Effective rabbit control in pine-buloke woodlands

Introduction:

Hattah-Kulkyne National Park was declared in 1980 and is a highly rabbit-prone environment. Before the arrival of rabbit haemorrhagic disease virus (RHDV), conventional control had reduced rabbit abundance to between two and eight rabbits per spotlight kilometre. After RHDV became established in 1996, spotlight counts dropped to fewer than one rabbit per kilometre. In areas where warrens were ripped, spotlight counts remained below 0.5 rabbits per kilometre.

This low level of rabbit abundance triggered a regeneration response from buloke (Allocasuarina leuhmannii)¹ — the first regeneration of this plant species recorded for at least 50 years. After 2004, the rabbit population recovered and a project was set up to demonstrate the most effective combination of conventional control techniques to protect regenerating native vegetation (Figure 1).

Aim:

To compare the effectiveness of two rabbit control strategies — one involving 1080 baiting followed by warren ripping and fumigation, and the other strategy using only ripping and fumigation.

Partners and management:

The project was a partnership between Parks Victoria, the Invasive Animals Cooperative Research Centre, the Victorian Department of Primary Industries and specialised contractors including Daryl Walters (Mallee Land Management Services), Jack Pryse (ripping contractor), Les Lidden (fumigation contractor) and Frann Sette (Sunrise 21 GIS services). Funding was provided by an Australian Government Caring for our Country grant.

Process:

Parks Victoria set up a demonstration site within the Hattah-Kulkyne National Park (see map at right). The site is an area of mixed pine-buloke woodland and riverine (black box) woodland within the Kramen management unit of the park. The western portion of the trial area (993 ha) contained six 50 ha treatment blocks. Two treatments were randomly allocated to the blocks and the areas surrounding each block. The two treatments were: 1) poison bait + rip + fumigate and 2) rip + fumigate. The eastern portion of the trial area (422 ha) was the experimental control area with no rabbit treatments applied. Three 50 ha blocks were set up in the control area for monitoring purposes. Outcomes were measured in terms of density of active warren entrances and spotlight counts.

The site generally had high levels of rabbit activity before treatment, as there had been no rabbit control works in the area for the previous five years. Warrens across the entire area were mapped before ripping and spotlight counts were completed. Other details recorded were: the total number of warren entrances, the number of active entrances and the situation of the warren (eg near tree, under log, in open area). Ongoing monitoring was done using spotlight counts and sampling of a subset of known warrens: 30 within each of the nine 50 ha blocks. The no-treatment site was only maintained for six months, after which time it was treated as part of the overall park control program.

Three of the six treatment blocks were poison baited over a two-week period during May 2009. Oat baits containing 1080 were used for poisoning rabbits after a round of three ‘free feeds’ (oats with no poison) had been laid at three-day intervals. Sterile oats were used to avoid residues germinating in the park and were dyed green to minimise bait take by non-target animals. Oats were laid during the afternoon, again to reduce non-target impacts, at a rate of 4 kg/km. A total of 60 km of bait trail was laid at a rate of about 20 linear km/100 ha. Trails were carefully placed in rabbit feeding areas and around warrens to ensure rabbits easily found baits. Dead rabbits were collected from the trail on the first and second mornings after poison baiting. On the third day, bait trails were covered with soil using a quadbike with a small sled attached.

A contractor ripped warrens in all the treatment blocks and immediate surrounds over two weeks in late June 20092. Two 140-horsepower log skidders were used; each with four rear-mounted tines 600 mm in length to ensure that ripping was deep enough to destroy each warren. The tines were modified at the tip with a ‘winged keel’ design that improved the lateral extent of the rip and aided manoeuvring (Figure 2). After each warren was cross-ripped using a zigzag pattern, the front blade on the skidder was used to smooth the ground surface.

The contractor recorded the location of each ripped warren using a GPS and the waypoints were downloaded at the end of the day. A single observer (or ‘spotter’) on a motorbike accompanied the contractor, searching for and marking warrens to be ripped. Any warrens that were inaccessible for ripping, or that were re-opened after ripping, were fumigated using chloropicrin gas in a pressure fumigator. A contractor was hired to carry out fumigation over six days in July 2009. The contractor’s dogs flushed rabbits into warrens and located rabbits in nearby fallen logs. A total of 326 warrens were located and fumigated2. Subsequent monitoring showed a high level of re-openings in Block 5 of the treatment area. This area was refumigated two weeks later and a further 16 warrens were treated.

Key finding

Best-practice rabbit control involves a strategic approach using multiple techniques to effectively reduce rabbit populations and their impacts.

Features of the study:

Key features of the project included training opportunities for staff onsite, and collection of detailed information to estimate comparative cost effectiveness of rabbit control strategies. A field day was also organised to demonstrate the effectiveness of integrated rabbit control to community and other interest groups.

Technical issues that were overcome included:

Emus were observed removing oats from the trail — to reduce bait uptake by non-target species, a light covering of soil was placed over the oats.
Some warrens were situated close to trees — to minimise damage to tree roots, warrens were ripped radially away from the tree.
Large logs prevented access to some warrens — the use of log skidders ensured that nearly all warrens were accessed, since such large obstructions could be pushed aside.

Results:

Overall, each strategy resulted in a major reduction in the rabbit population (Figure 3), and met the park’s management goal of 0.5 rabbits per hectare to allow plant regeneration (Figure 4).

Specific results from the study include:

Significant bait uptake was recorded on the second night after poisoned baits were laid. Some dead rabbits (42 individuals) were picked up off the trail but almost all rabbits died within warrens. No non-target deaths were recorded².
The total area ripped was 1227 ha. 2601 warrens were ripped over 11 days of operations by the two machines². This equated to an average of 2.1 warrens/ha.
342 warrens were fumigated in the western part of the study area: an average of 0.34 warrens/ha.
There was no measurable difference in spotlight counts between the poison-baited and non-baited sites one month after treatment. However, spotlight counts six months after treatment showed the number of rabbits per spotlight kilometre had dropped by 97% in the poison-baited area and by 87% in the non-baited area, compared to counts done before treatment was applied2. In addition, the treatment that included poisoning only left 0.8% of warren entrances open compared to 4.3% (five times as many) in the non-poisoned treatment. So, poisoning removed more rabbits in the early stages of control, leaving fewer rabbits to breed.
Six months after treatment, the level of rabbit activity at each site showed that populations had recovered at the same rate (ie regardless of the treatment). However, rabbits on the blocks that were poisoned as well as ripped and fumigated were going to take a longer time to recover to numbers where they inhibit plant regeneration.

Figure 1: A Parks Victoria ranger monitors the damage caused by rabbits to a young native pine. Photo: Brian Cooke.

Figure 2: Log skidder with modified tines. Photo: Peter Sandell.

Figure 3: Estimated number of active warren entrances per hectare in the areas of the study site before and after treatment2. The threshold line indicates the level at which rabbits cause loss or severe damage to buloke seedlings1.

Figure 4: A Parks Victoria ranger inspects a newly regenerated buloke in Hattah-Kulkyne NP. Photo: Peter Sandell.

What didn’t work:

Recent rainfall meant that soil moisture conditions were not ideal for warren ripping².

Conclusion:

Best-practice rabbit control involves a strategic approach using multiple techniques to effectively reduce rabbit populations and their impact on vegetation over the long term. Parks Victoria is now managing the time between rounds of treatments such that rabbit populations can be maintained below the threshold at which damage occurs. Including poisoning in the treatment regime (ie poison + rip + fumigate, rather than just rip + fumigate) means a longer time can pass before subsequent treatment rounds are needed. Poisoning with 1080 adds significant cost to initial rabbit control but is likely to be cost effective in the longer term. This approach provides a good basis for the development of a feasible, effective rabbit control program for the Hattah-Kulkyne National Park and may also be applicable to other areas.

Table 1. Costs of rabbit control

References:

Murdoch FA, McPhee SR and Cooke BD (2008). How Many Rabbits Are Too Many? Monitoring Grazing Pressure and Regeneration of Semi-Arid Woodlands. Unpublished report.
Sandell P (2010). Final Report to Invasive Animals Ltd: Demonstration of Effective Rabbit Control in Pine-Buloke Woodlands within Hattah-Kulkyne National Park, NW Victoria. Parks Victoria, Melbourne, Victoria.
Sandell P (2002). Implications of rabbit haemorrhagic disease for the short-term recovery of semi-arid woodland communities in north-west Victoria. Wildlife Research 29:591–598.

Acknowledgements:

Thanks to Peter Sandell (Parks Victoria) and Brian Cooke (IA CRC) for providing information for this case study.

More information:

Pestsmart Toolkit for European rabbits

Navigate our sites