Phil Bell (Innovation Director) – November 2017
INTRODUCTION/BACKGROUND
Zero Invasive Predators (ZIP) is focussed on the development of knowledge, tools, and techniques that contribute to the ‘toolbox’ to enable permanent and complete removal of rats, stoats, and possums, and the long-term protection against reinvasion, within large landscapes of the New Zealand ‘mainland’ (i.e. the North and South Islands). We call this model ‘Remove and Protect’.
An aerially-applied tool is expected to be required for the initial removal of these predators at large-scale sites. Some of the early work developing techniques for island eradication investigated sodium fluoroacetate (1080) as an option (McFadden & Towns, 1991; Moors, 1985). However, it was subsequently discounted as an eradication tool on account of its acute toxicity and the perception that some individuals of the target populations – particularly rodents – could detect it in the bait and avoid it (McFadden & Towns, 1991). There has been significant improvement in its use as a control tool since that work, namely prefeeding to increase toxicant uptake (Nugent et al., 2011), improved quality control of manufacture (Nugent et al., 2012; Nugent et al., 2010), and reductions in sowing rates to achieve the same results for less cost. Extensive use in suppression operations has refined aerial 1080 use, but those operations do not aim to remove all target individuals (Elliott & Kemp, 2016) nor do they currently attempt to limit reinvasion (Griffiths & Barron, 2016).
ZIP sought to test whether aerial 1080 could completely remove possums and rats from a site on the NZ mainland. This trial builds on the encouraging results from an earlier trial conducted at Mt Taranaki in 2016/17 (Bell et al, in press). In that trial, the first phase of baiting resulted in ‘functional extinction’ of possums (with only 1 possum detection over 36,430 detection nights), while there were 41 rat detections over 46,755 detection nights. Subsequent live capture trapping of the surviving rats using 1080 pellets as bait suggested that 1080 bait aversion may not be such a significant issue as previously thought (Morgan, 2004), but does require further testing in an aerial baiting operation (i.e. without the influence of cage traps).
OBJECTIVE
This trial sought to develop a technique (sometimes referred to as a ‘prescription’) for aerial 1080 to achieve ‘functional’ extinction of possums in South Westland. We dubbed the technique ‘1080 to Zero’. The threshold for achieving functional extinction was set at 1 possum per 400 ha (OSPRI, 2014), a level at which the population is no longer functionally viable.
It was initially intended to be a trial for rats too; however the rat numbers in the block (as measured in the pre-operational monitoring) were too low to be a significant test for the technique. We report the results of the rat aspects of the trial here, for completeness.
METHODOLOGY
The trial was carried out in South Westland, on a land parcel at the confluence of the Jackson and Arawhata Rivers (see Appendix One). The aerial 1080 operation took place over an area of approx. 2,300ha, while a detection core area of 394ha, assumed to be representative of the entire treatment area, was intensively monitored to assess the performance of the treatment.
The detection core area was also situated for maximum protection against reinvasion, given that there was no control of the rats and possums beyond the aerial 1080 operation. Based on the accepted reluctance of possums to enter waterways, it was considered that the Jackson and Arawhata Rivers would slow down the reinvasion rate such that the sides of the core would not be vulnerable if located approximately 500m from the rivers. The detection core was also set a minimum of 2km from the other/unprotected edges of the aerial 1080 treatment zone – 2km was used to improve the integrity of the trial, given that the 1km buffer in the Mt Taranaki trial appeared to be compromised during the first detection period of that trial.
Prior to the aerial 1080 application, pre-operational monitoring was conducted to confirm the presence of possums and rats, and to gauge their relative abundance. This monitoring was carried out on 5 lines within the detection core area (lines B, G, M, S, and Y – see Appendix Two for track line locations) using PCR WaxTags® and peanut butter chew cards that were installed in an alternating pattern at 50m intervals along each line (i.e. 100m between each ‘device’ type). There were 99 chew cards, and 95 wax tags used in total. Tracking tunnels were not used at this stage of the trial because there was not enough time for them to be installed and weathered, potentially compromising any population estimates.
The devices were deployed on 8 and 9 June 2017, and retrieved on 15 and 16 June; i.e. a 7 night deployment, in line with the National Pest Control Agencies (NPCA) best practice for possum population monitoring.
The aerial baiting prescription consisted of two prefeed applications, both of 2kg/ha, followed by a toxin application of 4kg/ha.
All bait applications were flown by GPS-guided helicopter with a bait spreading bucket. All bait was applied with a 50% overlap of the bait swath to ensure there were no gaps in the bait coverage – effectively bait was applied everywhere twice, with an application rate of 1kg/ha (for prefeed) and 2kg/ha (for toxin) out of the bucket which results in double that (e.g. the desired rate) on the ground. Bait was sown only when the wind speed was below 10 knots, to reduce wind drift of the bait once it left the bucket.
Permission was secured from the Medical Officer of Health (Ministry of Health) to aerially sow toxic bait to the river edge, thus removing any exclusion zones from the treatment area. As such, there were no areas where target animals could be present where they would not be able to gain access to bait (i.e. no animals were ‘exempt’ from the trial treatment).
The bait used was Wanganui #7 double orange-lured 6g cereal pellets (Orillion), with the toxic bait carrying a loading of 0.15% 1080 (concentration assumed as ordered, with no assays of bait conducted). This trial had provision to conduct a second aerial operation utilising RS5 cinnamon lured 6g cereal bait if surviving rats and/or possums were detected post-treatment 1. Given that Wanganui #7 is known to last longer in the field, it was decided to use it first when the target species numbers would be highest and they would be expected to consume the majority of the bait, leaving less bait available to be found by kea (a consideration in the kea risk mitigation planning).
The first prefeed application was applied on 5 June 2017. The second prefeed application was applied on 16 June.
Permission was secured from the Department of Conservation to undertake this trial. Prior to the application of toxin, an external contractor carried out kea observation work within the treatment area, as required by the approval conditions. The objective of this work was to confirm that no juvenile kea flocks (defined as 5 or more young birds) were present in the area immediately prior to toxic baiting. Juvenile kea were identified as the ‘high risk’ age class for likelihood of consuming bait. The work was conducted between 13 and 27 June.
The toxin was aerially applied on 5 July 2017. Five fine nights followed the application of toxic bait.
Five days post-toxin application, detection devices were deployed throughout the detection core area in an attempt to locate any target species survivors of the operation (see Appendix Two for locations). Tracking cards baited with Nutella were placed into pre-weathered tracking tunnels (Black Trakka) in a 50m x 100m grid – 964 tracking tunnels were used. Peanut butter-filled chew cards were placed in a 100m x 100m grid – 482 chew cards were used. All devices were serviced (and refreshed if needed) every week, or whenever river conditions allowed access if beyond a week. Each service lap was conducted by a team of five field rangers, and took 4 days to complete (working approx. 6 hours per person per day).
Twenty days post-toxin application, motion-activated cameras (Brownings), lured with peanut butter-filled chew cards, were deployed in a 200m x 200m grid throughout the detection core – 97 cameras were used (see Appendix Three for locations). The SD cards and batteries were replaced every week, as per the servicing regime of all devices.
Detection monitoring was to be carried out for at least 50 days post-toxin application. Fifty days was chosen for the following reasons:
to allow any surviving animals to recover from ‘1080 shock’ and become detectable (anecdotal accounts from 1080 operators in DOC and TB Free NZ suggest sub-lethally dosed animals avoid interactions during the initial recovery phase)
to ensure that any ‘near independent’ young that were still in the natal nest while the toxin was available would emerge and be of sufficient age to be detected
to ensure that any survivors would not be able to breed and birth independent young before the second application would be carried out (if a sufficient number of survivors were detected to trigger that response).
The 50-day window post-toxin application ended on 24 August 2017. However, the detection monitoring continued until 3 September (60 days post-toxin application) to enable completion of the final lap, following a weather delay.
The detection team was also required to log any observations of target or nontarget species e.g. carcasses of any animals, kea sightings.
RESULTS
Pre-operational monitoring for possum and rat presence and relative abundance
Possums were recorded at 51% of pre-operational monitoring devices, with all lines recording above 35%, suggesting a consistent abundance across the detection area (and presumably the treatment area). Rats were only recorded on 1 monitoring device, suggesting a very low abundance of rats in the treatment area.
Table 1: Pre-operational rat and possum indicative monitoring results, 8 to 16 June 2017
Pre-operational monitoring for kea presence:
The kea observational work found no flocks of juvenile kea. Only 3 kea ‘encounters’ (consisting of 2 adult and 3 sub adult birds, and all as a result of call playback) occurred during the survey. It was therefore decided by DOC that the toxin baiting could proceed (as per the approval secured for this trial).
During this kea observational work, the following native and exotic birds were also recorded as present:
Riversides: harrier hawk, magpie, fernbird, paradise duck, spur-winged plover, black shag, black-backed gull, pipit
Forest: tomtit, fantail, kereru, waxeye, brown creeper, riflemen (above 600m), kaka (heard), tui, kakariki (unknown sp.)
Post-operational possum and rat survivor detection:
Due to weather and subsequent river access restrictions, only six service laps of the detection area were able to be completed (i.e. one every ten days on average). However, given the nature of the detection devices, they were available to any surviving rats and possums throughout the detection period (55 days; to 3 September 2017).
Table 2: Type and numbers of detection devices deployed, and total number of detection nights per device and in total
Each service lap took approximately 120 person hours to complete. The total person hours for the entire detection period was 720 hours (or 90 person days, at 8 hours per day).
Throughout all those detection nights and person hours, no rat or possum were detected.
In that time, nine possum carcasses (and one ship rat carcass) were found by the field team.
Non-target species deaths
In the course of covering the detection core area, a small number of non-target species were found dead. No toxin residue testing was undertaken, with the assumption being that 1080 caused the deaths.
Table 3: Non-target species individuals found dead, and the date found.
Non-target species recorded
During the course of the detection monitoring, the field team recorded 31 instances of seeing or hearing kea within the block. A number of non-target species live encounters were also recorded on the motion-activated cameras throughout the detection core area, post-toxin application (i.e. individual seen in the footage upon review by the field team).
Table 4: Non-target species recorded alive on motion-activated cameras in the detection area, and the date of each recording (noting that camera detection began on 25 July 2017)
DISCUSSION
The complete removal of possums and rats appears to have been successfully achieved within the treatment area for at least 60 days post control. This is possibly the first time that aerial 1080 has been shown to successfully remove all rats and possums from a site.
While it is acknowledged that the rat numbers appeared to be very low in this trial, the possum numbers were high and therefore represent a ‘hard test’. In addition, it has been noted by local DOC staff that rat numbers can be variable between sites where they undertake predator control on the West Coast – therefore, the population dynamics (high possum numbers, low rat numbers) at Jackson-Arawhata might be similar to that elsewhere in the region and so the rat result in this trial does still have value.
The combination of no exclusion zones, double prefeeding, bait swath overlaps (i.e. no gaps in coverage), and higher than standard bait sowing rates are all considered to have contributed to this successful outcome. All of these factors are differences between the trial technique and the current standard technique for aerial 1080 operations for predator control.
A small number of non-target deaths were observed in this trial. This result is in line with that which might be expected from a standard 1080 operation. In addition, the non-target animals seen alive on the cameras during the detection period provides strong evidence that the impact on these species’ populations overall was very low. It is particularly encouraging that we did not observe any negative impact on kea during the trial.
The double prefeeding involved in this technique may increase the risk of bait acceptance by non-target species, but it does not appear to be a major increase in risk (based on the observations of this trial). This aspect may warrant further investigation.
Based on the results at Mt Taranaki, and the results in this trial, we believe that the 1080 to Zero prescription could be a technique to initially remove possums and rats from large-scale sites. ZIP is currently developing the other tools and techniques required to defend large-scale sites against predator invasion, utilising natural barriers or ‘virtual’ ones. In time, the repeated landscape-scale use of aerial 1080 may not be required in areas where we can completely remove the predators, using the ‘1080 to Zero’ aerial removal technique, and then maintain predator freedom in perpetuity through preventing reinvasion and responding to incursions. This aerial 1080 technique now needs to be tested at a scale that confirms its potential – we suggest this scale could be up to 20,000ha.
ACKNOWLEDGEMENTS
Thank you to the DOC South Westland team for their administrative, logistical, and operational support throughout this trial. DOC ‘Battle for our Birds’ staff and technical advisors provided advice and review, which helped to make this trial as robust as possible. Ripley Dean (team leader), Tim Sjoberg, Michael Tunnicliff, Nate St Hill, Chelsea Price, Cody Luckin, Matt Chisnall, Oscar Pollard, and Duncan Kay all undertook aspects of the field work during this trial. Nick Mulgan provided GIS support throughout the establishment, aerial baiting, and detection monitoring stages.
Thank you to James Russell and James Ross for their review of this report.
REFERENCES
Bell, P., Nathan, H., & Mulgan, N (2017) ‘Island’ eradication within large landscapes: the remove and protect model. Proceedings from Island Invasives Conference, Dundee. In press.
Elliott, G. & Kemp, J. (2016). ‘Large-scale pest control in New Zealand beech forests’. Ecological Management & Restoration 17:200–209.
Griffiths, J.W. & Barron, M.C. (2016). ‘Spatiotemporal changes in relative rat (Rattus rattus) abundance following large-scale pest control’. New Zealand Journal of Ecology 40:371–380.
McFadden, I. & Towns, D. (1991). Eradication campaigns against Kiore (Rattus exulans) on Rurima Rocks and Korapuki, Northern New Zealand. Science & Research Internal Report No 97. Wellington: Department of Conservation.
Moors, P.J. (1985). ‘Eradication campaigns against Rattus norvegicus on the Noises Islands, New Zealand, using brodifacoum and 1080’. ICBP Technical Publication 3:143–155.
Morgan, D.R. (2004). ‘Maximising the Effectiveness of Aerial 1080 Control for Possums (Trichosurus vulpecula)’. PhD thesis. Lincoln: Lincoln University.
Nugent, G., Morgan, D., Clayton, R. & Warburton, B. (2010). ‘Improving the efficacy of aerial poisoning of brushtail possums (Trichosurus vulpecula) through reduced fragmentation of bait’. International Journal of Pest Management 57:51–59.
Nugent, G., Twigg, L.E., Warburton, B., McGlinchy, A., Fisher, P., Gormley, A.M. & Parkes, J.P. (2012). ‘Why 0.02 %? A review of the basis for current practice in aerial 1080 baiting of rabbits in New Zealand’. Wildlife Research 39:89–103.
Nugent, G., Warburton, B., Thomson, C., Sweetapple, P. & Ruscoe, W.C. (2011). ‘Effect of prefeeding, sowing rate and sowing pattern on efficacy of aerial 1080 poisoning of small-mammal pests in New Zealand’. Wildlife Research 38:249–259.
OSPRI (2014). Our first year|2013/2014 OSPRI Annual Report and TBfree Research and Disease Control Report. Wellington: OSPRI.
APPENDIX ONE
Site of the trial, with map showing the aerial 1080 treatment area and the location of the detection area within it
APPENDIX TWO
Detection core area with track line lettering shown. The dots represent each 50m point along each line – tracking tunnels were deployed at every point, while chew cards were deployed at every second point along every line. [NB: the different colouring of dots was for internal use, as distance and line markers]
APPENDIX THREE
Detection area showing the location of the motion-activated cameras.