Individual study: The efficacy of feral cat Felis catus exclusion fence designs for threatened species protection at the Arid Recovery Project Reserve, South Australia
Moseby K.E. & Read J.L. (2006) The efficacy of feral cat, fox and rabbit exclusion fence designs for threatened species protection. Biological Conservation, 127, 429-437
Many government and private conservation organizations on mainland Australia rely on exclusion fences for the protection of threatened species from introduced non-native predators. Predation by feral cats Felis catus on native species such as the eastern barred bandicoot Perameles gunnii, mala Lagorchestes hirsutus, burrowing bettong Bettongia lesueur, greater bilby Macrotis lagotis and western barred bandicoot Perameles bougainville can be a major driver of population declines. Fence testing helps to maximize the effectiveness of fence exclosures for conservation purposes, however, most types of exclusion fence have not been trialed in Australia. Experimental pen trials were conducted to test the efficacy and cost effectiveness of cat exclusion fences at the Arid Recovery conservation project reserve, South Australia.
Study site: The Arid Recovery conservation project has removed feral cats Felis catus, European rabbits Oryctolagus cuniculus (see: www.conservationevidence.com/ViewSummary.aspx?ID=10421) and red foxes Vulpes vulpes (www.conservationevidence.com/ViewSummary.aspx?ID=10420) from a 60 km² fenced reserve in South Australia. Feral cat densities around the reserve were approximately 0.6-0.9 individuals per km². The habitat is dominated by lowland shrubland/chenopod scrub dominated by saltbush Atriplex spp. and bluebush Maireana spp. Three fence trials were set up in two test areas (20 x 20 m and 119 x 25 m), as well as a reserve-wide trial.
Cat capture and release: Feral cats were trapped from within 20 km of the trial pen. Trapped cats were weighed and most sexed (12 male, 15 female, 6 unknown) before being released into the trial pen, where their escape attempts were monitored. To establish whether the feral cats used were comparable to the local population, weights of cats used in the trials were compared with weights of 1,173 feral cats shot/trapped in the Roxby Downs region (an area adjacent to the reserve) between January 1998 and January 2004. There was no significant difference in weights recorded.
Trial 1: Trial 1 was conducted from March 1998 to August 1999.
Fence design - An L-shaped pen of 20 x 20 m was constructed, incorporating one external corner, the part of a fence usually targeted by animals attempting to access enclosures. Pine posts 1.8 m high (from ground level) were installed at 10 m intervals and at each corner of the pen. Seven selvage wires were attached to these posts at 30 cm vertical intervals. Wire mesh netting, 40 mm (mesh diameter) x 1.4 mm (wire thickness) x 900 mm (high) was clipped to them from ground level to 90 cm height using ring fasteners. A net foot apron (40 x 1.4 x 300 mm) was clipped to the ground level selvage wire, stretched into the enclosure to 300 mm, and buried. A non-ridged overhang was created by clipping wire netting (50 x 1.4 x 1200 mm) to the fence at a height of 90 cm, then vertically weaving 90 m lengths of fencing wire through the top 60 cm and bending it to an angle of 45º.
Fence improvements - Cat escape routes were monitored to identify where improvements could be made to the original fence design, leading to the development of nine designs (not all designs are discussed here but are described in the original paper). These incorporated additional elements including corrugated iron that was nailed to the fence posts to prohibit climbing, the incorporation of an extra 30 cm length of netting on the 45º corner overhangs, and the addition of electrical wires to deter the cats from scaling the fence. Metal posts (using bore casings to keep costs down) were also used.
Trial 2: Trial 2 was conducted from February 2004 to December 2004.
Fence design - A 119 x 25 m rectangular pen was constructed using the original design, plus corrugated iron around the fence posts and extra netting on the corners. The ‘floppy’ overhang was extended to 60 cm to form a rounded arc (Design 5 - material costs AUD$12,432/km). Within this rectangle, two perpendicular fences (Designs 8 and 9) were created around a central section of dense vegetation. Feral cats were released in the open outer sections and encouraged to breach the central fences where there was more vegetation cover.
Design 8 used 30 x 1.4 x 1200 mm netting, 30 cm buried as a foot apron with 90 cm held vertically. The top of the netting was reinforced with 75 mm of barbed wire, plus additional 10 cm lengths of PVC conduit strung 15 cm above. These freely rotating conduit lengths served to prevent feral cats from gaining a purchase to the upper wire if climbing or jumping onto the top of the fence.
Design 9 (material costs AUD$8814/km) was a shorter version (115 cm) of design 5. Steel droppers, 1.65 m high were spaced 10 m apart with five selvage wires strung at heights of 0, 30, 65, 90, and 115 cm. The netting (30 x 1.4 x 1200 mm), consisted of a 30 cm horizontal footing and 90 cm vertical netting. The top netting (50 x 1.4 x 900 mm) included 30 cm of vertical netting and a 60 cm ‘floppy’ overhang.
Trial 3 - in situ testing: Trial 3 was conducted from 1997 to 2005.
Design 5, plus two electric wires at heights of 1.3 and 1.6 m (Design 7) was used in the 14 km² external enclosure of the reserve. A minor alteration involved the creation of extra support for the electric wires by repositioning them adjacent to selvage wires at 1.2 and 1.5 m. The last feral cat was removed from the enclosure in February 1999, and incursions into the reserve were monitored between July 1999 and January 2005. The reserve’s 28 km perimeter fence utilized design 5 with steel posts, star droppers every 6 m and 120 cm wide netting on the bottom of the fence to create a 30 cm horizontal foot apron and a 90 cm vertical panel.
In January 2005, another part of the reserve, Red Lake, was fenced. This fence was created using design 9 plus steel droppers spaced every 7 m for extra rigidity of the 'floppy' overhang.
Cat escapes: The main methods of escape were climbing all the way from ground level, or a combination of jumping (1 to 1.7 m) onto the fence and climbing the rest. No cat escaped by digging under, chewing through or jumping clean over the top of the fence. Size and sex of the cats were not found to be significant factors in their escape capabilities during the pen trails.
Best fence features: The rounded arc 60 cm ‘floppy’ overhang successfully contained and excluded feral cats in the trials. The lower height fence (Design 9) of 1.15 m was a suitable barrier because the overhang curved back, made it difficult for the animals to jump over. The floppy nature of the overhang was not a factor in its success. However, ‘floppy’ netting was easier and cheaper to install than a rigid structure.
Steel posts were more effective than wooden ones, though large amounts of netting around these posts facilitated climbing. Because feral cats were able to climb up the wooden posts, a modification such as a wider overhang may be necessary when planning corners.
The non-electrified ‘floppy’ overhang (Design 5) proved just as effective as the electrified fence (Design 7) in the field trials. Electric fences with no overhang were ineffective at preventing access by cats as they were able to jump over the 30 cm high electrified wire and between wires placed more than 8 cm from the netting.
The 30 cm foot apron was effective at preventing cats from digging at the base of the fence.
The 30 mm (mesh diameter) hexagonal netting, which accounted for 57% of the cost of Design 9, proved unnecessary because the 40 mm (mesh diameter) netting was just as effective at barring cats and was less expensive. Material costs for design 9 were AUD$ 8,8814/km using 30 mm mesh but this was reduced to AUD$6939 when using 40 mm mesh.
Conclusions: In these trials, steel post, 40 mm mesh diameter non-electrified fences with a 60 cm rounded arc ‘floppy’ overhang, proved to be the cheapest and most effective feral cat exclusion fence design. Where netting joints do not overlap, it is recommended clips be placed at a maximum of 10 cm apart to prevent cats getting through any fence gaps.
Note: If using or referring to this published study, please read and quote the original paper.