Study

The efficacy of feral cat, fox and rabbit exclusion fence designs for threatened species protection

  • Published source details 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.

Summary

Study 1: Cat exclusion

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.

Study 2: Fox exclusion

Many government and private conservation organizations on mainland Australia rely on exclusion fences for the protection and reintroduction of threatened species. Predation by red foxes Vulpes vulpes 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 exclosure design, 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 fox exclusion fences at the Arid Recovery conservation project reserve.

Study site: The Arid Recovery conservation project has removed red foxes Vulpes vulpes, European rabbits Oryctolagus cuniculus (see: www.conservationevidence.com/ViewSummary.aspx?ID=10421) and feral cats Felis catus (www.conservationevidence.com/ViewSummary.aspx?ID=10419) from a 60 km² fenced reserve. Fox densities around the reserve at the time of the fence trials were approximately 0.6-0.9 individuals per km². The habitat is lowland shrubland/chenopod (dominated by saltbush Atriplex spp. and bluebush Maireana spp.). Two trials were set up, one (Trial 1) in a 119 x 25 m test area, and the second (Trial 2) on a much larger, reserve-wide scale.

Trial 1: Trial 1 was conducted from February to December 2004.

Fox capture & release - Foxes were caught using soft rubber leg-hold traps set around the reserve perimeter and upon capture, quickly released into the pens. Only uninjured foxes were used in the trial. Released animals were observed until their escape or retreat into the shelter of vegetation cover. Dents, rips and fur caught on the netting were used, as well as direct observation, to identify the exact escape points.

Fence design - A 119 x 25 m rectangular pen was constructed using pine posts 1.8 m high (from ground level) spaced at 10 m intervals, and at each corner of the pen. Corrugated iron was nailed around the fence posts to inhibit foxes from scaling them. Seven selvage wires were attached to these posts at 30 cm intervals. Wire mesh netting (40 mm mesh diameter) x 1.4 (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 exclosure to 30 cm, and buried. A non-rigid overhang was then 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 cm lengths of fencing wire through the top 60 cm and bending it to form a complete arch. Extra netting was added to reinforce the corners. (This was the fifth of a total of nine designs used in fence trials - see also Cases 221 and 223).

Within the pen, two perpendicular fences (designated 'Design 8' and 'Design 9') were fitted around a central section of dense vegetation. Foxes were released in the open outer sections and encouraged to breach the central fences to gain greater cover which was present on the opposite side of the fence.

Design 8 used 30 x 1.4 x 1200 mm netting, with 30 cm buried as a foot apron and 90 cm erected 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 foxes from gaining a purchase to the upper wire if climbing or jumping onto the top of the fence.

Design 9 was a shorter version (115 cm tall) of design 5 (see original paper). Steel droppers, 1.65 m high were spaced at 10 m intervals with five selvage wires strung at heights of 0, 30, 65, 90, and 115 cm. The netting was 30 x 1.4 x 1200 mm, consisting 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 2: The reserve scale trials were conducted between 1997-2005. The same design as the 119 x 25 m rectangular pen was incorporated into the 14 km² external fence of the reserve. However, two electric wires at heights of 1.3 and 1.6 m were added to this fence (which was designated 'Design 7'). A minor alteration was needed to add extra support for the electric wires by repositioning them adjacent to selvage wires at 1.2 and 1.5 m. The last fox was removed from the exclosure in February 1999, and incursions into the reserve were monitored between July 1999 and January 2005.

In January 2005, another part of the reserve, Red Lake, was fenced. The fencing followed design 9 but with the addition of steel droppers spaced every 7 m to give extra rigidity to the ‘floppy’ overhang.

Effectiveness of fence designs:
Design 7 - As of July 2005, no fox has been seen within the 14 km² external fence of the reserve for six years. A total of 130 foxes have been caught in leghold traps set outside the perimeter between 1999 and 2004. This suggests that foxes come into regular contact with the fence but are unable to scale or dig under it.

Design 8 - this fence design proved ineffective, as two male foxes escaped within two minutes of release. One pushed itself through a 75 mm gap between the barbed wire and the top of the netting. The other climbed over the rollers at the corner of the fence.

Design 9 - 27 foxes were tested in design 9. Biting and ramming were the most attempted ways of escape. However, no fox escaped, whether by digging under, chewing through or jumping on top of the fence.

Best fence features: The rounded arc 60 cm ‘floppy’ overhang successfully contained and excluded foxes in the trials. The lower height fence (Design 9) of 1.15 m was a suitable barrier because the overhang curved back, making it difficult for the animals to jump over it. 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 (recycled bore casings were used to minimise costs) were more effective than wooden ones, though large amounts of netting around the posts facilitated climbing. Foxes 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 in situ field trials.

The 30 cm foot apron was effective at preventing foxes 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 effective at barring foxes. Material costs using 30 mm diameter wire mesh was AUD$8,814/km, this was reduced to AUD$6,939/km using 40 mm diameter wire mesh.

Conclusions: In these trials, steel post, 40 mm diameter non-electrified fences with a 60 cm ‘floppy’ overhang proved to be the cheapest and most effective fox exclusion fence design. Where netting joints do not overlap, it is recommended that maximum spacing of clips to join net sections is 10 cm to prevent foxes from pushing through gaps and escaping. Electric fences that have no overhang above them will be ineffective at deterring foxes. The foxes were able to jump over the 30 cm high electrified wires and between wires placed more than 8 cm from the netting.

 

Study 3: Rabbit exclusion

Many government and private conservation organizations on mainland Australia rely on exclusion fences for the protection and reintroduction of threatened species. Competition from introduced European rabbit Oryctolagus cuniculus populations with 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 driver of population declines. Fence testing helps to maximize the effectiveness of fence exclosures, 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 rabbit exclusion fences at the Arid Recovery conservation project reserve in South Australia.

Study site: The Arid Recovery conservation project has removed European rabbits Oryctolagus cuniculus, feral cats Felis catus (see: www.conservationevidence.com/ViewSummary.aspx?ID=10419) and red foxes Vulpes vulpes (www.conservationevidence.com/ViewSummary.aspx?ID=10420) from a 60 km² fenced reserve. Rabbit densities around the reserve can exceed 200 individuals per km². The habitat is lowland shrubland/chenopod scrub dominated by saltbush Atriplex spp. and bluebush Maireana spp. The rabbit fence trial was undertaken from March 1998 to August 1999.
 
Fence design: An L-shaped pen of 20 x 20 m was constructed incorporating one external corner. Pine posts 1.8 m tall (from ground level) were installed at 10 m intervals and on each corner of the pen. Wire mesh netting 40 mm (mesh diameter) x 1.4 mm (wire thickness) x 900 mm (high) was clipped from ground level to 90 cm height using ring fasteners. A net (40 x 1.4 x 300 mm) foot apron was clipped to the ground level selvage wire, stretched into the exclosure to 30 cm and buried.

Rabbit observations: Two juvenile rabbits were caught by hand and placed in the enclosure. Video observations were made as they tried to breach the wire mesh fencing.

Manual testing: The effectiveness of three different mesh sizes (30, 37 and 40 mm diameter) was tested by trying to pull 21 dead juvenile rabbits, weighing from 148 to 918 g, through the netting.
 
Rabbit observations: The 40 mm mesh diameter netting was not adequate to exclude juvenile rabbits. The two placed within the exclosure escaped within 30 seconds of release by squeezing through the netting. One of these rabbits repeatedly jumped in and out of the pen through the netting.

At least 12 other rabbits were able to gain access to the exclosure by squeezing through the netting. Once additional 30 mm diameter netting was added to the fence, no rabbits breached the netting over the five year study period.

Manual testing: Of the 21 juvenile rabbits used, only the ones weighing up to 495g could be pulled through the 40 mm diameter netting. Only one small juvenile rabbit was successfully pulled through the 30 mm diameter netting.

Fence features: The 30 cm foot apron was effective at preventing rabbits from digging at the base of the fence. However, wider netting or heavy rubber matting may be necessary to stop rabbit incursions in areas of softer substrate such as sand dunes or watercourses.

Netting cost: The 30 cm mesh diameter hexagonal netting was very expensive and accounted for 57% of the fence cost. Fence contruction expense could be reduced by adding netting to existing stock fences.

Conclusions: Trials at the Arid Recovery conservation project reserve suggest that a combination of 30 mm mesh diameter wire netting and 30 cm horizontal foot aprons are necessary when trying to prevent rabbit incursions into a conservation area. Where netting joints do not overlap, it is recommended that clips be added, spaced at a maximum of 10 cm to prevent rabbits from squeezing through gaps.


Note: If using or referring to this published study, please read and quote the original paper.

Output references
What Works 2021 cover

What Works in Conservation

What Works in Conservation provides expert assessments of the effectiveness of actions, based on summarised evidence, in synopses. Subjects covered so far include amphibians, birds, mammals, forests, peatland and control of freshwater invasive species. More are in progress.

More about What Works in Conservation

Download free PDF or purchase
The Conservation Evidence Journal

The Conservation Evidence Journal

An online, free to publish in, open-access journal publishing results from research and projects that test the effectiveness of conservation actions.

Read the latest volume: Volume 21

Go to the CE Journal

Discover more on our blog

Our blog contains the latest news and updates from the Conservation Evidence team, the Conservation Evidence Journal, and our global partners in evidence-based conservation.


Who uses Conservation Evidence?

Meet some of the evidence champions

Endangered Landscape ProgrammeRed List Champion - Arc Kent Wildlife Trust The Rufford Foundation Save the Frogs - Ghana Mauritian Wildlife Supporting Conservation Leaders
Sustainability Dashboard National Biodiversity Network Frog Life The international journey of Conservation - Oryx Cool Farm Alliance UNEP AWFA Bat Conservation InternationalPeople trust for endangered species Vincet Wildlife Trust