«July 2008 ABSTRACT Concerns regarding nontarget and secondary poisoning from the use of compound 1080 to control browsing damage by native animals in ...»
Use of bait stations to minimise poisoning of nontarget herbivore species in Tasmania
M. Statham and Helen L Statham Tasmanian Institute of Agricultural Research, Mt Pleasant
Laboratories, PO Box 46, Kings Meadows, Tasmania 7249
Concerns regarding nontarget and secondary poisoning from the use of compound 1080 to
control browsing damage by native animals in Tasmania have led to public pressure to ban its
use. The use of specific bait stations combined with Feratox® (encapsulated cyanide) may allay public concerns and allow ongoing use of toxins as one method of control.
Animal access to a range of bait stations was tested with captive populations and in field trials using video and still cameras to collect data. The Kilmore bait station was found to be the most suitable for our conditions on the basis of ease of use, volume, and minimal spillage during feeding. Strategies were developed to allow target species to feed from the bait stations while minimising nontarget species access.
INTRODUCTIONIn 1949 Tasmania was the first State in Australia to consider the use of sodium monofluoroacetate (1080) as a pest control agent against the European rabbit (Oryctolagus cuniculus) (Statham 2005). Initial trials were in 1952 with widespread use soon after.
Reduction in rabbit numbers allowed for increased pasture development where practical, leaving poorer ground and rocky hills with native vegetation. This resulted in extensive pasture / bush interface and provided ideal conditions for native herbivores. Bennett’s wallaby (Macropus rufogriseus), Tasmanian pademelon (Thylogale billardierii) and brushtail possum (Trichosurus vulpecula) increased in abundance to become serious pests in agricultural and forest production areas and protocols to use 1080 for their control were developed. These species are partly protected under legislation and a permit can be sought if browsing damage has been proven. Under the permit a series of non toxic carrot free feeds are laid on the ground in areas where pest species are thought to be feeding. After 4 to 6 free feeds and when most of the free feed has been eaten, carrot dyed blue (Statham 1989) and mixed with 1080 (0.014%) is laid. While this is an effective method of control the required Bait Stations to Minimise Nontarget Poisoning Statham and Statham protocols make it time consuming and expensive. The use of poison to control native animals, its perceived effect on nontarget species, and occasional secondary poisoning of dogs, has led to public complaints and increasing opposition to 1080. Eventually the State Government adopted a plan to reduce 1080 usage with a phase out by 2015 (Community Leaders Group 2001). In 2004 the Federal Government offered $4 million to develop and implement a coordinated research and demonstration program into practical, effective and financially viable alternatives to 1080 for control of browsing native animals on private forest and agricultural land in Tasmania.
Bait stations have been used in New Zealand since the early 1990s for introduced possum control (Thomas and Hickling 1995) and are now routinely used there in pest control operations (Thomas et al. 2003). They can be made relatively species specific, are weatherproof and can be used in the field for a long period (Morgan 2004). Although the bait stations have only been used for possum control, anecdotal evidence suggests that introduced wallaby species will feed from at least some of them as individuals have been found dead near bait stations that used Feratox as the poison (S Boot, personal communication).
Feratox has been developed in New Zealand to overcome the problems of bait shyness (Warburton and Drew 1994), handling, and risk to native birds (Spurr 2000) which occurred with traditional cyanide paste used in possum control. Feratox is a pea-sized encapsulated pellet containing enough cyanide to give a possum a toxic dose. The pellet cannot be broken by hand but cyanide is released when the pellet is broken by chewing. Once released from the pellet the cyanide is rapidly degraded. The advantages of cyanide over 1080 are its humaneness (Gregory et al. 1998) and low risk of secondary poisoning (Eason and Wickstrom 2001). Recent New Zealand pen trials have shown that Bennett’s wallaby can also be rapidly killed by Feratox pellets (C Eason, personal communication).
This trial was designed to evaluate the possibility of using bait stations to allow target species access to bait while excluding nontarget species. The aim was to evaluate bait stations for their accessibility to the 3 target species, Bennett’s wallaby (head and body 800 mm, weight 16 kg), pademelon (head and body 600 mm, 8.5 kg) and possum (head and body 500 mm, 3.8 kg) (Green, 1973) while excluding non target species, principally Tasmanian bettong (Bettongia gaimardi), Long-nosed potoroo (Potorous tridactylus), Forester kangaroo (Macropus giganteus) and birds.
The Tasmanian bettong and Long-nosed potoroo are the 2 species of rat-kangaroos present in Tasmanian agricultural and forest areas. Neither is considered a pest species as their diet is primarily hypogean fungi (Claridge et al. 2007). Both are wholly protected species under 2 Bait Stations to Minimise Nontarget Poisoning Statham and Statham Tasmanian Legislation. They are of similar head and body size, 350 mm and 360 mm with weights of 2 and 1.6 kg respectively, with the bettong having longer legs and tail (Green 1973). Home range sizes have been estimated at up to 63 ha and 85 ha for female and male bettongs (Taylor 1993) and up to 11 ha 34 ha for female and male potoroos.(Kitchener 1973).
In addition, Mooney and Johnson (1979) reported that a bettong may travel up to 2 km to and from a feeding site in any given night.
MATERIALS AND METHODSStudy Areas Compound trials We conducted preliminary trials in a 0.85 ha compound containing breeding populations of Bennett’s wallaby and pademelon at Mt Pleasant Laboratories, Launceston, Tasmania (41.5°S, 147.1°E). Rye grass pasture with clumps of Phalaris sp. dominated the ground cover with scattered prickly box (Bursaria spinosa) shrubs and mature black peppermint (Eucalyptus amygdalina) trees.
Field trials We conducted further trials at 5 locations within an hour’s drive of Launceston, namely Bayview, Maitland, Security Road, Bloomfield, and Tasmania Zoo. It was not possible to test the response of non target species in captive trials because of the need to provise artificial diets forthese species in captivity. The aim was to trial feeders using an artificial bait in natural surroundings.
Bayview is a private pastoral property in the Tamar Valley (41.2°S, 146.9°E) with clay soils derived from a dolerite base. The trial was at the native bush and improved pasture interface.
The south facing native bush was dominated by tall E. amygdalina and E. viminalis with a sclerophyll under storey. Animal runways between bush and pasture were visible.
Maitland is a private pastoral property in the Tasmanian midlands (41.9°S, 147.3°E). The trial was in Greenhill Bush paddock, an area known for its population of bettongs. The dry sclerophyll forest on lateritic gravel was dominated by E. amygdalina with a sparse under storey of low heath species.
The State Forest E. nitens plantation at Security Road, north of Launceston (41.1°S,
147.1°E), on clay soils derived from mudstone, was also chosen for its known bettong population. Rows of E. nitens, planted on old pasture, were interspersed with drainage areas of grass, native shrub species and introduced weeds.
3 Bait Stations to Minimise Nontarget Poisoning Statham and Statham Bloomfield and Tasmania Zoo were both chosen for their populations of forester kangaroos.
Bloomfield is a private pastoral property near Ross in the Tasmanian midlands (42.0°S,
147.5°E). The clay soils were derived from dolerite and the area was in drought conditions having not received its average annual rainfall of 450 mm for several years. The trial was on the dry east facing slopes at the native bush and pasture interface. The sparse native bush was dominated by E. viminalis with an open native grass and sedge ground cover and the dry overgrazed pasture was dominated by Phalaris sp. Tasmania Zoo is a private wildlife park on the outskirts of Launceston (41.4°S, 146.9°E). The trial site was in a paddock with captive forester kangaroos. It was dominated by E. amygdalina with other native tree species and an under story of native grass species.
Methods Six different New Zealand designed bait stations were used. Three were commercial products used to dispense dry bait. The first was a plastic bait station with a lift up lid, opening of 60 mm diameter and capacity of 120 g (Ferafeed, Connovation Ltd.). The second and third were plastic boxes (Sentry and Kilmore Bait Stations, Pest Control Research).
Sentry had a capacity of 200 g and an opening of 85 x 70 mm. Kilmore had a capacity of 1.5 kg, an opening 85 x 90 mm, and a baffle to direct feed behind the opening. The fourth bait station was a 15 x 75 x 20 mm edible potato starch box, capacity 20 g with an open front (Striker, Connovation Ltd.). The final two bait stations were experimental ones developed by Landcare Research New Zealand to hold a single Feratox pellet. One was a hanging device (Hanger) and the other designed so that a spring loaded outer sleeve slides down to expose free feed and a toxic pellet when a possum holds it with its front paws and pulls down (Spring).
Compound Trials Between April and August 2007 the 6 bait stations were tested in the wallaby compound at Mt Pleasant Laboratories using infrared activated video (DVR Eye, Pixcontroller, PA, USA) and still cameras (Sony Cyber-shot DSC-S40 with Pixcontroller infrared detectors, BEAST digital camera traps, Tasmania, Australia) to monitor animal behavior and to see if Bennett’s wallabies and pademelons could and would feed from them. At times the number of working cameras available limited the number of feeder stations which could be monitored.
Commercial wallaby pellets (Barastock, Ridley Agri Products, Victoria, Australia) were used in the box style bait stations, viz Ferafeed, Sentry, and Kilmore and peanut butter in Spring
5 Bait Stations to Minimise Nontarget Poisoning Statham and Statham use the same height for all feeders at a set time as the size of bettong and potoroo home ranges would have meant that an individual could have accessed the lowest level feeder in any trial. This could have masked the fact that animals may have been able to reach a higher feeder but did not while more accessable pellets were available in the feeding range.
At Bayview 10 Kilmore bait stations were tested along the paddock / native bush edge at 350 mm height for 11 nights and at 450 mm for 3 nights.
At Maitland 10 Kilmore bait stations were used and they were spread throughout the bush set at 450 mm above ground for 9 nights.
At Security Rd. Kilmore bait stations in the E. nitens plantation were tested at varying heights. First, 9 bait stations each at 450 mm for 8 nights and at 510 mm for 3 nights, then 8 bait stations at 550 mm for 7 nights and at 600 mm for 4 nights. Finally we tested 9 door bait stations at 450 mm for 17 nights.
At Bloomfield 9 Kilmore bait stations at 600 mm and 2 Kilmore door bait stations at 500 mm for were trialed for14 days and at Tasmania Zoo we tested 3 Kilmore bait stations at 500 mm for 6 nights.
Photograph Recording and Analysis
Only photos and videos containing animal images were recorded. They were classified as:
Near the bait station:-Within camera range At the bait station:-Apparently actively investigating the bait station, within 10 cm.
Feeding:-Head actually in the bait station.
These groupings weren’t exclusive or quantitative. It was obvious from results where a video and still camera were used at the same bait station that there was an overlap in the classifications as some animals recorded at the bait station by the still camera had or subsequently did take a mouthful of feed while the camera was powered down. Although video clips varied in length from less than a minute to several minutes we only scored each clip once. For each clip we recorded an animal as feeding rather than near or at the bait station if it fed at any stage during the clip regardless of the length of the clip or the proportion of time it actually spent feeding. Also there was often no way to differentiate between different animals feeding and a single animal returning repeatedly, so records could only be of the presence of an animal.
There were several problems with replication which resulted in this trial not being statistically analysable as the number of records differed between cameras. Localised shade and vegetation conditions meant that position and sensitivity of individual cameras often had to be adjusted during a trial, affecting the numbers of photos recorded. Other inconsistencies 6 Bait Stations to Minimise Nontarget Poisoning Statham and Statham were caused by weather. For example in very cold weather camera batteries lost power more quickly and in windy weather a lot of daylight photos and clips were taken triggered by moving vegetation and changes in cloud cover. Also cameras sometimes failed for unexplained reasons.
In the field animal visitations to bait stations increased with the time the feeders were in position so it was not possible to replicate trials in the same area. Replication in different areas was also a problem because different populations of the same species behaved differently to the presence of bait stations.
Feeding from a Striker was defined as occurring when an animal had its mouth on top of the feeder. With the Spring the animal had to be pulling or pushing down on the device to be recorded as feeding. When testing Hangers an animal was defined as feeding if its paws were on the feeding device.
RESULTS Overall there were over 20,000 photos or video clips containing animals recorded and analysed in compound trials and 37,500 in the field.
Compound trials For these trials only data relating to Bennett’s wallaby and pademelon feeding are given.
Ferafeed.—At a bait station height of 250 mm on average there were 8.7 Bennett’s wallaby and 5.3 pademelons feedings per night and at 500 mm there were 24 and 4.3 respectively.
Spillage was 70.6 (± 24.0) g of 185 g offered (38.2%).