Jenny Dent - June 2020
EXECUTIVE SUMMARY
The Otago Biodiversity Peninsula Trust, alongside Predator Free Dunedin, is working to eradicate possums from the Otago Peninsula. Two members of the ZIP team, Al Bramley and Caroline Wallace, visited the site in 2019, and observed a modified electric farm fence that, based on anecdotal success, warranted further experimental testing as a possum barrier.
The original fence design, consisting of a 900 mm high wire-mesh fence fitted with three single hotwires, was developed by a Trustee of the Otago Peninsula Biodiversity Trust, Brendon Cross. It was then adapted into an octagonal testing pen at the ZIP Lincoln Facility. A total of 16 (8 female, 8 male) wild-caught brushtail possums (Trichosurus vulpecula) were tested against the fence. Each individual was released into the pen overnight and their interactions with the fence were monitored by continually recording video cameras.
Overall, the pen contained 62.5% of individuals (10/16). Male possums were more likely to escape the pen (62.5% escape, 5/8 individuals) than their female counterparts (12.5% escape, 1/8 individuals). The most common mode of escape was termed ‘jump and climb’. Individuals, having previously encountered the lower hotwire, jumped past the first hotwire onto the mesh, and climbed over or under the mid and upper hotwires. It is speculated that the sex bias observed in this trial may relate to a disparity in initial jump height between male and female possums.
Placing an individual in a pen generates a very high motivation to escape. Despite the high overall rate of escape in these trials, this low-cost fence may offer benefit to control operations as a means of slowing migration. The higher containment rate of females is a promising aspect of this tool as it suggests male-skewed invasion, which would limit the breeding potential of the population. Improvements to the fence design should aim at targeting these male individuals. Alterations that could be considered include raising the height of the mesh, or installing a fourth hotwire between the mid and upper hotwire.
INTRODUCTION
Predator fencing is a valuable wildlife management tool that has contributed to the success of eradication programmes worldwide (Bode et al. 2013, Parkes et al. 2010). In New Zealand, the application of multi-species exclusion fencing (typically 1.8-2m tall with narrow mesh and a hooded cap) remains scale-limited due to high cost (Scofield et al. 2011) and intrusive design. Practical application at a landscape scale is likely to require the development of low cost alternatives.
Fence design is constrained by both the behavioural and physical capabilities of target animals. Refining the target specificity of a fence is, therefore, one means of reducing cost and allowing for design flexibility e.g. reduced height if not targeting feral cats (Agnew & Nichols 2018). A possum (Trichosurus vulpecula) specific fence, using a combination of wire mesh and electrified hotwires is an option that has been explored by many groups around New Zealand (Cowan & Rhodes 1992; Aviss & Roberts 1994; Sanders et al. 2007). Anecdotal reports of effectiveness have been mixed, suggesting a need for further experimental testing.
Two prior studies have experimentally tested the efficacy of mesh hotwire fences for possum control in New Zealand. Clapperton & Matthews (1996) tested a low height, 600 mm mesh fence, fitted with 3-4 hotwires on two outriggers (110 - 200 mm). They concluded that, while the design had some merit, possums were easily able to cross the fence if sufficiently motivated (42% containment in ‘forced’ trials). Day and MacGibbon (2007) tested a modified variant of the Clapperton & Matthews (1996) fence, consisting of a 1200 mm mesh fence, fitted with 6 hotwires on two outriggers (300 mm). This design also had a low rate of containment when tested on captive possums (52%).
In this experimental trial, we tested the effectiveness of an alternative electric fence design, based on specifications provided by the Otago Peninsula Biodiversity Trust, as a possum-specific barrier. The design differs from previously tested fences in that it employs three shorter, single hotwire outriggers. Such a design, if effective, could be constructed by retrofitting to existing farm fencing with additional hotwires and, if necessary, a finer grade mesh. This may present a low-cost option for possum management at a landscape scale.
METHODOLOGY
Trial Pen
An octagonal pen, measuring approximately 20 m2, was constructed at the Zero Invasive Predators (ZIP) predator behaviour facility in Lincoln in December 2019. The pen was located within a pre-existing, 2-ha predator enclosure ensuring containment of trial subjects. The pen consisted of eight, 1100 mm tall timber posts spaced at 2000 mm intervals (Figure 1). Wire mesh (1060 mm wide, 45 mm hex aperture) was fixed directly on to the inside face of the timber post and run 160 mm out onto the ground. High tensile wire was used to reinforce the back of the wire mesh, preventing bagging. Three hotwires, supplied with electricity at a standard rate of 0.2A, 6kV, were installed on the inside face of the fence. The two lower hotwires (400 mm and 800 mm) were fitted on 100 mm fibreglass outriggers (Figure 2). This third, upper hotwire (1000 mm) was fixed directly to the inside face of the posts on small insulator units (Figure 2).
Figure 1: Trial pen used for testing
Figure 2: Hotwire and mesh configuration within trial pen
Animals
A total of 16 (8 female, 8 male) wild-caught brushtail possums (Trichosurus vulpecula) from the Canterbury region were supplied to ZIP by independent contractors for fence testing. Upon arrival at the facility, individuals were examined and determined to be in good physical condition. The animals were acclimatised to captivity for a two week period prior to fence testing. During trials, individuals were supplied with a regular diet of multi-feed pellets, fresh fruit or vegetables, and water daily. Shelter was also provided within the pen. Animal health was monitored during the trials and no injury was noted during, or resulting from, trialling.
Trial procedure
Trials were conducted on fine-weather nights between December 2019 and February 2020. Individual possums were transported to, and released inside, the trial pen in hessian sacks. This soft-release method minimised stress as possums did not encounter the fence in the presence of a human handler. Possums were left in the trial pen overnight. This single night approach was justified by pilot study work (n=6) which indicated that if escape was the outcome, it always occurred on the first night of activity.
Interactions with the fence were continuously recorded via three Techview QV-3140 cameras (full D1 resolution, 100 frames per second, IR illumination) positioned around the perimeter of the pen. Footage collected by these cameras was reviewed by researchers following conclusion of each individual trial.
The trial was carried out under Lincoln University Animal Ethics Approval (AEC2017-03)
Analyses
Footage was reviewed in full by researchers. Interactions with the fence and hotwires, as well as escape attempts, were noted and described. A binary logistic regression model was fitted to determine whether trial outcome was related to an individual’s sex, or body weight. Two analyses were performed, one with the dependent variable ‘escape success’, and the other ‘escape attempt’. Statistical analyses were carried out in SPSS version 20 (IBM Corp. 2011).
RESULTS
Overall result
The trial pen contained 62.5% of participants (10/16). The binary logistic regression model indicated that sex had a significant statistical effect on trial outcome, X2 = 4.26, p = 0.04. Male possums were more likely to escape the trial pen than female possums (62.5% and 12.5% escape respectively; Figure 3). Body weight was not found to be a significant predictor of trial outcome and therefore was excluded from the model (X2 = 0.97, p = 0.32).
Figure 3: Escape (%) of female and male possums from the modified electric fence trial pen
Observed behaviour
All individuals initially interacted with the lower hotwire. Contact with the lower hotwire was usually made with forepaws, and resulted in immediate recoil from the fence. The majority of possums (75%) continued to interact with the fence after the initial hotwire interaction. The maximum number of shocks received by an individual possum in the 24 hr period was 26.
Overall 62.5% of individuals went on to attempt to escape the pen (10/16). There was no significant difference between males and females in this regard (X2 = 1.06, p = 0.303; Table 1). All successful escapes occurred on the first attempt, but some ultimately unsuccessful individuals made up to three further attempts. Escapes all occurred within a few hours of release with the majority occurring within the first hour (range: 0.5 – 4.5 hr).
Two modes of escape attempt were observed. The most common (84.6%) was termed ‘jump and climb’. The individual, having previously interacted with and been shocked by the lower hotwire, jumps past the first hotwire onto the mesh, then climbs over or under the mid and upper hotwires. Five of six escapes were conducted in this manner (Table 1). Two escape attempts were classed as ‘climb’, in which an individual attempted to climb under or over all hotwires (Table 1). Only one successful attempt was carried out in this manner.
Table 1: Summary of individual trial behaviours
DISCUSSION
Overall Fence Feasibility
The electric fence utilised in this trial contained 62.5% of possums tested. This rate of containment is similar to that reported by Day & MacGibbon (2007), who conducted an equivalent experimental trial using an alternative electric fence design (52% contained, n=25). The high rate of breach (37.5%) suggests that, while the electric wires provided a behavioural deterrent for possums, the combination of mesh and electric wire fence could be crossed provided there is sufficient motivation to do so. A review of fencing in Australia indicated that electric fencing had limited effectiveness unless paired with a sufficiently challenging physical barrier (Rob & Longley 2005). It is therefore plausible, that the pairing of electric wires with a simple, low height mesh, limits the effectiveness of the design regardless of electric wire configuration.
Greater containment of females is a promising aspect of this tool as it suggests that an invading population may become male-skewed. Initial reinvasion following control typically comprises equal proportions of male and female possums from surrounding areas (Little & Cowan 1992; Efford et al. 2000, Ji et al. 2004). By limiting female dispersal relative to male dispersal, a reinvading population becomes male-skewed, limiting the growth potential and viability of the population (Wedekind 2002). Eradication of a male dominated colonising population may be easier as the population is likely smaller and slower growing. Male possums may also adopt a larger home range than female counterparts (Yockney et al. 2013), potentially making male invaders easier to target in follow up control. Given the low cost of this retrofitted electric fence design, slowed, male-dominated migration may be an acceptable outcome, provided the barrier is used in conjunction with other control measures. Previous experimental trials have not specified the sex of participants, and therefore it is unknown if sex bias is a shared feature of electric fence barriers.
Actual rates of migration across an electric fence may vary from that observed in our experimental scenario. This style of testing – an individual placed in a small pen with little else to do but test the fence – is designed to rapidly answer the containment question, but it does generate a very high motivation to escape (which may not be reflective of encounter/interaction behaviour in the field). Wild animals are likely to interact with a barrier fence over longer timescales, but likely face lower motivation to cross to the clean side when they do. In a landscape scale application, receiving an exploratory shock from the lower hotwire has alternative outcomes, including avoidance or edge following behaviour, which are not feasible for individuals within the confines of the trial pen. This may help to explain why some field applications of possum-specific electric fences report higher rates of efficacy than that observed in this trial (Cowan & Rhodes 1992).
Female exclusion
Greater containment of females is a beneficial feature of this barrier as it suggests a disproportionate effect on invader population viability. Female possums were not found to make significantly fewer escape attempts, indicating that variation in motivational drive was unlikely to be a contributing factor. Disparate experience of the electric hotwires during an escape attempt is a more likely explanation for the observed sex bias.
The main mode of escape attempt in both sexes was ‘jump and climb’. We propose that, despite the similarity of the approach, variation in the height of the initial ‘jump’ may affect the escape outcome. Review of trial footage suggests that male possums may ‘jump’ higher than their female counterparts, resulting in different interaction with the mid electric wire (Figure 2). Females tended to make contact with the mesh at or below the mid hotwire (Figure 2). This typically resulted in shocks to their forepaws, forelegs, or face. Males made contact with the mesh slightly higher, usually reaching above the mid hotwire. This resulted in shocks to the shoulders or back. Clapperton & Matthews (1996) observed that a possum’s response to electric shock was dependant on the part of the body that contacted the fence. In this fence configuration, the position of the body may have influenced whether the shock caused possums to bolt forward or jump back into the pen. If contact point is an important factor in determining the outcome of an escape attempt, a potential improvement to the fence design may be adding a fourth outrigger between the mid and upper hotwire at a height of c. 900 mm. Alternatively, the height of the mesh and existing mid hotwire could be raised slightly.
Figure 4: Maximum jump height of male and female possums during ‘jump and climb’ escape attempts. Blue lines denote possums which did not successfully pass the mid-hotwire; green lines denote individuals that did. First escape attempts are shown only.
ACKNOWLEDGEMENTS
We acknowledge and thank the following people for their help in designing and running this trial:
Brendon Cross of the Otago Peninsula Biodiversity Trust for providing us with the original fence design and specifications.
Nigel Broadbridge and his team at Central Fencing Ltd for construction of the trial pen for experimental testing.
Rhys Millar from Predator Free Dunedin, and Nathan McNally from Otago Peninsula Biodiversity Trust, for visiting our site and offering input to the trial design.
Zero Invasive Predators team members for their assistance reviewing footage and editing this report.
REFERENCES
Aviss, M. & Roberts, A. (1994) Pest fences: notes and comments. Threatened species occasional publication 5. Department of Conservation, Wellington.
Agnew, T. & Nichols, M. (2018) Low height predator fencing. ZIP technical report 6. https://zip.org.nz/technical-reports-feed/2020/2/technical-report-6-low-height-predator-fencing
Bode, M., Brennan, K.E., Helmstedt, K., Desmond, A., Smia, R. & Algar, D. (2013) Interior fences can reduce cost and uncertainty when eradicating invasive species from large islands. Methods in Ecology and Evolution, 4, 819-827.
Clapperton, B.K. & Matthews, L.R. (1996) Trials of electric fencing for restricting the movements of common brushtail possums, Trichosurus vulpecula. Wildlife Research, 23 , 571-579.
Cowan, P.E. & Rhodes, D.S. (1992) Restricting the movements of brushtail possums Trichosurus vulpecula on farmland with electric fencing. Wildlife Research, 19, 47-58.
Day, T. & MacGibbon, R. (2007) Multiple-species exclusion fencing and technology for mainland sites. Managing Vertebrate Invasive Species: Proceedings of an International Symposium (eds. G.W. Witmer, W.C. Pitt & K.A. Fagerstone), pp. 418 – 433. National Wildlife Research Centre, Fort Collins.
Efford, M., Warburton, B. & Spencer, N. (2000) Home-range changes by brushtail possums in response to control. Wildlife Research, 27, 117-127.
Ji, W., Sarre, S., White, P., & Clout, M. (2004) Population recovery of common brushtail possums after local depopulation. Wildlife Research, 31, 543-550.
Little, E.C.S., Cowan, P.E. (1992) Natural immigration of brushtail possums, Trichosurus vulpecula onto Aroha Island, Kerikeri Inlet, Bay of Islands, New Zealand. New Zealand Journal of Zoology, 19, 53-59.
Long K. & Robley A. (2004). Cost effective feral animal exclusion fencing for areas of high conservation value in Australia. Arthur Rylah Research Institute, Melbourne.
Parkes, J.P., Ramsey, D.S., Macdonald, N., Walker, K., McKnight, S., Cohen, B.S. & Morrison, S.A. (2010) Rapid eradication of Feral Pigs (Sus scrofa) from Santa Cruz Island, California. Biological Conservation, 143, 634–641.
Sanders, M., Brown, K.P. & Keedwell, R.J. (2007) Testing the effects of a predator-exclusion fence on predator abundance and wetland bird breeding success at Ruataniwha wetlands, Twizel. Department of Conservation, Christchurch.
Scofield, R.P., Cullen, R. & Wang, M. (2011) Are predator-proof fences the answer to New Zealand’s terrestrial faunal biodiversity crisis? New Zealand Journal of Ecology, 35, 312-317.
Wedekind, C. (2002) Manipulating sex ratios for conservation: short-term risks and long-term benefits. Animal Conservation, 5, 13-20.
Yockney, I. J., Nugent, G., Latham, M. C., Perry, M., Cross, M. L., & Byrom, A. E. (2013) Comparison of ranging behaviour in a multi-species complex of free-ranging hosts of bovine tuberculosis in relation to their use as disease sentinels. Epidemiology and infection, 141, 1407–1416.