Assessing the effectiveness of ground- based ...

Assessing the effectiveness of groundbased baiting for the control of wild dogs

A. Robley, L. Woodford, P. Lee, V. Kingston, W. Peters, D. Klippell, A. Gormley 2009

Arthur Rylah Institute for Environmental Research Technical Report Series No. 193

Arthur Rylah Institute for Environmental Research Technical Series No. 193

Assessing the effectiveness of ground-based baiting for the control of wild dogs Alan Robley, Luke Woodford, Peter Lee, Vaughn Kingston, Wayne Peters, David Klippell, and Andrew Gormley

Arthur Rylah Institute for Environmental Research 123 Brown Street, Heidelberg, Victoria 3084

October 2009

In partnership with:

Department of Primary Industries, Victoria

Arthur Rylah Institute for Environmental Research Department of Sustainability and Environment Heidelberg, Victoria

Report produced by:

Arthur Rylah Institute for Environmental Research Department of Sustainability and Environment PO Box 137 Heidelberg, Victoria 3084 Phone (03) 9450 8600 Website: www.dse.vic.gov.au/ari

© State of Victoria, Department of Sustainability and Environment 2008 This publication is copyright. Apart from fair dealing for the purposes of private study, research, criticism or review as permitted under the Copyright Act 1968, no part may be reproduced, copied, transmitted in any form or by any means (electronic, mechanical or graphic) without the prior written permission of the State of Victoria, Department of Sustainability and Environment. All requests and enquiries should be directed to the Customer Service Centre, 136 186 or email [email protected] Citation: Robley, A., Woodford, L., Lee, P., Kingston, V., Peters, W., Klippell D., and Gormley, A. (2009) Assessing the effectiveness of ground-based baiting for the control of wild dogs. Arthur Rylah Institute for Environmental Research Technical Report Series No. 193. Department of Sustainability and Environment, Heidelberg, Victoria ISSN 1835-3827 (print) ISSN 1835-3835 (online) ISBN 978-1-74242-294-7 (print) ISBN 978-1-74242-295-4 (online) Disclaimer: This publication may be of assistance to you but the State of Victoria and its employees do not guarantee that the publication is without flaw of any kind or is wholly appropriate for your particular purposes and therefore disclaims all liability for any error, loss or other consequence which may arise from you relying on any information in this publication. Front cover photo: Wild Dog (Alan Robley) Authorised by: Victorian Government, Melbourne Printed by: PRINTROOM 77 St Georges Rd, Preston 3072

ii

Contents List of tables and figures...................................................................................................................2 Acknowledgements............................................................................................................................3 Summary............................................................................................................................................4 1

Introduction.............................................................................................................................5

2 2.1

Methods....................................................................................................................................6 Study site...................................................................................................................................6

2.2 2.3

Wild dog capture.......................................................................................................................7 Baiting.......................................................................................................................................7

2.4

Cameras ....................................................................................................................................8

2.5

Wild dog area of use .................................................................................................................9

2.6

Movement rates.........................................................................................................................9

3 3.1 3.2

Results ......................................................................................................................................9 Wild dogs at Deptford and Merrijig..........................................................................................9 Bait take ..................................................................................................................................10

3.3

4

3.2.1

Deptford...................................................................................................................10

3.2.2

Probability of encountering and taking a bait..........................................................11

3.2.3 Merrijig....................................................................................................................12 Area of use and rates of movement.........................................................................................12 3.3.1

Deptford...................................................................................................................12

3.3.2

Merrijig....................................................................................................................12

3.3.3

Comparisons of area used........................................................................................14

Discussion ..............................................................................................................................15

References ........................................................................................................................................17 Appendix 1 Example images of the nine wild dogs identified by cameras at Deptford............19 Appendix 2 Estimated costs of wild dog control at Deptford (eastern Victoria) and Merrijig (north-eastern Victoria)..................................................................................................................21

Assessing the effectiveness of ground-based baiting for the control of wild dogs

List of tables and figures List of tables

Table 1. Details of wild dogs captured at Deptford and Merrijig....................................................... 9 Table 2. Number of baits taken and the number of detections of species taking baits or present at bait stations at Deptford.......................................................................................................... 10 Table 3. Wild dog home range estimates from studies in Australia ................................................. 14 List of figures

Figure 1. Deptford study location....................................................................................................... 6 Figure 2. Merrijig study location. ....................................................................................................... 6 Figure 3. Wild dog fitted with GPS/satellite collar............................................................................. 7 Figure 4. Digital heat-in-motion camera set at bait station................................................................. 8 Figure 5. Wild dogs recovered at Deptford. ..................................................................................... 11 Figure 6. Cumulative bait take by wild dogs at Deptford recorded from digital cameras................ 11 Figure 7. Area of activity (95% MCP) for the four wild dogs at Deptford, August 2008–January 2009 showing track network and location of cameras and bait stations................................. 13 Figure 8. Area of activity (95% MCP) for the six wild dogs at Merrijig, February–May 2009 showing track network and location of cameras and bait stations.......................................... 13

2



Acknowledgements We thank the Department of Primary Industries (DPI) for logistical support throughout this project. This project was funded by the DPI, Victoria. Wildlife Unlimited assisted in the collection of field data. We thank the Gippsland and North East Wild Dog Management groups for their continued support. This work was conducted under the Department of Sustainability and Environment (DSE) Animal Ethics Committees permit number 09/09. Our thanks to Dave Forsyth and Lindy Lumsden for improving earlier drafts of this report.


3


Summary Baits containing the poison 1080 (sodium fluoroacetate) are a common tool used to reduce wild dog numbers around Australia. In Victoria, the use of ground-based baiting is part of an integrated wild dog management strategy. This project aimed to quantify the effectiveness of buried baiting for the control of wild dogs in Victoria. We assessed the number of captured and released wild dogs that were subsequently killed by baiting. We also used cameras placed at bait stations that were activated by heat-in-motion to assess the visitation rate of wild dogs to bait stations. We captured and attached Global Positioning System (GPS) data-logging collars to four wild dogs at a site near Deptford in Gippsland during spring 2008 and six wild dogs at a site near Merrijig during autumn 2009 in the north-east of Victoria. The nightly capture probability at Deptford (0.73 ± 0.36 SE) was similar to Merrijig (0.95 ± 0.38 SE) indicating that underlying dog density was similar at the two sites. Digital heat-in-motion activated cameras placed at bait stations recorded the presence of a further five dogs at Deptford and three dogs at Merrijig. The combined method of recording collared dogs and individually identified dogs captured by cameras provided an estimate of the known population of wild dogs at each site. Poisoned baiting was undertaken by Department of Primary industries, Victoria (DPI) staff using standard operating procedures at both sites for seven weeks with predator meat baits containing 4.5 mg of 1080. Baiting killed six wild dogs at Deptford (70% of the known population) and one wild dog at Merrijig (11% of the known population). The probability that a dog would encounter a bait station with bait in it over the course of the operation was on average 43%, and having encountered the bait station, the probability of a dog taking bait was on average 25% at Deptford. Lace Goannas (Varanus varius) (at Deptford) and foxes (Vulpes vulpes) (at Merrijig) were the main species that consumed bait. A range of other species were recorded at bait stations, including nontarget wombats, lyrebirds, native and introduced rodents, and wallabies. Wild dog controllers’ assessment of bait take by various species differed from that recorded by digital cameras. The difference is likely to be due to the difficulty of reliably identifying tracks several days after the species has taken the bait. In light of this, we recommend that bait take not be used as a measure of reduction in wild dog numbers following a baiting operation. The results from these trials indicate that buried baiting is likely to be a cost-effective tool in the management of wild dogs. Variation in success between the two sites could in part be accounted for by differences in landscape features, e.g. road density, which contributed to a difference in the spatial arrangement of baits in relation to the area used by wild dogs.

4



1 Introduction Baits containing the poison 1080 (sodium fluoroacetate) are commonly used to reduce wild dog numbers around Australia. Despite this widespread use in other parts of Australia, baiting in Victoria is limited by a lack of quantitative data on its effectiveness under Victorian conditions. In temperate south-eastern Australia, the predation of livestock by wild dogs—wild-living members of the species Canis familiaris, including feral dogs (C. f. familiaris), dingoes (C. f. dingo) and their hybrids—affects the profitability of graziers whose holdings are adjacent to or within terrain inhabited by wild dogs (Fleming and Kom 1989). In accessible terrain in NSW, ground baiting is preferred by Wild Dog Control Association members to aerial baiting for the control of wild dogs (Fleming et al. 1996). Ground baiting requires fewer baits and the baiters have more control over the placement of baits. However, the efficacy of existing ground-baiting programs in Victoria remains unclear. The Victorian Wild Dog Management Strategy’s two main aims are to ensure best practices are used in the control of wild dogs and effective research is undertaken to fill gaps in our knowledge. The North East and Gippsland Wild Dog Management Plans both indicate that baiting should be incorporated into regular control programs. A key gap in our knowledge is a robust and generalised demonstration of the effectiveness of the buried-baiting technique under Victorian conditions. The targeted use of baits would greatly reduce the amount of toxin in the environment, and enable public land managers to establish clear guidelines for best management practices. A number of studies using changes in indices of abundance have investigated the effectiveness of baiting for the control of wild dogs in Australia. A ground-based baiting campaign with small meat baits containing 20 mg of 1080 in central Australia achieved a 69% reduction in dingo sign (Best et al. 1974). In the arid zone of South Australia, a single placement of 430 baits achieved a reduction of 10–13% of 300–400 dingoes that were watering at a single bore (Bird 1994). Fleming et al. (1996) assessed the efficacy of a baiting program that replaced taken baits on a daily basis. Wild dog abundance indices were calculated from visits to stations containing non-toxic baits before and after a replacement-baiting program. The program achieved a mean reduction of 76.1% in the index of wild dog abundance. Only one other study has assessed the actual kill rate of a baiting program. In New South Wales, using two placements of poisoned meat baits, two of nine (22%) radiocollared wild dogs were killed (McIlroy et al. 1986). We used both direct and indirect measures of changes in wild dog abundance to assess the effectiveness of buried baiting. We attached GPS collars to wild dogs to provide a direct measure of individuals killed from baiting. This also allowed for an assessment of wild dog movement in areas adjacent to private land. Heat-in-motion activated cameras have been used to survey the presence of a wide range of species both in Australia and elsewhere (Tobler et al. 2008; Towerton et al. 2008). Remote cameras can operate in all weather conditions, collect more reliable information on species’ identity, and for the information returned, they are cost effective. Images collected by heat-in-motion activated cameras can be used to assess the rate of bait take by target and non-target animals. This project aimed to quantify the effectiveness of buried baiting for the control of wild dogs in Victoria by assessing the number of captured and released wild dogs that succumbed to baiting and by assessing the visitation rate of wild dogs to bait stations via images recorded by heat-in-motion activated cameras placed at bait stations. This will provide critical information for future management of the control of wild dogs.


5


2 Methods 2.1 Study site We captured wild dogs in state forest north of Bairnsdale in eastern Victoria at a site near Deptford (147o 35′E, 37o 37′S; Fig. 1) in August 2008. A second site was north-east of the township of Merrijig in north-east Victoria (146o 9′E, -37o 2′S; Fig. 2), with trapping occurring in January 2009. Both sites were within 3–5 km of the public/private land interface, and had no wild dog control within the previous 12 months.

Ú Ê

Ú Ê

Deptford

LEGEND Deptford study area Roads Sealed road Unsealed track Land Tenure Freehold State Forest

Bullumwaal

N

0

2

4 Km

Figure 1. Deptford study location.

Ú Ê

LEGEND Study area Roads Sealed road Unsealed road Land Tenure Freehold State Forest Parks

N

Merrijig

0

Figure 2. Merrijig study location.

6


2

4Km


2.2 Wild dog capture At Deptford, 26 padded Lanes dog traps (Coast to Coast Vermin Traps, Baldivis, WA) were set for 21 consecutive days in August 2008 (i.e., a total of 546 trap nights). At Merrijig, 30 Lanes dog traps were set for 21 consecutive days in January and February 2009 (i.e., a total of 630 trap nights). All traps were operated by wild dog controllers from the Victorian Department of Primary Industries (DPI) following standard operating procedures (DCE 1991). Traps were set on dog sign identified by the wild dog controller (scat or scratching). All traps were set within 1 m of the side of tracks and roads. Traps were checked daily and captured dogs were initially subdued using a ketchall pole (Ketch-all Co., California, USA), and restrained on a holding board with broad straps fitted to their waist, shoulders and neck. Each dog had its capture location, sex, and weight recorded. Each dog was fitted with a collar housing a GPS data logger (Sirtrack, Haveloch North, New Zealand; Fig. 3) linked to the Argos satellite network (http://www.argos-system.org). The GPS unit estimated a location every 60 minutes. Data from GPS–Argos linked collars include a horizontal dilution of position (HDOP) as well as the number of satellites used to calculate the location. A low HDOP value represents better GPS positional accuracy due to the wider angular separation between the satellites used to calculate a GPS unit’s position. The maximum allowable error (MAE) in location accuracy was determined using the formula: A * HDOP, where A is the GPS device accuracy (MAE; http://www.developerfusion.co.uk/show/4652/3/). The GPS receiver was a Navman Callisto module and had an estimated accuracy of 2.5 m (Navman Wireless OEM Solutions 2006). Thus, the MAE for a HDOP of six is 15 m; we excluded all locations with a HDOP of > 6 in all analyses. Collars were also fitted with a Very High Frequency (VHF) transmitter and a timed-release mechanism programed to release 75 days after wild dogs were captured. Collars were retrieved after the release date by tracking the VHF transmitter on foot.

Figure 3. Wild dog fitted with GPS/satellite collar.

2.3 Baiting The DPI wild dog controllers conducted the baiting operation. Predator meat baits (150–200 g boneless red meat) containing 4.5 g of 1080 (sodium fluoroacetate) were prepared by Gippsland Environmental Services Pty, Bairnsdale, Victoria. Baits were placed in pre-prepared bait stations, to a depth of 5–10 cm. When changing baits, wild dog controllers from DPI attempted to identify species that either took bait or visited bait stations from sign. At Deptford, baits were checked and replaced every two weeks initially and then on a weekly basis for the final three weeks. The change in replacement rate was initiated to reduce bait decay after daily temperatures exceeded 30oC. At Merrijig, baits were checked and replaced on a weekly basis. A sample of liver was retrieved from dead dogs and sent to the Alan Fletcher Research Station,


7


Queensland Primary Industries and Fisheries Department for assays to determine the presence of residual 1080. We determined the nature of the distribution of bait stations at each site using the nearest neighbour method in Arcview 3.3 (ESRI California), which determines if a distribution is clumped, uniform or random. The distance of each individual bait station to its nearest neighbouring bait station was recorded. For two individual stations that are each other’s nearest neighbour, the distance is recorded twice, once for each station. The average distance between nearest neighbours is compared to the expected distance in the case of random distribution to give the ratio R: mean distance R= 1

2

density

If R is equal to 1 then the population is randomly dispersed. If R is significantly greater than 1 the population is evenly dispersed. Lastly, if R is significantly less than 1 the population is clumped.

2.4 Cameras Heat-in-motion triggered digital cameras were set at 26 bait stations from 17 September 2008 to 15 November 2008 (1539 camera days) at Deptford. At Merrijig, cameras were set at 23 bait stations from 2 February 2009 to 8 April 2009 (1495 camera days). Images of animals at bait stations or passing on the road were recorded using Reconyx RapidFireTM ProPC90 heat-in-motion activated digital cameras (Reconyx, LLP Wisconsin, USA; Fig. 4). These cameras record 3.1 mega-pixel colour images during daylight and 3.1 mega-pixel infra-red images at night. These cameras were set to record three images as fast as possible once motion was detected (on average two images every three seconds) and to keep recording images as long as motion was detected. Individual dogs were identified by first grouping images by broad coat colour, (e.g. yellow, sable, black, and brindle). Within these groups, individual dogs were identified by assessing multiple images and comparing markings (e.g. presence, location and size of socks, colouration around muzzles, and chest markings), signs of age (e.g. grey muzzle and greying of coats), sex (males v females where visible) and physical condition. Dogs that could not be assigned to an individual were not counted.

Digital camera

Bait station

Figure 4. Digital heat-in-motion camera set at bait station.

8



2.5 Wild dog area of use Area of use of each individual fitted with a GPS collar was calculated by the 95% minimum convex polygon (MCP) (Jenrich and Turner 1969) using the Home Range Analysis extension (Jenness Enterprises, Arizona, USA) in ArcView 3.2. We also compared the areas of use with estimates from previous Australian studies.

2.6 Movement rates Movement rate was calculated by first determining the distance (m) moved between each successive location and the time interval (h) between the two locations and then dividing distance moved by the time to derive mean hourly velocity (m h-1).

3 Results 3.1 Wild dogs at Deptford and Merrijig Four wild dogs were captured (0.73 captures per 100 trap nights ± 0.36 SE) and collared at Deptford (two males and two females; Table 1). One female had swollen nipples indicating recent weaning. Six wild dogs were captured and collared (0.95 captures per 100 trap nights ± 0.38 SE) at Merrijig (three females and three males; Table 1). Table 1. Details of wild dogs captured at Deptford and Merrijig. Location

Capture date

ID

Sex

Weight (kg)

Dog colour

Date found dead

Deptford

26/08/2008

D904

F

17

Brindle

13/11/2008

Deptford

3/09/2008

D905

F

19

Yellow / Brindle

–

Deptford

29/08/2008

D907

M

25

Yellow

–

Deptford

3/09/2008

D909

M

20

Yellow

15/11/2008

Merrijig

16/01/2009

D906

F

14

Black, white chest, white feet

–

Merrijig

15/01/2009

D911

F

13

Black, white legs, grey muzzle

–

Merrijig

18/01/2009

D914

F

14

Yellow/ginger, faint white feet

–

Merrijig

28/01/2009

D910

M

19

Black, white chest and socks, white tip tail

–

Merrijig

18/01/2009

D912

M

16

Ginger

–

Merrijig

17/01/2009

D915

M

17

Sable, yellow, white legs

–


9


3.2 Bait take 3.2.1

Deptford

Thirty bait stations were established at Deptford (26 with cameras). This represented one bait station for every eight kilometres of road in the area used by collared dogs or 0.10 baits/ km2. The average distance between bait stations was 1493 m. The nearest neighbour analysis indicated that bait stations were randomly distributed (z = 1.171 and R = 0.880). A total of 192 poison baits were laid over the 51 days of baiting. DPI wild dog controllers recorded bait take by wild dogs (7), wild dog/fox (13), and fox (9; Table 2). No other species were identified by sign at bait stations by wild dog controllers. Nine individual wild dogs were identified from images captured by the digital cameras, including three collared dogs (Appendix 1). Wild dogs were detected by cameras at 54% of bait stations. This included wild dogs walking past, investigating, and taking bait. Table 2. Number of baits taken and the number of detections of species taking baits or present at bait stations at Deptford DPI records

*

Camera Records

Species

Bait taken (including bait stations with no cameras)

Bait taken (only at bait stations with cameras)

Wild dog

11

7

6

7

Wild dog/fox

18

13

–

–

Fox

18

9

1

5

Unknown

8

3

–

–

Rodent spp

0

0

1

0

Goanna

0

0

10

19

Wombat

0

0

0

21

Total

55

32

18

52

Bait taken

Present at bait station*

Present—walking on, sniffing and /or scratching around bait station.

Digital cameras detected six wild dogs taking baits (70% of the identifiable dogs in the area; Table 2). Two of the four collared wild dogs were recorded taking baits and both dogs were subsequently recovered dead (Fig. 5). One of these was recovered two days after being photographed taking a bait. Analysis of a small portion of liver was unable to detect residual traces of 1080. The second dog was recovered in a decayed and mummified state and no tissue sample was available for analysis. Cameras also detected wild dogs at bait stations on seven other occasions during the baiting program.

10



Figure 5. Wild dogs recovered at Deptford.

Bait take by wild dogs tended to occur more towards the later stages of the 51 days of the program (Fig. 6). 7

Cumulative bait take

6 5 4 3 2 1 0 1

5

9 13 17 21 25 29 33 37 41 45 49 Day

Figure 6. Cumulative bait take by wild dogs at Deptford recorded from digital cameras.

Digital cameras recorded foxes at eight of the 26 bait stations on 17 separate occasions. Over the course of the entire baiting program (51 days) foxes walked past bait stations on 11 occasions (22%), investigated bait stations five times (10%) and were detected taking bait on one occasion (2%). Feral cats were detected at 19 of the 26 bait stations, but were never recorded investigating or taking bait. Over the course of the baiting program, goannas were recorded at 17 bait stations on 23 occasions (45%). Goannas were detected taking bait on 10 occasions and scratching at bait stations on 10 occasions. Wombats were detected scratching and walking over 13 bait stations but never taking bait. 3.2.2

Probability of encountering and taking a bait

The probability that a dog encountered a bait station on a single day at Deptford was 0.015 ± 0.0033 SE. However, there were times when bait stations did not contain bait due to another wild dog, a fox or goanna having already taken the bait. The corrected probability that a wild dog encountered a bait station with bait present on a single day was 0.0115 ± 0.0031 SE.


11


At Deptford, wild dogs were recorded at bait stations with baits present on 13 occasions, with six of these resulting in bait take; hence the corresponding probability of bait take given an encounter with a baited bait station was 0.429 ± 0.132 SE. The probability that a wild dog encountered a bait station with bait present and took the bait on any given day is 0.0115*0.429 = 0.00494 ± 0.0020. During the course of the Deptford trial (51 days), the overall probability that a bait was taken was 0.25 ± 0.10 or 25%. 3.2.3

Merrijig

Twenty-three bait stations were established at Merrijig. This represented one bait station for every four kilometres of road within the area used by all the collared dogs (0.19 baits / km2). The average distance between bait stations was 2960 m. The nearest neighbour analysis indicated that bait stations had a tendency towards clumping (z = 3.304 and R = 0.640). A total of 184 poison baits were laid over the 62 days of baiting. DPI wild dog controllers recorded bait take by wild dog (1), fox (3) and unknown (34). No other species were identified from sign at a bait station by wild dog controllers. Cameras operated for an average of 63 days (range 45–65 days). Six individual wild dogs were identified from images captured by the digital cameras, including three collared dogs. Wild dogs were detected by cameras at 35% of bait stations. This included wild dogs walking past, investigating, and taking bait. Digital cameras detected wild dogs at eight of the 23 bait stations on nine separate occasions, and one wild dog taking bait (11% of the identifiable dogs in the area) on the 9 February, 6 days into the baiting program. No collared wild dogs were recorded taking baits. Digital cameras recorded foxes at 16 of the 23 bait stations on 42 separate occasions and were detected taking bait on 15 occasions (65%). Feral cats were detected at 22 of the 23 bait stations on 23 separate occasions, but were never recorded investigating or taking bait. No goannas were recorded at bait stations. Wombats were detected at 17 bait stations on 46 separate occasions. Wombats were detected 21 times scratching and walking over bait stations but never taking bait. The low level of bait take at this site prevented us from determining the probability of bait being taken.

3.3 Area of use and rates of movement 3.3.1

Deptford

GPS data at Deptford were collected between August 2008 and January 2009. Wild dogs had areas of use ranging between 48 km2 and 94 km2 for the six month period of the study. All wild dogs had access to bait throughout the study (Fig. 7). Mean hourly distance moved by males was 241 m h-1 (range 1–19,559 m h-1; median 49 m h-1) and for females 230 m h-1 (range 1–9844 m h-1; median 37 m h-1). 3.3.2

Merrijig

GPS data at Merrijig were collected between February 2009 and May 2009. Wild dogs had areas of activity ranging between 17 km2 and 30 km2 for the 2.5 month period of the study. All wild dogs had access to some baits throughout the study (Fig. 8). Mean hourly distance moved by males was 243 m h-1 (range 1–3499 m h-1; median 92 m h-1) and for females 261 m h-1 (range 1–3219 m h-1; median 91 m h-1).

12



Ú Ê

# #

##

##

# #

Ú Ê

Deptford

# # # #

# #

Deptford

# #

#

#

# # # #

Home Range F-905 (94 km2)

# # # #

Bullumwaal

Bullumwaal

# #

##

#

#

# #

# # # #

# ##

# #

# # #

# #

#

LEGEND Camera locations Bait stations

# #

# #

##

F-904 (54 km2) M-909 (48 km2)

# #

# #

M-907 (73 km2)

#

Roads Sealed road Unsealed track

#

Land Tenure Freehold State Forest

#

##

N

0

2

4 Km

Figure 7. Area of activity (95% MCP) for the four wild dogs at Deptford, August 2008–January 2009 showing track network and location of cameras and bait stations.

#

#

#

#

#

#

# Ú Ê

#

#

#

# # # # #

#

# #

#

#

# #

#

LEGEND Bait & camera locations

Home Range F-906 (17 km2) M-910 (23 km2)

#

F-911 (22 km2) M-912 (30 km2) F-914 (25 km2) M-915 (23 km2) Roads Sealed road Unsealed road Land Tenure Freehold State Forest Parks and Reserves

Merrijig

0

2

N

4Km

Figure 8. Area of activity (95% MCP) for the six wild dogs at Merrijig, February–May 2009 showing track network and location of cameras and bait stations.


13


3.3.3

Comparisons of area used

Differences in data collection and analysis methodologies, time of year, length of study and sample size and composition (age and sex) make direct comparisons of areas of use between studies difficult. Bearing these differences in mind, the area of use at Merrijig was smaller than reported in other studies, while Deptford was comparable to studies from Western and Central Australia. Female home ranges have generally been reported as smaller than males, but at Deptford this was not the case. This may in part be explained by the fact that at Deptford only two females were tracked. Table 3. Wild dog home range estimates from studies in Australia Author(s)

Method of estimation

Location

Mean home range size (km2) All

Males

Females

Harden (1985)

100% MCP

South-east Australia

–

27

–

McIlroy et al. (1986)

100% MCP


22

25

20

Catling (pers. comm. in McIlroy et al. 1986)

Not reported

East Australia

10

10

11

Catling (pers. comm. in McIlroy et al. 1986)

Not reported

East Australia

18

20

12

Corbett (1995)

Not reported

Central Australia

67

98

47

Thomson (1992)

95% MCP

Western Australia

–

85

56

Eldridge et al. (2002)

95% MCP

Northern Territory

–

157

272

Claridge et al. (2009)

MCP


90

100

79

Robley et al. (in press)

100% MCP


100

156

55

78

124

45

95% MCP Deptford (this study)

95% MCP


67

60

74

Merrijig (this study)

95% MCP


23

25

21

14



4 Discussion The aims of this study were to assess the effectiveness of ground-based baiting as a technique for reducing wild dog numbers, and to determine the probability of a wild dog encountering a bait station, and once encountered, the probability of bait being taken. We assessed the fate of 10 wild dogs fitted with GPS–Argos satellite linked collars at two locations during a poison baiting operation. We also used digital cameras placed at bait stations to determine the number of dogs without collars visiting bait stations. Using information from the cameras, we were able to determine the probability that a wild dog would encounter a bait station, and subsequently take bait. The baiting operation at Deptford removed 70% of the identifiable wild dog population, including two of the four collared wild dogs, with bait take at this site tending to increase later in the baiting period. Non-target species (foxes and goannas) were also recorded taking baits, reducing the availability of baits to wild dogs. Non-target bait take contributed to a moderate probability of a bait being taken once found by a dog (25%). At Merrijig, 11% of the wild dog population, or one wild dog, was recorded taking bait. Foxes were a key non-target species recorded taking baits; no goannas were recorded at this site. The low level of bait take at this site prevented us from determining the probability of bait being taken. Results indicate that buried baiting is likely to be a useful and effective tool for killing wild dogs. However, the varied results indicate that its effectiveness may not be uniformly successful at all locations and at all times, and the moderate probability of bait take when a bait was encountered needs to be interpreted with care as it is based on data from only one site. There are a range of possible factors that may influence the success of a baiting program (e.g. degradation of 1080 in the baits over time, the removal of baits by non-target species, and/or the availability of natural prey; McIlroy et al. 1986). Other factors may include the underlying density of wild dogs in relation to the density of baits, and the density of baits in the landscape in relation to the movement patterns and area of use of wild dogs. A greater percentage of the population was killed by the baiting program at Deptford than at Merrijig. In part, this can be explained by the distribution of baits in the landscape. Baits at Deptford were randomly distributed across the area used by wild dogs compared to the clumped distribution at Merrijig. This difference was a result of the spatial arrangement of the roads at each site. Road density (length of road / total km2) was higher at Deptford compared to Merrijig (0.85/km2 v 0.75/km2). The higher road density provided the opportunity to spread baits more widely across the landscape than at Merrijig, allowing a greater opportunity for wild dogs to encounter a bait station. The timing of the baiting program differed at each site. Baiting at Deptford was undertaken in late spring to early summer, and at Merrijig it was undertaken in late summer to early autumn. Baiting campaigns are traditionally undertaken from spring to autumn, although generally later in autumn and earlier in spring than was the case in our study (Fleming et al. 2001). Baiting at this time of year is related to either wild dog ecology (e.g. breeding season—autumn—or possible dispersal and movement of pups—spring) or livestock protection (e.g. lambing in spring). Differences in season per se are unlikely to explain the differences in the results in this study. At Deptford, cameras identified nine individual wild dogs, with 70% of these killed by the 51 day buried baiting program compared to the 21 days of trapping that potentially removed 40%, thus trapping removed more dogs per week (0.19) than baiting (0.12) in that area at that time. However, the requirement under the Prevention of Cruelty to Animals Act 1986 is to check traps on a 24 hour basis, whereas current Government policy is that baits can remain unchecked for 14 days. In the program at Deptford, this means that trapping would be approximately five times more expensive than baiting (Appendix 2). At Merrijig, traps caught six wild dogs and baiting removed one, hence


15


trapping potentially removed more dogs per day than baiting (0.29 and 0.02 respectively) and cost 40% less (Appendix 2). At both sites, bait take by non-target species was an issue. At Deptford, goannas removed 5% of baits, potentially reducing the effectiveness of that program, while at Merrijig, foxes removed 8%. McIlroy et al. (1985) assessed the sensitivity of V. varinus to 1080 poisoning, and from limited data suggested a LD50 of 119 g / kg-1. They concluded that it is unlikely that reptiles face any direct poisoning risk from pest-poisoning campaigns involving 1080, given their high tolerance and the enormous amounts of poisoned bait that would have to be eaten. Thus it is extremely unlikely that goannas at Deptford would have suffered any mortality from consuming the predator meat baits which contain 4.5 mg/bait. Foxes are a known predator of stock and native wildlife and are targeted in control operations around Australia. Integrated pest management is a strategic aim of many control programs, thus the reduction of foxes, while impacting on programs specifically aimed at protection of stock from wild dog attacks, provides a benefit to both livestock and wildlife. The reason for the discrepancy between the number of bait takes and the species taking baits recorded by DPI and the digital cameras is not clear. It is likely to be a combination of the digital cameras failing to record bait takes on some occasions and wild dog controllers incorrectly assessing bait as being taken, when in fact it was still present in a bait station. Predator meat baits are 250 g of moist horse or kangaroo meat. Dirt from bait station adheres to the bait making it difficult to distinguish from sods of dirt. Placing the bait into a small hole and sieving dirt back on top of the bait may reduce the incidence of this happening. Differences in the identity of the species taking bait could be attributed to the difficulty in assessing species from sign several days after bait has been taken. In light of this we would recommend that bait take be discontinued as a measure of control program success. This work provides initial information on the efficacy of baiting as a control tool for wild dogs and non-target species encounters with bait stations. To assess the optimal strategies for the control of wild dogs adequately, we need meaningful a priori expectations of the level of population coverage achievable by differing densities of control devices (baits and/or traps). Spatially explicit models can simulate encounter rates between wild dogs and control devices at different road and control device densities. By applying a modelling approach, we can objectively control for confounding effects of differing bait and trap density, length of toxic bait availability, and spatial constraints in placement of baits and traps, and different population densities. Population density is an issue because home-range size, and hence the probability that a wild dog will interact with devices located at different points within its home range, is likely to vary with density. The result will be that for a given underlying dog and road density we will be able to estimate the required control device density for a given reduction in wild dogs. This approach has been applied to optimising bait station density for the control of possums in New Zealand (Tompkins and Ramsey 2007). We recommend that the next phase of work be undertaken using the approach outlined above.

16



References Best, L. W., Corbett, L. K., Stephens, D. R., and Newsome, A. E. (1974). Baiting trials for dingoes in Central Australia, with poison 1080’, encapsulated strychnine, and strychnine suspended in methyl cellulose. CSIRO Division of Wildlife Research, Technical Paper No. 30. Bird, P. (1994). Improved electric fences and baiting techniques — a behavioural approach to integrated dingo control. Unpublished Final Report on Project DAS 39 to the Wool Research and Development Corporation. Corbert, L. (1995). The Dingo in Australia and Asia. University of New South Wales Press Ltd, Sydney. Claridge, A., Mills, D., Hunt, R., Jenkins, D., and Bean, J. (2009). Satellite tracking of wild dogs in south-eastern mainland Australian forests: Implications for management of a problematic top-order carnivore. Forest Ecology and Management 258, 814–22. Department of Conservation and Environment (1991). VerminPac. Policies for vermin control. Department of Conservation and Environment, East Melbourne. Eldridge, S. R, Shakeshaft, B. J., and Nano, T. J. (2002). The impact of wild dog control on cattle, native and introduced herbivores and introduced predators in central Australia. Final Report to Bureau of Rural Sciences. Northern Territory Parks and Wildlife Commission, Alice Springs, Northern Territory, Australia. Fleming, P. (1996). Ground-placed baits for the control of wild dogs: evaluation of a replacementbaiting strategy in north-eastern New South Wales. Wildlife Research 23, 729–40. Fleming, P. and Kom, T. J. (1989). Predation of livestock by wild dogs in eastern New South Wales. Australian Rangelands Journal 11, 61–6. Fleming, P., Thompson, J. A., and Nicol, H. I. (1996). Indices for measuring the efficacy of aerial baiting for wild dog control in northeastern New South Wales. Wildlife Research 23, 665– 74. Fleming, P., Corbert, L., Harden, R., and Thomson, P. (2001). Managing the impact of dingoes and other wild dogs. Bureau of Rural Science, Canberra. Harden, R. (1985). The ecology of the dingo in north-eastern New South Wales. I. Movements and home range. Australian Wildlife Research 12, 25–38. Jenrich, R. I. and Turner, F. B. (1969). Measurement of non-circular home range. Journal of Theoretical Biology 22, 227–37. McIlroy, J. C., King, D. R., and Oliver, A. J. (1985). The Sensitivity of Australian Animals to 1080 Poison VIII. Amphibians and Reptiles. Australian Wildlife Research 12, 113-18. McIlroy, J. C., Cooper, R. J., Gifford, E. J., Green, B. F., and Newgrain, K. W. (1986). The effect on Wild Dogs, Canis f. familiaris, of 1080-poisoning campaigns in Kosciusko National Park, NSW. Australian Wildlife Research 13, 535–44. Navman Wireless OEM Solutions (2006). Jupiter 31 GPS receiver module data sheet. (http://www.navmanwirelessoem.com/uploads/TM/30/TM30qEQoinYQRhT58SZXMA/L A010811B_J31_DataSheet.pdf) Accessed November 2008. Person, J. (2005). Writing Your Own GPS Applications: Part 2. http://www.developerfusion.co.uk/show/4652/3/. Accessed November 2008. R Development Core Team (2008). R: A language and environment for statistical computing. R Foundation for Statistical Computing. Vienna, Austria.


17


Robley, A., Gormley, A., Forsyth, D. M., Wilton, A., and Stephens, D. (in press). Movement and habitat use by wild dogs in eastern Victoria. Australian Mammalogy in press. Thomson, P. C. (1992). The behavioural ecology of dingoes in north-western Australia. IV. Social and spatial organisation and movements. Wildlife Research 19, 543–63. Thomson, P. C., Rose, K., and Kok, N. (1992). The behavioural ecology of dingoes in northwestern Australia. VI. Temporary extraterritorial movements and dispersal. Wildlife Research 19, 585–95. Tobler, M. W., Carrillo-Percastegui, S. E., Pitman, L. R., Mares, R., and Powell, G. (2008). Further notes on the analysis of mammal inventory data collected with camera traps. Animal Conservation 11, 187–9. Tompkins, D. M., and Ramsey, D. (2007). Optimising bait-station delivery of fertility control agents to brushtail possum populations. Wildlife Research 34, 67–76. Towerton, L. A., Penman, D. T., Blake, E. M., Deane, T. A., Kavanagh, P. R., and Dickman, R. C. (2008). The potential for remote cameras to monitor visitation by birds and predators at Malleefowl mounds. Ecological Management & Restoration 9, 64–7.

18



Appendix 1 Example images of the nine wild dogs identified by cameras at Deptford. Identification of individuals was based on multiple images of animals from a variety of angles and times of day. These images are presented as examples only.


19


20



Appendix 2 Estimated costs of wild dog control at Deptford (eastern Victoria) and Merrijig (north-eastern Victoria) Activity

Task

Person days

Number of operational days

No. dogs killed

Dogs killed per day

Cost per dog killed* ($)

Deptford Baiting

Prepare bait stations and check and replace baits

7

52

6

0.12

583

Trapping

Set and check traps

25

21

4

0.19

3125

Baiting

Prepare bait stations and check and replace baits

7

52

1

0.02

3500

Trapping

Set and check traps

25

21

6

0.29

2083

Merrijig

* Assumes $500/day labour and material costs


21


ISSN 1835-3827 (print) ISSN 1835-3835 (online) 2 Arthur Rylah Institute for Environmental Research Technical Report Series No. 193 ISBN 978-1-74242-294-7 (print)

ISBN 978-1-74242-295-4 (online)