 
(3.6)
(3.7) F-measure
The F-measure (van Rijsbergen 1979), or F-score, is the harmonic mean, or the weighted
average, of precision and sensitivity, where the score value between 0 and 1 indicates worst
and best match performances, respectively:
F 


 1  TP
  1  TP   2  FN  FP


2

 
  2 1 
F1  2 
2
2
Precision  Sensitivit y
 Precision  Sensitivit y

Precision  Sensitivit y
Precision  Sensitivit y
81
(3.7)
(3.8) Youden's index
Youden's index (Youden 1950), or Youden‘s J statistic, is one of the single-measure indices
to evaluate the binary classification performance which takes into account the sensitivity and
specificity of bitemark match measures:
Youden ' s index 
TP
TN

1
TP  FN FP  TN
(3.8)

TP  TN  FN  FP
TP  FN  FP  TN 
3.3 Results
3.3.1 Bitemark features for identification
Observation of possum bitemarks under the stereomicroscope revealed fine striations on the
wax surface running parallel to the direction of tooth travel where the wax had contact with
the incisal edges including upper I1 to I3 (Figure 3-1). These striations closely resembled those
found on WaxTag® bitemarks of upper and lower incisors by possums. Ripple-like stripes of
identical profiles occurring at near-perpendicular angles to the tooth travel axis were also
present on some samples (Figure 3-5), which resembled nicks or breaks in bitemarks on the
WaxTag®. The parallel fine striations were considered to result from tooth imperfections at
the last point of contact with the wax while it slid on the substrate in biting, and left scrape
marks. Ripple-like perpendicular stripes were considered to represent uneven travel speeds of
teeth during the biting motion, such as loose teeth or uneven hand pull in the simulated
bitemarks, or uneven pull or hesitation by the possum on WaxTags®. Image enhancement
techniques were used on the cropped images to improve feature detectability and individual
matching probability.
82
3.3.2 Image capture and enhancement
Due to the low-contrast tone of the wax bearing bitemarks, various observation and image
capture settings were assessed for the enhancement of features suitable for individual
identification. The most effective illumination set-up to observe and enhance bite-mark
features was when the light source was approximately at a 45° angle from the camera‘s
optical axis, and from an azimuth of 0° or 180° to the axis of tooth travel. Illumination
settings were explored with raw light emitted from a gooseneck light source at differing
elevations and azimuths to the bitemark. A single light source enhanced the shadow-sourced
contrast. In addition to the tonal differences between the diffuse reflection of the wax and
shadows created by the light source, specular reflection from the wax highlighted areas with a
certain tilt while simultaneously placing dissimilar angled areas, e.g., scrapes, as shadows
enabling a high contrast against the striations.
Image processing tools in standard image editing software (Corel® PHOTO-PAINT versions
9 & 12) were capable of enhancing microscopic striations on captured bite-mark images to
reveal their fine details. Contrast, brightness and tone-curve adjustment options were available
to enhance the details, though fine-tuning of single and combined adjustments were difficult
to record and apply to a large number of samples as a standard filtering process. Among the
standard image adjustment tools, local equalization (Gonzalez and Woods 2002) was effective
in enhancing the bite-mark‘s characteristics (Figure 4). This filter adjusts the intensity of each
image pixel according to a histogram equalization applied locally from its surrounding
neighbourhood of n × m pixels (50 × 50 default) thereby greatly enhancing subtle tones or
intensity changes in the image.
83
3.3.3 Individual match trials
Match trials on simulated bitemarks resulted in 97.41% ± 2.07 (mean ± SD) of correct
matched pairs. Two enhanced images to be compared were aligned so that the scrape mark
striations‘ dark/light band location, width and sequence could be visually matched to judge if
the two samples came from the same individual (Figure 3-5). In some bitemarks, ripple
patterns that assisted in identification were observed, though this was unreliable because the
same individual may or may not reproduce this pattern, whereas scrape mark striations were
more consistently found in the same individual. Various observation methods were tested,
including rotation of the computer monitor by 90°, to try to reduce the strain on the observer‘s
eyes.
Microscopic characteristics on bitemark casts were sufficient to allow identification (Figure
3-6). Direct visual comparison of images was adequate to align, compare, and conclude
whether the two specimens were a match or not.
84
(a)
(b)
(c)
(d)
(e)
Figure 3-5. Sample bitemark images cropped (a, d), enhanced (b, c) and aligned (b–c). A
schematic sample generated from greyscale intensity values (e) demonstrate positive
identification possibility from a pair of unequal sized images.
85
Figure 3-6. Example of bitemark characteristics on casts used to identify individual animals.
86
3.3.4 Repeatability and performance
3.3.4.1
Uncontrolled observation
The novice observer who was instructed to classify 100 simulated bitemarks chose auto
equalisation to enhance bitemark characteristics. The trial successfully resulted in 95% correct
classification with no false positives, and 5% false negatives. Five samples were not matched
to any group and were considered equivalent to inconclusive samples. Of the 20 individuals
used to simulate bitemarks 25% (5/20) individuals were each found to have 20% (1/5) of
bitemarks that were not identifiable in this trial.
3.3.4.2
Controlled observation
Performance comparisons of controlled trials by trained observers are shown in Figure 3-7 to
Figure 3-10. The percentage correct identification, accuracy, specificity, sensitivity and
precision of the matching performances of three observers were evaluated and illustrated in
Figure 3-7 and Figure 3-8, which showed two contrastingly different results. While all
performance indicators of one of the observers (AS) did not significantly differ from the
original trial (KS), the other observer (CS) was significantly scoring lower performances in all
except specificity and accuracy. Overall performance evaluation scores calculated from the
trials, i.e., g-mean, Matthews correlation coefficient, F-measure and Youden's index are
shown in Figure 3-9, and the fluctuation of percentage correct identification with increasing
trial repeats are given in Figure 3-10. Figure 3-11 illustrates the lack of significant
performance differences between the four illumination azimuths when the images are
enhanced by local equalization.
87
Correct
identification
Positive
identification(%)
(%)
100
50
0
KS
AS
CS
Observer
Figure 3-7. Reference match trial identification performance by three observers — percent
correct identification.
88
Value
1
0.5
0
KS
AS
CS
Observer
Sensitivity
Specificity
Precision
Accuracy
Figure 3-8. Reference match trial identification performance by three observers — sensitivity,
specificity, precision and accuracy.
89
Value
1
0.5
0
KS
AS
CS
Observer
g-mean(1)
Matthews correlation coefficient
F-measure
Youden's index
Figure 3-9. Reference match trial identification performance indices by three observers —
g-mean, Matthews correlation coefficient, F-measure and Youden's index.
Correct
identification
(%)
identification(%)
Positive
100
50
0
1
2
3
4
5
6
7
8
9
10
Trial repeats
Figure 3-10. Reference match trial identification performance indices by three observers —
correct identification rate fluctuation with trial repeats.
90
Value
000
090
180
270
Azimuth
Sensitivity
Speficity
Precesion
Accuracy
Figure 3-11. Blind trial identification performance at differing azimuths (observer KS)
3.4 Discussion
3.4.1 Observation
Despite trialling various illumination schemes, bitemarks made on the semi-transparent wax
block consistently produced a monotone image with limited tone difference. Bitemarks were
therefore subject to enhancement to make microscopic features more observable and
comparable. Image enhancement algorithms built in the image processing software were used,
including contrast enhancement, edge detection, auto equalisation, histogram equalisation and
local equalisation. After trialling all methods available, local equalisation (Gonzalez and
Woods 2002) was chosen as most suitable because of its effectiveness in enhancing a small
tone difference between striations.
91
3.4.2 Identification
Observable characteristics found on bitemarks made by simulating bitemarks on wax were
sufficient to enable highly accurate identification. This was true in comparing a known set of
individuals as well as a random selection of known individuals. In this study microscopic
striations that result from the incisor‘s edge scraping on the wax were the most prominent
feature used to identify the individual animals. In contrast, human bitemark identification
commonly uses macroscopic features such as tooth alignment as the main feature to identify
the bitemarks (Miller et al. 2009). In human forensic cases, the medium on which bitemarks
are found is mostly skin, where it is subject to distortion due to the skin‘s flexibility creep
(DeVore 1971). In this possum study, the tooth and/or bitemark uniqueness was not
confirmed, though the results positively demonstrated that individual identification was
possible from bitemarks through visual analysis, comparison and evaluation.
In analysing human bitemarks, the forensic toolmark examination approach most commonly
used is the impression type toolmark comparison (AFTE 1992) where the feature pattern
distribution and locations on the tool, i.e., teeth, are compared with the toolmark, i.e., the
bitemark. In contrast, possum bitemarks in this study were analysed by the striation toolmark
comparison approach in forensic firearm and toolmark examination (AFTE 1992). This was
the first attempt to focus on mammal bitemarks made by teeth acting as tools to create
striation marks on wax. One aspect of bitemark creation that enabled this approach was the
possum‘s biting action of pulling on the material (Gilmore 1966) to cause drag marks.
Another variation source may have been the possum‘s incisors‘ likelihood to bear incidental
markings due to its natural wear during use.
92
Cognitive understanding of how pattern matches were achievable in striation patterns were
among intriguing issues faced during the trials, although further pursuit of the topic was
abandoned due to limited time and knowledge background.
3.4.3 Comparison of bitemark casts
Once it was established that the striations can be used to characterise simulated individual
bitemarks, actual bitemarks by live animals were used to develop a method to capture the
characteristics from WaxTags®. Due to the difficulty of imposing illumination inside a
concave surface of a bitemark, negative convex casts were found useful to adequately
illuminate and enhance the microscopic features. An elastometric silicone casting material
Forensic Sil was used to acquire the casts. The grey colour of Forensic Sil also enabled to
observe and photographically capture striation details without image enhancement techniques
used for wax blocks.
3.4.4 Repeatability
This newly developed method showed promise that it is a repeatable technique by other
observers, provided training is given and that the observer is keen on doing the task.
Unfortunately one of the novice observers (CS) showed limited interest in the project and
went as far as commenting it was questionable whether such method was of any use. It is not
surprising such attitude may have influenced the output.
Training methods of novice observers were not thoroughly examined after the trials nor were
compared with alternative approaches. Enhancement attempts of performance levels beyond
those that were achieved by each observer were not attempted.
93
Whilst the study took considerable time to develop the individual identification method up to
a satisfactory performance level, attempts to further raise performance were not made. Formal
examination of the training and learning processes may provide a valuable guide to
understand essential elements in achieving high identification performance, as well as to
streamline the method.
Additionally, the development of a purpose-oriented training schedule should assist in an
effective transfer of the technique, which may need a standardised level of achievement to
assure its proper application in identifying individuals from bitemarks. Collaboration with
forensic science training providers may assist in formulating the training programme.
3.4.5 Further study
The high performance scores in correctly matching individual possums from bitemarks
proved that it was possible to use the observable features to identify individual animals under
laboratory conditions. This allowed the study to proceed to the next stage to determine
whether individual identification of wild possum is possible from bitemarks on WaxTags®
under field conditions.
94
Chapter 4
Individual Identification of Wild Possums
From Bitemarks
4.1 Introduction
Various ecological research methods rely on correctly identifying the object of interest to
collect information. From a broad to specific focus; presence/absence and identification of the
species is needed to understand a community composition; demographic information from age
and sex identification assists in studying population dynamics; marking or identifying
individuals or groups allow population abundance estimates to be calculated as well as
conducting detailed behaviour studies by following each individual or group across the
landscape and/or time (Dinsmore and Johnson 2005, Lancia et al. 2005, Schroeder and Robb
2005, Silvy et al. 2005).
Bait-interference techniques have proved to be useful in determining possum presence, but
studies that have attempted to use bait-interference to determine abundance have had varied
results (Bamford 1970). Presence/absence information may provide useful abundance indices
when there is a known or assumed relationship between abundance and presence information
(Caughley 1977a). Though bait interference provides a positive presence indication where
baits have been interfered with, the unknown probability in time and space of an individual
interfering with multiple baits, or of multiple individuals interfering with a single bait, make
the results difficult to apply to indices of population abundance (McGlinchy and Warburton
2000, Fraser et al. 2002).
95
The latest development of the bait interference method presented in an operative protocol
(NPCA 2008a) investigates possum bitemarks left on a wax block (WaxTag®), and infers the
animal‘s activity by calculating the proportion of a set sample of blocks bitten. This field
protocol relies on training and accrediting qualified field operators to identify the species that
interfered with the tag and left its bitemarks. While a study obtained a high accuracy (99%) of
correctly identifying possum bitemarks made from possum skulls (Thomas et al. 2007), the
study did not extend to bitemarks made by live animals or to quantify the species‘ bitemark
characteristics.
In Chapter 2, species identification from quantifiable characteristics of bitemarks was shown
to be possible from live animals, and in Chapter 3, a novel method to identify known
individuals‘ bitemarks on WaxTags® under laboratory conditions was developed. The
question remains about the applicability of the new technique to the field environment i.e., to
determine whether the technique can be used to identify wild, free-ranging individuals.
In contrast to laboratory conditions where the biting species is known and its biting behaviour
can be checked any time, field application allows any combination of multiple biting species
to freely bite any number of WaxTag® at any location, and for an unknown duration
overnight. Possum bitemarks may be obscured by other species biting on the same tag.
Additionally, it would be impossible to monitor each individual‘s biting behaviour and
respond to some over-eager biters as were encountered in the captive trials, e.g., some
possums started biting the tags as soon as the WaxTag® was brought into the cage, even
before the tag was secured to the cage. In a worst-case scenario, complete removal of the wax
was expected by such individuals in the wild. Furthermore, the biting animals were purposely
not captured, compared to human forensics where it may be the usual case to identify the
96
suspect biter. This was to emulate the operational field conditions when using WaxTags® for
monitoring purposes. The NPCA WaxTag® protocol (2008a) is entirely free from the need to
confirm the actual biter‘s identity, as it does not require any direct encounter with animals.
Results from laboratory trials in the previous Chapters were considered sufficiently reliable to
overcome the need to confirm findings by actually capturing the animals that produced
bitemarks.
WaxTags® were used in a field environment to assess the utility of individual identification in
bait interference monitoring. Previous studies using bait interference encountered the
unknown probability of an individual interfering with more than one bait, or more than one
individuals interfering with a single bait (Bamford 1970, Jane 1979, 1981, Arulchelvam and
Brown 1995, Spurr 1995). This was an immeasurable variable that limited the use of a direct
count of the baits being interfered with to infer the monitored animal‘s abundance because it
was not necessarily proportional to the actual abundance. Bearman (2002) correctly described
the WaxTag® method‘s intrinsic problem as its inability to prove that if two baits were
interfered with, different individual animals interfered with them. Subsequently, if such
individual identification is possible, we should also be able to collect information of the
individual‘s spatial and temporal movements to observe behavioural patterns in visiting
multiple times or multiple tags.
Identification of wild individual possums from its bitemarks, as was found possible in
controlled laboratory trials, is the first attempt of its kind. It will not only add to the
enrichment of tools to study the possum, but also open a new door of opportunities where the
method can be applied to different animal species, especially when the non-invasive nature of
the WaxTag® may be advantageous, such as in endangered or protected species.
97
With the abovementioned background, the objectives of this Chapter were therefore to:
1. Undertake field trials to develop a WaxTag® methodology that allows the
identification of individual possums in the wild,
2. Determine the prevalence and extent of multiple visits by possums to WaxTags®, and
the influence this has on the reliability of the WaxTag® as a monitoring tool,
3. Determine if the addition of a food reward, in the form of an icing sugar/flour mix,
influences WaxTag® visitation by possums, and,.
4. Investigate the utility of the WaxTag® to elucidate new information about the home
range, distribution, and activity patterns of individual possums.
4.2 Methods
4.2.1 Study area
The work presented in this Chapter was undertaken at Bottle Lake Forest Reserve, an 845 ha
pine plantation near Christchurch City, New Zealand (43° 31' 60 S, 172° 37' 60 E, Figure
4-1). The area has multiple uses. It is a public recreational area, providing for walking,
running, mountain-biking and horse-trekking, (Christchurch City Council 2003). The forest is
owned by the Selwyn Plantation Board Limited (SPBL).
Radiata pine (Pinus radiata D. Don) comprise 99.5% of the trees in the plantation, the
remainder are Douglas fir (Pseudotsuga menziesii (Mirb.) Franco), macrocarpa (Cupressus
98
macrocarpa Hartw. ex Gordon), and Corsican pine (Pinus nigra var. maritima (Ait.) Melv.)
(Hewett and Banks 1999). Undergrowth is generally sparse, but where present, the most
conspicuous species are blackberry (Rubus fruticosus L.) and common gorse (Ulex europaeus
L.). The soil classification is Waikuku deep loamy sand for most of the inland area, with an
approximately one-kilometre wide band of Kairaki deep sand on the coastal edge (Manaaki
Whenua - Landcare Research 1997). The forest is suitable possum habitat (Nugent et al.
2000), and was free of possum control leading into this field study (SPBL personal
communication).
4.2.2 Site selection within the study area
As the Bottle Lake Forest contains patches of forest of mixed age structure, sites of older trees
were selected, because they were more likely to harbour possums. Preliminary site selection
was by combined satellite images and ortho-photographs of the surrounding area downloaded
from Google™ Earth version 4.2 (2007). Captured screen images were combined using Corel
PHOTOPAINT software (Versions 9 and 12) to produce whole area images of each forest
section. Ground-truthing with a Garmin® GPS II® was then undertaken to confirm tree
maturity and ground cover. Further information on planted species, maturity, logging scheme,
and physical boundaries was obtained from SPBL, and applied to refine area and block
definitions. For the purposes of this study it was decided to nominate suitable sample sites in
areas where species composition was homogenous, tree stands were mature, and where
undergrowth permitted reasonable access to the area. Trees in nominated areas were all over
15 years old (range 15 – 48 years).
Having defined suitable study habitat, a 100 × 100 m grid was generated and overlaid on a
forest image with numbered grid cells. Computer generated random numbers defined sample
99
grids where they fell on a suitable sample block. Paired sample plots of 110 × 110 m were
placed at 100 m distances i.e. a whole sample set of 110 × 320 m (Figure 4-1) were drawn on
the map within the nominated area to meet the following conditions: (1) sample set reference
point (north-west corner of first grid) was located in the randomly selected sample grid, (2)
the sample set of two grid plots fitted within 10 m inside the suitable sample block, and (3)
the sample grid orientation followed tree row orientation. The latest aerial photographs of the
selected sample sites were obtained from SPBL for final update and refinement of plot setting
on the map. Sample site nomination was continued until a suitable sample site was identified
in the Bottle Lake Forest (Figure 4-1).
4.2.3 WaxTag® layout design
The sample plot design comprised paired sub-plots with each sub-plot consisting of a 12 × 12
grid arrangement, separated by a 100 m distance, resulting in a 110 × 320 m area (Figure 4-2),
as described above. Each of the paired sample grids was given letters ‗A‘ and ‗B‘ without
specifying which lure protocol version it would follow. The grid nodes where WaxTags®
were deployed were allocated alpha-numeric codes to specify location within the grids. Each
WaxTag® was randomly assigned a unique four-digit identification (ID) number written on
the plastic tag with permanent ink before the trials. WaxTags® were taken from a pool of prenumbered tags and attached to a photo-luminescent tag, also colloquially known as a glo-tag,
as per the NPCA protocol (2005, 2008a). Each glo-tag was marked with its alpha-numeric
grid node location code, which assisted in confirmation and location recording at tag retrieval
because they were re-used at the same location during the trial regardless of interference of
the bait. Food lure application was done only once on the first day and was not renewed or
added to, in accordance with the NPCA protocol (2008a).
100
4.2.4 WaxTag® field deployment
WaxTag® deployment sites were located by firstly measuring the distance on a vehicle
odometer from distinct geographical features, e.g., road junctions, and were confirmed by a
GPS reading. Secondly, beyond the road reference point, off-road coordination was achieved
with compass readings and distance measurement. GPS reception was poor inside the wooded
area due to extensive canopy cover so was abandoned. Field distances were measured by a
―beltchain field measuring unit‖ (GeoSystems PO Box 8160, Riccarton, Christchurch 8440),
also known as field-chain or hip-chain, which measures the length of cotton thread reeled out
from a mobile device attached to one‘s belt, with one end of the thread tied to a fixed
reference point (Figure 4-3).
At every 10 m point on the hip chain while marking the nominal grid, the closest tree within a
5 m radius was selected. Where no tree was found, any suitable support, e.g., a stump, log,
shrub, or fallen tree, was substituted. WaxTag® attachment followed the operational protocol
to set the wax portion of the tag 30 cm above the ground, and a photo-luminescent strip
(GloTag) was attached with the tag to the trunk (Figure 4-4). One of the two grid plots (A, B)
was randomly chosen to follow previous protocol with no food lure (NPCA 2005) and the
other to follow the current protocol with food lure applied onto the trunk (NPCA 2008a). The
food lure was a 5:1 mixture of flour and icing sugar. The food lure was applied directly onto
the tree bark below the attached tag with the aid of a one-litre plastic container with a
perforated cap. WaxTags® were set up during daylight commencing 11 March 2008, left
overnight, and checked daily until 21 March 2008.
101
10 km
Christchurch
Lincoln University
100 km
Christchurch
43° 31' 60 S
172° 37' 60 E
Mature tree
stand (15+ yrs)
Sample site
(paired grids)
1 km
Figure 4-1. Study area: Bottle Lake Forest Reserve, Christchurch, New Zealand.
102
01 02 03 04 05 06 07 08 09 10 11 12
(a)
A
B
C
D
E
F
320 m
G
H
I
J
K
10 m
L
10 m
100 m
01 02 03 04 05 06 07 08 09 10 11 12
(b)
M
N
O
P
Q
R
110 m
S
T
U
V
W
X
110 m
Figure 4-2. Sampling grid arrangement. Grids were composed of 12 × 12 nodes at 10 m
intervals, resulting in two 110 m × 110 m square plots. Minimum distance between
closest nodes in grid (a) and (b) was 100 m. Node location notation was presented as a
row + column combination, e.g., C10 for 3rd row 10th column node. Node A01 was
designated as grid start reference.
103
Case lid with window
Thread spool
Thread drum, rotating trip meter as
thread is pulled out
Trip meter
0
0
Cotton thread
2
4
1cm
Mobile unit case
attached to belt
Distance travelled (not to scale) = Length of thread pulled out
= Trip meter reading
End of thread fixed on tree etc. (not to scale)
Figure 4-3. Components of hip-chain and its functional principle. Free end of thread is fixed
to a reference location, e.g. tree, while the mobile unit (yellow box) is carried by the
researcher along the pre-determined transect. Distance travelled from the reference
location is measured by length of thread pulled out from mobile unit recorded by the trip
meter.
104
Nail/screw
Glo-tag
WaxTag
30 cm
Food lure
Figure 4-4. WaxTag® field deployment. WaxTags® and glo-tags were attached together to tree
trunk by a wood screw at 30 cm above ground. In one grid plot, food lure (flour/sugar
mixture) was applied to the tree below the WaxTag®. Photograph taken after one night.
4.2.5 WaxTag® retrieval and recording
Each WaxTag® that showed signs of interference was retrieved, its identification (ID) number
and location recorded, and a new WaxTag® was attached in the identical location, together
with the same glo-tag. Any doubtful interference was recorded and the WaxTag® replaced to
minimise false negatives. A two-person recording/retrieval system with each observer
checking alternative rows was used to assist efficiency. Data were later combined from each
person‘s record sheet. The grid plot name, WaxTag® setup date, retrieval date, weather
conditions and observer were recorded on each sheet, and additional observations, e.g.,
detached WaxTag®, were also recorded. All retrieved WaxTags® were collected, carried and
stored in a paper pulp-based 5 × 6 mid-size egg tray with its wax portion suspended in the
105
depression by the plastic tag, to minimise physical contact with sources that could interfere
with the bitemark.
Laboratory observation to screen out false positives from questionable marks was done under
a stereo microscope fitted with a × 0.7–4.0 zoom objective lens and × 10 ocular lenses and
illuminated with a linear light source by a fluorescent tube (Figure 3-2). Negative casts were
taken for all confirmed and possible bitemarks with Forensic Sil (Armor Forensics,
Jacksonville, Florida USA) an addition-curing silicone elastomer, polyvinylsiloxane.
Micrographic images of the casts were captured on an Olympus DM12 digital camera fitted to
an Olympus stereoscope SZX12 at × 16 objective zoom and × 10 ocular lenses. Bite
orientation and illumination conditions were kept constant throughout all image captures. All
individual identification procedures followed the previously established observation method
with bitemarks from known captive possums, described in Chapter 3. Casts from WaxTag®
bitemarks were pooled and re-sorted according to the assigned tag ID. Blind test observation
was assured due to the random nature of each tag ID and its arbitrary deployment in the field.
No location information was made available during the individual identification process in
cast observation, and the identified individuals were later associated with the spatial and
temporal attributes at the post-identification stage. Identified individuals were given ID codes
identical to the first observed WaxTag® number on which its bitemark was recognised as an
individual.
4.2.6 Data collation
After all bitten WaxTags® were assessed for individual possum bitemarks, nightly bitten tag
locations were plotted according to individuals in distribution grid maps. The number of tags
bitten, its location, by which individual and for which night were the basic spatial and
106
temporal data collected for analysis. Information derived from raw data included that of each
individual, e.g., distance travelled; that of tags, e.g., proportion of tags bitten; and that of grid
location, e.g., multiple visits to same tag location.
4.2.7 NPCA protocol
The current operational standard to index possum activity is stipulated in the NPCA protocol
(2008a). It calculates the proportion of 20 WaxTags® arranged in a line at 10 m intervals that
are bitten by possums. A Bite Mark Index (BMI) is calculated as the mean percentage of
WaxTag® lines that are set up in the monitored area at a minimum distance of 100 m between
lines, along with its associated statistics of standard error and 95% confidence intervals. Two
duration options of field exposure, i.e., 3 nights and 7 nights, are provided in the protocol for
operational practicality. In this trial bitten tags form the grid were re-sampled as lines to
simulate the protocol, and emulated BMIs were calculated accordingly from the bitten grids
up to the 3rd, 7th and 10th nights.
4.3 Results
4.3.1 Retrieved WaxTags®
A total number of 490 tags was retrieved from the field with positive and questionable marks
for the two grids with different lure types, 95.3% (467) were bitten by possums, 1.6% (8) by
rats, and 3.1% (15) were false positives, i.e., not bitemarks. After excluding false positives,
the distribution of total tags with positive bitemarks was biased toward possum bitemarks
(Table 4-1). Considering the sample sizes, further data analyses focused only on possum
bitten WaxTags®.
107
Table 4-1. Number of WaxTags® retaining bitemarks by possums and rodents.
Treatment
Possum
Rodent
Total
Visual lure
192
1
193 (40.6%)
Visual + food lure
275
7
282 (59.4%)
Total
467 (98.3%)
8 (1.7%)
475
4.3.2 Influence of food lure on WaxTags® interference
The addition of food lure significantly increased (Analysis of Variance = ANOVA, p<0.05)
the cumulative number of WaxTag® setup locations being bitten at least once, i.e. number of
bitten tags without replacement analogous to field application and collection after the
respective nights and discounting any multiple visits (Figure 4-5). Both numbers of bitten
WaxTags® with and without food lure reached asymptote after five nights. The cumulative
number of bitten tags with replacement, which takes into account multiple visits to the same
tag, similarly showed a significant increase (ANOVA, p<0.05) with the addition of food lure
(Figure 4-6). Contrastingly, the number of bitten tags, with or without food lure, did not
approach asymptote and showed a continuous increase during the 10 nights. A nightly count
of bitten tags showed a significant increase (ANOVA, p<0.05) with application of food lure
on nights one, two and three, while no significant effect was detected from night four onwards
(Figure 4-7).
108
Cumulative visited tag locations (line)
Number WaxTags
10
5
No food lure
Food lure
0
1
2
3
4
5
6
7
8
9
10
Nights
Figure 4-5. Cumulative number of WaxTags® bitten per line without replacement. Both
numbers of bitten WaxTags® with and without food lure reach asymptote after five
nights. Error bar = SEM. Significantly more tags with food lure were bitten throughout
the 10 nights.
109
Cumulative tags bitten per line
30
Number WaxTags
20
10
No food lure
Food lure
0
1
2
3
4
5
6
7
8
9
10
Nights
Figure 4-6. Cumulative number of WaxTags® bitten per line with replacement. Both numbers
of bitten WaxTags® with and without food lure do not reach asymptote and continuously
increase. Error bar = SEM. Significantly more tags with food lure were bitten throughout
the 10 nights.
110
number tags bitten per line per night
Number WaxTags
10
5
No food lure
Food lure
0
1
2
3
4
5
6
7
8
9
10
Nights
Figure 4-7. Number of WaxTags® bitten per line each night. Error bar = SEM. Significantly
more tags with food lure were bitten up to the third night. No significant difference
between lure type was observed after the fourth night. (p<0.05)
111
4.3.3 Repeated visits to tags
The frequency and distribution of multiple visits to WaxTags® were not homogenous (Figure
4-8). Some grid locations were visited a maximum six times but others were not visited at all.
No clear pattern of visit frequency was observed, although a spatial gradient was loosely
visible. Patches with no bitemarks registered were more commonly bordered by single-night
biting activity, compared with frequently bitten areas which could be sharply bordered by
much less frequent visit area.
112
(a)
10 m
(b)
10 m
Number nights visited
1 2
3
4
5
1 2
3
4
5
6
Bubble size
No food lure
With food lure
®
Figure 4-8. Locations where WaxTags were repeatedly bitten during 10-night period. Grid
(a) with no food lure and grid (b) with food lure. Actual distance between grids was
100 m. Result shows that no saturation occurred after 10 nights.
113
4.3.4 Identification of individual possums
Following the method to identify individually-known captive possums from their bitemarks
(Chapter 3), it was possible to identify individuals from field specimens of a wild population
of possums. The interpretation central to the observations of field samples followed that of
captive possums: where positive identification was made on multiple bitemarks, i.e. the
striations and/or the impression marks of bitemarks of corresponding teeth matched, the
bitemarks were made by the same individual animal. Conversely, where the identification was
negative, it was concluded that the bitemarks were not made by the same individual.
In specific terms, where a bitemark on a WaxTag® from a certain grid location matched that
of another WaxTag® at a different location, the observation was interpreted as the same
individual possum bit one of the two tags, travelled to the other location, and bit the other tag.
When a bitemark on a WaxTag® from a certain night matched that from another night, the
interpretation was that the same individual animal bit the tags across a temporal span. Where
a bitemark on a WaxTag® matched another bitemark on the same tag, it was deduced that the
same individual possum bit the same tag multiple times on the same night. Such multiple bites
by the same individual on the same night were counted as a single individual identified on the
tag.
A sample identification illustrated in Figure 4-9 shows grid locations and microscopic images
of sample bitemark casts on WaxTags® 0077 and 4986 recovered from the first night, and
WaxTag® 6118 recovered from the second night. Matching striation patterns of tooth 31 were
observed in all three tags, i.e. positive identification was made on the bitemarks. The
interpretation from the sample was that an individual possum bit multiple WaxTags® across a
spatial span of grid locations C02, J10 and L11, and across a temporal span of one night.
114
A
1
2
3
4
5
6
7
8
9 10 11 12
...... 6
A
B
B
C
C
D
D
E
E
F
F
G
G
H
H
I
I
J
J
K
L
K
L
nd
2 night (partial)
st
1 night
7
8
9 10 11 12
10 m
®
WaxTags 0077,
4986 & 6118
Other WaxTags
with bitemarks
Tag 0077
Tag 4986
Tag 6118
1st night
location C02
tooth 31
1st night
location J10
tooth 31
2nd night
location L11
tooth 31
®
1 mm
Figure 4-9. Grid locations and microscopic images of sample bitemark casts on WaxTags®
0077 and 4986 recovered from the first night, and WaxTag® 6118 recovered from the
second night. Identification was made on the bitemarks from the three tags, interpreted as
the same individual possum biting across a spatial span of grid locations C02, J10 and
L11, and across a temporal span of one night.
115
4.3.5 Identified individuals
Examination of 3341 bitemark cast images from 467 possum-bitten WaxTags® yielded a total
of 504 identifiable bitemarks, i.e., those either positively identified as made by the same
individual or excluded as made by a different individual. Six bitemarks from 6 tags were
designated as inconclusive. The largest night number of identified bitemarks was 114 on night
one and smallest on night eight at 15. From the total 461 tags with at least one identifiable
bitemark, 38 (8.2%) were bitten by a second individual, and five (1.1%) were bitten by a third
individual (Table 4-2).
Table 4-2. Individual identification made from possum bitemarks on WaxTags®
Number of WaxTags® with identified individuals*
Night
Tags bitten
Individual #1
Individual #2
Individual #3
All
Inconclusive
1
92
91
20
3
114
1
2
92
92
6
0
98
0
3
70
70
4
0
74
0
4
46
46
4
2
52
0
5
34
33
0
0
33
1
6
36
36
1
0
37
0
7
27
26
1
0
27
1
8
16
15
0
0
15
1
9
35
33
2
0
35
2
10
19
19
0
0
19
0
467
461
38
5
504
6
Total
* Number of WaxTags® with nth individual identified. Individual #n refers to number of individuals identified on
same tag, e.g., on night four, two tags yielded three identified individuals
116
Twenty groups of distinctive bitemark patterns were identified from 504 bitemarks originating
from 461 WaxTags®. This was interpreted as detecting 20 possum individuals in the trial. Six
bitemarks were inconclusive due to limited observable characteristics to either include in or
exclude from any group, or to place in its own entity. A detailed observation output is given in
Appendix E.
4.3.6 Activity patterns of individual possums
All WaxTags® bitten by identified individuals were associated with grid location coordinates
and time attributes. The number of tags, the distance between the furthermost tags, and the
minimum convex polygon (MCP) area were calculated for each observation night from the
attributes and results plotted in Figure 4-10. The individuals were ranked according to the
maximum number of tags bitten per night and simultaneously indicated whether the
individual bit a tag with or without food lure. The presence of food lure did not have a
significant effect on the maximum number of bitten tags (p<0.05), maximum distance
between tags (p<0.05) or maximum MCP area covered by the individual (p<0.05). An output
summary is given in Appendix F.
The presence of a food lure on WaxTags® showed a significant effect (ANOVA, p<0.05) on
the mean number of individuals detected on sample lines on the first night (Figure 4-11).
There was no significant difference in the number of identified individuals biting after the
second night, but the rate of change in the number of individuals detected was significantly
higher (Restricted Maximum Likelihood = REML variance components analysis, p<0.05) in
food-lured tags area than that in non food-lured tags area.
117
35
(a)
Number of tags
30
25
20
15
10
5
(b)
0
350
300
Distance (m)
250
200
150
100
50
(c)
0
3.0
Individual animal number:
0077 (blue): Individuals that bit tags without food lure on first night.
0719 (pink): Individuals that bit tags with food lure on first night.
2.5
Area (ha)
2.0
1.5
1.0
0.5
0.0
8349
1885
3443
6990
0919
8355
4458
3082
1833
0375
4242
2272
0139
2353
2382
0866
0560
0077
0131
0719
8349
1885
3443
6990
0919
8355
4458
3082
1833
0375
4242
2272
0139
2353
2382
0866
0560
0077
0130
0719
Individual animal
Figure 4-10. Box-and-whisker plots of nightly (a) number of tags bitten, (b) distance between
furthest bitten tags and (c) minimum convex polygon area, of individual possums.
Individuals are sorted to the order of least to most number of tags bitten in all three charts
for comparison.
118
detected individuals per line
4
3
No food lure
Individuals
Food lure
2
1
0
1
2
3
4
5
6
7
8
9
10
Nights
Figure 4-11. Number of individual possums identified per line. Error bars = SEM.
The information obtained from plotting WaxTag® locations that were identified as bitten by
each of the 20 individual possums is described in the following sections including (1) the
number of tags, with and without food lure, identified with the individual‘s bitemarks; (2) the
number of nights that these bitemarks were observed i.e. ‗active nights‘; (3) the maximum
number of tags bitten on a single night; (4) the maximum distance between the furthest bitten
tags on a single night; (5) the maximum area of minimum convex polygon (MCP) covered on
a single night; (6) the centre of nightly activity calculated as an arithmetic centre of mass
(centroid) from number and location of bitten tags; (7) the centroid confinement or expansion
from food-lured or non food lured grid where the first bitten tag was observed; (8) the nightly
119
movement of centroids (centroid shift) describing the shift of activity centre; (9) the number
of tags bitten by other individuals i.e. visits by multiple individuals; and (10) the number of
tags bitten on multiple nights by the same individual.
Individual 0077 (Figure 4-12) left bitemarks on a total of 69 WaxTags® in eight active nights.
The maximum number of tags bitten on one night was 16 on the first night followed by 15 on
the third night. It was most spatially active on night three covering 1.40 ha of MCP and 216 m
travelled. It was largely active in the grid with no food lure (8/8 active nights) with 88% (7/8
active nights) of its centroid confined within the grid. Centroid shifts totalled to 396 m. Foodlured tags interfered with were limited to those closest to the non-lured tags and were 9%
(6/69) of total bitten tags. Ten tags (14% = 10/69) bitten by individual 0077 were also bitten
by other individuals on the same night. WaxTags® set on 56 grid locations yielded 69 tags
bitten by this individual, broken down to 44 locations (79% = 44/56) visited once, 11
locations visited twice and one location visited three times by the individual i.e. 21% (12/56)
of locations visited multiple times.
Individual 0375 (Figure 4-13) bit a total of 10 tags on five active nights with a peak of three
tags on the first night. The largest distance travelled was 71 m on night three. Calculation of
nightly MCPs was not possible because bitten tag locations were linear or single. All bitten
tags were in the grid with no food lure, with centroids all confined within the grid. Centroids
shifted a total 72 m. Other individuals, on the same night, bit 30% (3/10) of tags bitten by this
individual. Of the nine grid locations where individual 0375 left bitemarks, one location (11%
= 1/9) was visited twice.
120
Individual 2353 (Figure 4-14) was active on six nights, biting a maximum seven tags on the
first night from a total of 19 tags. It travelled a maximum 215 m distance on night four and
covered 0.52 ha of MCP area on nights one and four. Activity was mostly in the grid where
tags did not have food lure (6/6 active nights). Three tags with food lure (75%) and one tag
without food lure were bitten on night four. These three tags were the only ones bitten with
food lure resulting in 16% (3/19) of all bitten tags. Centroids were confined within the gird
without food lure in 83% (5/6) of active nights, from a shifting total of 569 m. Four tags (21%
= 4/19) were bitten by other individuals. The 19 tags bitten by this individual were at 17 grid
locations, i.e., 15 locations visited once and two locations visited twice by the individual.
Individual 0866 (Figure 4-15) was one of three most temporally active possums (10/10 nights)
biting 58 tags. It bit a maximum number of 10 tags on nights three and nine; the minimum
was one tag on night 10. The individual travelled 218 m and covered 1.30 ha MCP on night
three. Number 0866 was active in the non food-lured grid on all nights. It also bit one, four
and one tags with food lure on nights two, three and four, respectively, comprising 10%
(6/58) of bitten tags and 30% (3/10) of active nights. All bitten food-lured tag locations were
confined to those closest to the non food-lured grid. Centroids shifted a total of 480 m with
20% (2/10) of active nights extending outside the non food-lured grid. Nine tags (16%) were
bitten by other individuals on same nights. The 58 bitten tags were from 46 locations, i.e., 34
locations bitten once and 12 locations bitten twice.
Individual 1885 (Figure 4-16) was identified from its first and only bitten tag on the sixth
night. No other tag with or without food lure was bitten by this individual, nor did any other
individual interfere with the bitten tag.
121
Individual 0139 (Figure 4-17) was first identified on night two with five bitten tags, and was
active on 40% (4/10) of observed nights with a total 13 bitten tags. The largest distance
between tags was 40 m and it covered an MCP area of 0.06 ha on night two. Centroid shifts
totalled 42 m. All bitten tags were with no food lure with none was interfered with by other
individuals on the same nights. The 13 bitten tags were retrieved from nine grid locations of
which six were bitten once, two bitten twice and one bitten three times.
Individual 2272 (Figure 4-18) was first identified on night three. On night four it bit four tags,
travelled 243 m and covered 0.36 ha MCP area, all maximum measurements. It was active on
60% (6/10) of nights interfering with a total of 17 tags. On all active nights a total of 15 tags
with no food lure were bitten, whereas two tags with food lure (12% = 2/17) were bitten only
on night four (17% = 1/6). Centroids were within the non food-lure grid in 83% (5/6) of active
nights shifting a total of 388 m. Two tags (12% = 2/17) were interfered with by other
individuals on nights four and nine. All 17 tags retrieved with bitemarks were from 17
different grid locations, i.e., all 17 tags were bitten once.
Individual 6990 (Figure 4-19) was first identified on night two travelling 10 m between the
two bitten tags. Four non food-lured tags were bitten on nights two to four resulting in 30%
(3/10) activity. One tag (25% = 1/4) was bitten by another individual, 0335, on the same
night. MCP polygon areas cannot be calculated; its centroid shift totalled 90 m. The four
bitten tags were retrieved from four locations, i.e., all tags were bitten once.
Individual 3443 (Figure 4-20) interfered with one tag on nights one and four (20% activity),
both tags were without food lure. One of these tags (50% = 1/2) was bitten by individual 0077
122
on the same night. The centroid shifted 82 m between the nights. No tag was bitten multiple
times by the individual.
Individual 0131 (Figure 4-21) left bitemarks on a total of 110 tags over 10 active nights. It
interfered with the greatest total number tags among all identified individuals and was one of
the three most temporally active individuals (No. 0866, 0131 & 2382). The maximum number
tags bitten per night was 31 covering its maximum 0.82 ha MCP area on night one. Its
maximum distance 310 m was travelled on night four. Except for two tags, all were lured with
food (98% = 108/110). Most activity areas were confined to the food-lured grid with nine
centroids (90%) confined within the grid area and the centroid of night four extending outside.
The total centroid shift was 419 m. Fourteen tags (13% = 14/110) were interfered with by
other individuals on the same night, of which 86% (12/14) were on night one. Of the 77 tag
locations where tags were bitten by individual 0131, 50 were bitten once, 23 were bitten
twice, two were bitten three times and two were bitten four times.
Individual 0560 (Figure 4-22) interfered with 41 tags of which 98% (40/41) were lured with
food. It bit one non-lure tag on night four travelling its 180 m maximum distance on that
night. The maximum number of bitten tags per night was 12 on night one, the maximum MCP
area was 0.34 ha on night two. The individual was active on nine nights (90%) with
interference observed on food-lure tags on all active nights. Centroids were mostly (89%
= 8/9) within the food-lure grid except night four, when it shifted 477 m. Ten tags (24%
= 10/41) were bitten by other individuals, of which 70% (7/10) were on night one. The 41
tags bitten by individual 0560 were retrieved from 34 grid locations, broken down to 27
locations visited once and seven locations visited twice.
123
Individual 0719 (Figure 4-23) was 90% (9/10) active biting 92 total tags. The maximum, 33
tags, was on night two, and the maximum distance, 311 m, was travelled on night four. This
distance was the furthest travelled among all 20 identified individuals. The maximum MCP
area was 2.8 ha on night four. Of the 92 bitten tags 80% (74/92) had food-lure and were bitten
on all nine active nights. The individual extended its foraging to interfere with 18 (20%
= 18/92) non food-lure tags on 44% (4/9) of active nights. Centroids were within food-lure
grid area on 78% (7/9) of active nights; it extended outside that area on nights four and ten.
The centroid shifted a total 617 m. Twenty-percent (18/92) of total tags were interfered with
other individuals. The 92 tags were retrieved from 72 grid locations; tags were bitten once at
56 locations, twice at 12 locations and three times at four locations.
Individual 2382 (Figure 4-24) was active on all 10 nights with maximum of seven bitten tags
on the first night. It travelled a maximum of 290 m and covered a 0.92 ha MCP area on night
six. The number of tags bitten totalled 37 among which 33 (89%) were food-lured, and four
(11 %) were without on nights four, five and six. Eighty-percent (8/10) of centroids were
within the food-lure grid, with centroids for nights four and five extending outside the grid.
The total centroid shift amounted to 254 m. This individual shared five tags (14% = 5/37)
with other individuals. The 37 bitten tags were from 30 grid locations; at 23 locations tags
were bitten once and at seven locations were bitten twice.
Individual 3082 (Figure 4-25) left two bitemarks on food-lured tags on night one. Both tags
were bitten by other individuals. The distance between the bitten tags was 130 m.
124
Individual 8355 (Figure 4-26) was 30% (3/10) active, nights one, four and eight, biting five
food-lured tags. The highest number of tags bitten on one night was two, on nights one and
eight. The maximum distance was 22 m between tags on night one, nightly MCP area cannot
be calculated. The centroid shifted a total 155 m. Other individuals interfered with 40% (2/5)
of tags. All five tags were bitten once each at five grid locations.
Individual 4242 (Figure 4-27) showed 60% (6/10) activity with a maximum four tags on night
four from a total 10 bitten tags. All six nights observed a total seven food-lured tags bitten,
while three tags (30% = 3/10) on night four were from the grid without food lure. Largest
MCP area 0.09 ha was on night four. The centroid was all within the food-lure grid except
night four, shifting a total 394 m. No tags were bitten by other individuals. Most tags (80%
= 8/10) were bitten once at eight locations while one tag was bitten twice at one location.
Individual 8349 (Figure 4-28) bit one tag with food lure on the first night. This tag was also
bitten by individual 0719.
Individual 1833 (Figure 4-29) bit nine tags, of which 89% (8/9) were with food lure. Its first
active night was night two, with six active nights. The largest distance travelled (64 m) and
MCP area covered (0.02 ha) were on night three. Five centroids (83% = 5/6) were within the
food-lure grid, with one centroid outside the grid. The total centroid shift was 433 m. No tags
were bitten by other individuals on the same night. The nine tags were bitten once by the
individual at seven grid locations and twice at one location.
125
Individual 0919 (Figure 4-30) bit two food-lured tags spaced 20 m apart on night three. One
of the tags was also bitten by individual 0719.
Individual 4458 (Figure 4-31) also bit two food-lured tags; they were 41 m apart on night
nine. No other individual interfered with the tags.
In total, 226 grid locations were bitten by 20 possums during 10 nights, giving a mean of 11.3
(range: 1–77) grid locations bitten by an individual possum. This was further broken down to
a total 176 grid locations bitten by 18 individuals up to 3 nights averaging 9.8 (range: 1–52)
grids per possum, and a total 219 grid locations bitten by 19 individuals up to 7 nights with a
mean 11.5 (range: 1–72) grids bitten per possum (Table 4-3). Inverse calculations showed that
the mean number identified possum per bitten grid location were 0.10, 0.09 and 0.09 for
cumulative 3, 7 and 10 nights respectively.
Table 4-4 summarises possum biting activity during the 10-night trial. The mean number of
WaxTags® bitten per night by the 20 individuals was 2.5 tags/night, when calculated without
consideration to multiple individuals biting the same tag (overlap). The maximum number
tags bitten by an individual ranged from 1 to 33 tags on a single night. The mean nightly
distance travelled by the 20 individual possums was 40.3 m/night, ranging from 0 m to 311 m
in one night. Mean nightly Minimum Convex Polygon (MCP) area was 0.1 ha with the range
between 0 ha and 2.8 ha covered by an individual per night.
126
Table 4-3. Cumlative bitten WaxTags® and grid locations up to 3, 7 and 10 nights.
3 nights
7 nights
Individual #
Total tags
Total grids
0719
59
0131
78
0077
10 nights
Total tags
Total grids
Total tags
Total grids
52
86
62
100
69
92
72
72
110
77
39
37
62
53
69
56
0560
28
25
35
30
41
34
0866
2382
25
22
41
34
58
46
14
13
31
28
37
30
2353
8
8
17
15
19
17
0139
9
7
11
9
13
9
2272
3
3
13
13
17
17
4242
1
1
7
6
10
9
0375
7
6
10
9
10
9
1833
4
4
7
5
9
8
3082
2
2
2
2
2
2
4458
0
0
0
0
2
2
8355
2
2
3
3
5
5
0919
2
2
2
2
2
2
6990
3
3
4
4
4
4
1885
0
0
1
1
1
1
3443
1
1
2
2
2
2
8349
1
1
1
1
1
1
Total (with overlap)
286
251
435
358
504
403
Total (actual tag/grid)
254
176
397
219
467
226
Total possums
18
18
19
19
20
20
Possum per tag/grid
0.07
0.10
0.05
0.09
0.04
0.09
Tag/grid per possum
14.1
9.8
20.9
11.5
23.4
11.3
The detailed measurements of all individual possums‘ activities by each night are summarised
in Appendix F.
127
128
(1,8)
(3)
(2)
(6)
(1,4)
(1)
2
2
2
1
1
1
8355
0919
6990
1885
3443
8349
0.1
0.2
0.1
0.4
0.2
0.5
0.2
0.2
0.9
1.0
1.0
1.7
1.3
1.9
3.7
5.8
4.1
6.9
11.0
9.2
Mean
0
0
0
0
0
0
1
0
1
0.5
1
3
0
1.5
3.5
6.5
2
7
5.5
3.5
Median
1
2
1
4
2
5
2
2
8
9
9
17
9
17
30
46
34
56
77
72
Total grids
1.0
1.7
1.0
3.7
2.0
3.3
2.0
2.0
5.1
6.8
4.7
10.5
8.1
12.3
20.8
29.2
27.2
43.6
63.9
60.4
Mean
1
2
1
4
2
3
2
2
5
6.5
5
11.5
9
13
22.5
31
29
50
68
67
Median
Bitten-tag grid locations (multiple
visits not counted)
-
-
-
10
20
22
41
130
64
71
180
243
40
215
290
218
180
216
310
311
(2)
(3)
(1)
(9)
(1)
(3)
(3)
(4)
(4)
(2)
(4)
(6)
(3)
(4)
(3)
(4)
(4)
Maximum (night #)
1.0
2.0
3.2
4.1
13.0
9.6
16.1
19.0
43.5
9.2
57.2
113.9
109.2
55.7
106.4
112.0
131.1
Mean
0
0
0
0
0
0
0
0
0
0
5
92
114
33
124.5
94.5
102
Median
Distance travelled per night (m)
-
-
-
-
-
-
-
-
-
-
0.10
0.40
0.10
0.50
0.90
1.30
0.30
1.40
0.80
2.80
(4)
(4)
(2)
(1,4)
(6)
(3)
(1,2)
(3)
(1)
(4)
Maximum (night #)
0.01
0.05
0.01
0.11
0.22
0.41
0.06
0.40
0.30
0.72
Mean
Area covered per night (ha)
Mean (overlapped)
128
7.4
2.5
1.8
20.1
15.5
16.7
128.1
40.3
33.2
0.4
0.11
NB: Where only one tag was bitten per night, distances cannot be calculated. Similarly the area cannot be calculated where bitten tags were single or arranged linearly.
(9)
2
4242
4458
(4)
4
2272
(1)
(4)
4
0139
2
(2)
5
2353
3082
(1)
7
2382
1833
(1)
7
0866
(3)
(3,9)
10
0560
(1)
(1)
12
3
(1)
16
0077
3
(1)
31
0131
0375
(2)
33
Maximum (night #)
0719
Individual #
Number tags bitten per night
Table 4-4. Summary WaxTag® biting activity of individual possums over 10 nights.
0
0
0
0
0.05
0.3
0
0.3
0.25
0.3
Median
The following plots (Figure 4-12 to Figure 4-31) graphically illustrate each individual‘s
change in biting behaviour over time and space. Keys to each colour plot are given in Legend
1, showing interfered tags with and without food lure, those visited by other possum(s),
centroids of bitten tag distribution, and spatial nightly distribution in MCP. Cumulative MCP
over all 10 nights and maximum nightly distance between tags are also plotted. Keys to each
greyscale plot are given in Legend 2, showing cumulative number of visits by the individual
animal (size = number of visits) with centroid (size = number of tags visited) on each of 10
nights, and its spatial shift between nights.
129
320 m
Night 1
2
3
4
5
7
8
9
10
10
m
6
Legend 1
Legend 2
Visit to tag with no food lure
1 2 3 4 5
Visit to tag with food lure
Visit by multiple individuals
20
5
Centroid (centre of mass)
1–10
Minimum convex polygon
(MCP)
MCP where bitten tag
locations are linear
MCP (cumulative over all
10 nights)
Maximum distance
between tags
30
10
Number of visits
to each location
(10-night total)
Centroid size =
number of visited
tags per night
Centroid shift
10 m
Scale
Figure 4-12. Distribution of WaxTags® bitten by individual possum 0077 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative number of
visits and nightly shift of centroid (greyscale plots). Legend for colour plots is shown in
―Legend 1‖. Legend for greyscale plot is shown in ―Legend 2‖.
130
Night 1
2
3
4
5
7
8
9
10
10 m
6
Legend 1
Legend 2
Visit to tag with no food lure
1 2 3 4 5
Visit to tag with food lure
Visit by multiple individuals
20
5
Centroid (centre of mass)
1–10
Minimum convex polygon
(MCP)
MCP where bitten tag
locations are linear
MCP (cumulative over all
10 nights)
Maximum distance
between tags
30
10
Number of visits
to each location
(10-night total)
Centroid size =
number of visited
tags per night
Centroid shift
10 m
Scale
Figure 4-13. Distribution of WaxTags® bitten by individual possum 0375 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative number of
visits and nightly shift of centroid (greyscale plots). Legend for colour plots is shown in
―Legend 1‖. Legend for greyscale plot is shown in ―Legend 2‖.
131
Night 1
2
3
4
5
7
8
9
10
10 m
6
Legend 1
Legend 2
Visit to tag with no food lure
1 2 3 4 5
Visit to tag with food lure
Visit by multiple individuals
20
5
Centroid (centre of mass)
1–10
Minimum convex polygon
(MCP)
MCP where bitten tag
locations are linear
MCP (cumulative over all
10 nights)
Maximum distance
between tags
30
10
Number of visits
to each location
(10-night total)
Centroid size =
number of visited
tags per night
Centroid shift
10 m
Scale
Figure 4-14. Distribution of WaxTags® bitten by individual possum 2353 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative number of
visits and nightly shift of centroid (greyscale plots). Legend for colour plots is shown in
―Legend 1‖. Legend for greyscale plot is shown in ―Legend 2‖.
132
Night 1
2
3
4
5
7
8
9
10
10 m
6
Legend 1
Legend 2
Visit to tag with no food lure
1 2 3 4 5
Visit to tag with food lure
Visit by multiple individuals
20
5
Centroid (centre of mass)
1–10
Minimum convex polygon
(MCP)
MCP where bitten tag
locations are linear
MCP (cumulative over all
10 nights)
Maximum distance
between tags
30
10
Number of visits
to each location
(10-night total)
Centroid size =
number of visited
tags per night
Centroid shift
10 m
Scale
Figure 4-15. Distribution of WaxTags® bitten by individual possum 0866 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative number of
visits and nightly shift of centroid (greyscale plots). Legend for colour plots is shown in
―Legend 1‖. Legend for greyscale plot is shown in ―Legend 2‖.
133
Night 1
2
3
4
5
7
8
9
10
10 m
6
Legend 1
Legend 2
Visit to tag with no food lure
1 2 3 4 5
Visit to tag with food lure
Visit by multiple individuals
20
5
Centroid (centre of mass)
1–10
Minimum convex polygon
(MCP)
MCP where bitten tag
locations are linear
MCP (cumulative over all
10 nights)
Maximum distance
between tags
30
10
Number of visits
to each location
(10-night total)
Centroid size =
number of visited
tags per night
Centroid shift
10 m
Scale
Figure 4-16. Distribution of WaxTags® bitten by individual possum 1885 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative number of
visits and nightly shift of centroid (greyscale plots). Legend for colour plots is shown in
―Legend 1‖. Legend for greyscale plot is shown in ―Legend 2‖.
134
Night 1
2
3
4
5
7
8
9
10
10 m
6
Legend 1
Legend 2
Visit to tag with no food lure
1 2 3 4 5
Visit to tag with food lure
Visit by multiple individuals
20
5
Centroid (centre of mass)
1–10
Minimum convex polygon
(MCP)
MCP where bitten tag
locations are linear
MCP (cumulative over all
10 nights)
Maximum distance
between tags
30
10
Number of visits
to each location
(10-night total)
Centroid size =
number of visited
tags per night
Centroid shift
10 m
Scale
Figure 4-17. Distribution of WaxTags® bitten by individual possum 0139 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative number of
visits and nightly shift of centroid (greyscale plots). Legend for colour plots is shown in
―Legend 1‖. Legend for greyscale plot is shown in ―Legend 2‖.
135
Night 1
2
3
4
5
7
8
9
10
10
m
6
Legend 1
Legend 2
Visit to tag with no food lure
1 2 3 4 5
Visit to tag with food lure
Visit by multiple individuals
20
5
Centroid (centre of mass)
1–10
Minimum convex polygon
(MCP)
MCP where bitten tag
locations are linear
MCP (cumulative over all
10 nights)
Maximum distance
between tags
30
10
Number of visits
to each location
(10-night total)
Centroid size =
number of visited
tags per night
Centroid shift
10 m
Scale
Figure 4-18. Distribution of WaxTags® bitten by individual possum 2272 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative number of
visits and nightly shift of centroid (greyscale plots). Legend for colour plots is shown in
―Legend 1‖. Legend for greyscale plot is shown in ―Legend 2‖.
136
Night 1
2
3
4
5
7
8
9
10
10 m
6
Legend 1
Legend 2
Visit to tag with no food lure
1 2 3 4 5
Visit to tag with food lure
Visit by multiple individuals
20
5
Centroid (centre of mass)
1–10
Minimum convex polygon
(MCP)
MCP where bitten tag
locations are linear
MCP (cumulative over all
10 nights)
Maximum distance
between tags
30
10
Number of visits
to each location
(10-night total)
Centroid size =
number of visited
tags per night
Centroid shift
10 m
Scale
Figure 4-19. Distribution of WaxTags® bitten by individual possum 6990 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative number of
visits and nightly shift of centroid (greyscale plots). Legend for colour plots is shown in
―Legend 1‖. Legend for greyscale plot is shown in ―Legend 2‖.
137
Night 1
2
3
4
5
7
8
9
10
10 m
6
Legend 1
Legend 2
Visit to tag with no food lure
1 2 3 4 5
Visit to tag with food lure
Visit by multiple individuals
20
5
Centroid (centre of mass)
1–10
Minimum convex polygon
(MCP)
MCP where bitten tag
locations are linear
MCP (cumulative over all
10 nights)
Maximum distance
between tags
30
10
Number of visits
to each location
(10-night total)
Centroid size =
number of visited
tags per night
Centroid shift
10 m
Scale
Figure 4-20. Distribution of WaxTags® bitten by individual possum 3443 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative number of
visits and nightly shift of centroid (greyscale plots). Legend for colour plots is shown in
―Legend 1‖. Legend for greyscale plot is shown in ―Legend 2‖.
138
10 m
Night 1
2
3
4
5
6
7
8
9
10
Legend 1
Legend 2
Visit to tag with no food lure
1 2 3 4 5
Visit to tag with food lure
Visit by multiple individuals
20
5
Centroid (centre of mass)
1–10
Minimum convex polygon
(MCP)
MCP where bitten tag
locations are linear
MCP (cumulative over all
10 nights)
Maximum distance
between tags
30
10
Number of visits
to each location
(10-night total)
Centroid size =
number of visited
tags per night
Centroid shift
10 m
Scale
Figure 4-21. Distribution of WaxTags® bitten by individual possum 0131 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative number of
visits and nightly shift of centroid (greyscale plots). Legend for colour plots is shown in
―Legend 1‖. Legend for greyscale plot is shown in ―Legend 2‖.
139
10 m
Night 1
2
3
4
5
6
7
8
9
10
Legend 1
Legend 2
Visit to tag with no food lure
1 2 3 4 5
Visit to tag with food lure
Visit by multiple individuals
20
5
Centroid (centre of mass)
1–10
Minimum convex polygon
(MCP)
MCP where bitten tag
locations are linear
MCP (cumulative over all
10 nights)
Maximum distance
between tags
30
10
Number of visits
to each location
(10-night total)
Centroid size =
number of visited
tags per night
Centroid shift
10 m
Scale
Figure 4-22. Distribution of WaxTags® bitten by individual possum 0560 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative number of
visits and nightly shift of centroid (greyscale plots). Legend for colour plots is shown in
―Legend 1‖. Legend for greyscale plot is shown in ―Legend 2‖.
140
10
m
Night 1
6
2
3
4
5
7
8
9
10
Legend 1
Legend 2
Visit to tag with no food lure
1 2 3 4 5
Visit to tag with food lure
Visit by multiple individuals
20
5
Centroid (centre of mass)
1–10
Minimum convex polygon
(MCP)
MCP where bitten tag
locations are linear
MCP (cumulative over all
10 nights)
Maximum distance
between tags
30
10
Number of visits
to each location
(10-night total)
Centroid size =
number of visited
tags per night
Centroid shift
10 m
Scale
Figure 4-23. Distribution of WaxTags® bitten by individual possum 0719 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative number of
visits and nightly shift of centroid (greyscale plots). Legend for colour plots is shown in
―Legend 1‖. Legend for greyscale plot is shown in ―Legend 2‖.
141
10 m
Night 1
2
3
4
5
6
7
8
9
10
Legend 1
Legend 2
Visit to tag with no food lure
1 2 3 4 5
Visit to tag with food lure
Visit by multiple individuals
20
5
Centroid (centre of mass)
1–10
Minimum convex polygon
(MCP)
MCP where bitten tag
locations are linear
MCP (cumulative over all
10 nights)
Maximum distance
between tags
30
10
Number of visits
to each location
(10-night total)
Centroid size =
number of visited
tags per night
Centroid shift
10 m
Scale
Figure 4-24. Distribution of WaxTags® bitten by individual possum 2382 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative number of
visits and nightly shift of centroid (greyscale plots). Legend for colour plots is shown in
―Legend 1‖. Legend for greyscale plot is shown in ―Legend 2‖.
142
10 m
Night 1
2
3
4
5
6
7
8
9
10
Legend 1
Legend 2
Visit to tag with no food lure
1 2 3 4 5
Visit to tag with food lure
Visit by multiple individuals
20
5
Centroid (centre of mass)
1–10
Minimum convex polygon
(MCP)
MCP where bitten tag
locations are linear
MCP (cumulative over all
10 nights)
Maximum distance
between tags
30
10
Number of visits
to each location
(10-night total)
Centroid size =
number of visited
tags per night
Centroid shift
10 m
Scale
Figure 4-25. Distribution of WaxTags® bitten by individual possum 3082 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative number of
visits and nightly shift of centroid (greyscale plots). Legend for colour plots is shown in
―Legend 1‖. Legend for greyscale plot is shown in ―Legend 2‖.
143
10 m
Night 1
2
3
4
5
6
7
8
9
10
Legend 1
Legend 2
Visit to tag with no food lure
1 2 3 4 5
Visit to tag with food lure
Visit by multiple individuals
20
5
Centroid (centre of mass)
1–10
Minimum convex polygon
(MCP)
MCP where bitten tag
locations are linear
MCP (cumulative over all
10 nights)
Maximum distance
between tags
30
10
Number of visits
to each location
(10-night total)
Centroid size =
number of visited
tags per night
Centroid shift
10 m
Scale
Figure 4-26. Distribution of WaxTags® bitten by individual possum 8355 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative number of
visits and nightly shift of centroid (greyscale plots). Legend for colour plots is shown in
―Legend 1‖. Legend for greyscale plot is shown in ―Legend 2‖.
144
10
m
Night 1
6
2
3
7
8
4
5
9
Legend 1
10
Legend 2
Visit to tag with no food lure
1 2 3 4 5
Visit to tag with food lure
Visit by multiple individuals
20
5
Centroid (centre of mass)
1–10
Minimum convex polygon
(MCP)
MCP where bitten tag
locations are linear
MCP (cumulative over all
10 nights)
Maximum distance
between tags
30
10
Number of visits
to each location
(10-night total)
Centroid size =
number of visited
tags per night
Centroid shift
10 m
Scale
Figure 4-27. Distribution of WaxTags® bitten by individual possum 4242 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative number of
visits and nightly shift of centroid (greyscale plots). Legend for colour plots is shown in
―Legend 1‖. Legend for greyscale plot is shown in ―Legend 2‖.
145
10
m
Night 1
6
2
3
4
5
7
8
9
10
Legend 1
Legend 2
Visit to tag with no food lure
1 2 3 4 5
Visit to tag with food lure
Visit by multiple individuals
20
5
Centroid (centre of mass)
1–10
Minimum convex polygon
(MCP)
MCP where bitten tag
locations are linear
MCP (cumulative over all
10 nights)
Maximum distance
between tags
30
10
Number of visits
to each location
(10-night total)
Centroid size =
number of visited
tags per night
Centroid shift
10 m
Scale
Figure 4-28. Distribution of WaxTags® bitten by individual possum 8349 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative number of
visits and nightly shift of centroid (greyscale plots). Legend for colour plots is shown in
―Legend 1‖. Legend for greyscale plot is shown in ―Legend 2‖.
146
10 m
Night 1
2
3
4
5
6
7
8
9
10
Legend 1
Legend 2
Visit to tag with no food lure
1 2 3 4 5
Visit to tag with food lure
Visit by multiple individuals
20
5
Centroid (centre of mass)
1–10
Minimum convex polygon
(MCP)
MCP where bitten tag
locations are linear
MCP (cumulative over all
10 nights)
Maximum distance
between tags
30
10
Number of visits
to each location
(10-night total)
Centroid size =
number of visited
tags per night
Centroid shift
10 m
Scale
Figure 4-29. Distribution of WaxTags® bitten by individual possum 1833 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative number of
visits and nightly shift of centroid (greyscale plots). Legend for colour plots is shown in
―Legend 1‖. Legend for greyscale plot is shown in ―Legend 2‖.
147
10 m
Night 1
2
3
4
5
6
7
8
9
10
Legend 1
Legend 2
Visit to tag with no food lure
1 2 3 4 5
Visit to tag with food lure
Visit by multiple individuals
20
5
Centroid (centre of mass)
1–10
Minimum convex polygon
(MCP)
MCP where bitten tag
locations are linear
MCP (cumulative over all
10 nights)
Maximum distance
between tags
30
10
Number of visits
to each location
(10-night total)
Centroid size =
number of visited
tags per night
Centroid shift
10 m
Scale
Figure 4-30. Distribution of WaxTags® bitten by individual possum 0919 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative number of
visits and nightly shift of centroid (greyscale plots). Legend for colour plots is shown in
―Legend 1‖. Legend for greyscale plot is shown in ―Legend 2‖.
148
10
m
Night 1
6
2
3
4
5
7
8
9
10
Legend 1
Legend 2
Visit to tag with no food lure
1 2 3 4 5
Visit to tag with food lure
Visit by multiple individuals
20
5
Centroid (centre of mass)
1–10
Minimum convex polygon
(MCP)
MCP where bitten tag
locations are linear
MCP (cumulative over all
10 nights)
Maximum distance
between tags
30
10
Number of visits
to each location
(10-night total)
Centroid size =
number of visited
tags per night
Centroid shift
10 m
Scale
Figure 4-31. Distribution of WaxTags® bitten by individual possum 4458 for each of 10
nights, and cumulative MCP for all nights (colour plots), with cumulative number of
visits and nightly shift of centroid (greyscale plots). Legend for colour plots is shown in
―Legend 1‖. Legend for greyscale plot is shown in ―Legend 2‖.
149
An overlay diagram of all 20 individuals‘ MCPs (Figure 4-32a) failed to show any distinctive
pattern of high and/or concentrated activity in the study area, while there were some
recognisable areas that showed higher activity compared to lower activity, both in a relative
sense. It was possible to recognise a faint inclination in overlap density towards the centre of
the grid in each lure types. When the MCP overlays are done separately by origin of first
bitemark i.e. whether the first bitten WaxTag® was with food lure or not (Figure 4-32b &
Figure 4-32c), it was observed bilaterally that some individual possums that initially
encountered one type of lure both showed expansion beyond areas without any WaxTags®
and biting those in the neighbouring grid with the opposite lure setting. From the patterns‘
darker, higher overlap areas there appears to be a central wedge style expansion of bitten tags
from non food-lured grid to the food-lured grid (Figure 4-32b), compared to the peripheral V
shaped expansion in the opposite direction (Figure 4-32c). A combination of Figure 4-32b &
Figure 4-32c illustrates the more homogenous distribution of MCP overlap (Figure 4-32d).
150
(a)
(b)
(c)
(d)
ssss
Degree of polygon overlap
ssss
Low
High
ssss
ssss
- All individuals
ssss
st
- No food lure on 1 night
ssss
st
- With food lure on 1 night
ssss
ssss
ssss
Figure 4-32. Overlay of all 20 identified individual possums‘
MCP plots in Figure 4-12 to
ssss
Figure 4-31, summarising individuals‘ nightly records
over
10 nights. From left to right:
ssss
(a) all individuals in greyscale, possums that visitedssss
WaxTags® (b) with no food lure, and
ssss of two lure-type initiated
(c) with food lure on first night, respectively, and overlap
groups (d). Area colour intensity levels correspond ssss
to degree of polygon overlap.
ssss
ssss
ssss
ssss
4.3.7 BMI emulation
ssss
ssss
ssdd
The emulated BMI values from all grid data for nights 3,
7 and 10 were 61.1%, 76,6% and
ddd
ddd
78.5%, respectiveley, and were positively rrelated to theddd
number of indidividuals identified
ddd
from bitemarks as summarised in Table 4-5. Positive correlation
was observed in WaxTags®
ddd
ddd
®
ddd
with and without food lure, as well as all WaxTags combined.
Figure 4-33 shows that
ddd
ddd
regression analysis of the observed number of possumsddd
with BMIs from tags with and without
ddd
ddd (r2 = 0.85–0.87).
food lure showed considerable level of positive correlation
ddd
ddd
ddd
ddd
ddd
151 ddd
ddd
ddd
ddd
ddd
Table 4-5. BMI emulation from WaxTag® grids.
Nights
All WaxTags®
BMI (%) 95% CI
No food lure
BMI (%) 95% CI
Food lure
BMI (%) 95% CI
Possums
3
61.1
8.4
47.2
8.4
75.0
9.6
18
7
76.0
6.5
66.7
9.4
85.4
5.0
19
10
78.5
6.5
69.4
10.1
87.5
4.3
20
Emulated BMI (error bars = ± 95% CI)
1.0
Night 3
Night 7
Night 10
0.5
All WaxTags
All WaxTags regression (r2 = 0.85)
No food lure
No food lure regression (r2 = 0.84)
Food lure
Food lure regression (r2 = 0.87)
0.0
18
19
20
Possum abundance
Figure 4-33. Relationship of emulated BMI with observed possum abundance.
4.4 Discussion
4.4.1 Individual identification
There were four objectives to the work presented in this Chapter. The first of these objectives
was to undertake field trials to test whether the WaxTag® methodology can allow the
152
identification of individual possums in the wild, from their bitemarks. This objective was
achieved. In this field trial, 504 bitemarks, obtained from 467 tags deployed in the field, gave
sufficient information in 3341 bitemark images to allow the identification of 20 individual
possums. To the best of my knowledge, this is the first time that this has been done for freeranging wild possums. It may, in fact, be the first time that any free-ranging wild animal
species was individualised from its bitemark in wax.
.
This new methodology has numerous possible applications, three of which address the next
three objectives of this Chapter, outlined below.
4.4.2 Multiple visits
The second objective in this Chapter was to determine the prevalence of multiple visits by
possums to WaxTags®. It was clear that possums will visit WaxTags® on multiple occasions.
There were 18 individual possums (90%) that bit more than one tag over a period of 10
nights. Clearly, therefore, WaxTags® pose a potential risk of over-estimating the prevalence
of possums from this aspect. On the other hand, there were also many incidences of individual
tags being bitten by more than one possum, in which case may a cause an under-estimation of
possum abundance. Whilst multiple visits to WaxTags® may make the direct interpretation of
its index value to the possum abundance difficult, the high probability that an individual may
bite at least one tag likely offers a highly sensitive method for determining its presence in
areas where numbers are thought to be low.
153
4.4.3 Food reward
The third objective in this Chapter was to determine if the addition of a food reward, in the
form of an icing sugar/flour mix, influences WaxTag® visitation by wild possums. It was clear
that a food reward did cause an enhanced visitation rate by wild possums. There was a
significant enhancement in the first three nights after WaxTags® were deployed. It is likely
that possums were better able to find the WaxTags® in the presence of the food reward. It is
also possible that possums that obtained a food reward at a WaxTag® site were more likely to
continue to look for, and bite, further WaxTags®. The two individual possums that bit the
most WaxTags® both obtained a food reward the first time they encountered a WaxTag®. This
finding has important implications in terms of using WaxTags® to monitor possum
populations. With a food reward present, it is likely that an over-estimation of possum
prevalence will occur. However, depending on the desired sensitivity for any particular
possum monitoring activity, the use of a food reward could be beneficial to enhance the
likelihood that possums are detected, particularly in cases where numbers are thought to be
low. This could be of particular use in Australia, at sites where possums are thought to be
locally extinct. A prominent change from the NPCA protocol 2005 version to the current
2008 version was the addition of food lure, i.e., sugar blaze to the existing visual lure. While
pre-protocol studies (e.g., Thomas et al. 2003, Thomas et al. 2004) on the WaxTag® method
used a food lure imitating the contemporary trap-catch method, concerns were raised whether
the animal‘s food reward had any effect on its behaviour, e.g., seeking more tags to interfere
with (Commins 2005). The 2005 protocol circumvented this assumed effect by using only
visual lures.
4.4.4 Further ecological information
The fourth objective in this Chapter was to investigate the utility of WaxTags® to elucidate
information about the home range, distribution, and activity patterns of individual possums. It
154
was clear that WaxTags® can be used to gain new information about the home range,
distribution and activity patterns of individual possums. In Figure 4-12 to Figure 4-31 this
type of information is given for 20 individual possums.
In summary of findings from field trials, individual identification of wild, free-ranging
possums was proven possible using the bitemarks that individuals leave on WaxTags®,
deployed within their environs. This novel methodology has many potential applications. It
was used to determine that individual possums can bite multiple WaxTags®, and multiple
possums can bite single tags. It was also used to determine that a food lure will increase the
incidence of WaxTags® being bitten. Finally, by deploying a grid of WaxTags® in the field,
information on the movement of individual possums can be obtained.
4.4.5 NPCA protocol
Although the grid arrangement setting in this study does not follow the NPCA WaxTag®
method protocol (2008a) where lines of 20 tags are set with a minimum distance of 100 m
between lines, this finding on multiple visits prevalence suggest that the protocol may need to
be further investigated and refined to function as an abundance index as suggested in the next
section.
The proportion of WaxTags® that were bitten over 3, 7 and 10 nights, i.e., the BMI as
stipulated in the NPCA protocol (2008a), was positively related to the possum abundance. A
crude regression analysis was also performed to estimate the possum population from the
emulated BMI (Figure 4-33). A high correlation value was achieved (r2 = 0.85–0.87) although
only three data-points were used and was therefore not considered a robust measure. It should
155
be reiterated that this was a very crude attempt to apply the principles of NPCA protocol in
calculating a BMI from the proportion of bitten tags to index the possum abundance, and its
significance is therefore limited and should be regarded with extreme caution.
4.4.6 Further directions
Methods to individually identify free-ranging animals have provided a valuable tool in
collecting a vast range of ecological, behavioural, demographical, and biological information
(e.g., McGregor and Peake 1998). Individualisation of field signs or traces have been
attempted in wildlife species from its calls, footprints, or genetic material contained in faeces
and hair follicles (e.g., Smallwood and Fitzhugh 1993). This study demonstrated the first case
that bitemarks on wax can be used to identify the biting animal, which has the potential to be
applied into a vast spectrum of ecological methods that need individuals to be identified.
Consultation with potential users of this method may allow for a wider scope to develop a
new application. Transfer of the identification method, also mentioned in Chapter 3, may also
be a necessary step to ensure the method is usable to other interested parties.
Furthermore, this study provided definitive evidence that multiple visits to WaxTags® do
occur, which confirms the concern of researchers who use this method to monitor possums.
The NPCA protocol (2008a) acknowledged the possibility of the phenomenon, and restricted
the interpretation of its derived BMI as an activity index. If the index is to be used as a
measure of possum abundance (Thomas et al. 2007), thorough consideration is necessary to
take the multiple visit phenomenon into account. The positive correlation between observed
possum abundance and emulated BMI does, however, suggest a possible direction for
research to refine and/or develop the method to obtain a more robust measure of abundance
from bitemarks on WaxTags®. Further studies may include comparison of population
156
abundance measures by independent methods, e.g., mark-recapture (Lancia et al. 2005) with
BMI derived from full compliance to the NPCA (2008a) protocol. Where monitoring was
done with simultaneous application of the trap-catch method and WaxTag® method in the
same habitat, a crude means to directly compare the proportion of bitten WaxTags® in the tag
line with those caught in the trap-catch line may be possible by discarding counts of those
individuals repeatedly detected on multiple tags.
Considerable labour was involved in analysing and comparing the large number of specimens.
Automation of the method may not only speed up the process, but also allow a larger sample
size to be analysed. This aspect is further discussed in Chapter 5.
Another future consideration may be to determine whether there is an upper limit to the
number of individuals that can be reliably identified from the limited examination area of
bitemarks on the wax block. In this study 20 individuals were identified, although it is not
known if the same procedure was applicable if 100 individuals were present in the study area.
157
Chapter 5
General Summary and Conclusions
5.1 Introduction
This study was centred on the core issue of using a forensic approach to identify the
individuals leaving bitemarks on the WaxTag®, a bait interference method device, to monitor
the New Zealand pest species. Chapters 1 and 2 introduced the principles with which the
WaxTag® method operates to collect information from the animals. Although the method
fundamentally depends on the ability to identify the species that interfered with the WaxTag®,
bitemarks of probable species were not defined in previous studies. The method's intrinsic
weakness lays in its unknown probability of a single individual interfering with multiple
devices, and likewise, a single device interfered with by multiple individuals. Therefore,
making an inference of population abundance from the number of WaxTags® interfered with,
or the proportion thereof, was impossible if not confounded.
One of the principle theories in identification relates closely to fields of forensic science
disciplines, i.e., forensic bitemark analysis, forensic firearm and toolmark identification. The
basic premises are that; (1) there exists features on the investigated item that is sufficiently
unique to enable assignment to a particular individual or group of individuals, and (2) the
unique features are transferred to the marked substance in a discernable characteristic form to
allow comparison of the tool against the mark, or a mark against another mark.
158
5.2 General discussions
5.2.1 Forensics
The ―striated scratch mark‖ by teeth biting into foodstuff were recognised by Solheim and
Leidal (1975) to be made by the incisal edge or ―vestibular‖ (labial) surface of the incisor
moving across the material during the biting process, similar in principle with possum incisor
edges producing striation marks on the wax surface as it moves along the surface. The method
of analysis and comparison in their work was from profiles of the striation, unlike this possum
study which was visual matching of striation marks Profile comparison of possum striation
bitemarks on wax and its casts were attempted although identification was likewise not
possible in this study, and further pursuit was abandoned. This study was the first case the
scratch mark by the incisal edge was compared and matched from its striated surfaces using a
toolmark examination approach.
A comparison microscope normally used in toolmark analysis was used in analysing and
matching impression type bitemark in cheese (Bernitz and Kloppers 2002), demonstrating an
unusual case of bitemark pattern comparison from microscopic features. This study also
focused on microscopic characteristics of the bitemarks to identify possums, although lack of
immediate access to a comparison microscope confined the observation and comparison
procedures to a computer screen. A striation match trial using an actual comparison
microscope under arrangement with a forensic laboratory may yield additional findings,
particularly in relation to the speed and ease of comparison.
5.2.2 Bitemark stability
Wear and attrition of the possum‘s incisor occlusal facet and resulting labial-occlusal facet
edge variation were most likely the causes of individuality. This raised the question of how
159
long does this variable remains stable to enable identification or how fast does it change with
time. Nicks and marks on soft tissue, e.g., sperm whale flukes, appear to be sufficiently stable
to allow reliable identification after a time span of 4.5 years (Childerhouse et al. 1996) In this
study the maximum trial time span of 10 days resulted in no observable change detected in
any individual, although the detectability of individuals beyond this time duration remains
unknown. Tooth-wear according to the possum‘s age has been evaluated on the upper left first
molar (M1; FDI:27) (Cowan and White 1989) although no study has been made on the
incisors. Diets are known to have effect on the tooth wear (Kaiser et al. 2009, Kaiser et al.
2010) and would likely to have effect on possum incisor wear depending on the population‘s
geographical and/or seasonal change in diet composition.
5.2.3 Wildlife management
Identification of individual animals allows a wide range of research methods to understand its
biology and ecology, besides estimating its abundance and distribution. This novel technique
is essentially non-invasive, enabling the detection of elusive and nocturnal individuals with
presumed minimal effect on its behaviour, especially for endangered species as its application
is not restricted to pest species. Compared to other available identification methods, this
method‘s spatial and temporal detection resolution can be easily set at a desired density: to
gather detailed information at a much higher precision than VHS/GPS collars, or to cover a
much larger area.
5.3 Study components
The current study was comprised of components illustrated in Figure 5-1, and described
below:
160
5.3.1 Species identification
The study's objective of defining bitemarks made by possums and rodents was described in
Chapter 2. Captive animals of known species were exposed to WaxTags® to collect bitemark
samples of known species. Blind measurements of tooth-mark widths from the bitemarks
showed a distinctive distribution. Tooth-mark width measurements were found useful to
define quantitatively the species responsible for the interference. Qualitative descriptions of
the species' bitemarks were also recorded, which may serve as a key to define a species from
its bitemarks.
5.3.2 Individual identification — laboratory
In Chapter 3 it was determined that under laboratory conditions, examination of microscopic
features on traces of WaxTag® interference, i.e. bitemarks, enabled individualisation of the
interfering animal. Using the novel approach of visually comparing individual characteristics
on the bitemark made in wax, it was possible to determine whether a particular bitemark was
made by the same individual animal that made another bitemark. Simulated bitemarks from
possum skulls and WaxTag® bitemarks from captive individual animals were assigned
random reference numbers to make a blind-test comparison of bitemarks from animals of
unknown (masked) identity. Although this study did not attempt to relate bitemarks to
possums i.e., identify the actual individual animal, information from identifying bitemarks
alone sufficed to conclude that the same or different unknown individual animal interfered
with one or more WaxTags®.
161
The new method is, to the best of my knowledge, the first example of successfully identifying
individuals from visually matching striations from its bitemarks. This new tool gave a means
to individually trace animals that bite WaxTags® over time or space.
5.3.3 Individual identification — field
Chapter 4 described field trials conducted to evaluate the utility of the new method. Although
no attempt was made to directly obtain any information from the wild animals, it was possible
to identify 20 groups of bitemarks on field-applied WaxTags®. This result was interpreted as
detecting 20 individual animals' bait-interference activity in the study area. Presence of the
target species in the study was confirmed by determining the species as was defined in
Chapter 2. The possum abundance was determined as a minimum of 20 live animals.
Individual identification from bitemarks by wild possums has illustrated, for the first time, the
most confounding intrinsic issues of the original WaxTag® method, i.e. multiple visits. The
previously unknown extent of multiple visits was observed spatially and temporally, the effect
of food lure on WaxTags® was determined, and activity areas were detected as minimum
convex polygons.
5.4 Conclusions
From the study components (5.2) summarised above, the following conclusions are made:
1. Measureable single tooth-mark widths and qualitatively observable characteristics
of bitemarks made by the brushtail possum, ship rat, Norway rat and mouse
defined the species that interfered with the WaxTag®. Possum tooth-mark width
measurements were significantly larger than all rodent species‘ measurements.
162
2. Microscopic observation and comparison of striation marks on bitemarks produced
under controlled laboratory environment permit individual identification of
possum bitemarks on WaxTags®.
3. Field application of this identification method makes it possible to differentiate
between individual wild possums.
4. Occurrence of multiple visits to WaxTags® by individual possums in the field was
widely varied — from extensively large number (33 tags in one night), distance
(311 m in one night), area (2.8 ha in one night) and temporal repetition (all 10
nights), to nil (one tag bitten on one night).
5. Food lure applied to the WaxTag® setup significantly increased the number of tags
bitten per night by individuals up to three nights, and showed no difference
afterwards. Presence of food lure did not significantly affect the number of
individuals detected during the study period.
6. This new method of non-invasive individual identification from bitemarks in
WaxTags® offered spatial and temporal information of the possums‘ behaviour at a
precision never achieved by any previous method, and has vast potential to
generate novel information in wild animals.
5.5 Future directions
5.5.1 Quantified identification
Quantification of identification method was attempted to mathematically compare and
evaluate similarity or correlation (Appendix G). Although attempts were made for fine-tuned
standardised image capture, the level of artefact inclusion was high. Therefore the results‘
data/noise ratio was not sufficient to allow mathematical comparison. Purpose-designed
163
image capture is necessary to investigate the method‘s potential for mathematical comparison,
and ultimately, automatic comparison.
5.5.2 Objective identification criteria
Establishment of objective criteria for identification would greatly enhance the reliability and
repeatability of identification. The subjective nature of pattern analysis in forensic science is
currently undergoing a paradigm shift to establish objective criteria which will be more robust
in its evidential value in court (Pretty and Sweet 2001a, 2010). The NAS report (2009) has
been particularly critical on the scientific basis of all identification methods except DNA
profiling, providing for a healthy stimulus to strengthen research efforts. In a broader scale,
the application of subjective traditional pattern analysis methods in this thesis raised the
intriguing question of how observers can recognise, extract and compare / identify an object.
This spills into the field of neurosensing and cognition and was not pursued in this study,
although it was felt excitingly interesting to look into.
5.5.3 Computer assisted identification
Together with research in the abovementioned topics to provide reliable methods and criteria
to quantify and identify individuals, automating the identification process through
computerised recognition has the potential to greatly enhance the efficiency by improving
accuracy and precision, increasing the ability to process a large number of specimens and
reducing labour/time costs.
Dental identification was achieved with commingled remains of Viet Nam war casualties with
assistance from Computer Assisted Postmortem Identification (CAPMI) system (Dailey 1997).
164
Similarly, Balasubramaniam et al. (2005) presented an individual identification system from
dental radiograph images. An intelligent dental identification system (IDIS) was developed by
Chomdej et al. (2006) to support dental identification from dental records, while also
accommodating incomplete and altered dental information input. In bitemark analysis, a shape
similarity index (SI) was generated from a shape comparison interactive program (SCIP) to
quantify the bitemark similarity between its biting dental cast, and was successfully used
under experimental conditions to provide the likelihood of a bitemark on skin or foodstuff to
be sourced to the biting set of teeth (Nambiar et al. 1995a, b). Naru and Dykes (1997)
attempted computerised dental–bitemark comparison from cross-correlation scores, although
have concluded that ―[the] method may have value when applied to certain mark types‖. They
also conceded that the when the transfer of dental features onto skin is not faithful enough to
replicate its details to a sufficient level, any comparison system may risk producing a false
result, which is believed to be an accurate statement regardless of an computerised or manual
comparison. Martin-de las Heras and Tafur (2011) observed that 3D image analysis of
bitemarks on wax provide a highly reliable result when the subjective decision of examiners
were forced to make a dichotomous response, and cautioned that a lower level of accuracy
could be expected if the method was applied to cutaneous bitemarks.
Likewise, many automated recognition systems have been developed with varied levels of
success in fingerprint recognition (e.g., Nyongesa et al. 2004, Meuwly 2009, Kumar and
Deva Vikram 2010, Langenburg 2011, Moses 2011), firearm and toolmark identification (e.g.,
Blackwell and Framan 1980, Geradts et al. 1995b, a, Alexandre 1996, Geradts and Keijzer
1996, Banno 2004, Banno et al. 2004, 2005, Roberge and Beauchamp 2006, Senin et al.
2006, Brinck 2008, Hamby et al. 2009), and wildlife identification (e.g., Burghardt et al.
2004b, Gope et al. 2005, Burghardt and Ćalić 2006, Burghardt and Campbell 2007, Anderson
et al. 2010), with some systems available commercially.
165
Adaptation of 3D laser scanning techniques to this study‘s bitemark surface on wax, as was
successfully done in bullet striation marks (Bachrach et al. 2010), may eliminate the need to
make a cast or apply standardised illumination, which was necessary in this study to visualise
3D features on a 2D image format. The applicability to the bitemark‘s concave surface as
opposed to the bullet‘s convex surface remains unknown.
5.5.4 Other applications of WaxTags®
The WaxTag® is a passive non-toxic detection device that poses no potential harm to target or
non-target animals. Therefore the non-invasive technique can target protected species if
discernable characteristics can be identified from the animal's interference, or for other target
pest species where it shares the habitat with a protected species.
Provided the collected data type, detectability and other assumptions are met, individually
identified animal occurrences can be used to measure its abundance e.g., (Efford 2004, Efford
et al. 2004, Efford et al. 2009), pattern and spatial analysis (Rangel et al. 2006, Rangel et al.
2010, Rosenberg and Anderson 2011). Home range studies, behaviour studies may also make
use of WaxTags® and enhance its findings from individual identification of the biter. Output
oriented field sampling design will further allow for a wide range of spatial analysis of
individual detection (Perry et al. 2002).
166
Chapter 1
WaxTag® operational principles
and forensic science
Chapter 2
Species identification
WaxTag® as a
monitoring tool
Chapter 3
Individual identification
— laboratory
Chapter 4
Individual identification
— field
Key Outputs





Single tooth-mark widths defined possum and rodent species' bitemarks.
Striation on bitemarks produced under laboratory environment enabled individual
identification of possum bitemarks
Individual wild possums were identified from field-applied WaxTags®.
Widely varied occurrences of multiple visits to WaxTags® by individual possums
were proven.
Although food lure significantly increased the number of tags bitten per night by
individuals up to three nights, no difference was observed afterwards, nor did it
increase the sensitivity to detect more individuals.
Figure 5-1. Relationship between study components, and the key outputs.
167
5.6 Contribution to knowledge
The question of identifying individual pest animals from bitemarks on WaxTags® to monitor
the species was addressed in this thesis by defining the interfering species‘ bitemarks, and
developing a novel method to identify individual animals from characteristics found on the
bitemarks. Although initially the bait interference method arose from the need to understand
the population abundance, particularly in wildlife management, the main approach used to
identify the individual characteristics were from the domain of forensic science from which
principles of individual identification were applied.
Interestingly, the forensic science disciplines are currently undergoing a major shift to review
the scientific foundations of what was previously assumed as a fact, e.g., the infallibility of
fingerprint identification or firearm identification (NAS 2009). Although this research used
the traditional cognitive method of pattern analysis and matching, the results add to empirical
evidence that the identification principles are useful to reliably identify individuals.
Conducting further research using cross-disciplinary approaches in this direction may advance
knowledge in both domains by identifying and strengthening the gaps in the methods and
their application.
168
References
Animal Welfare Act 1999. Public Act 1999 No 142. Statues of New Zealand.
Aboshi, H., J. A. Taylor, T. Takei and K. A. Brown. 1994. Comparison of bitemarks in
foodstuffs by computer imaging: a case report. The Journal of Forensic OdontoStomatology 12:41-44.
Abrams, J. L., D. A. McEntee, A. M. Paterson, S. C. Ogilvie, D. N. McLane and K. L.
Campbell. 2004. A comparative assessment of lures for the possum (Trichosurus
vulpecula). Lincoln University Contract Report, (unpubl.), Lincoln, N.Z. 17 p.
Acevedo, M. A., C. J. Corrada-Bravo, H. Corrada-Bravo, L. J. Villanueva-Rivera and T. M.
Aide. 2009. Automated classification of bird and amphibian calls using machine
learning: a comparison of methods. Ecological Informatics 4:206-214.
Adams, B. J. 2003. The diversity of adult dental patterns in the United States and the
implications for personal identification. Journal of Forensic Sciences 48:497-503.
von Ahn, L., M. Blum, N. Hopper and J. Langford. 2003. CAPTCHA: Using hard AI
problems for security. In: E. Biham (ed.) Advances in cryptology — EUROCRYPT
2003: International Conference on the Theory and Apploications of Cryptographic
Techniques, Warsaw, Poland, May 2003, Proceedings. 294-311. Springer, Berlin.
von Ahn, L., B. Maurer, C. McMillen, D. Abraham and M. Blum. 2008. reCAPTCHA:
human-based character recognition via web security measures. Science 321:14651468.
Aksu, M. N. and J. P. Gobetti. 1996. The past and present legal weight of bite marks as
evidence. The American Journal of Forensic Medicine and Pathology 17:136-140.
Al-Diri, B., A. Hunter, D. Steel and M. Habib. 2010. Automated analysis of retinal vascular
network connectivity. Computerized Medical Imaging and Graphics 34:462-470.
Al-Talabani, N., N. D. Al-Moussawy, F. A. Baker and H. A. Mohammed. 2006. Digital
analysis of experimental human bitemarks: application of two new methods. Journal
of Forensic Sciences 51:1372-1375.
Alberink, I. B. and A. C. C. Ruifrok. 2007. Performance of the FearID earprint identification
system. Forensic Science International 166:145-154.
Alexandre, G. 1996. Computerized classification of the shoeprints of burglars' soles. Forensic
Science International 82:59-65.
Allen, A., B. Golden, M. Taylor, D. Patterson, D. Henriksen and R. Skuce. 2008. Evaluation
of retinal imaging technology for the biometric identification of bovine animals in
Northern Ireland. Livestock Science 116:42-52.
Allen, L. R., R. Engeman and H. Krupa. 1996. Evaluation of three relative abundance indices
for assessing dingo populations. Wildlife Research 23:197-206.
American Board of Forensic Odontology. 1997. ABFO guidelines and standards. In: C. M.
Bowers and G. L. Bell (eds.) Manual of forensic odontology. 299-357. American
Society of Forensic Odontology, Saratoga Springs, N.Y.
169
Anderson, C. J. R., J. D. Roth and J. M. Waterman. 2007. Can whisker spot patterns be used
to identify individual polar bears? Journal of Zoology 273:333-339.
Anderson, C. J. R., N. D. V. Lobo, J. D. Roth and J. M. Waterman. 2010. Computer-aided
photo-identification system with an application to polar bears based on whisker spot
patterns. Journal of Mammalogy 91:1350-1359.
Articus, K., J.-E. Mattsson, L. Tibell, M. Grube and M. Wedin. 2002. Ribosomal DNA and tubulin data do not support the separation of the lichens Usnea florida and U.
subfloridana as distinct species. Mycological Research 106:412-418.
Arulchelvam, I. M. and J. A. Brown. 1995. The relative accuracy, precision, and costeffectiveness of the bait-consumption and bait-interference monitoring techniques.
Landcare Research Contract Report LC9495/159, (unpubl.), Lincoln, N.Z. 8 p.
Arzoumanian, Z., J. Holmberg and B. Norman. 2005. An astronomical pattern-matching
algorithm for computer-aided identification of whale sharks Rhincodon typus. Journal
of Applied Ecology 42:999-1011.
Association of Firearm and Tool Mark Examiners. 1992. Theory of identification, range of
striate comparison reports and modified glossary definitions - an AFTE Criteria for
Identification Commitee report. Association of Firearm and Tool Mark Examiners
Journal 24:336-340.
Association of Firearm and Tool Mark Examiners. 1998. Theory of identification as it relates
to toolmarks. Association of Firearm and Tool Mark Examiners Journal 30:86-88.
Association of Firearm and Toolmark Examiners Training and Standardization Committee.
2001. Glossary of the Association of Firearm and Tool Mark Examiners, 4th edition.
Auger-Méthé, M. and H. Whitehead. 2007. The use of natural markings in studies of longfinned pilot whales (Globicephala melas). Marine Mammal Science 23:77-93.
Avon, S.-L., C. Victor, J. T. Mayhall and R. E. Wood. 2010. Error rates in bite mark analysis
in an in vivo animal model. Forensic Science International 201:45-55.
Bachrach, B., A. Jain, S. Jung and R. D. Koons. 2010. A statistical validation of the
individuality and repeatability of striated tool marks: screwdrivers and tongue and
groove pliers. Journal of Forensic Sciences 55:348-357.
Baker, A. N., A. N. H. Smith and F. B. Pichler. 2002. Geographical variation in Hector's
dolphin: recognition of new subspecies of Cephalorhynchus hectori. Journal of the
Royal Society of New Zealand 32:713-727.
Baker, C. S., F. Cipriano and S. R. Palumbi. 1996. Molecular genetic identification of whale
and dolphin products from commercial markets in Korea and Japan. Molecular
Ecology 5:671-685.
Baker, C. S., D. Steel, Y. Choi, H. Lee, K. S. Kim, S. K. Choi, Y.-U. Ma, C. Hambleton, L.
Psihoyos, R. L. Brownell and N. Funahashi. 2010. Genetic evidence of illegal trade in
protected whales links Japan with the US and South Korea. Biology Letters 6:647-650.
Balasubramaniam, S. R., G. V. Kumar, R. A. Narayanan, S. Raju, A. Banumathi and A.
Kannammal. 2005. Person identification by matching dental images. In: Proceedings
of the International Conference on Cognition and Recognition. 857-862. P E S College
of Engineering, Maysore, India.
170
Baldi, P., S. Brunak, Y. Chauvin, C. A. F. Andersen and H. Nielsen. 2000. Assessing the
accuracy of prediction algorithms for classification: an overview. Bioinformatics
16:412-424.
Bamford, J. 1970. Evaluating opossum poisoning operations by interference with non-toxic
baits. Proceedings of the New Zealand Ecological Society 17:118-125.
Banno, A. 2004. Estimation of bullet striation similarity using neural networks. Journal of
Forensic Sciences 49:500-504.
Banno, A., T. Masuda and K. Ikeuchi. 2004. Three dimensional visualization and comparison
of impressions on fired bullets. Forensic Science International 140:233-240.
Banno, A., T. Masuda and K. Ikeuchi. 2005. Forensic retrieval of striations on fired bullets by
using 3D geometric data. IAPR Conference on Machine Vision Applications
(MVA2005), May 2005.
Bashir, K., T. Xiang and S. Gong. 2010. Gait recognition without subject cooperation. Pattern
Recognition Letters 31:2052-2060.
Batcheler, C. L., J. H. Darwin and L. T. Pracy. 1967. Estimation of opposum (Trichosurus
vulpecula) populations and results of poison trials from trapping data. New Zealand
Journal of Science 10:97-114.
Bearman, R. 2002. Wax blocks: developing an alternative possum monitoring method.
Internal Department of Conservation Report TWIAO-12415, (unpubl.), Wellington,
N.Z. 27 p.
Beausoleil, N. J., D. J. Mellor and K. J. Stafford. 2004. Methods for marking New Zealand
wildlife: amphibians, reptiles and marine mammals. Department of Conservation,
Wellington, N.Z. 147 p.
Beier, P. and S. C. Cunningham. 1996. Power of track surveys to detect changes in cougar
populations. Wildlife Society Bulletin 24:540-546.
Bernitz, H. and B. A. Kloppers. 2002. Comparison microscope identification of a cheese
bitemark: a case report. The Journal of Forensic Odonto-Stomatology 20:13-16.
Bernitz, H., W. F. P. van Heerden, T. Solheim and J. H. Owen. 2006. A technique to capture,
analyze, and quantify anterior teeth rotations for application in court cases involving
tooth marks. Journal of Forensic Sciences 51:624-629.
Bernstein, M. L. 2005. Nature of bitemarks. In: R. B. J. Dorion (ed.) Bitemark evidence. 5980. Marcel Dekker, New York.
Besana, J. L. and T. L. Rogers. 2010. Personal identification using the frontal sinus. Journal
of Forensic Sciences 55:584-589.
Bianucci, G., B. Sorce, T. Storai and W. Landini. 2010. Killing in the Pliocene: shark attack
on a dolphin from Italy. Palaeontology 53:457-470.
Biasotti, A. A. 1959. A statistical study of the individual characteristics of fired bullets.
Journal of Forensic Sciences 4:34-50.
Biasotti, A. A. and J. Murdock. 1984. "Criteria for identification" or "state of the art" of
firearm and toolmark identification. Association of Firearm and Tool Mark Examiners
Journal 16:16-24.
171
Biasotti, A. A. 1989. Methods applied to the comparison of class and individual
characteristics in firearms and toolmark identification. Association of Firearm and
Tool Mark Examiners Journal 21:260-263.
Bishara, S. E., J. R. Jakobsen, J. E. Treder and M. J. Stasl. 1989. Changes in the maxillary and
mandibular tooth size–arch length relationship from early adolescence to early
adulthood : a longitudinal study. American Journal of Orthodontics and Dentofacial
Orthopedics 95:46-59.
Bishara, S. E., P. Khadivi and J. R. Jakobsen. 1995. Changes in tooth size–arch length
relationships from the deciduous to the permanent dentition: a longitudinal study.
American Journal of Orthodontics and Dentofacial Orthopedics 108:607-613.
Black, N. A., A. Schulman-Janiger, R. L. Ternullo and M. Guerrero-Ruiz. 1997. Killer whales
of California and Western Mexico: a photograph catalog of photo-identified
individuals. NOAA Technical Memorandum NMFS NOAA-TM-NMFS-SWFSC-247,
NOAA. 174 p.
Blackwell, G. L., M. A. Potter and J. A. McLennan. 2002. Rodent density indices from
tracking tunnels, snap-traps and Fenn traps: do they tell the same story? New Zealand
Journal of Ecology 26:43-51.
Blackwell, R. and E. Framan. 1980. Automated firearms identification system (AFIDS):
Phase I. Association of Firearm and Tool Mark Examiners Journal 12:11-37.
Blackwell, S. A., R. V. Taylor, I. Gordon, C. L. Ogleby, T. Tanijiri, M. Yoshino, M. R.
Donald and J. G. Clement. 2007. 3-D imaging and quantitative comparison of human
dentitions and simulated bite marks. International Journal of Legal Medicine 121:917.
Bodziak, W. J. 2005a. Forensic footwear evidence. In: S. H. James and J. J. Nordby (eds.)
Forensic science: an introduction to scientific and investigative techniques. 361-376.
CRC Press, Boca Raton, Fla.
Bodziak, W. J. 2005b. Forensic tire impression and tire track evidence. In: S. H. James and J.
J. Nordby (eds.) Forensic science: an introduction to scientific and investigative
techniques. 377-390. CRC Press, Boca Raton, Fla.
Boersen, M. R., J. D. Clark and T. L. King. 2003. Estimating black bear population density
and genetic diversity at Tensas River, Louisiana using microsatellite DNA markers.
Wildlife Society Bulletin 31:197-207.
Bollongino, R., J. Burger and S. Hummel. 2003. DNA typing from elephant tusks and
rhinoceros horn - Strategies and applications in wildlife conservation and animal
forensics. Forensic Science International 136:374-375.
Bookhout, T. A., (ed.) 1994. Research and management techniques for wildlife and habitats,
5th edition. The Wildlife Society, Bethesda. 740 p.
Borgula, L. M., F. G. Robinson, M. Rahimi, K. E. Chew, K. R. Birchmeier, S. G. Owens, J.
A. Kieser and G. R. Tompkins. 2003. Isolation and genotypic comparison of oral
streptococci from experimental bitemarks. The Journal of Forensic OdontoStomatology 21:23-30.
Bouchrika, I., M. Goffredo, J. Carter and M. Nixon. 2011. On using gait in forensic
biometrics. Journal of Forensic Sciences 56:882-889.
172
Bowers, C. M. 1997. Determining age from teeth: the estimation of age from dental
development. In: C. M. Bowers and G. L. Bell (eds.) Manual of forensic odontology.
74-85. American Society of Forensic Odontology, Saratoga Springs, N.Y.
Bowers, C. M. and R. J. Johansen. 2002. Photographic evidence protocol: the use of digital
imaging methods to rectify angular distortion and create life size reproductions of bite
mark evidence. Journal of Forensic Sciences 47:178-185.
Bowers, C. M. 2006. Problem-based analysis of bitemark misidentifications: the role of DNA.
Forensic Science International 159:S104-S109.
Bowers, C. M. 2008. Compelling cases for change. Science 320:1722-1722.
Bowers, C. M. and I. A. Pretty. 2009. Expert disagreement in bitemark casework. Journal of
Forensic Sciences 54:915-918.
Bowers, C. M. and G. L. Bell, (eds.) 1997. Manual of forensic odontology, 3rd edition.
American Society of Forensic Odontology, Saratoga Springs. 357 p.
Brinck, T. B. 2008. Comparing the performance of IBIS and BulletTRAX-3D technology
using bullets fired through 10 consecutively rifled barrels. Journal of Forensic
Sciences 53:677-682.
Brown, C. G. and W. W. Bryant. 1995. Consecutively rifled gun barrels present in most crime
labs. Association of Firearm and Tool Mark Examiners Journal 27:254-258.
Brown, J. A. and M. D. Thomas. 2000. Residual trap-catch methodology for low-density
possum populations. Part 1, Improving the precision of possum population density
estimates. Research report no. UCDMS2000/6, University of Canterbury. Dept. of
Mathematics and Statistics, Christchurch, N.Z. 29 p.
Brown, J. A., R. J. McAuliffe and G. J. Hickling. 2004. A new trap-catch index of patchiness
in animal distribution. Wildlife Research 31:433-436.
Brown, K. P., H. Moller, J. Innes and N. Alterio. 1996. Calibration of tunnel tracking rates to
estimate relative abundance of ship rats (Rattus rattus) and mice (Mus musculus) in a
New Zealand forest. New Zealand Journal of Ecology 20:271-275.
Brown, T. and G. N. Rutty. 2005. Identification: prints, footprint. In: J. Payne-James, R. W.
Byard, T. S. Corey and C. Henderson (eds.) Encyclopedia of forensic and legal
medicine. 013-018. Elsevier, Amsterdam.
Brundage, D. J. 1998. The identification of consecutively rifled gun barrels. Association of
Firearm and Tool Mark Examiners Journal 30:438-444.
Burghardt, T., J. Ćalić and B. T. Thomas. 2004a. Tracking animals in wildlife videos using
face detection. European Workshop on the Integration of Knowledge, Semantics and
Digital Media Technology, 2004. EWIMT 2004.
Burghardt, T., B. Thomas, P. J. Barham and J. Ćalić. 2004b. Automated visual recognition of
individual African penguins. Fifth International Penguin Conference, Ushuaia, Tierra
del Fuego, Argentina, September 2004, Ushuaia, Tierra del Fuego, Argentina.
Burghardt, T. and J. Ćalić. 2006. Analysing animal behaviour in wildlife videos using face
detection and tracking. IEEE Proceedings - Vision, Image, and Signal Processing
153:305-312.
173
Burghardt, T. and N. Campbell. 2007. Individual animal identification using visual biometrics
on deformable coat patterns. In: Proceedings of the 5th International Conference on
Computer Vision Systems. University Library of Bielefeld.
Bush, M. A., H. I. Cooper and R. B. J. Dorion. 2010a. Inquiry into the scientific basis for
bitemark profiling and arbitrary distortion compensation. Journal of Forensic Sciences
55:976-983.
Bush, M. A., K. Thorsrud, R. G. Miller, R. B. J. Dorion and P. J. Bush. 2010b. The response
of skin to applied stress: investigation of bitemark distortion in a cadaver model.
Journal of Forensic Sciences 55:71-76.
Bush, M. A., P. J. Bush and H. D. Sheets. 2011a. Statistical evidence for the similarity of the
human dentition. Journal of Forensic Sciences 56:118-123.
Bush, M. A., P. J. Bush and H. D. Sheets. 2011b. A study of multiple bitemarks inflicted in
human skin by a single dentition using geometric morphometric analysis. Forensic
Science International 211:1-8.
Byers, R. E. 1981. Field method for evaluation of rodenticides for control of Microtus
pinetorum in apple orchards. In: E. W. Schafer, Jr. and C. R. Walker (eds.) Vertebrate
pest control and management materials: proceedings of the Third Symposium on Test
Mehods for Vertebrate Pest Control and Management Materials. 77-85. American
Society for Testing and Materials, Philadelphia.
Carey, P. W., C. O'Connor, R. M. McDonald and L. R. Matthews. 1997. Comparison of the
attractiveness of acoustic and visual stimuli for brushtail possums. New Zealand
Journal of Zoology 24:273-276.
Casanova-Rosado, J. F., C. E. Medina-Solis, A. A. Vallejos-Sanchez, A. J. Casanova-Rosado,
G. Maupome and L. Avila-Burgos. 2005. Dental attrition and associated factors in
adolescents 14 to 19 years of age: a pilot study. International Journal of
Prosthodontics 18:516-519.
Cattaneo, C. and D. De Angelis. 2009. Visual recognition systems in identification. In: A.
Jamieson and A. A. Moenssens (eds.) Wiley encyclopedia of forensic science. John
Wiley & Sons, Ltd, New York.
Caughley, G. 1977a. Analysis of vertebrate populations. Wiley, London. 234 p.
Caughley, G. 1977b. Sampling in aerial survey. Journal of Wildlife Management 41:605-615.
Caughley, G. and A. R. E. Sinclair. 1994. Wildlife ecology and management. Blackwell
Scientific Publications, Boston. 334 p.
Ceugniet, M. and A. Izumi. 2004. Vocal individual discrimination in Japanese monkeys.
Primates 45:119-128.
Chattah, N. L.-T. and P. Smith. 2006. Variation in occlusal dental wear of two Chalcolithic
populations in the southern Levant. American Journal of Physical Anthropology
130:471-479.
Chen, C., J. Liang and X. Zhu. 2011. Gait recognition based on improved dynamic Bayesian
networks. Pattern Recognition 44:988-995.
Cheng, J., Y. Sun and L. Ji. 2010. A call-independent and automatic acoustic system for the
individual recognition of animals: a novel model using four passerines. Pattern
Recognition 43:3846-3852.
174
Chesmore, E. D. and E. Ohya. 2004. Automated identification of field-recorded songs of four
British grasshoppers using bioacoustic signal recognition. Bulletin of Entomological
Research 94:319-330.
Chew, M. and H. S. Baird. 2003. BaffleText: a human interactive proof. In: T. Kanungo, E.
H. B. Smith, J. Hu and P. B. Kantor (eds.) Document Recognition and Retrieval X,
Santa Clara, CA, USA, 20 January 2003, Proceedings. 305-316. SPIE.
Childerhouse, S. J., S. M. Dawson and E. Slooten. 1996. Stability of fluke marks used in
individual photoidentification of male sperm whales at Kaikoura, New Zealand.
Marine Mammal Science 12:447-451.
Chomdej, T., W. Pankaow and S. Choychumroon. 2006. Intelligent dental identification
system (IDIS) in forensic medicine. Forensic Science International 158:27-38.
Christchurch City Council. 2003. Bottle Lake Forest Park fact sheet. Christchurch City
Council, Christchurch.
Clout, M. and K. Ericksen. 2000. Anatomy of a disastrous success: the brushtail possum as an
invasive species. In: T. L. Montague (ed.) The brushtail possum: biology, impact and
management of an introduced marsupial. 1-9. Manaaki Whenua Press, Lincoln, N.Z.
Cole, S. A. 2009. Forensics without uniqueness, conclusions without individualization: the
new epistemology of forensic identification. Law, Probability and Risk 8:233-255.
Coleman, J. and P. Caley. 2000. Possums as a reservoir of bovine Tb. In: T. L. Montague
(ed.) The brushtail possum: biology, impact and management of an introduced
marsupial. 92-104. Manaaki Whenua Press, Lincoln, N.Z.
Commins, P. 2005. Possum monitoring report: Matea WaxTag trial. Commins Enterprises,
(unpubl.).
Committee on Identifying the Needs of the Forensic Science Community and National
Academy of Sciences. 2009. Strengthening forensic science in the United States: a
path forward. National Academies Press, Washington, D.C.
Conner, M. C., R. F. Labisky and D. R. Progulske, Jr. 1983. Scent-station indices as measure
of population abundance for bobcats, raccoons, gray foxes, and opossums. Wildlife
Society Bulletin 11:146-152.
Cooper, J. E. and M. E. Cooper. 1998. Future trends in forensic veterinary medicine.
Seminars in Avian and Exotic Pet Medicine 7:210-217.
Corbett, M. E. and D. Spence. 1984. A forensic investigation of teeth marks in soap. British
Dental Journal 157:270-271.
Cork, D. L., J. E. Rolph, E. S. Meieran and C. V. Petrie, (eds.) 2008. Ballistic imaging.
National Academies Press, Washington, D.C. 322 p.
Cowan, E. J. and R. Koppl. 2011. An experimental study of blind proficiency tests in forensic
science. The Review of Austrian Economics 24:251-271.
Cowan, P. 2000. Biological control of possums: prospects for the future. In: T. L. Montague
(ed.) The brushtail possum: biology, impact and management of an introduced
marsupial. 262-270. Manaaki Whenua Press, Lincoln, N.Z.
Cowan, P. E. and A. J. White. 1989. Evaluation of a tooth-wear age index for brushtail
possums, Trichosurus vulpecula. Australian Wildlife Research 16:321-322.
175
Cowan, P. E. 1990. Brushtail possum. In: C. M. King (ed.) The handbook of New Zealand
mammals. 68-98. Oxford University Press, Auckland, N.Z.
Coward, R. C. 2007. The stability of lip pattern characteristics over time. The Journal of
Forensic Odonto-Stomatology 25:40-56.
Cunado, D., M. S. Nixon and J. N. Carter. 2003. Automatic extraction and description of
human gait models for recognition purposes. Computer Vision and Image
Understanding 90:1-41.
Curran, J. M. and T. N. Hicks. 2009. Glass evidence: Bayesian approach to. In: A. Jamieson
and A. A. Moenssens (eds.) Wiley encyclopedia of forensic science. John Wiley &
Sons, Ltd, New York.
Cuthbert, R. 2003. Sign left by introduced and native predators feeding on Hutton's
shearwaters Puffinus huttoni. New Zealand Journal of Zoology 30:163-170.
Dailey, J. C. 1997. Computer assisted identification of vietnam war dental remains. In: C. M.
Bowers and G. L. Bell (eds.) Manual of forensic odontology. 227-231. American
Society of Forensic Odontology, Saratoga Springs, N.Y.
Dailey, J. C. 2005. The comparison. In: R. B. J. Dorion (ed.) Bitemark evidence. 423-451.
Marcel Dekker, New York.
Dailey, M. N. and G. W. Cottrell. 1999. Organization of face and object recognition in
modular neural network models. Neural Networks 12:1053-1074.
Daugman, J. 2009. How iris recognition works. In: A. C. Bovik (ed.) The essential guide to
image processing (second edition). 715-739. Academic Press, Boston.
Davis, D. E. and R. L. Winstead. 1980. Estimating the numbers of wildlife populations. In: S.
D. Schemnitz (ed.) Wildlife management techniques manual. 221-245. The Wildlife
Society, Washington, D.C.
Davis, J. E. 1958. An introduction to tool marks, firearms and the striagraph. Charles C.
Thomas, Springfield (Illinois). 282 p.
Day, T., C. O'Connor and L. Matthews. 2000. Possum social behaviour. In: T. L. Montague
(ed.) The brushtail possum: biology, impact and management of an introduced
marsupial. 35-46. Manaaki Whenua Press, Lincoln, N.Z.
De Barba, M., L. P. Waits, E. O. Garton, P. Genovesi, E. Randi, A. Mustoni and C. Groff.
2010. The power of genetic monitoring for studying demography, ecology and
genetics of a reintroduced brown bear population. Molecular Ecology 19:3938-3951.
De Kinder, J., P. Prévot, M. Pirlot and B. Nys. 1998. Surface topology of bullet striations: an
innovating technique. Association of Firearm and Tool Mark Examiners Journal
30:294-299.
Deedrick, D. W. and S. L. Koch. 2004. Microscopy of hair part II: a practical guide and
manual for animal hairs. Forensic Science Communications 6.
Deinet, W. 1981. Studies of models of striated marks generated by random processes. Journal
of Forensic Sciences 26:35-50.
Deng, W., J. Hu, J. Guo, W. Cai and D. Feng. 2010. Emulating biological strategies for
uncontrolled face recognition. Pattern Recognition 43:2210-2223.
176
Dessimoz, D. and C. Champod. 2008. Linkages between biometrics and forensic science. In:
A. K. Jain, P. Flynn and A. A. Ross (eds.) Handbook of biometrics. 425-459. Springer
US, New York.
DeVore, D. T. 1971. Bite marks for identification? — A preliminary report. Medicine,
Science, and the Law 11:144-145.
DeYoung, R. W. and R. L. Honeycutt. 2005. The molecular toolbox: genetic techniques in
wildlife ecology and management. Journal of Wildlife Management 69:1362-1384.
Dinsmore, S. J. and D. H. Johnson. 2005. Population analysis in wildlife biology. In: C. E.
Braun (ed.) Techniques for wildlife investigations and management. 154-184. The
Wildlife Society, Bethesda.
Drennan, J. E., P. Beier and N. L. Dodd. 1998. Use of track stations to index abundance of
sciurids. Journal of Mammalogy 79:352-359.
Dror, I. E. 2005. Perception is far from perfection: the role of the brain and mind in
constructing realities. Brain and Behavioural Sciences 28:763-763.
Dror, I. E., A. E. Péron, S.-L. Hind and D. Charlton. 2005. When emotions get the better of
us: the effect of contextual top-down processing on matching fingerprints. Applied
Cognitive Psychology 19:799-809.
Dror, I. E. and D. Charlton. 2006. Why experts make errors. Journal of Forensic
Identification 56:600-616.
Dror, I. E., D. Charlton and A. E. Péron. 2006. Contextual information renders experts
vulnerable to making erroneous identifications. Forensic Science International
156:74-78.
Dror, I. E. 2009. On proper research and understanding of the interplay between bias and
decision outcomes. Forensic Science International 191:e17-e18.
Dror, I. E. and S. A. Cole. 2010. The vision in "blind" justice: expert perception, judgment,
and visual cognition in forensic pattern recognition. Psychonomic Bulletin & Review
17:161-167.
Dror, I. E., C. Champod, G. Langenburg, D. Charlton, H. Hunt and R. Rosenthal. 2011.
Cognitive issues in fingerprint analysis: inter- and intra-expert consistency and the
effect of a 'target' comparison. Forensic Science International 208:10-17.
Du Pasquier, E., J. Hebrard, P. Margot and M. Ineichen. 1996. Evaluation and comparison of
casting materials in forensic sciences: applications to tool marks and foot/shoe
impressions. Forensic Science International 82:33-43.
Dutton, G. 2009. Firearms: bullet and cartridge case identification. In: A. Jamieson and A. A.
Moenssens (eds.) Wiley encyclopedia of forensic science. John Wiley & Sons, Ltd,
New York.
Efford, M. 2004. Density estimation in live-trapping studies. Oikos 106:598-610.
Efford, M. G., D. K. Dawson and C. S. Robbins. 2004. DENSITY: software for analysing
capture-recapture data from passive detector arrays. Animal Biodiversity and
Conservation 27:217-228.
Efford, M. G., D. L. Borchers and A. E. Byrom. 2009. Density estimation by spatially explicit
capture-recapture: likelihood-based methods. In: D. L. Thomson, E. G. Cooch and M.
177
J. Conroy (eds.) Modeling demographic processes in marked populations. 255-269.
Springer, New York.
Ekdahl, M. O., B. L. Smith and D. F. L. Money. 1970. Tuberculosis in some wild and feral
animals in New Zealand. New Zealand Veterinary Journal 18:44-45.
Erickson, G. M. and K. H. Olson. 1996. Bite marks attributable to Tyrannosaurus rex:
preliminary description and implications. Journal of Vertebrate Paleontology 16:175178.
Everhart, M. J. 2005. Bite marks on an elasmosaur (Sauropterygia; Plesiosauria) paddle from
the Niobrara Chalk (Upper Cretaceous) as probable evidence of feeding by the
lamniform shark, Cretoxyrhina mantelli. Vertebrate Palaeontology 2:14-24.
Everhart, M. J. 2009. A bitten skull of Tylosaurus kansasensis (Squamata: Mosasauridae) and
a review of mosasaur-on-mosasaur pathology in the fossil record. Transactions of the
Kansas Academy of Science 111:251-262.
Faden, D., J. A. Kidd, J. Craft, L. S. Chumbley, M. D. Morris, L. J. Genalo and e. al. 2007.
Statistical confirmation of empirical observations concerning tool mark striae.
Association of Firearm and Tool Mark Examiners Journal 39:205-214.
Fageth, R., W. G. Allen and U. Jäger. 1996. Fuzzy logic classification in image processing.
Fuzzy Sets and Systems 82:265-278.
Frank, L. G. 1986. Social organization of the spotted hyaena (Crocuta crocuta). I.
Demography. Animal Behaviour 34:1500-1509.
Franklin, D. 2010. Forensic age estimation in human skeletal remains: Current concepts and
future directions. Legal Medicine 12:1-7.
Fraser, K. W., J. A. Brown and B. Warburton. 2002. Vector (possum and ferret) population
monitoring: a review. Landcare Research Contract Report LC0203/040, (unpubl.),
Lincoln, N.Z. 48 p.
Freeman, R. A. 1978. Consecutively rifled polygon barrels. Association of Firearm and Tool
Mark Examiners Journal 10:40-42.
Garrett, B. L. and P. J. Neufeld. 2009. Invalid forensic science testimony and wrongful
convictions. Virginia Law Review 95. 97 p.
Geradts, Z. J. M. H., J. Keijzer and I. Keereweer. 1995a. A new approach to automatic
comparison of striation marks. In: B. Jacob and W. Bonte (eds.) Advances in forensic
sciences - Proceedings of the 13th Meeting of the IAFS, Düsseldorf 1993. 266-269.
Verlag Dr. Köster, Berlin.
Geradts, Z. J. M. H., J. Keijzer and I. Keereweer. 1995b. Automatic comparison of striation
marks and automatic classification of shoe prints. In: L. I. Rudin and S. K. Bramble
(eds.) Investigative and trial image processing. 151-164. The International Society for
Optical Engineering (SPIE), San Diego.
Geradts, Z. J. M. H. and J. Keijzer. 1996. TRAX for toolmarks. Association of Firearm and
Tool Mark Examiners Journal 28:183-190.
Gilmore, D. P. 1966. Studies on the biology of Trichosurus vulpecula Kerr. Thesis, Doctor of
Philosophy. University of Canterbury, Christchurch. 300 p.
Glass, R. T. 2002. Body identification by forensic dental means. General Dentistry 50:34-38.
178
Glass, R. T. 2005. Forensic odontology. In: S. H. James and J. J. Nordby (eds.) Forensic
science: an introduction to scientific and investigative techniques. 79-98. CRC Press,
Boca Raton, Fla.
Goddard, K. 2011. Wildlife forensic science. In: A. Mozayani and C. Noziglia (eds.) The
forensic laboratory handbook: procedures and practice. 563-579. Humana Press, New
York.
González-Colmenares, G., M. C. Botella-López, G. Moreno-Rueda and J. R. FernándezCardenete. 2007. Age estimation by a dental method: a comparison of Lamendin's and
Prince & Ubelaker's technique. Journal of Forensic Sciences 52:1156-1160.
Gonzalez-Rodriguez, J., J. Fiérrez-Aguilar, J. Ortega-Garcia and J. J. Lucena-Molina. 2006.
Biometric identification in forensic cases according to the Bayesian approach. In: M.
Tistarelli, J. Bigun and A. K. Jain (eds.) Biometric Authentication. 177-185. Springer
Berlin / Heidelberg, Berlin.
Gonzalez, R. C. and R. E. Woods. 2002. Digital image processing, 2nd edition. Prentice Hall,
Upper Saddle River. 793 p.
Gope, C., N. Kehtarnavaz, G. Hillman and B. Würsig. 2005. An affine invariant curve
matching method for photo-identification of marine mammals. Pattern Recognition
38:125-132.
Gowans, S. and H. Whitehead. 2001. Photographic identification of northern bottlenose
whales (Hyperoodon ampullatus): sources of heterogeneity from natural marks.
Marine Mammal Science 17:76-93.
Graham, E. A. M. 2009. DNA: an overview. In: A. Jamieson and A. A. Moenssens (eds.)
Wiley encyclopedia of forensic science. John Wiley & Sons, Ltd, New York.
Green, W. Q. 1984. A review of ecological studies relevant to management of the common
brushtail possum. In: A. P. Smith and I. D. Hume (eds.) Possums and gliders. 483-499.
Surrey Beatty in association with The Australian Mammal Society, Chipping Norton,
Australia.
Grigione, M. M., P. Burman, V. C. Bleich and B. M. Pierce. 1999. Identifying individual
mountain lions Felis concolor by their tracks: refinement of an innovative technique.
Biological Conservation 88:25-32.
Grzybowski, R., J. Miller, B. Moran, J. Murdock, R. Nichols and R. Thompson. 2003.
Firearm/toolmark identification: passing the reliability test under federal and state
evidentiary standards. Association of Firearm and Tool Mark Examiners Journal
35:209-241.
Hall, E. E. 1983. Bullet markings from consecutively rifled Shilen DGA barrels. Association
of Firearm and Tool Mark Examiners Journal 15:33-53.
Hall, J. 1992. Consecutive cuts by bolt cutters and their effect on identification. Association of
Firearm and Tool Mark Examiners Journal 24:260-272.
Hall, L. J. and E. Player. 2008. Will the introduction of an emotional context affect fingerprint
analysis and decision-making? Forensic Science International 181:36-39.
Haltenorth, T. and H. Diller. 1980. A field guide to mammals of Africa including Madagascar.
Collins, London.
179
Hamby, J. E. and J. W. Thorpe. 1999. The history of firearm and toolmark identification.
Association of Firearm and Tool Mark Examiners Journal 31:266-284.
Hamby, J. E., D. J. Brundage and J. W. Thorpe. 2009. The identification of bullets fired from
10 consecutively rifled 9mm Ruger pistol barrels: a research project involving 507
participants from 20 countries. Association of Firearm and Tool Mark Examiners
Journal 41:99-110.
Hamby, J. E. 2010. The examination of bullets fired from 10 consecutively manufactured
9 mm barrels — A project involving 502 participants from 20 countries. Science &
Justice 50:30-30.
Harari, D., G. Hermolin and O. Harari. 2005. The effect of age on morphology and eruption
of the lower incisors in mature rats. Archives of Oral Biology 50:953-958.
Harvey, W., P. Millington, J. C. Barbenel and J. H. Evans. 1976. Experimental human bitemarks. In: W. Harvey (ed.) Dental identification and forensic odontology. 124-135.
Kimpton, London.
Hay, J. R. 1981. The kokako. Forest Bird Research Group, Wellington. 168 p.
Haynes, G. 1983. A guide for differentiating mammalian carnivore taxa responsible for gnaw
damage to herbivore limb bones. Paleobiology 9:164-172.
Hayward, W. G., G. Rhodes and A. Schwaninger. 2008. An own-race advantage for
components as well as configurations in face recognition. Cognition 106:1017-1027.
Herzog, C. J., R. W. Kays, J. C. Ray, M. E. Gompper and W. J. Zielinski. 2007. Using
patterns in track-plate footprints to identify individual fishers. Journal of Wildlife
Management 71:955-963.
Hewett, D. and E. Banks. 1999. Bottle Lake Forest Park management plan. Christchurch City
Council, Christchurch.
Hodge, E. E. 1988. Guarding against error. Association of Firearm and Tool Mark Examiners
Journal 20:290-293.
Hoffman, K. L. and N. K. Logothetis. 2009. Cortical mechanisms of sensory learning and
object recognition. Philosophical Transactions of the Royal Society B: Biological
Sciences 364:321-329.
Howitt, D., F. Tulleners, K. Cebra and S. Chen. 2008. A calculation of the theoretical
significance of matched bullets. Journal of Forensic Sciences 53:868-875.
Hueske, E. E. 2011. Firearms and toolmarks. In: A. Mozayani and C. Noziglia (eds.) The
forensic laboratory handbook: procedures and practice. 227-264. Humana Press, New
York.
Hunter, A. E. 2004. An assessment of visual lures for the possum Trichosurus vulpecula.
Thesis, Master of Applied Science. Lincoln University, Lincoln, N.Z. 57 p.
Innes, J. G. 1990. Ship rat. In: C. M. King (ed.) The handbook of New Zealand mammals.
206-225. Oxford University Press, Auckland, N.Z.
Jackson, M. 2009. Firearm examination: ballistics. In: A. Jamieson and A. A. Moenssens
(eds.) Wiley encyclopedia of forensic science. John Wiley & Sons, Ltd, New York.
180
Jane, G. T. 1979. Opposum density assessment using the bait interference method. New
Zealand Journal of Forestry 24:61-66.
Jane, G. T. 1981. Application of the Poisson model to the bait interference method of possum
Trichosurus vulpecula assessment. In: B. D. Bell (ed.) Proceedings of the First
Symposium on Marsupials in New Zealand. 185-195. Zoology Publications from the
Victoria Univeristy of Wellington, Wellington, N.Z.
Janson, G., P. V. P. Oltramari-Navarro, R. B. S. de Oliveira, C. L. Quaglio, S. H. d. C. SalesPeres and B. Tompson. 2010. Tooth-wear patterns in subjects with Class II Division 1
malocclusion and normal occlusion. American Journal of Orthodontics and
Dentofacial Orthopedics 137:14.e11-14.e17.
Jewell, Z. C., S. K. Alibhai and P. R. Law. 2001. Censusing and monitoring black rhino
(Diceros bicornis) using an objective spoor (footprint) identification technique.
Journal of Zoology 254:1-16.
Ji, W., C. R. Veitch and J. L. Craig. 1999. An evaluation of the efficiency of rodent trapping
methods: the effect of trap arrangement, cover type, and bait. New Zealand Journal of
Ecology 23:45-51.
Johnson, T. L. 2005. The suspect. In: R. B. J. Dorion (ed.) Bitemark evidence. 415-421.
Marcel Dekker, New York.
Jolly, G. M. 1969. Sampling methods for aerial census of wildlife populations. East African
Agricultural and Forestry Journal (special issue) 34:46-49.
Kaiser, T. M., J. Brasch, J. C. Castell, E. Schulz and M. Clauss. 2009. Tooth wear in captive
wild ruminant species differs from that of free-ranging conspecifics. Mammalian
Biology - Zeitschrift für Säugetierkunde 74:425-437.
Kaiser, T. M., J. Fickel, W. J. Streich, J. Hummel and M. Clauss. 2010. Enamel ridge
alignment in upper molars of ruminants in relation to their natural diet. Journal of
Zoology 281:12-25.
Kannikeswaran, N. and D. Kamat. 2009. Mammalian bites. Clinical Pediatrics 48:145-148.
Katona, S. K., P. M. Harcourt, J. S. Perkins and S. D. Kraus. 1980. Humpback whales: a
catalogue of individuals identified by fluke photographs, 2nd edition. College of the
Atlantic, Bar Harbor. 169 p.
Katterwe, H. 2009. Toolmarks. In: A. Jamieson and A. A. Moenssens (eds.) Wiley
encyclopedia of forensic science. John Wiley & Sons, Ltd, New York.
Kauffman, E. G. 2009. Mosasaur predation on Upper Cretaceous nautiloids and ammonites
from the United States Pacific coast. Palaios 19:96-100.
Kavermann, M. J. 2004. Increasing the attractiveness of WaxTags® to possums (Trichosurus
vulpecula) using visual and olfactory lures. Thesis, Bachelor of Science with Honours.
Lincoln University, Lincoln, N.Z. 47 p.
Kelly, M. J. 2001. Computer-aided photograph matching in studies using individual
identification: an example from Serengeti cheetahs. Journal of Mammalogy 82:440449.
Kenna, J., M. Smyth, L. McKenna, C. Dockery and S. D. McDermott. 2011. The recovery
and persistence of salivary DNA on human skin. Journal of Forensic Sciences 56:170175.
181
Kennedy, D. 2011. Forensic dentistry and microbial analysis of bite marks. Australian Police
Journal 2011:6-15.
Kevin Zhou, S., R. Chellappa and N. Ramanathan. 2009. Unconstrained face recognition from
a single image. In: A. C. Bovik (ed.) The essential guide to image processing (second
edition). 677-713. Academic Press, Boston.
Kiely, T. F. 2005. Forensic evidence. In: S. H. James and J. J. Nordby (eds.) Forensic science:
an introduction to scientific and investigative techniques. 649-666. CRC Press, Boca
Raton, Fla.
Kieser, J. A., H. T. Groeneveld and C. B. Preston. 1985. Patterns of dental wear in the Lengua
Indians of Paraguay. American Journal of Physical Anthropology 66:21-29.
Kieser, J. A. and H. T. Groeneveld. 1987a. A metric analysis of Cercopithecus aethiops
dentition. International Journal of Primatology 8:261-280.
Kieser, J. A. and H. T. Groeneveld. 1987b. Static intraspecific allometry of jaws and teeth in
Cercopithecus aethiops. Journal of Zoology 212:499-510.
Kieser, J. A. and H. T. Groeneveld. 1988a. Allometric relations of teeth and jaws in man.
American Journal of Physical Anthropology 77:57-67.
Kieser, J. A. and H. T. Groeneveld. 1988b. Patterns of metric variability in the dentition of
Papio ursinus. American Journal of Primatology 14:141-151.
Kieser, J. A. and H. T. Groeneveld. 1989a. Patterns of sexual dimorphism and of variability in
the dentition of Otolemur crassicaudatus. International Journal of Primatology
10:137-147.
Kieser, J. A. and H. T. Groeneveld. 1989b. Allocation and discrimination based on human
odontometric data. American Journal of Physical Anthropology 79:331-337.
Kieser, J. A., N. A. Firth and H. Buckley. 2001. Dental misidentification on the basis of
presumed unique features. The Journal of Forensic Odonto-Stomatology 19:36-39.
Kieser, J. A. 2005. Weighing bitemark evidence: a postmodern perspective. Forensic Science,
Medicine, and Pathology 1:75-80.
Kieser, J. A., G. Tompkins, D. Buckingham, N. A. Firth and M. Swain. 2005. Bitemarks:
presentation, analysis, and evidential reliability. In: M. Tsokos (ed.) Forensic
pathology reviews. 157-179. Humana Press, Totowa.
Kieser, J. A., V. Bernal, J. N. Waddell and S. Raju. 2007. The uniqueness of the human
anterior dentition: a geometric morphometric analysis. Journal of Forensic Sciences
52:671-677.
Kim, Y.-K., H. S. Kho and K. H. Lee. 2000. Age estimation by occlusal tooth wear. Journal
of Forensic Sciences 45:303-309.
Kindelan, J. D., G. Rysiecki and W. P. Childs. 1998. Hypodontia: genotype or environment?
A case report of monozygotic twins. British Journal of Orthodontics 25:175-178.
King, C. M. and R. L. Edgar. 1977. Techniques for trapping and tracking stoats (Mustela
erminea); a review, and a new system. New Zealand Journal of Zoology 4:193-212.
King, C. M., (ed.) 1990. The handbook of New Zealand mammals. Oxford University Press,
Auckland, N.Z. 600 p.
182
King, C. M. and P. Barrett, (eds.) 2005. The handbook of New Zealand mammals, 2nd edition.
Oxford University Press, Melbourne. 610 p.
Kittelson, J. M., J. A. Kieser, D. M. Buckingham and G. P. Herbison. 2002. Weighing
evidence: quantitative measures of the importance of bitemark evidence. The Journal
of Forensic Odonto-Stomatology 20:31-37.
Kovach, A. I., M. K. Litvaitis and J. A. Litvaitis. 2003. Evaluation of fecal mtDNA analysis
as a method to determine the geographic distribution of a rare lagomorph. Wildlife
Society Bulletin 31:1061-1065.
Kumar, R. and B. R. Deva Vikram. 2010. Fingerprint matching using multi-dimensional
ANN. Engineering Applications of Artificial Intelligence 23:222-228.
Lancia, R. A., J. D. Nichols and K. H. Pollock. 1994. Estimating the number of animals in
wildlife populations. In: T. A. Bookhout (ed.) Research and management techniques
for wildlife and habitats. 215-253. The Wildlife Society, Bethesda.
Lancia, R. A., W. L. Kendall, K. H. Pollock and J. D. Nichols. 2005. Estimating the number
of animals in wildlife populations. In: C. E. Braun (ed.) Techniques for wildlife
investigations and management. 106-153. The Wildlife Society, Bethesda.
Lange, N. D., R. P. Thomas, J. Dana and R. M. Dawes. 2011. Contextual biases in the
interpretation of auditory evidence. Law and Human Behavior 35:178-187.
Langenburg, G. 2009. Friction ridge skin: comparison and identification. In: A. Jamieson and
A. A. Moenssens (eds.) Wiley encyclopedia of forensic science. John Wiley & Sons,
Ltd, New York.
Langenburg, G. 2011. Scientific research in the forensic discipline of friction ridge
individualization. In: Scientific Working Group on Friction Ridge Analysis Study and
Technology and National institute of Justice (eds.) SWGFAST - The fingerprint
sourcebook. 1-31. Scientific Working Group on Friction Ridge Analysis, Study and
Technology, Washington, D.C.
Lasser, A. J., A. J. Warnick and G. M. Berman. 2009. Three-dimensional comparative
analysis of bitemarks. Journal of Forensic Sciences 54:658-661.
Laws, R. M. 1966. Age criteria for the African elephant, Loxodonta a. africana. East African
Wildlife Journal 4:1-37.
Layton, J. J. 1966. Identification from a bite mark in cheese. Journal of the Forensic Science
Society 6:76-80.
Lewin, C. and A. Herlitz. 2002. Sex differences in face recognition—Women's faces make the
difference. Brain and Cognition 50:121-128.
Lewison, R., E. L. Fitzhugh and S. P. Galentine. 2001. Validation of a rigorous track
classification technique: identifying individual mountain lions. Biological
Conservation 99:313-321.
Linhart, S. B. and F. F. Knowlton. 1975. Determining the relative abundance of coyotes by
scent station lines. Wildlife Society Bulletin 3:119-124.
Litonjua, L. A., S. Andreana, P. J. Bush and R. E. Cohen. 2003. Tooth wear: attrition, erosion,
and abrasion. Quintessence International 34:435-446.
183
Lovelock, T. J. 2009. Glass. In: A. Jamieson and A. A. Moenssens (eds.) Wiley encyclopedia
of forensic science. John Wiley & Sons, Ltd, New York.
Lukacs, P. M. and K. P. Burnham. 2005. Estimating population size from DNA-based closed
capture-recapture data incorporating genotyping error. Journal of Wildlife
Management 69:396-403.
Lukoschek, V., N. Funahashi, S. Lavery, M. L. Dalebout, F. Cipriano and C. S. Baker. 2009.
High proportion of protected minke whales sold on Japanese markets is due to illegal,
unreported or unregulated exploitation. Animal Conservation 12:385-395.
Lynch, Á. B., M. J. F. Brown and J. M. Rochford. 2006. Fur snagging as a method of
evaluating the presence and abundance of a small carnivore, the pine marten (Martes
martes). Journal of Zoology 270:330-339.
Lynnerup, N. and J. Vedel. 2005. Person identification by gait analysis and photogrammetry.
Journal of Forensic Sciences 50:112-118.
Lyver, P. O. B. 2000. Identifying mammalian predators from bite marks: a tool for focusing
wildlife protection. Mammal Review 30:31-44.
MacFarlane, T. W., D. G. MacDonald and D. A. Sutherland. 1974. Statistical problems in
dental identification. Journal of the Forensic Science Society 14:247-252.
MacKenzie, D. I., J. A. Royle, J. A. Brown and J. D. Nichols. 2004. Occupancy estimation
and modeling for rare and elusive species. In: W. L. Thompson (ed.) Sampling rare or
elusive species: concepts, designs, and techniques for estimating population
parameters. 149-172. Island Press, Washington, D.C.
Maiden, N. R. 2009. Firearms: scene investigation. In: A. Jamieson and A. A. Moenssens
(eds.) Wiley encyclopedia of forensic science. John Wiley & Sons, Ltd, New York.
Manaaki Whenua - Landcare Research. 1997. Soil map of Christchurch City. Manaaki
Whenua - Landcare Research, Lincoln, N.Z.
Manjunath, B. C., B. R. Chandrashekar, M. Mahesh and R. M. Vatchala Rani. 2011. DNA
Profiling and forensic dentistry — A review of the recent concepts and trends. Journal
of Forensic and Legal Medicine 18:191-197.
Martin-de las Heras, S., A. Valenzuela, A. J. Valverde, J. C. Torres and J. D. Luna-delCastillo. 2007. Effectiveness of comparison overlays generated with DentalPrint©
software in bite mark analysis. Journal of Forensic Sciences 52:151-156.
Martin-de las Heras, S. and D. Tafur. 2009. Comparison of simulated human dermal
bitemarks possessing three-dimensional attributes to suspected biters using a
proprietary three-dimensional comparison. Forensic Science International 190:33-37.
Martin-de las Heras, S. and D. Tafur. 2011. Validity of a dichotomous expert response in
bitemark analysis using 3-D technology. Science & Justice 51:24-27.
Masson, J. J. 1997. Confidence level variations in firearms identifications through
computerized technology. Association of Firearm and Tool Mark Examiners Journal
29:42-44.
Matty, W. 1985. A comparison of three individual barrels produced from one button rifled
barrel blank. Association of Firearm and Tool Mark Examiners Journal 17:64-69.
184
McGlinchy, A. and B. Warburton. 2000. Review of the potential of wax-blocks and Feratox®
cyanide pellets as an alternative to traps for monitoring possum populations. Landcare
Research Contract Report LC9899/93, (unpubl.), Lincoln, N.Z. 12 p.
McGregor, P. K. and T. Peake. 1998. The role of individual identification in conservation
biology. In: T. M. Caro (ed.) Behavioral ecology and conservation biology. 31-55.
Oxford University Press, New York.
McKenna, C. J., M. L. Haron, K. A. Brown and A. J. Jones. 2000. Bitemarks in chocolate: a
case report. The Journal of Forensic Odonto-Stomatology 18:10-14.
Meekan, M. G., C. J. A. Bradshaw, M. Press, C. McLean, A. Richards, S. Quasnichka and J.
G. Taylor. 2006. Population size and structure of whale sharks Rhincodon typus at
Ningaloo Reef, Western Australia. Marine Ecology-Progress Series 319:275-285.
Meijerman, L., A. Thean and G. J. R. Maat. 2009. Earprints: interpretation of. In: A. Jamieson
and A. A. Moenssens (eds.) Wiley encyclopedia of forensic science. John Wiley &
Sons, Ltd, New York.
Meuwly, D. 2009. Automated fingerprint identification system. In: A. Jamieson and A. A.
Moenssens (eds.) Wiley encyclopedia of forensic science. John Wiley & Sons, Ltd,
New York.
Miller, J. and M. McLean. 1998. Criteria for identification of toolmarks. Association of
Firearm and Tool Mark Examiners Journal 30:15-61.
Miller, J. 2001. An examination of the application of the conservative criteria for
identification of striated toolmarks using bullets fired from ten consecutively rifled
barrels. Association of Firearm and Tool Mark Examiners Journal 33:125-132.
Miller, R. G., P. J. Bush, R. B. J. Dorion and M. A. Bush. 2009. Uniqueness of the dentition
as impressed in human skin: a cadaver model. Journal of Forensic Sciences 54:909914.
Montague, T. L. and B. Warburton. 2000. Non-toxic techniques for possum control. In: T. L.
Montague (ed.) The brushtail possum: biology, impact and management of an
introduced marsupial. 164-174. Manaaki Whenua Press, Lincoln, N.Z.
Moors, P. J. 1990. Norway rat. In: C. M. King (ed.) The handbook of New Zealand mammals.
192-206. Oxford University Press, Auckland, N.Z.
Morgan, D. and G. Hickling. 2000. Techniques used for poisoning possums. In: T. L.
Montague (ed.) The brushtail possum: biology, impact and management of an
introduced marsupial. 143-153. Manaaki Whenua Press, Lincoln, N.Z.
Morgano, S. M., P. Milot, P. Ducharme and L. Rose. 1995. Ability of various impression
materials to produce duplicate dies from successive impressions. The Journal of
Prosthetic Dentistry 73:333-340.
Moses, K. R. 2011. Automated Fingerprint Identification System (AFIS). In: Scientific
Working Group on Friction Ridge Analysis Study and Technology and National
institute of Justice (eds.) SWGFAST - The fingerprint sourcebook. 1-33. Scientific
Working Group on Friction Ridge Analysis, Study and Technology, Washington, D.C.
Moss, C. J. 1983. Oestrous behaviour and female choice in the African elephant. Behaviour
86:167-196.
185
Moss, C. J. 1996. Getting to know a population. In: K. Kangwana (ed.) Studying elephants.
58-74. African Wildlife Foundation, Nairobi.
Mowat, G. and D. Paetkau. 2002. Estimating marten Martes americana population size using
hair capture and genetic tagging. Wildlife Biology 8:201-209.
Murdock, J. E. 1981. A general discussion of gun barrel individuality and an empirical
assessment of the individuality of consecutively button rifled .22 caliber rifle barrels.
Association of Firearm and Tool Mark Examiners Journal 13:84-95.
Murie, O. J. 1954. A field guide to animal tracks. Houghton Mifflin, Boston. 374 p.
Murmann, D. C., P. C. Brumit, B. A. Schrader and D. R. Senn. 2006. A comparison of animal
jaws and bite mark patterns. Journal of Forensic Sciences 51:846-860.
Murphy, E. C. and C. R. Pickard. 1990. House mouse. In: C. M. King (ed.) The handbook of
New Zealand mammals. 225-242. Oxford University Press, Auckland, N.Z.
Murphy, E. C. and J. E. Dowding. 1995. Ecology of the stoat in Nothofagus forest: Home
range, habitat use and diet at different stages of the beech mast cycle. New Zealand
Journal of Ecology 19:97-109.
Nambiar, P., T. E. Bridges and K. A. Brown. 1995a. Quantitative forensic evaluation of bite
marks with the aid of a shape analysis computer program: Part 1; The development of
"SCIP" and the similarity index. The Journal of Forensic Odonto-Stomatology 13:1825.
Nambiar, P., T. E. Bridges and K. A. Brown. 1995b. Quantitative forensic evaluation of bite
marks with the aid of a shape analysis computer program: Part 2; "SCIP" and bite
marks in skin and foodstuffs. The Journal of Forensic Odonto-Stomatology 13:26-32.
Naru, A. S. and E. Dykes. 1996. The use of a digital imaging technique to aid bite mark
analysis. Science & Justice 36:47-50.
Naru, A. S. and E. Dykes. 1997. Digital image cross-correlation technique for bite mark
investigations. Science & Justice 37:251-258.
National Possum Control Agencies. 2004. Protocol for possum population monitoring using
the trap-catch method, 3rd version edition. National Possum Control Agencies,
Wellington, N.Z. 30 p.
National Possum Control Agencies. 2005. Protocol for low-density possum monitoring using
the WaxTag® method. National Possum Control Agencies, Wellington, N.Z. 22 p.
National Possum Control Agencies. 2008a. Protocol for low-density possum monitoring using
the WaxTag® method. National Possum Control Agencies, Wellington, N.Z. 21 p.
National Possum Control Agencies. 2008b. Protocol for possum population monitoring using
the trap-catch method. National Possum Control Agencies, Wellington, N.Z. 34 p.
Nelson, K. J., N. Bowman-Fowler, S. R. Berkowitz and E. F. Loftus. 2009. Eyewitness
testimony. In: A. Jamieson and A. A. Moenssens (eds.) Wiley encyclopedia of
forensic science. John Wiley & Sons, Ltd, New York.
Neufeld, P. J. and B. C. Scheck. 2010. Making forensic science more scientific. Nature
464:351-351.
186
Norton-Griffiths, M. 1975. Counting animals. Handbook - Serengeti Ecological Monitoring
Programme No. 1, African Wildlife Leadership Foundation, Nairobi. 110 p.
Norton, D. 2000. Benefits of possum control for native vegetation. In: T. L. Montague (ed.)
The brushtail possum: biology, impact and management of an introduced marsupial.
232-240. Manaaki Whenua Press, Lincoln, N.Z.
Nouvellet, P., C. D. Buesching, H. L. Dugdale, C. Newman and D. W. Macdonald. 2011.
Mouthing off about developmental stress: individuality of palate marking in the
European badger and its relationship with juvenile parasitoses. Journal of Zoology
283:52-62.
Nugent, G., P. Sweetapple, J. Coleman and P. Suisted. 2000. Possum feeding patterns: dietary
tactics of a reluctant folivore. In: T. L. Montague (ed.) The brushtail possum: biology,
impact and management of an introduced marsupial. 10-19. Manaaki Wheuna Press,
Lincoln, N.Z.
Nyongesa, H. O., S. Al-Khayatt, S. M. Mohamed and M. Mahmoud. 2004. Fast robust
fingerprint feature extraction and classification. Journal of Intelligent & Robotic
Systems 40:103-112.
Ogilvie, S. C., M. D. Thomas, H. Fitzgerald and D. R. Morgan. 1996. Sodium
monofluoroacetate (1080) bait-shyness in a wild brushtail possum (Trichosurus
vulpecula) population. Proceedings of the New Zealand Plant Protection Conference
49:143-146.
Ogilvie, S. C., M. D. Thomas, G. A. Morriss, D. R. Morgan and C. T. Eason. 2000.
Investigation of sodium monofluoroacetate (1080) bait shyness in wild brushtail
possum (Trichosurus vulpecula) populations. International Journal of Pest
Management 46:77-80.
Ogilvie, S. C., A. M. Paterson, J. G. Ross and M. D. Thomas. 2006a. Improving techniques
for the Waxtag® possum (Trichosurus vulpecula) monitoring index. New Zealand
Plant Protection 59:28-33.
Ogilvie, S. C., K. Sakata, M. D. Thomas and F. Maddigan. 2006b. Novel visual lures for the
management of brushtail possums. In: R. M. Timm and J. M. O'Brien (eds.)
Proceedings of the Twenty-Second Vertebrate Pest Conference, (VPC-2006). March
6-9, 2006. Berkeley CA, arranged by the Vertebrate Pest Conference. 479-482.
University of California, Davis, Davis.
Oltramari-Navarro, P. V. P., G. Janson, R. B. S. de Oliveira, C. L. Quaglio, J. F. Castanha
Henriques, S. H. de Carvalho Sales-Peres and J. A. McNamara, Jr. 2010. Tooth-wear
patterns in adolescents with normal occlusion and Class II Division 2 malocclusion.
American Journal of Orthodontics and Dentofacial Orthopedics 137:730.e731730.e735.
Otis, D. L., K. P. Burnham, G. C. White and D. R. Anderson. 1978. Statistical inference from
capture data on closed animal populations. Wildlife Monographs No. 62, The Wildlife
Society, Washington, D.C. 135 p.
Page, M., J. Taylor and M. R. B. Blenkin. 2011. Uniqueness in the forensic identification
sciences — Fact or fiction? Forensic Science International 206:12-18.
187
Palma, A. R. T. and R. Gurgel-Gonçalves. 2007. Morphometric identification of small
mammal footprints from ink tracking tunnels in the Brazilian Cerrado. Revista
Brasileira De Zoologia 24:333-343.
Payton, I. 2000. Damage to native forests. In: T. L. Montague (ed.) The brushtail possum:
biology, impact and management of an introduced marsupial. 111-125. Manaaki
Whenua Press, Lincoln, N.Z.
Perry, J. N., A. M. Liebhold, M. S. Rosenberg, J. Dungan, M. Miriti, A. Jakomulska and S.
Citron-Pousty. 2002. Illustrations and guidelines for selecting statistical methods for
quantifying spatial pattern in ecological data. Ecography 25:578-600.
Peterson, T. and E. F. Loftus. 2009. Memory: reconstructive. In: A. Jamieson and A. A.
Moenssens (eds.) Wiley encyclopedia of forensic science. John Wiley & Sons, Ltd,
New York.
Preatoni, D. G., M. Nodari, R. Chirichella, G. Tosi, L. A. Wauters and A. Martinoli. 2005.
Identifying bats from time-expanded recordings of search calls: comparing
classification methods. Journal of Wildlife Management 69:1601-1614.
Pretty, I. A. and D. Sweet. 2001a. The scientific basis for human bitemark analyses - a critical
review. Science & Justice 41:85-92.
Pretty, I. A. and D. Sweet. 2001b. Digital bite mark overlays - an analysis of effectiveness.
Journal of Forensic Sciences 46:1385-1391.
Pretty, I. A. and M. D. Turnbull. 2001. Lack of dental uniqueness between two bite mark
suspects. Journal of Forensic Sciences 46:1487-1491.
Pretty, I. A., R. J. Pretty, B. R. Rothwell and D. Sweet. 2003. The reliability of digitized
radiographs for dental identification: a web-based study. Journal of Forensic Sciences
48:1325-1330.
Pretty, I. A. 2005a. Unresolved issues in bitemark analysis. In: R. B. J. Dorion (ed.) Bitemark
evidence. 547-563. Marcel Dekker, New York.
Pretty, I. A. 2005b. Reliability of bitemark evidence. In: R. B. J. Dorion (ed.) Bitemark
evidence. 531-545. Marcel Dekker, New York.
Pretty, I. A. and D. Sweet. 2010. A paradigm shift in the analysis of bitemarks. Forensic
Science International 201:38-44.
Rahimi, M., N. C. K. Heng, J. A. Kieser and G. R. Tompkins. 2005. Genotypic comparison of
bacteria recovered from human bite marks and teeth using arbitrarily primed PCR.
Journal of Applied Microbiology 99:1265-1270.
Rangel, T. F., J. A. F. Diniz-Filho and L. M. Bini. 2010. SAM: a comprehensive application
for spatial analysis in macroecology. Ecography 33:46-50.
Rangel, T. F. L. V. B., J. A. F. Diniz-Filho and L. M. Bini. 2006. Towards an integrated
computational tool for spatial analysis in macroecology and biogeography. Global
Ecology and Biogeography 15:321-327.
Ratz, H. 1997. Identification of footprints of some small mammals. Mammalia 61:431-441.
Ratz, H., H. Moller and D. Fletcher. 1999. Predator identification from bite marks on penguin
and albatross chicks. Marine Ornithology 27:149-156.
188
Rawson, R. D., R. K. Ommen, G. Kinard, J. Johnson and A. Yfantis. 1984. Statistical
evidence for the individuality of the human dentition. Journal of Forensic Sciences
29:245-253.
Rawson, R. D., G. L. Vale, E. E. Herschaft, N. D. Sperber and S. Dowell. 1986. Analysis of
photographic distortion in bite marks: a report of the Bite Mark Guidelines
Committee. Journal of Forensic Sciences 31:1261-1268.
Reed, J. E., R. J. Baker, W. B. Ballard and B. T. Kelly. 2004. Differentiating Mexican gray
wolf and coyote seats using DNA analysis. Wildlife Society Bulletin 32:685-692.
Reid, B. 1983. Kiwis and possums, traps and baits. Fur Facts 4:17-26.
Reisser, J., M. Proietti, P. Kinas and I. Sazima. 2008. Photographic identification of sea
turtles: method description and validation, with an estimation of tag loss. Endangered
Species Research 5:73-82.
van Rijsbergen, C. J. 1979. Information retrieval, 2nd edition. Butterworths, London. 208 p.
Roberge, D. and A. Beauchamp. 2006. The use of BulletTRAX-3DTM in a a study of
consecutively manufactured barrels. Association of Firearm and Tool Mark Examiners
Journal 38:166-172.
Robertson, H. A. and B. D. Heather. 1999. The hand guide to the birds of New Zealalnd.
Penguin Books, Auckland.
Rollins, C. E. and D. E. Spencer. 1995. A fatality and the American mountain lion: bite mark
analysis and profile of the offending lion. Journal of Forensic Sciences 40:486-489.
Rosenberg, M. S. and C. D. Anderson. 2011. PASSaGE: Pattern Analysis, Spatial Statistics
and Geographic Exegesis. Version 2. Methods in Ecology and Evolution 2:229-232.
Roughton, R. D. and M. W. Sweeny. 1982. Refinements in scent-station methodology for
assessing trends in carnivore populations. Journal of Wildlife Management 46:217229.
Ruifrok, A. C. C. 2009. Facial comparison. In: A. Jamieson and A. A. Moenssens (eds.)
Wiley encyclopedia of forensic science. John Wiley & Sons, Ltd, New York.
Saks, M. J. 1994. Implications of the Daubert test for forensic identification science.
Shepard's Expert and Scientific Evidence Quarterly 1:427-434.
Saks, M. J., D. M. Risinger, R. Rosenthal and W. C. Thompson. 2003. Context effects in
forensic science: a review and application of the science of science to crime laboratory
practice in the United States. Science & Justice 43:77-90.
Saks, M. J. and J. J. Koehler. 2005. The coming paradigm shift in forensic identification
science. Science 309:892-895.
Saks, M. J. and J. J. Koehler. 2008. The individualization fallacy in forensic science evidence.
Vanderblit Law Review 61:199-219.
Salmon, J. T. 1968. Field guide to the alpine plants of New Zealand. Reed, Wellington. 326 p.
Santoro, V., G. Smaldone, P. Lozito, M. Smaldone and F. Introna. 2011. A forensic approach
to fatal dog attacks. A case study and review of the literature. Forensic Science
International 206:e37-e42.
189
Sargeant, G. A., D. H. Johnson and W. E. Berg. 1998. Interpreting carnivore scent-station
surveys. Journal of Wildlife Management 62:1235-1245.
Sargeant, G. A., D. H. Johnson and W. E. Berg. 2003. Sampling designs for carnivore scentstation surveys. Journal of Wildlife Management 67:289-298.
Schroeder, M. A. and L. A. Robb. 2005. Criteria for gender and age. In: C. E. Braun (ed.)
Techniques for wildlife investigations and management. 303-338. The Wildlife
Society, Bethesda.
Seber, G. A. F. 1982. The estimation of animal abundance and related parameters, 2nd
edition. Charles Griffin, London. 654 p.
Senin, N., R. Groppetti, L. Garofano, P. Fratini and M. Pierni. 2006. Three-dimensional
surface topography acquisition and analysis for firearm identification. Journal of
Forensic Sciences 51:282-295.
Sharma, S., Y. Jhala and V. B. Sawarkar. 2005. Identification of individual tigers (Panthera
tigris) from their pugmarks. Journal of Zoology 267:9-18.
Siemers, B. M. and G. Kerth. 2006. Do echolocation calls of wild colony-living Bechstein's
bats (Myotis bechsteinii) provide individual-specific signatures? Behavioral Ecology
and Sociobiology 59:443-454.
da Silva, R. H. A., J. de Oliveira Musse, R. F. H. Melani and R. N. Oliveira. 2006. Human
bite mark identification and DNA technology in forensic dentistry. Brazilian Journal
of Oral Sciences 5:1193-1197.
Silvy, N. J., R. R. Lopez and M. J. Peterson. 2005. Wildlife marking techniques. In: C. E.
Braun (ed.) Techniques for wildlife investigations and management. 339-376. The
Wildlife Society, Bethesda.
Simon, A., H. Jordan and K. Pforte. 1974. Successful identification of a bitemark in a
sandwich. International Journal of Forensic Dentistry 2:17-21.
Simsik, D., V. Porada, A. Galajdova and Z. Dolna. 2010. Experimental gait parameters
analysis for identification purposes in criminalistics. Science & Justice 50:37-37.
Singh, R. P. and H. R. Aggarwal. 1984. Identification of wires and the cutting tool by
scanning electron microscope. Forensic Science International 26:115-121.
Skolrood, R. W. 1975. Comparison of bullets fired from consecutively rifled Cooey .22
calibre barrels. Journal of the Canadian Society of Forensic Science 8:49-52.
Smallwood, K. S. and E. L. Fitzhugh. 1993. A rigorous technique for identifying individual
mountain lions Felis concolor by their tracks. Biological Conservation 65:51-59.
Sognnaes, R. F., R. D. Rawson, B. M. Gratt and N. B. Nguyen. 1982. Computer comparison
of bitemark patterns in identical twins. Journal of the American Dental Association
105:449-451.
Solheim, T. and T. I. Leidal. 1975. Scanning electron microscopy in the investigation of bite
marks in foodstuffs. Forensic Science 6:205-215.
Solheim, T. 1993. A new method for dental age estimation in adults. Forensic Science
International 59:137-147.
190
Souviron, R. R. 2005. Animal bites. In: R. B. J. Dorion (ed.) Bitemark evidence. 275-291.
Marcel Dekker, New York.
Speed, C. W., M. G. Meekan and C. J. A. Bradshaw. 2007. Spot the match - wildlife photoidentification using information theory. Frontiers in Zoology 4:(11).
Speed, C. W., M. G. Meekan, D. Rowat, S. J. Pierce, A. D. Marshall and C. J. A. Bradshaw.
2008. Scarring patterns and relative mortality rates of Indian Ocean whale sharks.
Journal of Fish Biology 72:1488-1503.
Spurr, E. B. 1994. Bait interference for measuring relative possum abundance. Landcare
Research Contract Report LC9495/018, (unpubl.), Lincoln, N.Z. 16 p.
Spurr, E. B. 1995. Evaluation of non-toxic bait interference for indexing brushtail possum
density. New Zealand Journal of Ecology 19:123-130.
Steadman, D. W., B. J. Adams and L. W. Konigsberg. 2006. Statistical basis for positive
identification in forensic anthropology. American Journal of Physical Anthropology
131:15-26.
Steinmetz, R. and D. L. Garshelis. 2008. Distinguishing Asiatic black bears and sun bears by
claw marks on climbed trees. Journal of Wildlife Management 72:814-821.
Stoddart, T. J. 1973. Bite marks in perishable substances. A method of producing accurate
permanent models. British Dental Journal 135:285-287.
Stols, G. and H. Bernitz. 2010. Reconstruction of deformed bite marks using affine
transformations. Journal of Forensic Sciences 55:784-787.
Stoney, D. A. 2001. Measurement of fingerprint individuality. In: H. C. Lee and R. E.
Gaensslen (eds.) Advances in fingerprint technology. 327-387. CRC Press.
Ström, F. 1963. Investigation of bite-marks. Journal of Dental Research 42:312-316.
Stroud, R. K. and W. J. Adrian. 1996. Forensic investigational techniques for wildlife law
enforcement investigations. In: A. Fairbrother, L. N. Locke and G. L. Hoff (eds.)
Noninfectious diseases of wildlife. 3-18. Iowa State University Press, Ames.
Stroud, R. K. 1998. Wildlife forensics and the veterinary practitioner. Seminars in Avian and
Exotic Pet Medicine 7:182-192.
Su, C. and S. N. Srihari. 2009. Probability of random correspondence for fingerprints. In: Z. J.
M. H. Geradts, K. Y. Franke and C. J. Veenman (eds.) Computational forensics. 5566. Springer Berlin
Sweet, D. and D. Hildebrand. 1999. Saliva from cheese bite yields DNA profile of burglar: a
case report. International Journal of Legal Medicine 112:201-203.
Sweet, D. J. 1997. Human bitemarks: examination, recovery and analysis. In: C. M. Bowers
and G. L. Bell (eds.) Manual of forensic odontology. 148-169. American Society of
Forensic Odontology, Saratoga Springs, N.Y.
Sweet, D. J. and C. M. Bowers. 1998. Accuracy of bite mark overlays: a comparison of five
common methods to produce exemplars from a suspect's dentition. Journal of
Forensic Sciences 43:362-367.
Sweet, D. J. and G. G. Shutler. 1999. Analysis of salivary DNA evidence from a bite mark on
a body submerged in water. Journal of Forensic Sciences 44:1069-1072.
191
Sweetapple, P. J. and G. Nugent. 2011. Chew-track-cards: a multiple-species small mammal
detection device. New Zealand Journal of Ecology 35:153-162.
Taylor, D. V. 1963. The law and the dentist. British Dental Journal 114:389-389.
Thali, M. J., M. Braun, T. H. Markwalder, W. Brueschweiler, U. Zollinger, N. J. Malik, K.
Yen and R. Dirnhofer. 2003. Bite mark documentation and analysis: the forensic
3D/CAD supported photogrammetry approach. Forensic Science International
135:115-121.
Thomas, M. D. and G. J. Hickling. 1990. A trial to determine an efficient method of
controlling possums around bait feeder stations. Forest Research Institute Contract
Report FWE 90/62, (unpubl.), Christchurch, N.Z. 7 p.
Thomas, M. D. and D. Meenken. 1995. Comparison of ground-laid 1080 paste baits and 1080
cereal baits in bait stations for possum control on farmland (40/90). Landcare
Research Contract Report LC9596/36, (unpubl.), Lincoln, N.Z. 12 p.
Thomas, M. D. 1998. Feasibility of using wax-blocks to measure rodent and possum
abundance and changes in population size. Landcare Research Contract Report
LC9798/109, (unpubl.), Lincoln, N.Z. 10 p.
Thomas, M. D., D. A. Wright, J. Mason and K. W. Briden. 1998. Use of bait stations for
possum and feral cat control. Science for Conservation 86, Department of
Conservation, Wellington, N.Z. 24 p.
Thomas, M. D., J. A. Brown and R. J. Henderson. 1999. Feasibility of using wax blocks to
measure rat and possum abundance in native forest. Proceedings of the New Zealand
Plant Protection Conference 52:125-129.
Thomas, M. D. and J. A. Brown. 2001. Calibration of possum density estimates from raised
leg-hold traps. New Zealand Plant Protection 54:71-75.
Thomas, M. D., J. A. Brown, F. W. Maddigan and L. A. Sessions. 2002. Comparison of trapcatch, orange and WaxTag methods for estimating possum densities in Northland.
Pest Control Research Contract Report 2002/7, (unpubl.), Christchurch, N.Z.
Thomas, M. D., J. A. Brown, F. W. Maddigan and L. A. Sessions. 2003. Comparison of trapcatch and bait interference for estimating possum densities. New Zealand Plant
Protection 56:81-85.
Thomas, M. D., J. A. Brown and F. W. Maddigan. 2004. Possum monitoring in the presence
of ground birds: (objective 1) calibrating TSI with RTCI. Pest Control Research
Contract Report 2004/6, (unpubl.), Christchurch, N.Z. 16 p.
Thomas, M. D. and F. W. Maddigan. 2004. Visual lures for possums. Pest Control Research
Contract Report 2004/5, (unpubl.), Christchurch, N.Z. 15 p.
Thomas, M. D., D. R. Morgan and F. W. Maddigan. 2007. Accuracy of possum monitoring
using WaxTags. Pest Control Research Contract Report 2006/7, (unpubl.),
Christchurch, N.Z. 50 p.
Tompkins, G. R. 2005. Genotypic comparison of oral bacteria isolated from bitemarks and
teeth. In: R. B. J. Dorion (ed.) Bitemark evidence. 593-599. Marcel Dekker, New
York.
192
Tonn, E. M., D. J. Long and D. Dechant. 2004. A comparison of intercanine widths among
various animals and humans. In: Proceedings of the 56th Annual Scientific Meeting of
the American Academy of Forensic Sciences; Feb 16–21, 2004. Dallas, TX. 206-206.
Tonn, E. M., J. P. Dumbacher and D. Dechant. 2005. Methods to identify various mammalian
bite marks. In: Proceedings of the 57th Annual Scientific Meeting of the American
Academy of Forensic Sciences; Feb 21–26, 2005; New Orleans, LA. 224-224.
Travaini, A., R. Laffitte and M. Delibes. 1996. Determining the relative abundance of
European red foxes by scent-station methodology. Wildlife Society Bulletin 24:500504.
Tsujita, C. J. and G. E. G. Westermann. 2001. Were limpets or mosasaurs responsible for the
perforations in the ammonite Placenticeras? Palaeogeography, Palaeoclimatology,
Palaeoecology 169:245-270.
Uchiyama, T. 1986. Similarity among breech face marks fired from guns with close serial
numbers. Association of Firearm and Tool Mark Examiners Journal 18:15-52.
Uchiyama, T. 1988. A criterion for land mark identification. Association of Firearm and Tool
Mark Examiners Journal 20:236-251.
Uchiyama, T. 1992. The probability of corresponding striae in toolmarks. Association of
Firearm and Tool Mark Examiners Journal 24:273-290.
Vale, G. L. 2005. History of bitemark evidence. In: R. B. J. Dorion (ed.) Bitemark evidence.
1-29. Marcel Dekker, New York.
Van Dyke, F. G., R. H. Brocke and H. G. Shaw. 1986. Use of road track counts as indices of
mountain lion presence. Journal of Wildlife Management 50:102-109.
Vanderkolk, J. R. 1995. Identifying consecutively made garbage bags through manufactured
characteristics. Journal of Forensic Identification 45:38-50.
Verma, S. K., K. Prasad, N. Nagesh, M. Sultana and L. Singh. 2003. Was elusive carnivore a
panther? DNA typing of faeces reveals the mystery. Forensic Science International
137:16-20.
Verma, S. K. and L. Singh. 2003. Novel universal primers establish identity of an enormous
number of animal species for forensic application. Molecular Ecology Notes 3:28-31.
Verner, J. 1985. Assessment of counting techniques. Current Ornithology 2:247-302.
Waits, L. P. and D. Paetkau. 2005. Noninvasive genetic sampling tools for wildlife biologists:
A review of applications and recommendations for accurate data collection. Journal of
Wildlife Management 69:1419-1433.
Wan, Q.-H., S.-G. Fang, G.-F. Chen, Z.-M. Wang, P. Ding, M.-Y. Zhu, K.-S. Chen, J.-H. Yu
and Y.-P. Zhao. 2003. Use of oligonucleotide fingerprinting and faecal DNA in
identifying the distribution of the Chinese tiger (Panthera tigris amoyensis
Hilzheimer). Biodiversity and Conservation 12:1641-1648.
Warburton, B. 2000. Monitoring possum populations. In: T. L. Montague (ed.) The brushtail
possum: biology, impact and management of an introduced marsupial. 132-138.
Manaaki Whenua Press, Lincoln, N.Z.
193
Warburton, B. and I. Yockney. 2000. Comparison of two luring methods for traps use to
monitor possum population in tussock country. Landcare Research Contract Report
LC9900/126, (unpubl.), Lincoln, N.Z. 9 p.
Warburton, B., R. Barker and M. Coleman. 2004. Evaluation of two relative-abundance
indices to monitor brushtail possums in New Zealand. Wildlife Research 31:397-401.
Warrick, G. D. and C. E. Harris. 2001. Evaluation of spotlight and scent-station surveys to
monitor kit fox abundance. Wildlife Society Bulletin 29:827-832.
Webster, G. and D. G. MacDonald. 1979. Forensic dental analysis of experimental bite marks
in food. Journal of Dental Research 58:1273-1273.
Webster, G. 1982. A suggested classification of bite marks in foodstuffs in forensic dental
analysis. Forensic Science International 20:45-52.
Western, D., C. J. Moss and N. Georgiadis. 1983. Age estimation and population age structure
of elephants from footprint dimensions. Journal of Wildlife Management 47:11921197.
Wetton, J. H., C. S. F. Tsang, C. A. Roney and A. C. Spriggs. 2004. An extremely sensitive
species-specific ARMs PCR test for the presence of tiger bone DNA. Forensic Science
International 140:139-145.
Whittaker, D. K., K. Watkins and J. Wiltshire. 1975. An experimental assessment of the
reliability of bite mark analysis. International Journal of Forensic Dentistry 3:3-7.
Whittaker, D. K. 1995. Forensic dentistry in the identification of victims and assailants.
Journal of Clinical Forensic Medicine 2:145-151.
Whittaker, D. K., L. Evans and M. R. Brickley. 1997. ROC analysis of the ability to
distinguish bite marks made by adults and children. Journal of Dental Research
76:1064-1064.
Williams, C. L., K. Blejwas, J. J. Johnston and M. M. Jaeger. 2003. A coyote in sheep's
clothing: predator identification from saliva. Wildlife Society Bulletin 31:926-932.
Wilson, G. J. and R. J. Delahay. 2001. A review of methods to estimate the abundance of
terrestrial carnivores using field signs and observation. Wildlife Research 28:151-164.
Wolff, D., R. Kennedy, E. Lounsbury, G. MacMaster, K. Olthuis, D. Sweet, A. Tutt and B.
Yamashita. 1998. Dental casting materials for the recovery of toolmark impressions.
Forensic Identification Research Services Bulletin No. 43, Royal Canadian Mounted
Police, Ottawa.
Wood, J. D., C. E. O'Connell-Rodwell and S. L. Klemperer. 2005. Using seismic sensors to
detect elephants and other large mammals: a potential census technique. Journal of
Applied Ecology 42:587-594.
Woolley, A. W., C. F. Chabris, A. Pentland, N. Hashmi and T. W. Malone. 2010. Evidence
for a collective intelligence factor in the performance of human groups. Science
330:686-688.
Würsig, B. and T. A. Jefferson. 1990. Methods of photo-identification for small cetaceans. In:
P. S. Hammond, S. A. Mizroch and G. P. Donovan (eds.) Reports of the International
Whaling Commission Special issue 12: Individual recognition of cetaceans: use of
photo-identification and other techniques to estimation population parameters. 43-52.
International Whaling Commission.
194
Xiaohu, X., H. P. Philipsen, N. G. Jablonski, K. M. Pang and Z. Jiazhen. 1992. Age
estimation from the structure of adult human teeth: review of the literature. Forensic
Science International 54:23-28.
Yoon, S., S.-S. Choi, S.-H. Cha, Y. Lee and C. C. Tappert. 2005. On the individuality of the
iris biometric. In: M. Kamel and A. Campilho (eds.) Image analysis and recognition.
1118-1124.
Youden, W. J. 1950. Index for rating diagnostic tests. Cancer 3:32-35.
Yuan, G., J. Russell, R. Klette, B. Rosenhahn and S. Stones-Havas. 2005. Understanding
tracks of different species of rats. Proceedings of the Image and Vision Computing
New Zealand Conference. University of Otago, Dunedin, N.Z.
Yun, J.-I., J.-Y. Lee, J.-W. Chung, H.-S. Kho and Y.-K. Kim. 2007. Age estimation of Korean
adults by occlusal tooth wear. Journal of Forensic Sciences 52:678-683.
Zheng, J., Z. R. Zhou, J. Zhang, H. Li and H. Y. Yu. 2003. On the friction and wear behaviour
of human tooth enamel and dentin. Wear 255:967-974.
Zheng, J. and Z. R. Zhou. 2006. Effect of age on the friction and wear behaviors of human
teeth. Tribology International 39:266-273.
Zielinski, W. J. and H. B. Stauffer. 1996. Monitoring Martes populations in California: survey
design and power analysis. Ecological Applications 6:1254-1267.
195
Appendix A
Wildlife abundance assessment methods
The relationship among population estimation approaches extracted from Lancia et al. (2005).
Figure A–1. The relationship among population estimation approaches: extracted from Lancia
et al. (2005).
196
Appendix B
Forensic odontology terminology (excerpts)
Selected terms relating to forensic odontology were extracted from the American Board of
Forensic Odontology bitemark methodology guidelines (ABFO 1997)
B.1 Bite mark vs. bitemark
The noun bite mark (two words) is used more frequently in the literature than bitemark (one
word) and was preferred 3: 1 in the ABFO survey. Dr. S. Miles Standish presented a cogent
rationale for the single term, bitemark, as the preferred grammatical form. A professor of
English at the University of Louisville concluded that, because language is a living thing,
either term is acceptable. Bitemark implies a type of mark whereas bite mark connotes an
entity unto itself and recognizable as such. Bitemark would be considered the more
progressive term, signifying that odontologists have a sufficient body of work and evolved in
similar fashion. Dr. Standish also adds that, when used as a compound adjective, bite mark is
hyphenated as in Bitemark analysis. After evaluating all these opinions, it is the feeling of the
ABFO that the meaning of the word in any of its forms is clear and there is no need for the
ABFO to endorse a particular form.
B.2 A characteristic (as it pertains to bitemarks)
A characteristic, as applied to a bitemark, is a distinguishing feature, trait or pattern within
the mark. Characteristics are two types, class characteristics and individual characteristics.
Class characteristic: a feature, trait or pattern preferentially seen in, or reflective of, a given
group. For example, the finding of linear or rectangular contusions at the midline of a
Bitemark arch is a class characteristic of human incisor teeth. "Incisors" represent the class in
this case. The value of identifying class characteristics is that, when seen, they enable us to
identify the group from which they originate. For instance, the class characteristics of incisors
(rectangles) differentiates them from canines (circles or triangles). If we define the class
characteristics of human bites, we can differentiate them from animal bites. Via class
characteristics, we differentiate the adult from the child bite or mandibular from maxillary
arch. The original term "class characteristic" was applied to toolmarks and its definition has
been modified to make it more applicable to bitemarks.
Individual characteristic: a feature, trait or pattern that represents an individual variation
rather than an expected finding within a defined group. An example of this is a rotated tooth.
The value of individual characteristics is that they differentiate between individuals and help
identify the perpetrator. The number, specificity and accurate reproduction of these individual
characteristics determine the confidence level that a particular suspect made the bitemark.
197
B.3 Bitemark definition
Bitemark
A physical alteration in a medium caused by the contact of teeth. A representative pattern left
in an object or tissue by the dental structures of an animal or human.
Cutaneous human bitemark
An injury in skin caused by contacting teeth (with or without the lips or tongue) which shows
the representational pattern of the oral structures.
198
Appendix C
Forensic toolmark examination glossary (excerpts)
Selected terms relating to the forensic examination of toolmarks were extracted from the
Glossary of the Association of Firearm and Tool Mark Examiners, 4th edition (AFTE
Training and Standardization Committee 2001).
C.1 Toolmark examination
Toolmark Identification
Toolmark identification is a discipline of forensic science which has as its primary concern to
determine if a toolmark was produced by a particular tool.
Tool
An object used to gain mechanical advantage. Also thought of as the harder of two objects
which when brought into contact with each other, results in the softer one being marked. See
TOOLMARK, IMPRESSED and TOOLMARK, STRIATED.
Striations
Contour variations, generally microscopic, on the surface of an object caused by a
combination of force and motion where the motion is approximately parallel to the plane
being marked. These marks can contain CLASS and/or INDIVIDUAL
CHARACTERISTICS.
Toolmark, Impressed
Marks produced when a tool is placed against another object and enough force is applied to
the tool so that it leaves an impression. The class characteristics (shape) can indicate the type
of tool used to produce the mark. These marks can contain CLASS and/or INDIVIDUAL
CHARACTERISTICS of the tool producing the marks. Also called COMPRESSION
MARKS.
Toolmark, Striated
Marks produced when a tool is placed against another object and with pressure applied, the
tool is moved across the object producing a striated mark. FRICTION MARKS, ABRASION
MARKS and SCRATCH MARKS are terms commonly used when referring to striated
marks. These marks can be either CLASS and/or INDIVIDUAL CHARACTERISTICS.
199
Class Characteristics
Measurable features of a specimen which indicate a restricted group source. They result from
design factors, and are therefore determined prior to manufacture.
Subclass Characteristics
Discernible surface features of an object which are more restrictive than CLASS
CHARACTERISTICS in that they are:
1. Produced incidental to manufacture.
2. Are significant in that they relate to a smaller group source (a subset of the class to
which they belong).
3. Can arise from a source which changes over time.
Examples would include: bunter marks, extrusion marks on pipe, etc.
Caution should be exercised in distinguishing subclass characteristics from INDIVIDUAL
CHARACTERISTICS.
Individual Characteristics
Marks produced by the random imperfections or irregularities of tool surfaces. These random
imperfections or irregularities are produced incidental to manufacture and/or caused by use,
corrosion, or damage. They are unique to that tool and distinguish it from all other tools.
Theory of Identification as it Relates to Toolmarks
The theory of identification as it pertains to the comparison of toolmarks enables opinions of
common origin to be made when the unique surface contours of two toolmarks are in
―sufficient agreement.‖
This ―sufficient agreement‖ is related to the significant duplication of random toolmarks as
evidenced by the correspondence of a pattern or combination of patterns of surface contours.
Significance is determined by the comparative examination of two or more sets of surface
contour patterns comprised of individual peaks, ridges and furrows. Specifically, the relative
height or depth, width, curvature and spatial relationship of the individual peaks, ridges and
furrows within one set of surface contours are defined and compared to the corresponding
features in the second set of surface contours. Agreement is significant when it exceeds the
best agreement demonstrated between toolmarks known to have been produced by different
tools and is consistent with agreement demonstrated by toolmarks known to have been
produced by the same tool. The statement that ―sufficient agreement‖ exists between two
toolmarks means that the agreement is of a quantity and quality that the likelihood another
tool could have made the mark is so remote as to be considered a practical impossibility.
Currently the interpretation of individualization/identification is subjective in nature, founded
on scientific principles and based on the examiner‘s training and experience.
Contour Variation
Variations in the elevations of the ridges and valleys in striated marks and of forms and
shapes or depressions in the impression mark.
200
C.2 Firearm terminology
Breech
The part of a firearm at the rear of the bore into which the cartridge or propellant is inserted.
Barrel
That part of a firearm through which a projectile or shot charge travels under the impetus of
powder gasses, compressed air, or other like means. May be rifled or smooth.
Bore
The interior of a barrel forward of the chamber.
Bore Diameter
1. Rifled barrels: the minor diameter of a barrel which is the diameter of a circle formed
by the tops of the lands.
2. Shotguns: the interior dimensions of the barrel forward of the chamber but behind the
choke.
Breech Face
That part of the breechblock or breech bolt which is against the head of the cartridge case or
shotshell during firing.
Cartridge
A single unit of ammunition consisting of the case, primer, and propellant with one or more
projectile(s). Also applies to a shotshell.
Cartridge Case
The container for all the other components which comprise a cartridge.
Chamber
The rear part of the barrel bore that has been formed to accept a specific cartridge. Revolver
cylinders are multi-chambered.
C.3 Firearm anatomy and sources of toolmarks
Core firearm components from the viewpoint of striation- and impression-type toolmarks
sources are illustrated in Figure C–1. Figure C–1(a): longitudinal section of the barrel and
breech block in a loaded firearm at battery; Figure C–1(b): numbered firing sequence as given
in detail below; close-up illustrations of (c) striation-type toolmarks, and of (d) impressiontype toolmarks: Transverse sections of (e) bullet and (d) barrel.
201
The firing sequence illustrated in Figure C–1(b) proceeds as follows: (1) force strikes the
firing pin, which (2) hits the primer transferring the physical impact into a thermal flare
causing propellant (gunpowder) to ignite (3) resulting in high pressure from a rapid expansion
of gasses, which (4) propels the bullet through the barrel and out of the muzzle, and
simultaneously forces the cartridge case onto the chamber wall and breech face. During the
bullet‘s travel through the barrel, its contact with twisted grooves and lands (rifling) forces it
to rotate longitudinally to obtain flight stabilisation.
Because the bullet diameter is larger than the bore diameter, its surface is forcibly made to
contact the microscopic rifling imperfections created during the barrel‘s production process.
The imperfections on the steel barrel are transferred on to the softer lead bullet, or in the case
of high-power cartridges, copper alloy jacket material.
Simultaneous to the firing process which drives the bullet forward, rapidly expanding gas
presses the brass cartridge case to the chamber wall and breech face (usually steel) with great
pressure. Therefore the harder material‘s imperfections that may have occurred during
manufacture process or defects that may arise from normal use of the firearm can be
transferred to the case in the form of impression toolmark. Chamber wall impressions,
however may also contain striation toolmarks due to the necessity to extract by pulling out the
empty case to make room for a fresh cartridge.
202
Rifling
(a)
Barrel
Breech (slide)
Propellant
Cartridge case
Bullet
Firing pin
2
3
1
4
(d) Impression toolmarks
(white dots)
(c) Striation toolmarks
(e)
(f)
Barrel
Breech
Primer cap
Gas pressure
Cartridge case
Bullet
(e)
(e)
Firing pin
(f)
Groove mark
Bullet diameter
(> bore diameter)
Groove
Land
(f)
Land mark
Bore
diameter
Core (lead)
Bore
diameter
Jacket
(copper alloy)
Groove width
Land width
Groove width
Land width
Figure C–1 Nomenclature of core firearm components, and sources of striation- and
impression-type toolmarks; Longitudinal sections of (a) loaded firearm at battery, (b)
numbered firing sequence, and close-up illustrations of (c) striation-type toolmarks, and
of (d) impression-type toolmarks; Transverse sections of (e) bullet and (d) barrel.
203
Appendix D
Site selection (random plots)
Compliant to the image capture scale (1000 m horizontal bar in screen) a 100 × 100 m grid
mesh was generated and overlaid to the forest image with numbered grid cells. Computer
generated random numbers defined sample grids where it fell on a suitable sample block.
Sample site selection was continued until 10 sample grids were identified in Bottle Lake
Forest. Paired sample plots of 110 × 110 m placed at 100 distance i.e. a whole sample set of
320 × 110 m were drawn into the map so as to fit in the following conditions: (1) sample set
reference point (north-west corner of first grid) is located in the abovementioned randomly
selected sample grid, (2) sample set of two plots fit within 10 m inside the suitable sample
block, and (3) sample plot orientation follows tree row orientation as seen in images. Latest
aerial photographs of the selected sample sites were obtained from Selwyn Plantation Board
Limited for final update and refinement of plot setting on the map.
204
Figure D–1. Bottle Lake Forest Reserve field sample site selection procedure, illustrating
randomly selected grid locations and sample plot designation for each candidate location.
205
Appendix E
Individuals identified on WaxTags® from each night
The following tables (Table E–1 to Table E–10) list WaxTag® identification numbers bitten
by each individual animal for each night. Each individual identification number is taken from
the smallest WaxTag® number from which the bitemark was first identified. Tag numbers in
bold represent those bitten by more than one individual on respective nights. WaxTags® in the
inconclusive category lacked sufficient observable characteristics to identify with any other
bitemark, or to classify as an independent individual, i.e., it was not possible to either
positively include in or exclude from previously identified individual.
206
207
0375
0375
1819
6815
2353
2353
3844
4661
6420
7768
8931
9133
0866
0866
1819
4167
4806
7008
7932
8865
1885
0139
2272
6990
Total
16
3
7
7
0
0
0
0
* Indiv. =Individual; ? =Inconclusive
®
WaxTag numbers in bold were bitten by multiple individuals.
Indiv.* 0077
Tag # 0077
0291
0862
1819
2353
2801
2895
3012
3443
4986
5969
6167
7877
8503
8577
8645
1
3443
3443
Table E–1. Individuals identified on tags from night one. (n=13)
0131
0131
1213
1729
1943
2333
3082
3799
4490
4564
4687
4877
5121
5133
5501
5773
6087
6266
6542
6574
6634
7046
7380
7661
7797
7835
8107
8355
8935
9015
9049
9548
31
207
12
0560
0560
2691
3658
4490
5637
6087
6584
6735
8538
9386
9761
9852
24
0719
0719
1283
2026
2177
2473
2491
2691
2957
3082
3658
3799
4564
5397
5637
5880
6017
6735
6744
7046
7244
7380
7797
7981
8349
7
2382
2382
4144
5773
6592
6634
8348
8538
2
3082
3082
8538
2
8355
8355
9983
1
4242
4242
1
8349
8349
0
1833
0
0919
0
4458
1
?*
0298
208
0375
6990
7424
2353
4320
0866
1727
1983
3245
5951
7424
7554
8840
8975
1885
0139
0139
0475
2213
8458
9206
2272
6990
6990
8476
Total
8
2
1
8
0
5
0
2
* Indiv. =Individual; ? =Inconclusive
®
WaxTag numbers in bold were bitten by multiple individuals.
Indiv.* 0077
Tag # 1956
1983
2325
5593
6118
6982
8901
9492
0
3443
Table E–2. Individuals identified on tags from night two. (n=11)
25
0131
0458
0459
0589
0936
1075
1921
2540
2550
2598
3018
4047
5039
6193
6440
6675
7923
7939
8539
8655
8711
8761
9021
9361
9508
9577
208
10
0560
0218
1741
2342
3431
3899
5011
5725
6294
7554
9946
0719
0132
0524
1164
1192
1431
1445
1660
1663
3571
3969
5076
5110
6189
6337
6982
7182
7356
7516
7738
7780
7911
8263
8363
8541
8876
8896
8901
8974
9190
9225
9656
9692
9803
33
3
2382
0105
5966
7802
0
3082
0
8355
0
4242
0
8349
1
1833
1833
0
0919
0
4458
0
?*
209
0375
3073
6117
2353
0866
0177
0533
2681
3801
5484
5679
6646
7103
7248
8810
1885
0139
1736
3412
4161
5226
2272
2272
3370
3381
6990
5732
Total
15
2
0
10
0
4
3
1
* Indiv. =Individual; ? =Inconclusive
®
WaxTag numbers in bold were bitten by multiple individuals.
Indiv.* 0077
Tag # 1162
1180
1713
6038
6092
6618
7103
7248
7323
7896
8360
8399
9197
9270
9627
0
3443
Table E–3. Individuals identified on tags from night three. (n=12)
0131
0169
0370
1416
1889
2169
2478
2771
3856
4611
5386
5480
6456
6475
6722
7703
7776
8273
8942
9103
9203
9339
9979
22
209
6
0560
2200
4656
4830
5132
8192
9198
2
0719
0919
8488
4
2382
2270
7499
7648
9744
0
3082
0
8355
0
4242
0
8349
3
1833
3776
4965
7703
2
0919
0919
3875
0
4458
0
?*
210
0375
2353
2689
5363
8139
9843
0866
1136
5318
1885
0139
2272
1673
2279
2824
6522
6990
2921
0375
2353
0866
1369
3180
4686
5539
6003
7566
9769
Total
6
0
0
7
* Indiv. =Individual; ? =Inconclusive
Indiv.* 0077
Tag # 1238
1284
4866
6931
7698
8125
0
1885
2
0139
3716
9861
0
2272
0
6990
0
3443
1
3443
5819
Table E–5. Individuals identified on tags from night five. (n=8)
Total
7
0
4
2
0
0
4
1
* Indiv. =Individual; ? =Inconclusive
®
WaxTag numbers in bold were bitten by multiple individuals.
Indiv.* 0077
Tag # 0395
1173
2824
4508
4941
5119
6863
Table E–4. Individuals identified on tags from night four. (n=13)
0131
0519
0648
1349
3267
5420
6969
8883
9223
8
5
0131
0387
1121
1951
2720
7978
210
2
0560
1571
1924
2
0560
6903
9530
5
0719
0227
0931
1482
2001
4170
0719
0026
2173
2247
2720
3002
3222
4021
4221
4266
5108
6558
6903
7900
7965
8139
15
2
2382
1099
4306
5
2382
0573
0981
1776
6188
6903
0
3082
0
3082
0
8355
1
8355
2720
1
4242
0538
4
4242
1888
2573
2679
3590
0
8349
0
8349
0
1833
1
1833
3808
0
0919
0
0919
0
4458
0
4458
1
?*
8704
0
?*
211
0139
0
1885
1885
1
3
2272
5083
6828
8899
0
6990
0
3443
0375
6514
9616
2353
7276
7793
0866
3955
4827
1885
0139
2272
4878
6623
7592
6990
0375
2353
0866
2965
4472
8276
8665
9033
9138
Total
0
0
0
6
* Indiv. =Individual; ? =Inconclusive
Indiv.* 0077
Tag #
0139
0
1885
0
0
2272
0
6990
0
3443
0
3443
Table E–8. Individuals identified on tags from night eight. (n=7)
Total
3
2
2
2
0
0
3
0
* Indiv. =Individual; ? =Inconclusive
®
WaxTag numbers in bold were bitten by multiple individuals.
Indiv.* 0077
Tag # 0655
5919
9957
Table E–7. Individuals identified on tags from night seven. (n=11)
Indiv.* 0077 0375 2353 0866
Tag # 2252 2531 0082 0082
3332
6359 2219
3822
6839 3595
5899
7073
7364
9722
7490
9618
Total
7
1
3
5
* Indiv. =Individual; ? =Inconclusive
Table E–6. Individuals identified on tags from night six. (n=11)
1
0131
8167
3
0131
1153
3349
8086
6
0131
0741
3237
4708
5097
5493
6539
211
1
0560
7387
2
0560
2627
3177
1
0560
5774
3
0719
1028
1595
3259
3
0719
1903
7401
8582
4
0719
1374
1831
2941
4695
1
2382
7282
2382
0365
3177
4871
5006
8380
5
5
2382
1587
2774
4969
5520
6885
0
3082
0
3082
0
3082
2
8355
7755
8214
0
8355
0
8355
1
4242
0868
1
4242
1046
0
4242
0
8349
0
8349
0
8349
0
1833
1
1833
1004
1
1833
9089
0
0919
0
0919
0
0919
0
4458
0
4458
0
4458
1
?*
5959
1
?*
7825
0
?*
212
2353
5
0
0375
2353
5914
6987
0866
7888
Total
0
0
2
1
* Indiv. =Individual; ? =Inconclusive
Indiv.* 0077
Tag #
0
1885
0
0139
3
2272
2179
4996
9322
0
6990
0
3443
0131
5428
5594
6953
7602
4
0131
2894
6606
6975
7774
9364
3443
Table E–10. Individuals identified on tags from night ten. (n=8)
0375
0866 1885 0139 2272 6990
0853
1140 3727
1455
5844
1691
2487
3727
4152
4304
4839
5134
9902
Total
7
0
0
10
0
2
1
0
* Indiv. =Individual; ? =Inconclusive
®
WaxTag numbers in bold were bitten by multiple individuals.
Indiv.* 0077
Tag # 0028
0425
3565
4309
6677
8326
9902
Table E–9. Individuals identified on tags from night nine. (n=8)
212
0
0560
5
0560
0159
1361
2190
4487
8718
3
0719
0930
4670
7285
0
0719
2
2382
1025
3103
3
2382
0317
1319
7905
0
3082
0
3082
0
8355
0
8355
2
4242
3779
6173
0
4242
0
8349
0
8349
2
1833
5115
7022
0
1833
0
0919
0
0919
0
4458
2
4458
4458
5522
0
?*
2
?*
6362
8730
Appendix F
Nightly activity by individual possums.
The following tables list the nightly number of WaxTags® bitten by individual possums, the
maximum distance between bitten WaxTags®, i.e., minimum distance travelled by individual
possums, and the minimum convex polygon (MCP) area of individuals‘ activity range. A
summary of these values were given in Table 4-4.
Table F–1. Nightly number of WaxTags® bitten by individual possums.
Night
Individual
0719
0131
0077
0560
0866
2382
2353
0139
2272
4242
0375
1833
3082
4458
8355
0919
6990
1885
3443
8349
1
24
31
16
12
7
7
7
1
3
2
2
1
1
2
33
25
8
10
8
3
1
5
0
2
1
0
0
2
0
0
3
2
22
15
6
10
4
0
4
3
0
2
3
0
0
2
1
0
0
4
15
5
7
2
2
5
4
0
4
4
0
1
0
1
0
1
1
0
5
5
8
6
2
7
2
0
2
0
1
0
0
0
0
0
0
0
0
6
4
6
7
1
5
5
3
0
3
0
1
1
0
0
0
0
1
0
0
7
3
3
3
2
2
5
2
0
3
1
2
1
0
0
0
0
0
0
0
8
3
1
0
1
6
1
0
0
0
1
0
0
0
2
0
0
0
0
0
9
0
5
7
5
10
3
0
2
1
0
0
0
0
2
0
0
0
0
0
0
10 Total
3
92
4 110
0
69
0
41
1
58
2
37
2
19
0
13
3
17
2
10
0
10
2
9
0
2
0
2
0
5
0
2
0
4
0
1
0
2
0
1
Active
nights
Max.
9
10
8
9
10
10
6
4
6
6
5
6
1
1
3
1
3
1
2
1
33
31
16
12
10
7
7
5
4
4
3
3
2
2
2
2
2
1
1
1
Mean (all)
9.2
11.0
6.9
4.1
5.8
3.7
1.9
1.3
1.7
1.0
1.0
0.9
0.2
0.2
0.5
0.2
0.4
0.1
0.2
0.1
Mean
(active)
10.0
11.0
8.6
4.6
5.8
3.7
3.2
3.3
2.8
1.7
2.0
1.5
2.0
2.0
1.7
2.0
1.3
1.0
1.0
1.0
NB: Individuals in pink and blue were first identified from bitemarks on WaxTags® with and
without food lure, respectively. Bold numbers are maximum nightly numbers of
WaxTags® bitten by each individual. Underlined numbers show maximum values in each
summarised category, e.g., total, active nights.
213
Table F–2. Nightly distance (m) travelled, i.e. distance between WaxTags® bitten by
individual possums.
Night
Individual
0719
0131
0077
0560
0866
2382
2353
0139
2272
4242
0375
1833
3082
4458
8355
0919
6990
1885
3443
8349
1
122
156
128
130
122
120
135
0
0
0
40
0
130
0
22
0
0
0
0
0
2
3
4
5
6
7
256 10 311 200 76 82
120 122 310 81 76 108
128 216 212 98 121 20
99 60 180 30
0 22
165 218 139 104 114 22
67 40 282 122 290 61
0
0 215
0 114 10
0 20
0
0
40 22
0 51 243
0 51 90
0
0 180
0
0
0
14 71
0
0
0 36
0 64
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0 20
0
0
0
0
0
0
0
0
0
10
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
8
9 10 Total
22
0 232 1312
0 80 67 1120
0 141
0 1065
0 36
0 558
94 114
0 1094
0 117 40 1138
0
0 98 572
0 10
0
92
0
0
0 435
0
0 10 190
0
0
0 161
0
0 32
96
0
0
0 130
0 41
0
41
10
0
0
32
0
0
0
20
0
0
0
10
0
0
0
0
0
0
0
0
0
0
0
0
Active
nights
9
10
8
9
10
10
6
4
6
6
5
6
1
1
3
1
3
1
2
1
Max.
311
310
216
180
218
290
215
40
243
180
71
64
130
41
22
20
10
0
0
0
Mean
Mean (all) (active)
131
112
106
56
109
114
57
9.2
44
19
16
9.6
13
4.1
3.2
2
1
0
0
0
146
112
133
62
109
114
95
23
73
32
32
16
130
41
11
20
3.3
0
0
0
NB: Individuals in pink and blue were first identified from bitemarks on WaxTags® with and
without food lure, respectively. Bold numbers are maximum nightly distances between
WaxTags® bitten by each individual. Underlined numbers show maximum values in each
summarised category, e.g., total, active nights.
Table F–3. Nightly MCP area (ha) covered by individual possums.
Night
Individual
0719
0131
0077
0560
0866
2382
2353
0139
2272
4242
0375
1833
3082
4458
8355
0919
6990
1885
3443
8349
1
0.8
0.8
0.6
0.3
0.3
0.2
0.5
0
0
0
0
0
0
2
2.1
0.6
0.3
0.3
0.9
0
0
0.1
0
0
0
0
0
0
0
0
3
0
0.5
1.4
0
1.3
0
0
0
0
0
0
0
0
0
0
0
0
0
4
2.8
0.6
0.8
0
0
0.8
0.5
0
0.4
0.1
0
0
0
0
0
0
0
0
5
0.5
0.2
0.3
0
0.5
0
0
0
0
0
0
0
0
0
0
0
0
0
6
0.1
0.3
0.2
0
0.3
0.9
0.1
0
0
0
0
0
0
0
0
0
0
0
0
7
0
0
0
0
0
0.2
0
0
0
0
0
0
0
0
0
0
0
0
0
8
0
0
0
0
0.2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
9
0
0
0.4
0
0.6
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
10 Total
0.9 7.2
0 3.1
0 3.8
0 0.6
0
4
0 2.3
0 1.2
0 0.1
0.1 0.6
0 0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Active
nights
9
10
8
9
10
10
6
4
6
6
5
6
1
1
3
1
3
1
2
1
Max.
2.8
0.8
1.4
0.3
1.3
0.9
0.5
0.1
0.4
0.1
0
0
0
0
0
0
0
0
0
0
Mean
Mean (all) (active)
0.7
0.3
0.4
0.1
0.4
0.2
0.1
0.0
0.1
0.0
0
0
0
0
0
0
0
0
0
0
0.8
0.3
0.5
0.1
0.4
0.2
0.2
0.0
0.1
0.0
0
0
0
0
0
0
0
0
0
0
NB: Individuals in pink and blue were first identified from bitemarks on WaxTags® with and
without food lure, respectively. Bold numbers are maximum nightly MCP areas covered
by each individual. Underlined numbers show maximum values in each summarised
category, e.g., total, active nights.
214
Appendix G
Striation quantification trial of bitemark cast image
Attempts were made during the study to quantify and objectively match bitemark images by
wavelet transform, and correlating greyscale values. The wavelet transform approach was a
collaborate effort with Auckland University of Technology, although it was abandoned
because the image capture requirements were too stringent to provide with correctly
referenced, standardised, relatively noise-free images necessary for computer processing.
While the greyscale correlation of cast images initially appereared promising, time
requirements to acquire and process the images were prohibitively large. In addition to the
usual effort to search, locate and align a full-width bitemark containing teeth I1 or I1, extra
effort in setting and re-setting critical illumination environments for each bitemark to
standardise the greyscale values was beyond the timeframe allowed to capture all specimens.
Additionally, the analySIS® image database and analysis software used in the study captured
greyscale values in its native format which needed to be transferred to a different format for
further use. The process was likewise abandoned because the image preparation and capture
requirements were too time consuming to obtain useful quantified data at the capture stage,
and that there were extra intermediate steps for the available software to allow quantitative
analysis trials in this study. Figure G–1 shows identical images of bitemark images with (a)
single-pixel and (b) band mean greyscale intensity aqcuisition, to illustrate part of the
challenge to capture a usable image in quantified correlation.
Further research in bitemark image preparation, capture and analysis may allow development
of automated process in quantitative comparison.
215
(a)
(b)
Greyscale value
(a)
(b)
Figure G–1. Trial to quantify the grayscatle intensity in striation patterns for individual
identification: identcal images‘ grayscale intensity at (a) single-pixel transect and (b)
band mean.
216