Feral Pigs:Pest Status and

Prospects for Control

Similar works published by The CooperativeResearch Centre for Tropical Rainforest Ecologyand Management:

Patch Deaths in Tropical Queensland Rainforests:association and impact of Phytophthora cinnamomi

and other soil borne organisms.Paul Gadek

Remote Sensing Requirements for ManagementAgencies Responsible for Forest and Water QualityMonitoring in the Wet Tropics.Stuart Phinn, Michael Stanford, Alex Held

The Comparative Assessment of Arthropod and TreeBiodiversity in Old-World Rainforests: The RainforestCRC / Earthwatch Protocol Manual.Roger Kitching, Guy Vickerman, Melinda Laidlaw,

Karen Hurley

Impacts of Roads and Powerlines on the Wet TropicsWorld Heritage Area.Miriam Goosem, Steve Turton

Impacts of Visitation and Use: Psychosocial andBiophysical Windows on Visitation and Use in theWet Tropics World Heritage Area.

Joan Bentrupperbäumer, Joseph Reser

Feral Pigs:Pest Status and

Prospects for Control

Proceedings of a Feral Pig WorkshopJames Cook University • Cairns

March 1999

Editor: C. N. Johnson

Cooperative Research Centre forTropical Rainforest Ecology

and Management

Cooperative Research Centre for TropicalRainforest Ecology and Management.

ISBN 086443 6815

This work is copyright. The Copyright Act 1968permits fair dealing for study, research, newsreporting, criticism or review. Selected pas-sages, tables or diagrams may reproducedfor such purposes provided acknowledge-ment of the source is included. Major extractsof the entire document may not be reproducedby any process without written permission ofthe Chief Executive Officer, CRC for TropicalRainforest Ecology and Management.

Published by the Cooperative ResearchCentre for Tropical Rainforest Ecology andManagement.

Further copies may be requested from theCooperative Research Centre for TropicalRainforest Ecology and Management, JamesCook University, PO Box 6811 Cairns, QLD,Australia 4870.

This publication should be cited as:Johnson, C.N. (Ed). (2001) Feral Pigs: Pest

Status and Prospects for Control. Proceed-

ings of a Feral Pig Workshop, James Cook

University, Cairns, March 1999. CooperativeResearch Centre for Tropical RainforestEcology and Management. Cairns. (75pp)

June 2001

v

The Feral Pig Workshop and production of this Proceedings were undertaken with the support of the sponsors listed below:

CONTRIBUTORS

Mr Brad Dorrington

Land Protection Branch, Queensland Department of Natural Resources and Mines

Dr John HerbohnSchool of Natural Systems Management, The University of Queensland

Dr Mike HollandCRC for the Biological Control of Pest Animals

Dr Jim HoneFaculty of Applied Science, University of Canberra

Dr Jack GilesDirector, Scientific Policy and Research, Zoological Parks Board of NSW

Dr Chris JohnsonDepartment of Zoology and Tropical Ecology, James Cook University

Mr Jonathan LeeAustralian Quarantine Inspection Service

Mr Reece LuxtonLand Protection Branch, Queensland Department of Natural Resources and Mines

Dr John McIlroyRetired, CRC for the Biological Control of Pest Animals

Mr Jim MitchellTropical Weeds Research Centre, Queensland Department of Natural Resources and Mines

Dr Scott RitchieTropical Public Health Unit

Dr Bob SeamarkChief Executive Officer, CRC for the Biological Contol of Pest Animals

Mr Jack ShieldQueensland Department of Primary Industries

Mr Trevor StanleyTropical Weeds Research Centre, Queensland Department of Natural Resources and Mines

Professor Nigel StorkChief Executive Officer, CRC for Tropical Rainforest Ecology and Management

Dr Karl VernesDepartment of Zoology and Tropical Ecology, James Cook University

vii

CONTENTS

CONTRIBUTORS ..... vii

INTRODUCTION ..... 1

C. N. Johnson and N. E. Stork

PART 1 - IMPACT OF FERAL PIGS ..... 5

Overview of the impact of feral pigs, Sus scrofa, ..... 7on the Australian EnvironmentJ. C. McIlroy

Economic impacts of feral pigs in the Wet Tropics ..... 11R. Luxton

Feral Pigs - Indigenous Perspectives ..... 14C. Roberts, R. Aken, A. Taisman and J. Chevathon

PART 2 - FERAL PIGS AND DISEASE THREATS ..... 17

The Role of the Northern Australian Quarantine Strategy ..... 19J. Lee

Japanese encephalitis - a case study of exotic animal disease incursion ..... 21J. Lee

Mosquitoes and Pigs: the critical mix for Japanese encephalitis ..... 27S. Ritchie, C Johansen, A. Van den Hurk, J Lee, and J. MacKenzie

Veterinary status of pigs on Cape York ..... 30J. Shield

COLOUR PLATES ..... 33

ix

PART 3 - POPULATION BIOLOGY AND CONTROL ..... 37

The dynamics of feral pig populations in Australia:Implications for management ..... 39J. R. Giles

Ecology of feral pigs in the Wet Tropics ..... 43J. Mitchell

Monitoring impacts and control of feral pigs:A case study in Namadgi National Park, A.C.T. ..... 48J. Hone

The effectiveness of trapping in reducing pig abundancein the Wet Tropics of north Queensland ..... 51K. Vernes, C. N, Johnson and J. Mitchell

PART 4 - IDENTIFYING MANAGEMENT OPTIONS ..... 57

Community based feral pig trapping in the Wet Tropics .... 59B. Dorrington

Prospects for pig control: A biotechnology perspective ..... 63R. F. Seamark

Fertility control for feral pigs: Options and status ..... 65M. K. Holland

Control of feral pigs in the Wet Tropics ..... 68

N. E. Stork and T. Stanley

REFERENCES ..... 71

INTRODUCTION

JOHNSON, C.N.1,2 and STORK, N.E.2

1Department of Zoology and Tropical Ecology, James Cook University, Townsville2Cooperative Research Centre for Tropical Rainforest Ecology and Management

Introduction 1

This volume reports the proceedings of the FeralPig Workshop held at James Cook University inCairns in March 1999 and hosted by the Rainfor-est CRC and the CRC for the Biological Controlof Pest Animals. Contributors to the workshoprepresented the range of issues related to feralpig management in North Queensland. It isprobably true that the feral pig problem is morecomplex in this region than anywhere else inAustralia. Consider first that the Wet Tropics WorldHeritage Area (WTWHA) preserves some of therichest, most complex and ancient ecosystemsin Australia. Feral pigs may be the most signifi-cant, and are certainly the most visible, introducedanimal species in this ecosystem. Their distur-bance of soil may affect ecosystem processes inthe WTWHA, they prey upon and may competewith a range of native plant and animal species,and they almost certainly contribute to the spreadof weeds and exotic fungi. There is some evidencethat pigs have caused the spread of feral earth-worms, and theyimpact on seedling set. Rootingby pigs of soil along roadsides and streams, andthe sight of pigs themselves, is a significant aes-thetic impact in some of the most pristine andbeautiful environments in Australia.

Neighbouring the Wet Tropics World HeritageArea are a range of valuable crops that suffereconomic damage from pigs. The economicimpacts of this interaction are significant inthemselves, but they also create tensions overthe management of the WTWHA, which is oftenperceived to be a safe refuge and breeding groundfor marauding (‘government’) pigs. Of growingconcern is the potential of the feral pig herd innorth Queensland to host and transmit a range ofexotic diseases threatening human and animalhealth. Recent incursions of Japaneseencephalitis, a potentially fatal human diseasespread by mosquitoes, illustrate this: Japaneseencephalitis is also hosted by pigs, which amplifythe disease and increase the likelihood of

transmission to humans. A range of otheremerging exotic diseases of humans couldbecome endemic in northern Australia via the feralpig herd. This also applies to livestock diseasessuch as Foot-and-Mouth, the arrival andestablishment of which in the feral pig herd wouldhave crippling consequences for Australia’s meatexport industries.

Feral pigs therefore pose ecological, economic,aesthetic, medical and veterinary threats, actualor potential, in north Queensland, and theirmanagement creates tensions in the relationshipbetween the community and government.However, it is clear that eradication of feral pigsover wide areas is not practical, and evensustained reduction of population size is difficultto achieve. Pigs have flexible behaviour andhabitat choice, they are highly mobile, and theyhave very high reproductive potential. They aregood at persisting in inaccessible habitats, andthey can recover quickly from reductions in popu-lation size, either by immigration or reproduction.

Pig control by conventional means is expensive,especially if the aim is to hold the population atvery low numbers : experience from elsewhere inAustralia shows that the cost of trapping pigs mayrange up to approximately $100 per pig(Choquenot et al. 1996). Typically, the cost ofcontrol of a pest population increases as thepopulation becomes sparser, because animalsbecome harder to find and the remaining indi-viduals may hold out in the most inaccessiblesites. Hunting down or trapping the last few pigsin any large area of rainforest would be extremelydifficult, and unless total eradication was achievedsimultaneously over very large areas any localsuccesses would be quickly reversed by immi-gration. Since an average litter of pigs is around10 individuals, repopulation of areas could be veryrapid.

Introduction2

Moreover, perceptions of the feral pig ‘problem’vary widely. Pig hunting is a significant recrea-tional activity, that contributes to the economiesof some small communities. Many recreationalhunters, though they believe themselves to bedoing a service by reducing the impact of pigs,would probably be unhappy to see pigsdisappear entirely. Hunting of feral pigs forhuman consumption is significant to many remoteAboriginal communities on Cape York: it providesa high-quality food source at low cost and hasbecome culturally significant in the maintenanceof traditional forms of wildlife harvesting andinteractions of people with country. Eradicationof pigs, even if it were practical, could dosignificant economic and cultural harm to thesecommunities and would be opposed by them.

The papers in this volume consider this range ofissues, reviewing our state of knowledge of thebiology, impact and control of pigs in northQueensland, and identifying important gaps in ourunderstanding. The volume opens with a reviewfrom John McIlroy of the environmental impactsof pigs in Australia, followed by reviews of theeconomic impacts of feral pigs from Reece Luxtonand their economic and cultural significance toindigenous people in north Queensland from ChrisRoberts and others. Papers by Jonathon Lee,Scott Ritchie and Jack Shield then examineaspects of the disease threats posed by the feralpig herd in north Queensland, using recentincursions of Japanese encephalitis as a casestudy. Jack Giles discusses our current under-standing of the population dynamics of feral pigsand their ecology and impact on ecosystem proc-esses in the Wet Tropics World Heritage Area aresummarised by Jim Mitchell. Jim Hone providesan example of the effectiveness of control andmonitoring of feral pigs in a conservation area inthe ACT, and Karl Vernes and others evaluate theeffectiveness of trapping in reducing abundanceof feral pigs in one location in the Wet TropicsWorld Heritage Area.

The final section considers current and futurecontrol of feral pigs in the area. Brad Dorringtondescribes the existing pig trapping program, andBob Seamark and Mike Holland consider thecontributions that biotechnology might make tothe control of feral pigs in the future. Nigel Storkand Trevor Stanley conclude with a summationof strategies and needs for pig control in the WetTropics World Heritage Area.

The Workshop shows that understanding of thepig problem in north Queensland, though solid inplaces, is decidedly patchy. Few people doubt thatferal pigs have the potential to do environmentaldamage to the rainforests of the Wet TropicsWorld Heritage Area and to key habitats, such aswetlands, elsewhere in north Queensland. JimMitchell’s work shows that pig feeding hasmeasurable effects on processes such as nativeplant regeneration in heavily-disturbed sites.However, such impacts are surprisingly subtle.Some potentially important impacts, such as onwater quality and in-stream fauna in uplandrainforest streams, have either yet to beexamined or have not been reported here. Moststudies of the effects of pigs on rainforestecosystems have been conducted at smallspatial scales. In future, we should be attemptingexperiments over larger areas, in which theeffects of reducing pig abundance on populationsof key taxa or indicators of ecosystem function(such as water quality) are examined at some-thing like a whole-catchment scale. Although feralpigs are capable of threatening the survival ofnative species – they almost exterminated theLord Howe Island woodhen, for example, andprobably had more to do with the extinction of thedodo than did hunting by Dutch and Portuguesesailors (Caughley & Gunn 1996) – there is as yetno clear statistically validated evidence that theyare causing declines of any native species in theWet Tropics World Heritage Area. However,studies of the effects of pigs on the demographyof potentially vulnerable species (such as ground-nesting birds, narrowly endemic earthworms, orstream-dwelling frogs) are needed. It is ofparticular interest that the density of pigs inrainforest in north Queensland is low, perhapssurprisingly so. Densities of around three pigs persquare kilometre have been estimated forlowland swamp/forest habitats, and abundanceappears much lower than this in upland rainfor-ests. Elsewhere in Australia, pig populations tendto be limited by high temperatures and low wateravailability (Choquenot et al. 1996); what limitspig abundance in the cool wet environmen underthe rainforest canopy?

Similarly, there is no doubt that feral pigs doeconomic damage to horticultural industries innorth Queensland, but the full extent of this hasnot been systematically measured. There aresome subtleties in the interaction between feralpigs and horticultural industries. For example, pigs

3Introduction

represent an economic cost to banana growerswhen banana prices are high, but when pricesare low pigs provide a service to growers bycleaning up banana dump areas and scrap pileswhich would otherwise harbour fruit flies. WhileAboriginal people recognise the benefits of pigs,they also realise that pigs do damage to somenative bush resources. What is needed in thesecases is an understanding of the relationshipbetween different levels of reduction of pigabundance and the magnitude of the economicand environmental benefits gained. It is quiteconceivable that there is a threshold populationsize of pigs, above which they do significant harm,but below which their impact can be tolerated andtheir services enjoyed. Discovering the positionof this threshold, and evaluating the cost ofholding a population below it, can only beachieved by large-scale experiments in controlwithin the framework of adaptive management.Certainly, given the near-impossibility of eradicat-ing pigs, we must learn to live with them.

Particularly in the rainforest of the Wet TropicsWorld Heritage Area, few options for pig controlare available. Pig trapping and hunting (by shoot-ing and dogging) are widely practiced, especiallyin the coastal lowlands adjacent to the WorldHeritage Area. The community based pig trap-ping program provides a coordinated approachto pig control over this area, and small-scaleevaluations of pig trapping show that it canproduce major reductions (around 80%) in localabundance of pigs. However, it is not clear whatlevels of population reduction are currently beingachieved by the prevailing regime of large-scalepig trapping. Although there is strong communitysupport for shooting and dogging as controltechniques, the effectiveness of these forms ofhunting have not been compared with trapping.

The papers on biotechnology and fertility controlmake it clear that such approaches, used incombination with conventional techniques, couldmake pig control more effective, more flexible andcheaper. Better understanding and manipulationof the pheremonal biology of pigs might enhancepig trapping rates. Strategic field immunisation ofpigs could reduce the need for massive (andperhaps unachievable) population reductions tolimit the spread of disease epidemics; andimmuno-contraceptive methods similar to onescurrently under development for rabbits and housemice could slow the rate of population growth and

make it more feasible to hold populations belowthresholds of acceptable damage. There is nowenough experience in the development of suchapproaches in Australia to make their applicationto feral pigs seem an achievable goal, but onlyafter a considerable research effort.

Finally, the colossal damage that the accidentalintroduction of Foot-and-Mouth has caused toBritish farming and possibly to the rest of Europe,is a timely warning of the potential threat that feralpigs offer to Australia’s livestock industry. If Foot-and-Mouth were accidentally introduced then itmight be impossible to eradicate it in the nearfuture.

PART 1

IMPACT OF FERAL PIGS

5

Overview of the impact of feral pigs, Sus scrofa,on the Australian environment

McILROY. J.C.

38 Hempleman Drive, AkaroaCanterbury, New Zealand

ABSTRACT

Feral pigs, Sus scrofa, with their large robust bodies, specially developed snouts forrooting up the ground, omnivorous diet and opportunistic feeding habits, adaptabilityto a wide range of habitats and gregarious behaviour, have the potential to detrimen-tally affect the Australian environment. Few studies, however, have been undertakento identify and measure the types of impacts they are having, including their severity,extent and location. Their most likely impacts are habitat degradation and predationon, or competition with, native animals. More quantitative information is needed onthe relationships between pig abundance and the degree of impact (and effort spenton control) in different natural areas to determine if, where and at what cost controlmeasures are justified.

INTRODUCTIONPigs, Sus scrofa, introduced into Australia by theearly European settlers, quickly established feralpopulations, particularly in New South Wales andQueensland (Choquenot et al. 1996). Today thereare possibly about 13.5 million of them inhabiting38% or so of Australia (Hone 1990). The size ofthe population, however, probably varies consid-erably each year depending on environmentalconditions, such as floods and droughts.

Human attitudes towards feral pigs range fromregarding them as a resource, such as forrecreational hunting or export of their meat, toconsidering them a serious pest because of theirdamage to crops and livestock or their potentialrole in spreading exotic diseases. This paperfocuses on the impact that feral pigs are knownto have, or could be having, on the Australianenvironment.

DESIGNED FOR IMPACTThe biology and ecology of feral pigs are majorpredisposing factors in the impact they may behaving on the environment. Their large robustbodies, specially developed snouts for rooting upthe ground, omnivorous diet and flexible activitypatterns allow them to live in a wide range of habi-tats. These include subalpine grasslands andforests, dry woodlands, tropical rainforests, semi-

arid and monsoonal floodplains, swamps andother wetlands in many parts of Australia. Theiropportunistic feeding habits and omnivorous dietallow them to exploit various temporarily abun-dant food sources, such as fruits and seeds,foliage and stems, rhizomes, bulbs and tubers,fungi and animal material. Apart from adult males,feral pigs are mostly social, gregarious animals.Group sizes vary considerably, ranging from1-12 up to 40-50 in different seasons and areas.Mobs of more than 100 can gather aroundremaining waterholes in dry seasons (Choquenotet al. 1996). Densities range from less than onepig per square kilometre in semi-arid rangelands,woodlands and open forests to 10-17.5 km-2 inswamps and other wetlands.

All these factors mean feral pigs have the capac-ity to adversely affect the environment but the typeand severity of their impact may vary in both spaceand time.

IDENTIFYING AND MEASURINGENVIRONMENTAL IMPACTS BY PIGSIdentifying and measuring the impacts that feralpigs have on the environment is a crucial part offeral pig management plans. The type, severity,extent and location of their impacts determineswhether control should be carried out, and if so,when and how much control is necessary to

Impact of feral pigs on the Australian environment 7

reduce or eliminate these impacts. If the impactof pigs in any area is not clearly known, then ex-pensive control measures may not be justified.Measuring or identifying environmental impactsby feral pigs can be difficult and affected by per-ceptions. In some cases their impacts may ap-pear direct or obvious, such as predation on tur-tle eggs or rooted up leaf litter and soil in a rain-forest. However, while such ‘damage’ may be dra-matic evidence of the pigs’ activities, they maynot necessarily have a significant effect onpopulations or ecosystems. Juvenile mortalityamongst turtles, for example, is intrinsically highwhile patchy disturbance of the rainforest floormay not affect the long-term processes of plantdynamics or community structure. In other casesthe impacts may be indirect, such as pig rootingon hillsides causing siltation in streams, whichmay ultimately affect aquatic life. Impacts of pigscan also vary over time. They may be acute, suchas if pigs suddenly rooted up the last known clumpof the subalpine herb Gentiana baeuerienii inNamadgi National Park near Canberra. Alterna-tively, they may be chronic such as wallowing bypigs causing the gradual degradation of a smallswamp. Pig impact can also be roughly constant

in intensity (eg. rooting up of leaf litter) orperiodic such as feeding on water lilies or otherplants in a northern lagoon during the “dry”season or rooting up valley floors during winter insouthern hill country areas.

The most important environmental impacts thatferal pigs are likely to have are habitat degrada-tion and predation. Habitat degradation couldoccur through selective feeding, trampling orrooting by pigs and affect plant species composi-tion and density, nutrient cycling, rates of erosionand nutrient losses, plant succession and thediversity of fauna (particularly soil invertebrates)present. Pigs may also have a role in the disper-sal of exotic plant seeds and the destruction ofnative plant seeds.

HABITAT DEGRADATIONThe most obvious signs of ‘damage’ by feral pigsare patches of ground, grassland or forest litterrooted up by them in their search for undergroundfood. Such disturbances can be locally exten-sive, especially around swamps, lagoons andwatercourses or after rain when the ground issofter. For example, Alexiou (1983) found thatmost pig rooting in a subalpine area near NamadgiNational Park occurred along drainage lines, in

depressions and around grassy flats. Altogether,32% of such sites showed evidence of pig dam-age, particularly a large reduction in the abun-dance of the dominant grassy vegetation andsome small native herbs. Several native plants,however, had become vigorous colonisers of therooted up areas. Hone (1998) found that plantspecies richness in a nearby area of grasslanddeclined as pig rooting increased. He concludedthat if the pig rooting covered more than about25% of the area, then species richness coulddecline rapidly, but at this stage the long-term andlarge-scale effects of this impact were unknown.

There is a similar lack of knowledge about thelong-term effects of feral pigs on soil nutrient andwater cycling, rates of erosion and nutrient losses,soil micro-organisms and invertebrate populationsand plant succession in Australia. Statham andMiddleton (1987) described how extensiverooting by feral pigs in moist gullies in StrzeleckiNational Park on Flinders Island led to erosion,loss of regenerating forest plants and theirreplacement by thick stands of bracken fern,Pteridium esculentum, but provided no quantita-tive details.

The extent to which feral pigs eat or disperseseeds is also unknown. Feral pigs are likely toeat a much greater range of fruit and seeds inAustralia than has been reported (eg. McIlroy1993) but the viability of the seeds in pig faecesmay depend on the size of the seeds, the feedingbehaviour of the pigs and where the faeces aredeposited. The ingestion by pigs of fruit contain-ing small (less than five millimetres diameter) seedfrom plants such as trunk-fruiting figs, Ficus

variegata, umbrella trees, Schefflera actinophylla,and guavas, Psidium guajava, appears to causeno physical damage to most of the seeds, butthere are conflicting reports on the fate of larger,softer seeds (McIlroy 1993).Guava and otherunidentified seeds have been observed germinat-ing in pig faeces but their viability appears to below (Pav Ecol. 1992, Pavlov et al. 1992, Mitchell1993). Weeds may also be spread to new areas,particularly freshly dug up or trampled areas,through the attachment of their seeds to pigs’coats or in soil clinging to the pigs’ snouts or feet.

There is growing evidence that feral pigs may helpspread rootrot fungus, Phytophthora cinnamomi,

responsible for dieback disease in native vegeta-tion. Although there is still no evidence of spread

Impact of Feral Pigs8

by the gut following ingestion of infected material(Masters 1979), the pigs can carry the organismin soil on their hooves (Kliejunas and Ko 1976).Pigs could also carry infected material on otherparts of their body, particularly after wallowingduring warmer conditions when sporulation of thefungus may occur (Masters 1979). The spread ofthe fungus has also been associated with soildisturbance and reduction of litter cover by pigs(Brown 1976). Pigs also chew or tusk the bark onbuttress roots and lower trunks of trees, whichmight allow the entry of fungi.

PREDATION, COMPETITION ANDDISTURBANCE OF OTHER ANIMALSFeral pigs can eat a range of animals, includingearthworms, amphipods, centipedes, beetles andother arthropods, snails, frogs, lizards, the eggsof the freshwater crocodile, Crocodylus johnstoni,

turtles and their eggs, small ground-nesting birdsand their eggs and young rabbits, Oryctolagus

cuniculus (Pullar 1950, Tisdell 1984, McIlroy 1990,Mitchell 1993, Roberts et al. 1996).

Earthworms are one of the most common sourcesof animal protein in the diet of feral pigs and it ispossible that pigs could significantly reduce thenumbers of worms in some localities. Pav Ecol(1992) found that feral pigs harvested over 95%of the available worms at sites in lowlandephemeral swamps near Cape Tribulation duringApril-July 1992. Although the number of wormsat different sites varied greatly, few adult wormsoccurred in freshly rooted up areas. Mitchell(1993), in contrast, found identical numbers ofearthworms in feral pig diggings and surroundingareas in the same general region south of CapeTribulation.

Frogs may also be a common food item for pigsin some areas. Richards et al. (1993) suggest thatferal pigs, through either direct predation orhabitat disturbance, may have contributed to thedeclines in some populations of endemic tropicalrainforest frogs. At the moment we don’t knowwhat effect rooting or wallowing by feral pigs maybe having on amphibians, such as the corroboreefrogs, Pseudophryne spp. which are nowendangered species very close to extinction inthe Australian Capital Territory and SnowyMountains. W. Osborne (pers. comm. 1996) hasfound no evidence so far of pigs destroying thespecies’ breeding sites in Kosciusko NationalPark, but he has witnessed their tadpoles being

splashed out of a pool by a wallowing pig.The pigs’ extensive rootings could cause patchesof the frogs’ habitat to dry out, but equally maycreate ponds for the frogs. W. Osborne has foundcorroboree frog eggs under rooted up tussockgrasses but there is no information on whethereggs in such situations survive.

The effect of pig predation on other invertebratesand small vertebrates in Australia is not known.Without data on what prey are actually eaten, therates of predation, the density and status of theprey, and whether or not predation by pigs isdensity dependent, it is premature to judgewhether pigs are a serious threat to the animalsconcerned. This also applies to their impact onlarger ground-nesting birds, such as cassowaries,Casuarius casuarius, scrubfowl, Megapodius

reinwardt, and brush-turkeys, Alectura lathama,despite reports of pigs destroying their nests andeating their eggs and young (Hopkins andGraham 1985, Crome and Moore 1990, Mitchell1993). Instead, in north Queensland rainforestsand secondary forest, opportunistic, omnivorousrodents, especially white-tailed rats, Uromys

caudimaculatus, may be the dominant predatorsof some bird nests (Laurance et al. 1993,Laurance and Grant 1994).

Despite conjecture, there is no evidence thatferal pigs (Sus scrofa) adversely affect thesurvival of cassowaries in the wet tropics regionof Queensland. The greater ability of pigs to switchor move to alternative food supplies once theseasonal flush of rainforest fruits wanes, however,may provide them with some competitiveadvantage over the more sedentary, frugivorouscassowaries, particularly sub-adult birds forcedto fend for themselves away from parental terri-tories. This may have repercussions in terms ofrainforest regeneration, dispersal of exotic weedsand the continued presence of cassowaries inmany areas of the wet tropics. Feral pigs mayalso compete with the northern bettong, Bettongia

tropica, brush-tailed bettong, B. penicillata, rufousbettong, Aepyprymnus rufescens, long-footedpotoroo, Potorous longipes, long-nosed potoroo,P. tridactylus, and musky rat-kangaroo,Hypsiprymnodon moschatus for undergroundfruiting bodies of mycorrhizal fungi, undergroundstems of grasses, roots, bulbs and tubers, seeds,rainforest fruit and insects in different parts ofAustralia.

Impact of feral pigs on the Australian environment 9

CONCLUSIONIt is clear that there is very little known about theimpact of feral pigs on the environment inAustralia. More often than not there are percep-

tions that they are detrimentally affecting theenvironment. While rooted up forest floors, grassyvalleys or swamps may not be aestheticallyacceptable, particularly in major tourist areas,such activities by pigs may have little ecologicalimpact. What is required before any managementplans are contemplated is a clear assessment ofwhat actual damage the pigs are doing, particu-larly the severity of the damage, its extent andlocation and how acceptable this is, publicly orscientifically. This is fundamental for decidingwhether the pigs are a problem or not and whichimpacts or particular areas should receivepriority control. The relationship between pigabundance and degree of impact also needs tobe measured to determine what levels ofpopulation reduction must be achieved andmaintained to obtain acceptable levels of impact.If, at the same time, the relationship between theabundance of pigs and efforts spent on control indifferent areas is quantified, it would then bepossible to determine how much would be neededto be spent on control to prevent or stop anyunacceptable impacts by feral pigs on theenvironment.

Impact of Feral Pigs10

Economic impact of feral pigs in the Wet Tropics

LUXTON, R.

Land Protection, Department of Natural Resources and MinesSouth Johnstone

The Wet Tropics region is categorised by the large tracts of World Heritage protected ar-eas. Feral pigs are regarded as a significant agricultural pest in the region. This is reflectedin management practices to reduce numbers where a direct economic loss is involved. Forsome primary producers, farming would not be economically viable without an effective/intensive feral pig control program. Recently however the trend is changing to allow areduction in impact, as seen through the success gained with the Community-based FeralPig Trapping Project.

This paper intends to draw together various sources of current information on economicimpacts of feral pigs in the Wet Tropics. This is a topic without specific statistical datafurther than reports from McIlroy (1993) and Mitchell (1993) but I hope to draw a definedpicture. Economically we can look at the direct costs/losses, indirect cost/losses includingthe management costs and the benefits of control techniques available.

In Queensland it is estimated that $1.1 million was spent on feral pig control in 1984, whichequates to $2.2 million in today’s dollar values (McGaw & Mitchell 1998). This amountincludes both government and private expenditure on control, although it does not includeamounts spent by recreational hunters. Despite the pressure of legal requirements to con-trol feral pigs, landholders control pigs principally because they are seen to have a nega-tive economic impact.

ABSTRACT

DIRECT LOSSES

Sugar CaneFeral pigs in cane can cause severe but local-ised damage. Cane in the Wet Tropics is usuallygrown in close proximity to feral pig habitat. Themain problem occurs with trampling of cane settsand plant cane, and the physical destruction inpaddocks.

This industry accounts for the majority of primaryproduction in the region, with numerous follow-on benefits for local communities. Damage bypigs is certainly a direct cost, with various CaneProtection and Productivity Boards (CPPB)reporting on hectares damaged and tonnes loston an annual basis. Some boards also cover a50% subsidy on the provision of trap material, onthe basis of 1 trap per farmer (Edwards pers comm1999).

To give an example of the damage wrought oncane farmers, figures are provided from theexpanded cane areas of the Upper Murray andWarrami districts, 30 km south west of Tully. Onthe South Johnstone Mill areas, 100 pigs havebeen trapped in the last six month period in a 2000ha area. This is in an area highly regarded andutilised for recreational hunting pursuits as well.A farmer relatively new to the area was alsoastonished with figures of 40 pigs shot on hisproperty in a 2 month period (Barnes perscomm.1999). The damage is severe in discreetlocations, and ranks third after cane grub and ratdamage. Damage is difficult to measure, costedon a district basis. In 1991 the cost to sugar canecrops was in the vicinity of $628 000, equating toa reduction in yield of 25 510 tonnes (McIlroy1993). However economic hardship to individualsis the problem, not losses to the industry as awhole.

Economic impact of feral pigs in the Wet Tropics 11

Impact of Feral Pigs

BananasFeral pigs are also regarded as an economic pestto this industry. Particularly when banana cropprices are high, significant losses are regardedas sustained on a localised basis. For example,average losses in the Tully area include up to 20bunches lost per month, at a cost between $600to $1800 (Noble 1996). When prices are low, thecleaning up of banana dump areas and scrap bypigs is considered a benefit.

Tropical FruitDamage to trees, fruit loss and irrigation equip-ment is reported regularly. Fruit loss is significantlyhigh due to the access to low hanging fruit – apreferred practice for farmers is to have specificfencing and trimming lower hanging branches.Pawpaw plants are easily damaged when tuskedor rubbed.

Small crop productionAlthough not a large industry in the Wet Tropicsarea, damage does occur. Irrigation piping andmulching seems to be disturbed in areas close toprime pig habitat, where losses anywhere up to50% losses are reported.

LivestockFor this industry, the impact of feral pigs isnegligible. However pig activities reduce pastureavailability and can promote weed spread(McGaw 1998). Diggings can also contribute todamage to farm equipment, with the example ofrocks being brought up by pigs damaging slasherblades.

Damage to infrastructureFeral pigs are also regarded as damaginginfrastructure such as roads, drains, easementsetc. Damage to roads through pig digging isparticularly evident in the wet season.

INDIRECT LOSSES

Recreational huntingHunting is seen to be a highly regarded activityfor people in the region, particularly when chillerboxes are open, and for some individuals, pighunting is a viable commercial business.

CommercialisationWhile seen as a possibility, the biggest problemwith major commercialisation is the consistencyof supply of pigs – this apparently is not possiblegiven the difficult terrain and limited access.

The chiller boxes in the region are only open forindefinite periods of time. The biggest constraintto commercialisation is the requirement forhunters to be licensed game meat harvesters,the Queensland Meat and Livestock Associationis regarded as having unreasonable requirementsfor accreditation.

Transmission of diseasesFeral pigs are regarded as significant vectors fornumerous diseases impacting on agriculturalproduction and human health. Pigs are alsoimplicated as vectors in transmission of plantdiseases such as Phytophthora cinnamoni (rootrot fungus) (Mitchell 1993).

Potential impact of diseasesThere are feral pig populations within a 2 kmradius of international air and sea ports in the WetTropics, and the risk of entry of exotic diseasethrough these areas is quite significant. Along withthis, local tourist resorts can be affected thusimpacting one of the largest industries in theregion (Pavlov et al. 1992).

Economics of controlFencing is used predominantly to protect sugarcane crops. However this is only regarded aseffective if fences are erected prior to feral pigsmoving into crops. The Tully CPPB have spent$8000 to protect a seed block adjacent to a pighabitat area (Noble 1996). Electric fencing is alsowidely used in the cane industry to inhibitdamage from feral pigs.

Trapping is regarded as best practice for the WetTropics area (Mitchell 1993). Coordination is thekey. To give a perspective on the costs for controlin the Wet Tropics, funding and in-kind contribu-tions can be looked at with the Community BasedFeral Pig Trapping Project. To run the programsuccessfully, the funding from various sourcesincluding the Wet Tropics Management Author-ity, Department of Defence, EnvironmentalProtection Agency, Department of NaturalResources and Mines, and various Cane Protec-tion and Productivity Boards. The in-kind contri-butions from landholders and trappers equatesto $30 000 each per year – with 30 trappers in-volved in this totals nearly $1 million (Dorringtonand James, 1999).

The Cairns City Council under their PestManagement Plan has an operations budget of

12

$10,000, which goes towards trapping in responseto requests from residents, and to monitor pigactivity and trap in key areas (Murray pers. Comm.1999).

Poisoning using 1080 baits is seen as appropri-ate where high levels of damage occur in a local-ised area, but is not regarded as effective for longterm management objectives. Also it is seen asimpossible on a broad scale level because of thepresence of endangered fauna and tourists.Shooting is highly inappropriate and discourageddue to the high vegetative cover in the area.Dogging is effective for the capture of trap-shyanimals, but on the whole is regarded asineffective for control.

Bounties are used by CPPB throughout the area,and are seen to generate goodwill between farm-ers and the board (Clarke pers. comm. 1999).This is seen as one of the best ways to inducecontrol of feral pigs and thus reduce numbers.However the Department of Natural Resourcesand Mines is against the use of bounties, and hasa particular policy regarding the issue. A studyby the Bureau of Resource Sciences (BRS) in1998 also concluded that ‘the use of a financialincentive in the form of a bounty payment as ageneral tool to reduce vertebrate pest damage isinappropriate’. Hence control methods used in theWet Tropics vary with habitat, size and location.The best option is to balance all factors for themost efficient control plan.

KEY POINTS

Since the reports produced by McIlroy andMitchell in 1993 to the Wet Tropics ManagementAuthority, and Noble’s report to the BRS in 1996,perceptions have not changed dramatically. Theimpact of pigs and their damage varies accord-ing to seasonal conditions and the perceptions oflandholders. On the whole the World HeritageArea is still regarded as a harbour for feral pigs.

For agricultural industries, maximising economicreturn is the key objective. Attitudes to control inthe Wet Tropics is driven largely by perceptionsof the problem. There are four different controlstrategies available to landholders (Choquenot et

al. 1996) - 1) no pig control as no net economicgain; 2) restrain pigs at a moderate level; 3) re-strain pigs at a low level; and 4) local eradication.Each of these options is practised in the WetTropics area, dependent on the perception of the

landholder. There is danger in allowing the man-agement of feral pigs in the Wet Tropics to steertowards commercialisation. In this, there wouldbe a change in perspective from hunting being arecreational pursuit to a likely income supplement;further to this, the hunters will feel they aren’tcovering the costs (Edwards pers. comm. 1999).

CONCLUSION

Reduction in impact is the key, and this is reflectedby landholders affected economically by feral pigs.An understanding of keeping the impact downthrough improved management practices willhopefully bring a change in the perception oflandholders.

Also, the continued coordination of managementof feral pigs is critical, as population levels arelikely to return to pre-management levels if fund-ing is withdrawn. Communication is vital betweenthose bodies involved in the policy, managementand on-ground control of feral pigs in the Wet Trop-ics. There is a need to investigate an incentivesprogram for expanding the Community BasedFeral Pig Trapping Project to a wider range of ruralindustry groups.

The issue of how to compare environmental, ag-ricultural and other values on an equal basis isstill unresolved, and requires further study for thebenefit of structuring sound planning and man-agement.

Strategic sustained control is the most likely sce-nario utilised by primary producers, wherebypopulations are reduced to a level where ben-efits are maximised compared to costs. Constantfunding and maintenance of coordinated pro-grams is crucial, particularly on the perimeter ofthe World Heritage Area in order to reduce theeconomic impact to adjacent primary production(Mitchell 1993).

Economic impact of feral pigs in the Wet Tropics 13

Feral Pigs - Indigenous perspectives

ROBERTS, C.1. AKEN, R.2 TAISMAN, A3., CHEVATHON, J. 4

1Author - Balkanu Cape York Development Corporation2Reader - Cape York Land Council3Wik Presentation - Aurukun4Translator - Cairns

Feral animals and plants are a contentious issueon Cape York. Most people of a developmentalmind want them removed. Most conservationists(including national parks) see extermination ofpigs as part of their mission. Roving shooters thinkthey are doing the world a favour as they stitchup their dogs and clean their knives or guns.Those in the health profession have little patiencefor the pig. Even the government gets emotionalabout these so called uninvited guests. For thesereasons, we see substantial amounts of fundingbeing allocated to getting rid of pigs.

It is relevant to compare the origins and evolutionof the views of Aboriginal people with the viewsof non-indigenous people as far as feral animalsand plants are concerned. If we take both viewsa long way back they would probably be quitesimilar (even 200 years ago). Food. It seems thatthe essential difference is that Aboriginal peopleare still in survival mode, literally. At the same timeAboriginal culture accords respect to the animals,plants and cultural practices that feed them.

At this point we refer you to a study that wasproduced by Bruce Rose and the Aboriginalpeople represented by the Central Land Councilin Alice Springs. The work provides some inter-esting insights into the perspectives of indigenouspeople in relation to the feral animals that arefound in that area. Attitudes relating to camels,rabbits, cats, donkeys, are easily transportableto the Cape York situation. Several of theseanimals have been incorporated into “story” ormythology as it is called by White Fellas.

We have previously embarked on the beginningsof a study with the Department of NaturalResources and Mines in Brisbane. We wereinvited to participate in a study of pigs on CapeYork after the body of NHT proposal concernedhad already been written. This sequence of eventsis not ideal, and we did address the matter withDNRM. The bottom line is:

Do not assume that everyone wants pigs

dead: they have value and that value

varies from place to place.

This concern along with others precipitated aworkshop in Cooktown where some relevantpoints were made by all participants. We encour-age further intentions to address the pig issue torevisit them. We have been over some of thisground before.

It is very important to realise that the perspectivefrom which Aboriginal people see feral pigs is verydifferent from the Western view and this is prima-rily because of the cultural base from which thisvision comes. “White people” have been educatedusing Darwinian principles as their main paradigmor “creation story” and with that the concept ofbiodiversity or “plenty different tucker”. Aboriginalpeople understand how things work in a practicalway. They are more effected by this Darwinianreality than non-indigenous people are. The wayin which they arrive at caring for this biodiversityis really quite different from the way in which whitepeople arrive at that end. Scientists seem to beof the opinion that feral animals disturb the great

Impact of Feral Pigs14

ABSTRACT

The views of feral pigs held by indigenous and non-indigenous people differ widely.While conservationists and non-indigenous landholders might like to see pigseradicated, pigs are a highly regarded food source for indigenous people, andthey provide an important outlet for the maintenance of aspects of traditionalculture. Planning for pig management, especially on Cape York, should take thisdiversity of interests into account.

scheme. The Aboriginal belief is very similarexcept that it is not based on the Darwinianprinciples supporting biodiversity as such, but themore urgent survival instinct based on food forfamily.

Food is a mighty powerful motivation and it isdifficult to appreciate bioregional theory whenone’s very culture is threatened by other arms ofthe same mind set. Politics in general does notfavour the maintenance of Aboriginal culture.There are no sporting shooters in indigenousCape York that I know of.

In central Australia, it seems from the report doneby Bruce Rose (1995), that Aboriginal people hadincorporated new arrivals into their landscapeand into their resource base. As a consequencethese new animals have in fact been elevated tothe spiritual realm of Aboriginal existence. This isvery different from the way these animals areviewed in the scientific world. Ferals are seenessentially as a nuisance, as destroyers of envi-ronment, carriers of disease etc. For Aboriginalpeople feral animals often provided somethingnew to eat. The main reason that there is anydispute about “what should be done about pigs”is that many of these communities still rely on pigsfor food. They have no argument about where thisfood came from, it is a resource that they can useand to some extent they are even keen to lookafter that resource, to preserve that option.

There is a spectrum of responses to feral animalsamong the Aboriginal and Islander communitieson Cape York. In general feral pigs are a highlyregarded food and are commonly used in festivi-ties. In fact people who shoot pigs withoutpermission from the local Aboriginal people canfind themselves in trouble. This issue revisits theownership question. Who owns these things onthe land? Who has the right to kill them? Do out-side people have a right to come and harvest themwithout getting permission? There are a numberof complex issues here.

Feral pigs do dig up the ground. There is noargument there. They dig yams which is animportant activity of Aboriginal women. In areassuch as Lockerbie scrub (the northern most rainforest on Cape York) you will find extensivedamage from pigs along the waterways.However; the Lockerbie scrub is a prime pighunting area for the local people at the top of Cape

York. Further south along the east coast atLockhart River people hunt pigs for food also.There does not appear to be any “recreational”hunting of pigs by Aboriginal people, they arehunting specifically for food.

Whilst the more fortunate population sectors inAustralia can pontificate about the awkwardnessand embarrassment of having pigs in nationalparks and other areas of Australia, we really needto get down to reality when considering subsist-ence and survival on the land. There are feweconomic enterprises that provide a living and lifeis hard. The truth of the matter is that Aboriginalpeople still rely heavily on hunted food.

Other protein sources such as beef have beenconsidered, however, the logistics, carryingcapacity of the country, fencing, environmentalcost, feasibility and absence of “the hunt”, aresignificant barriers to what might appear to be asimple solution.

Pig hunting is not only about food. It provides anoutlet for men to practise an important part of theirculture. I have some figures provided by AndrewRoberts which illustrate the dynamics of huntingat the top of Cape York. Hunting pigs can take agood deal of pressure off other species.

Recently two graduates did work in the Cooktown/Hope Vale area and managed to terrify the localsby explaining the various diseases that occur inpigs. Now this is all very well but theseexplanations need to be measured against socialeffects, and they also need to be realisticallymeasured against the health consequences.Before rushing in and telling people how unhealthyor healthy such and such an activity might be, wereally need to get our facts straight. For examplein the last two decades, how many people fromHope Vale have been hospitalised because of apig-related illness? This same logic might beapplied to the remainder of Cape York. We arenot saying that people should carry on regard-less, but we do need to consider the statistics.Perhaps selection and cooking processes reducerisks. Humans are after all one of the fewpredators that pigs have. Enquiries to BamagaHospital, Cooktown Hospital and Tropical Healthdid not reveal a single disease case directlyattributable to pigs within the experience of thestaff spoken to.

Feral Pigs - Indigenous perspectives 15

Impact of Feral Pigs

Human predation appears to be a relativelyimportant control factor in the number of pigs onthe Cape. At Aurukun, there appears to be ameasured approach to pig numbers and throughbrief conversation with people from there it seemsthat when there are a number of good wetseasons one after the other, pig numbers can getquite high and in such cases the community maywell be in favour of a culling program. On the otherhand when the wet seasons are not so good,having a bit of extra meat running around onAboriginal country is quite useful. The importanceof lily bulbs as a cultural food it relevant here.

The new spectre of Japanese encephalitisappearing on the Cape is of great concern to allthe Cape communities and their service agencies(eg Cape York Land Council, Apunipima CapeYork Health Council, Balkanu etc). Aboriginalpeople use dogs for hunting and generally use afirearm only to kill the pig once it has been bailedup. A pig of course can be a very dangerousanimal and for this reason it is necessary to usefirearms at the end of the hunt to be on the safeside. There have been some rather simplisticarguments as to whether Aboriginal people shouldbe using spears or guns during their hunting, butessentially it is the hunt itself that is most impor-tant. Pig numbers on Cape York haven’t been wellestablished. Somewhere there is a balancebetween having too few and too many pigs onthe Cape.

Pigs armed with strength, a powerful sense ofsmell and high intelligence are very good atfinding food. They eat an astonishing array of fooditems (beach crabs, bird eggs, carrion includingpig skulls etc). With a wide variety of dentalequipment (teeth) they create several problemsfor predators and prey. Their ability to learn abouttraps is well recorded by bush people.

For Aboriginal people the effects of pigs on bushtucker - digging up of turtle eggs, yams, bulbs,water lilies - are an issue in some cases. Forconservationists predation of ground nesting birdeggs such as the Red Bellied Pitta is a concern.We need to discover what effect feral pigs haveon traditional resources of Aboriginal people if wewant support for culling. At the moment there areno statistics, no figures that will allow for anobjective comparison between having this newanimal on the landscape and what losses it mightbe causing in relation to cultural resources.

16

Unless these things are addressed, managementauthorities will have a difficult task convincingAboriginal people that feral pigs do more harmthan good.

There are several research directions necessaryto provide this information and we should worktowards that end together so that we not onlypursue the “precautionary principle” but alsoprovide Aboriginal people with both the culturaland physical benefits that feral pigs provide onthe Cape.

PART 2

FERAL PIGSAND

DISEASE THREATS

17

BACKGROUNDThe Northern Australia Quarantine Strategy(NAQS) is a program conducted by the AustralianQuarantine and Inspection Service (AQIS), andwas established in 1989 to assume responsibilityfor quarantine surveillance in northern Australia.This was due to the recognition of the clear andpresent dangers of incursions of exotic human,animal and plant diseases and pests into northernAustralia from countries to our north.

NAQS has three arms: NAQS Operations, NAQSScientific and NAQS Public Awareness.• NAQS Operations provides border security in

the Torres Strait and northern Cape YorkPeninsula region, and uses indigenouspersonnel to address the specific problemsposed by the close proximity to PNG and thetraditional movement of indigenous peoplebetween the two countries, which is allowedunder the Torres Strait Treaty.

• NAQS Scientific is the investigative arm andprovides the epidemiological information onexotic diseases and pests.

• NAQS Public Relations provides support tothe other two arms by producing publicawareness materials and resources.

NAQS SCIENTIFICThe mission of NAQS - Scientific is to identify andevaluate quarantine risks to northern Australia,and to provide early warning of quarantine pestsand diseases through a program of monitoring,surveillance and public awareness acrossnorthern Australia and in neighbouring areas ofPapua New Guinea, East Timor and Indonesia.NAQS has most recently been reviewed in 1995and again as part of the AQIS review in 1996. Asa result of these reviews. It’s role and importancein the surveillance and early warning of exoticdiseases, has been recognised and this hasresulted in increased funding and an expansionof capabilities.

The role of the Northern Australian QuarantineStrategy

LEE, J

Australian Quarantine Inspection ServiceAirport Administration Building, Cairns

STRATEGIES

The NAQS - Scientific Strategic Plan outlines thestrategies designed to address the missionoutlined above. These are:

(1) Identification and assessment of risk, and

(2) Surveillance activities.

Risk assessment is a formal process involvingscientific working parties evaluating the risk ofentry and resulting impact of a range of exoticanimal diseases and pests. This process resultsin the production of a NAQS animal disease targetlist. The provision of a target list enables availableresources to be focused on the most significantthreats to enhance the likelihood of providing earlydetection and early warning of potentialincursions.

Using the resulting risk assessments and NAQSexotic disease target lists as a reference, NAQSveterinarians and scientists develop operationalplans to provide comprehensive surveillance forthese target list diseases both onshore andoffshore. Although specifically aimed at target listdiseases, these surveillance activities are alsodesigned to detect emergent or aberrant exoticdiseases or strains. The NAQS veterinarians andscientists use their findings and experience to finetune these surveillance activities to optimise theresults obtained.

Surveillance activities covered by the animalprogram are structured around two mainconcepts. These are(a) the use of sentinel animal herds andstrategically sited insect traps both onshore andoffshore, and: (b) a series of targeted surveys (survey frequencyis correlated to risk levels for the geographic areascovered).

Role of Northern Australian Quarantine Strategy 19

Feral Pigs and Disease Threats20

The sentinel animal herds utilise serologicallynaive pigs or cattle, from which blood samplesare regularly analysed for evidence of exposureto subclinical exotic diseases. These sentinelherds are also regularly checked for evidence ofclinical disease due to other exotic pathogens ormyiasis due to screw worm fly.

Targeted surveys are conducted along thenorthern Australian coastline from Cairns toBroome and are coordinated by the NAQS staffin each of these States and Territory. Surveys arealso regularly conducted in Papua New Guinea,East Timor and Indonesia with joint participationby scientific staff from those countries.

NAQS Scientific also funds scientific researchaimed at developing new techniques to improverisk assessment and enhanced surveillance.

The NAQS Japanese encephalitis surveillanceprogram is a good example of how activitiesundertaken by NAQS contribute to identification,assessment and response to an exotic diseasethreatening Australia and all those who live here.

Japanese encephalitis - a case study of exoticanimal disease incursion

LEE, J

Australian Quarantine Inspection ServiceAirport Administration Building, Cairns

EPIDEMIOLOGY OF JAPANESEENCEPHALITISJapanese encephalitis (JE) is an exotic animaland zoonotic disease threatening the NAQS areaof responsibility. It is a mosquito-transmittedflavivirus, closely related to several existingendemic Australian flaviviruses, which includeMurray Valley encephalitis (MVE), Kunjin, Alfuy,Stratford, Edge Hill and Kokobera (Hall et al,1986).

The most common natural hosts of JE are Ardeidwater birds (herons and egrets). Pigs howeveract as amplifying hosts by developing high levelsof viraemia when infected. They can readily actas a reservoir for mosquitoes to transmit JE virusto susceptible humans and horses, which are endhosts and may develop fatal encephalomyelitisin a small percentage of infected individuals. JEis currently endemic throughout most of SouthEast Asia including Indonesia (J. Mackenzie, UQ,pers. Comm. 1999).

RISKS OF ESTABLISHMENT INAUSTRALIASerological studies indicate that JE is probablynow endemic in the Western Province of PapuaNew Guinea, Irian Jaya and Timor (J. Macken-zie, UQ, pers. Comm. 1999). NAQS is concernedthat there is a significant risk of an incursion fromPNG, via the Torres Strait islands, or directly fromTimor or other parts of Indonesia into northernAustralia.

If JE becomes established in Australia, therewould be significant social and economic impli-cations. The mechanism of spread of JE is un-clear, but potential vectors include wind bornespread of infected mosquitoes, or movement ofinfected avian hosts.

Factors favouring JE gaining a foothold andbecoming endemic include the presence of viablevectors and vertebrate amplifying hosts. Culex

annulirostris, an endemic mosquito and viablevector for JE, is common through northern Aus-tralia and its summer range extends as far as Vic-toria. Another important mosquito vector, Culex

gelidus, has recently been identified in Queens-land and the Northern Territory (Whelan, P., NTDept. of Health and Haseler, B., AQIS, pers.comm. 2000). Feral pigs are considered to be thevertebrate host of primary importance and thegreatest population densities occur in the Gulf ofCarpentaria, Cape York Peninsula and the top endof the Northern Territory (Choquenot, McIlroy &Korn 1996). Once established the disease wouldpose an immediate threat to the human popula-tion of northern Australia, which is primarilyconfined to coastal regions where there is favour-able habitat for vectors and amplifying hosts. Fromhere JE could potentially spread to encompassthe same geographical range as Murray ValleyEncephalitis virus.

Japanese encephalitis - a case study 21

ABSTRACT

As outlined in the previous paper, the Northern Australia Quarantine Strategy (NAQS)conducts surveillance activities as part of its role in the detection and provision ofearly warning of exotic human, animal and plant diseases and pests. Sentinel herdsare an integral part of the strategies used to fulfil this role. The value of the NAQSapproach has been demonstrated by the success of the Japanese encephalitis (JE)sentinel pig program, which has provided essential early detection and epidemiologyof the multiple JE incursions into Australia since 1995.

Feral Pigs and Disease Threats22

PAST JAPANESE ENCEPHALITISINCURSIONS INTO AUSTRALIA1995

An epidemic of JE erupted through the northernhalf of the Torres Straits in March/April 1995,causing three human clinical cases, of which twowere fatal. A relatively high number of people (35/215) sampled were seropositive for JE indicatingwidespread exposure to the virus throughout thetop, western, central and some of the easternislands of the Torres Strait (Hanna, Ritchie, et al.

1996).

In May 1995, an extensive serological survey ofdomestic pigs located in the Torres Strait wasundertaken by the Australian Quarantine andInspection Service (AQIS), the QueenslandDepartment of Health (Q Health), the Queens-land Department of Primary Industries (QDPI) andthe University of Queensland Department ofMicrobiology. Eight outer Torres Strait islands,three inner islands and two mainland comm-unities were surveyed. Between 33-100% of pigsin the outer islands were seropositive as were 7out of10 horses on Badu Island (at this stage onlyBadu Island had a population of horses), althoughno horses were observed displaying any clinicalsigns of JE. Following this, it was decided to uti-lise pigs on the Torres Strait islands as sentinelanimals to monitor for any future incursions.

As a result of these findings, a joint operation wasdeveloped by the four organisations noted above,to set up a network of sentinel herds in the TorresStraits and the northern peninsula area. Therationale behind this move was to attempt todetect any incursion thought to originate in theWestern Province of PNG where the disease wasbelieved to be recently established, and trans-mitted by either infected mosquito vectors ormigratory waterbirds.

1996

Sentinel pigs were established on Badu, Saibaiand Darnley Islands in 1996 by NAQS, in co-operation with Q Health, as a response to theprevious year’s widespread outbreak of JE. Theherds were supplemented by survey visits byNAQS staff to the other islands where pigs, horsesand poultry were tested as well. In March 1996,12/13 sentinel pigs on Saibai Island had sero-converted. No other sentinel pigs or animalstested during surveys of the Torres Straits orNorthern Peninsula Area sero-converted in 1996.

As the majority of human inhabitants of the north-ern Torres Straits had been vaccinated by QldHealth in 1995 in a response to the initial out-break, only the sero-conversion of sentinel ani-mals could provide the necessary indications ofJE activity in the area.

In preparation for the 1996/97 wet season, NAQSestablished a testing regime of domestic pig herdson Saibai and Badu Islands as well as in fivemainland communities located in the NPA. Asentinel cattle herd was established at Bamagaand used to provide serological data on JE,although the sensitivity of cattle as sentinels forthis disease has not yet been established. Thesentinel animal herd operations were supple-mented by a series of surveys covering the TorresStraits, Northern Peninsula Area and Cape YorkPeninsula.

This was a collaborative program with samplescollected from the NAQS sentinel herds, while theserological testing was performed by Q Health.Duplicate samples were tested for other exoticanimal pathogens by the Australian Animal HealthLaboratory (AAHL), Geelong. Results weredistributed to all participating and interestedagencies and feed back was provided to thelocal inhabitants. In addition to these activities,samples of serum from wild migratory waterbirdscollected by cannon netting were tested for JE aswell as other avian diseases.

1997

Results from the 1996/97 NAQS programindicated sero-conversion of sentinel pigs onSaibai Island only. These pigs sero-converted inMarch 1997, showing that infection of the topwestern islands was a repetitive event probablylinked to the seasonal changes present at thistime of year. A general animal survey wasconducted by NAQS staff of the PNG borderregion with Irian Jaya in July/August 1997. Sero-logical results indicated widespread exposure ofdomestic and feral pigs to JE, particularly inWestern Province. This information was sharedwith Q Health and John Mackenzie’s group at TheUniversity of Queensland who were conductinghuman serosurveys, mosquito trapping and virusisolation studies in this area. Their results alsoindicated that a potential epizootic of JE wasoccurring in southern PNG (Siba, P. and Macken-zie, J. pers. comm. 1998).

Japanese encephalitis - a case study 23

Due to the threat posed by the huge potentialreservoir of JE virus in PNG, consultation betweenall participating agencies resulted in the decisionthat NAQS should increase JE surveillance forthe 1997/98 program, to provide better earlywarning of any major incursion into the TorresStraits and northern Australia.

It was generally acknowledged that usually theJE virus tended to cycle a couple of times inpreviously naive pigs, before spilling over into thehuman population (Chu & Joo 1992). However,due to the lag in the development of protectiveantibodies by the pigs and processing times forthe samples, diagnostic results may not give muchadvanced warning prior to development ofclinical disease in unprotected humans. In spiteof this, results from sentinel pigs are invaluablefor confirming an incursion and delineating itsdistribution. This information is essential for theplanning of response actions by health authori-ties as well as State animal health authorities.

1998

The 1997/98 NAQS operational plan expandedthe sentinel pig herds to include Saibai, Boigu,Badu, Moa (St. Pauls community) Mabuiag andYam Islands. Other islands were covered by sixmonthly surveys. However since the 1995outbreak on Badu, island communities had beensteadily reducing numbers of domestic pigs onmost islands, partly due to fears of pigs harbour-ing JE, which could be transmitted to humans.This caused some problem with low numbers ofpigs of a suitable age available for sentinel herdsor bleeding as part of surveys. Yam Island had tobe dropped from the sentinel herd program asthere were no suitable pigs present, while islandssuch as Boigu had a total of only four pigs of asuitable age (3 -12 months) for JE testing.

In spite of these difficulties, sufficient numbers ofpigs were tested in December 1997 and January1998 to ensure that there were no indications ofJE in the Torres Straits at that time. Themonsoonal wet season commenced somewhatlater than usual, in February with a tropical lowcentred in the Gulf of Carpentaria. This producedstrong north-westerly winds to sweep across theTorres Straits and Cape York Peninsula accom-panied by torrential rain. PNG had previously beensuffering from an extended drought for much of1997.

A popular hypothesis used to explain the sporadicincursions of JE into the Torres Straits and theAustralian mainland is that of wind borne infectedmosquitoes. Optimum conditions for an incursionare thought to include high populations ofmosquito vectors in Western province of PNGfollowing stagnation of water due to drought anda strong low - pressure weather system estab-lished in the Gulf of Carpentaria. These condi-tions are thought to enable large quantities ofinfected mosquitoes to be carried from southernPNG and to be deposited over host-rich portionsof the Torres Straits and Cape York Peninsula(Ritchie, S. pers. comm. 1998).

The first indication of an incursion of JE into TorresStraits was the nearly simultaneous sero-conversion of sentinel pigs on Badu and MoaIslands and the diagnosis of a clinical case in aneleven year old unvaccinated boy resident onBadu in March 1998 (Hanna, Ritchie, et al. 1999).This was rapidly followed by seroconversions ofsentinel pigs on Saibai, Mabuiag Islands and atSeisia on the Australian mainland (March – April1998).

This evidence from the NAQS sentinel herdsindicated that an incursion of unprecedentedproportions was occurring. NAQS personnelimmediately initiated a response aimed atdetermining the extent of the incursion and therate of spread. Surveys were conducted tosample all pigs on islands and regional mainlandcommunities not containing sentinel animals.Results indicated that all of the islands in TorresStraits contained pigs which had seroconvertedexcept for Warraber and Kadel Islands. Pigs onHammond Island were also seropositive whentested, causing some alarm there, and on nearbyThursday Island.

As only residents on the northern islands hadpreviously been vaccinated, the results from theNAQS surveillance resulted in prompt action fromQ Health to extend its human sero-surveillanceand response actions beyond Badu Island to othercommunities. Within weeks of the initial humanclinical case and seroconversion of pigs, asecond human case was diagnosed (23/3/98) ina fisherman in the Mitchell River region of CapeYork (Hanna, Ritchie, et al. 1999).

Feral Pigs and Disease Threats

The NAQS response activities were extended asa result of discussions with Q Health and QDPIto include surveillance of areas on the west coastof Cape York and the Gulf of Carpentaria. Thisresponse centred on the rapid collection of serafrom domestic and feral pigs in the Pormpuraawand Kowanyama area to confirm the incursion andto discover the extent of its spread. ConcurrentlyQ Health conducted a sero-survey of human resi-dents of this area while other NAQS staff con-ducted cannon netting operations in the Karumbaregion to sample waterbirds for JE and other aviandiseases. The joint NAQS / QDPI helicopter andground based survey of the Mitchell River area,commenced at the beginning of April 1998, andcollected 114 feral pigs and 20 domesticated pigs,from which sera was obtained. This was sent toQ Health’s laboratory in Brisbane and to AAHL.Results from the adult feral pigs were complicatedby concurrent exposure to endemic flaviviruses,but 6/20 of the domestic pigs showed serologicalevidence of exposure to JE only. This result cou-pled with the recovery of JE virus from sentinelpigs on Mabuiag Island and at Seisia confirmedthe extent of the incursion.

Human serological testing failed to detect anyfurther unequivocal human cases, however theseriousness of the incursion prompted the deci-sion by Q Health and Commonwealth Health tohold a conference to discuss the current situationand determine future courses of action (Hanna,J. pers. comm. 1998). The conference, which washeld in Cairns on the 8th and 9th July 1998, in-cluded all the major stakeholders. The participantsof the conference recognised the value of theAQIS / NAQS sentinel animal and survey activi-ties and requested a major extension of the sen-tinel animal program to cover coastal mainlandareas south of the existing sentinel herds. AQISagreed to implement new sentinel pig herds atthe following locations; Badu Island, Bamaga, OldMapoon, Wathaneen (Aurukun), Baas Yard(Pormpuraaw), Normanton and Hopevale inQueensland; two pig herds in the Darwin regionof NT and pig herds at Kununurra and Broome inWA. These new herds were established in late1998, in time for the 1999 wet season.

1999

No evidence of any Japanese encephalitis (JE)activity was detected from serological monitoringof the NAQS sentinel herds in QLD, NT or WAduring the 1999 wet season. However there is

serological evidence that an epidemic of Kunjincaused sero-conversion of most sentinel pig herdsin Queensland during this period. Summary re-sults from testing conducted by Q Health fromsamples obtained from the NAQS sentinel herdsare presented in Table 1.

Table 1. Seroconversions due to Kunjin in

Queensland sentinal pig herds, 1999

Site Date of InitialSeroconversion

Normanton 2 March 1999

Baas Yard 23 March 99

Wathaneen 18 April 99

Old Mapoon 5 May 99

Injinoo 31 May 99

Badu Island Not detected.

Hopevale Not detected.

Serological data suggests that the initial focus ofthe Kunjin epidemic originated in the Gulf(Normanton) and progressed steadily northwardsalong the west coast of Cape York Peninsula butfailed to cause seroconversion of the sentinel pigson Badu Island. However there is further evidenceof flavivirus activity in pigs in Umagico (nearBamaga) and on Moa Island (Kubin) recordedfrom NAQS survey conducted in June 99. Theexact cause of these serological results cannotbe determined but it is likely that this was afurther extension of the previously describedKunjin epidemic.

2000

Flavivirus activity was detected very early in 2000,with the most significant event being the sero-conversion of sentinel pigs on Badu Island to JE.These pigs were negative on the 7th January andyet by the time of the next bleed on the 17th

January, all had significant haem-agglutinationinhibition (HAI) titres indicating exposure to JE.Concurrent testing of sera at AAHL confirmed theincursion through positive competitive enzymelinked immuno-sorbent assay (C-ELISA) andserum neutralisation test (SNT) results. Howeverpolymerase chain reaction (PCR) results werenegative. Further sampling of the sentinel pigsover subsequent weeks showed greater than four-fold increases in the JE titres indicating that JE

24

was the most likely cause of the seroconversions.Testing of samples by serum neutralisation andvirus isolation confirmed the diagnosis with threeisolates of the virus being detected.

Subsequent serological sampling of domestic pigsfrom Moa Island was conducted in May 2000, aspart of the annual NAQS survey of Torres Strait.Results indicate that these pigs had also beenexposed to the JE virus. Analysis of theseresults is continuing.

On the mainland, all other sentinel pigs (exceptthose at Injinoo airport and those near Hopevale)also seroconverted over February and March.However, serological analysis indicated that theseseroconversions were most likely due to theendemic flaviviruses; Kunjin and MVE rather thanto JE. HAI titres in the order of 320 - 640 wererecovered from the sentinel pigs to these viruses,indicating considerable vector activity over thisperiod.

CONCLUSIONSSentinel animals form an essential part of exoticanimal disease surveillance and complementsurvey activities conducted by NAQS. Resultsfrom these activities can provide early warningand monitoring of exotic disease incursions,allowing the early implementation of responseactivities.

There is a need to further elucidate the epidemi-ology of Japanese encephalitis, in epidemicregions, particularly with reference to the roleplayed by arthropod vectors and non-human hostsincluding wild birds and feral pigs. Little is knownabout the migratory habits of various waterbirdsand their potential to act as amplifying hosts forJE. This data will need to be added to the exist-ing information about the epidemiology of thedisease involving humans, domestic pigs, horsesand mosquitoes, in endemic areas.

The role of JE resistant animals, such as cattle inproviding potential zooprophylaxis, is unknown,but could be an important factor in the epidemiol-ogy of a JE incursion into northern Australia.

The development of high serological titres to MVE,Kunjin and other domestically occurring flaviviruses by sentinel pigs, certainly complicatesserological diagnosis of potential JE incursions,and better serological techniques are urgently

required to clearly distinguish between theseviruses. It is unknown whether exposure todomestic flaviviruses produces any crossprotectivity in animals and humans to JE.

While these issues are being investigated, NAQSis continuing to conduct surveillance for JE usingsentinel animals and survey activities. Thisinformation will be crucial to understand how thedisease spreads and to assess the risks ofpotential incursions of JE into Australia.

GLOSSARYamplifying hosts. Hosts which, when infected bya virus, have a limited immune response whichallows the virus to replicate in vast numbers withinthe animal’s tissues. These hosts then become avirus reservoir and allow vectors such as bloodsucking mosquitoes to pick up an infectious virusload when obtaining a blood meal from theseanimals.

clinical disease. The animal has been infected (oraffected) with the disease and is showing signsindicative of the disease (eg. fever, coma).

cross reactive serology. A serological test whichmay produce a positive result when it detectsantibodies to a number of closely relateddiseases. The ELISA test for JE will also producepositive responses to antibodies which protectagainst MVE or Kunjin. Hence a positive resultcan not differentiate which disease the animal hasbeen exposed to.

flavivirus. A member of the genus Flavivirus inthe family Flaviviridae. These are Group B arthro-pod-borne viruses. Most are transmitted bymosquito vectors. There are approximately 80 dif-ferent members but include the important arthro-pod-borne viruses. Most are transmitted by mos-quito vectors. There are approximately 80 differ-ent members but include the important virusesof medical and veterinary interest, such as - Japa-nese encephalitis (JE), Murray Valley encephali-tis (MVE), Kunjin, West Nile, louping ill &Wesselbron viruses.

neurological sequelae. Damage to the centralnervous system (brain & spinal cord) whichresults in histo-pathological or clinical signs (eg.with JE, sequelae may include loss of memory,fine motor control, blindness, or coma).

Japanese encephalitis - a case study 25

Feral Pigs and Disease Threats26

seroconversion. The infected anima’s immunesystem has responded to the infection by produc-ing antibodies which can be detected in theserum by serological testing.

serological data. A collection of results from sero-logical testing of animals or humans.

serologically naive. The same as immunonaive -the animal has not been exposed to the diseaseand has not developed protective antibodies tothe disease and is negative when tested by sero-logical techniques.

seropositive. An animal which give a positiveresult to serological tests used to detect specificantibodies.

subclinical exotic disease. The animal or humanhas been infected with the disease but is notshowing any overt signs of the disease.

viraemia. When a virus is circulating in the bloodand tissues of an infected animal or human.

virus isolation. Techniques used to attempt torecover live intact virus from animal (or human)blood or tissues. The techniques usually attemptto grow these recovered viruses in tissue culturesor embryonated eggs, so that they can be identi-fied either by characteristic lesions produced orserology or electron-microscopy.

zoonotic disease. A disease which can be trans-mitted from animals to humans.

zooprophylaxis. The use of animals to control orprevent infection (of humans) by a specificdisease. For example, cattle are attractive targetsfor potentially JE infected mosquito vectors butas they do not develop high levels of viraemiaand are not affected clinically by JE. They there-fore can act as immunological “sponges” to “soakup” many of the potentially infective doses of JEbefore they are transmitted to susceptible ampli-fying hosts such as pigs or susceptible end hostssuch as humans or horses.

INTRODUCTIONJapanese encephalitis is a severe neurologicaldisease that afflicts nearly 50,000 people a year,primarily in tropical and temperate Asia (Burke &Leake 1988). JE is fatal in 25% of clinical humancases, with a remaining 50% sufferingneurological sequelae ranging from coma andparalysis to psychological disturbances (Burke &Leake 1988). JE is caused by a flavivirus, withbirds, especially ardeid waders, as the naturalhost. Mammals tend to be incidental hosts,although pigs develop a high transient viraemia,serving as the major amplifying host (Burke &Leake 1988). Although there are outbreaks inareas with low numbers of pigs, most severeoutbreaks have been in rural areas with largepopulations of domestic pigs. JE is transmittedby mosquitoes, especially Culex tritaeniorhynchus

in SE Asia.

Prior to 1995, JE was unknown in Australia. InApril of 1995, three human cases of JE, twoultimately fatal, occurred on Badu Island in theTorres Strait (Hanna et al. 1996, Lee this volume).Subsequent investigation showed that mostTorres Strait Islands from Badu to the Papua NewGuinea border had JE activity that year (Hannaet al. 1996), and that the mosquito C. annulirostris

was the probable vector (Ritchie et al. 1996).

The intensity of the outbreak on Badu Is. was, inpart, due to the large numbers of domestic pigs(n = 200) that lived in close contact with theresidents (Ritchie et al. 1997). On Badu Is. JEvirus infected 17% of the human population while1 in 300 C. annulirostris carried JE virus. Nearlyevery other house had a backyard piggery and itwas estimated that all 200 pigs became infectedwith JE (Hanna et al. 1996, Ritchie et al. 1997).Furthermore, clogged drains, stagnant swampsand overflowing septic tanks created idealbreeding grounds for the vector mosquito withinthe community. Thus the residents of Badu Is.were literally living within a maelstrom of JE virus(Ritchie et al. 1997).

After a two year hiatus (excepting thenorthernmost islands of Saibai and Boigu), JEvirus returned in 1998 with the largest outbreakto date. Sentinel pigs throughout the Torres Straittested positive for JE virus, and the outbreakextended as far south as the Mitchell River inwestern Cape York, where the first human caseof JE on the Australian mainland was recorded(Hanna et al. 1999, Lee this volume). Again,activity on Badu Is. was intense, with all pigstested seroconverting, 43 isolates of JE virus frommosquitoes and, despite vaccination of most of

the residents, a human case.

Mosquitoes and pigs: the critical mix forJapanese encephalitis

RITCHIE, S., JOHANSEN, C., VAN DEN HURK, A., LEE, J., MACKENZIE, J.

1Tropical Public Health Unit, Queensland Health, Cairns2Department of Microbiology and Parasitology, The University of Queensland, St Lucia3Australian Quarantine Inspection Service, Airport Administration Building, Cairns

Mosquitoes and Pigs - the critical mix 27

ABSTRACT

This paper compares rates of infection of pigs by Japanese encephalitis in theTorres Strait and on Cape York Peninsula. Pigs may become infected with JE ifbitten by mosquito vectors, and this may then increase the risk that the disease willbe transmitted to humans. During 1995 and 1998 when rates of infection of domes-tic pigs on Badu Island was high, seroconversion of pigs on Cape York was low.The difference may be due to the high exposure of domestic pigs on Torres StraitIslands to mosquito attack, relative to wild pigs on Cape York, and suggests thatferal pig populations may not induce epidemics of JE on Cape York.

Feral Pigs and Disease Threats

Table 1. Comparison of Japanese encephalitis activity in Badu Is. and selected regions of Cape York.

Parameter Badu Is. Badu Is. Cape York Cape York

(1995) (1998) (Kowanyama, (NPA 1998)

Pormpuraaw 98-99)

Pig seroconversions (ca.%) 100% 100% 65% 50%Mosquito JE virus isolates 8 43 0 0Duration of seroconversion < 2 months < 2 months ? > 2 months

Human cases (ca. % infections) 3 (17%) 1 (50%) 1 (0.2%) 0 (0%)

Table 2. Comparison of key parameters associated with JE virus transmission in Badu Is. and selectedregions of Cape York. Note, it is assumed that there are numerous dogs and humans as “alternative hosts”

to swine in all communities.

Parameter Badu Is. Badu Is. Cape York Cape York(1995) (1998) (Kowanyama, (NPA 1998)

Pormpuraaw 98-99)

C. annulirostris population 100 – 500/trap 500 – 2000/trap 500 – 1000/trap 100 – 500/trapDomestic pig population Ca. 200 Ca. 120 Ca. 5 Ca. 50Feral pigs Yes, ? number Yes, ? number Yes, locally high Yes, high locallyLivestock Horses Horses Cattle Cattle, horsesOther alternative hosts No wallabies No wallabies Many wallabies Many wallabies% feeding on pigs byC. annulirostris 33%1 33%1 1% 20%

1Based on data from mosquitoes collected in 1996.

However, the story was quite different on CapeYork. The Mitchell River case was the onlyevidence of JE virus infection among 700 peoplewhose blood was tested (Hanna et al. 1999).Although domestic pigs bled from Baa’s Yard nearthe Mitchell River and the sentinel pigs in thenorthern peninsula area of Cape York (NPA) didindeed seroconvert for JE virus, the level of JEvirus activity waned in comparison to Badu(Table 1). Pigs seroconverted slowly, withsporadic seroconversions among NPA pigs overa three month period from March – May 1998.Some pig herds in the area (e.g., Bamaga) didnot show a single seroconversion! Why was theJE activity in Cape York so different, so muchlower, than on Badu Is.? This paper addressesthis issue and, in doing so, explores the role thatferal pigs may play in the natural history of JEvirus on the Australian mainland.

FERAL PIGS, DOMESTIC PIGS ANDMOSQUITOES

The potential key to the difference in JE activityin Cape York and Badu Is. can perhaps best beelucidated by contrasting JE activity and the major

components associated with JE virustransmission for Badu Is. and Cape York (Tables1, 2). We can clearly see that on Badu Is. thevirus activity was greatest in 1998, reflecting thefourfold increase in mosquito populations over1995. Because of the vaccination program (ca.90% of population of 800 received at least onedose of JE vaccine), humans were not bled toestimate the human infection rate. However, asa single unvaccinated person did develop clinicalJE, the rate of human infection can be estimatedusing the clinical/subclinical ratio for JE in man.

In the 1995 outbreak, it was estimated that only1/50 infections resulted in clinical symptoms. Soin 1998 the single clinical JE case suggests that50 of the unvaccinated people (80) may havebeen exposed to JE virus, a total of 62%. To beconservative, we suggest that 50% may havebeen infected. This is backed up by the highnumber of virus isolations (43) obtained in 2 nightsof mosquito trapping on Badu.

28

In the NPA, mosquito numbers were lower (Table2) but comparable to those on Badu Is. duringthe 1995 outbreak. Thus mosquito abundancedoes not seem to account for the low JE activityin the NPA. However, there are some criticaldifferences in vertebrate host populations.Domestic pig populations were high on Badu, withmost of the pigs in backyard piggeries. NPA pigswere in larger communal piggeries (ca. 10-20/site), with many located outside urban areas(Bamaga, Umagico). This may account for thelack of human infection, but does not explain why

JE virus activity itself was low in the NPA.

Perhaps a high level of alternative hosts on theNPA dampened JE transmission, a situationtermed zooprophylaxis. In Sri Lanka, Pieris et al.(1993) observed that in rural areas with highpopulations of domestic pigs, pigs are subject tosynchronous JE transmission: all pigs seroconvertwithin ca. 2 cycles of the virus in pigs (ca. 1month), a situation comparable to Badu Is.However, in areas where cattle are common, pigseroconversion is spotty and staggered (termedasynchronous), much like the situation in the NPAin 1998.

Could zooprophylaxis have reduced JEtransmission in the NPA? The Badu communitydid have many horses but no other largevertebrates except man, dogs and pigs. In theNPA, cattle were common and horses andwallabies were also observed. Perhaps mosttelling is the study of feeding preferences of C.

annulirostris in Kowanyama by Kay et al. (1979)where in stable traps baited with pigs, dogs,calves, chickens, kangaroo and man, pigs were2nd in preference after calves to bloodfeeding C.

annulirostris. However, bloodmeals from wildmosquitoes contained few (ca. 3%) feedings onpigs, with most (ca. 60%) on dogs. This, to alarge extent, reflects the fact that mosquitoes werecollected within the community where dogs werecommon.

Bloodmeal analysis of mosquitoes collected in theCape during the 1998 JE outbreak demonstratedthat C. annulirostris did not feed predominantlyon pigs. Less than 2% of the C. annulirostris

collected in rural areas near Kowanyama andPormpuraaw had fed on pigs, with a majority offeeds upon marsupials (39% and 90% for therespective communities). However, wheremosquitoes were trapped near penned domestic

pigs, such as Baa’s Yard near the Mitchell Riverand Seisia in the NPA, feedings on swine werehigher (86% and 20%, respectively).

These data, together with the low rate ofseroconversion in the NPA in 1998 despite animmunonaive pig population, suggest that feralpig populations may not induce epidemics of JEin Cape York. Why? Certainly populations of feralpigs are high in the NPA and Cape York area.Also crossreacting flaviviruses common to thearea, such as Kokobera, Alfuy and Kunjin virus,may serve to protect pigs from infection by JE.Zooprophylaxis induced by other fauna such ascattle and, especially, wallabies may reduce JEvirus transmission. But perhaps the mostimportant factor is the relationship between feralpig populations and C. annulirostris. While thismosquito readily feeds on pigs, few mosquitoescollected away from penned swine contained pigblood. It may be that feral pigs are not as good ahost as domestic pigs. Feral pigs are generallyactive at night (J. Shield and J. Lee, personnelcommunication) and may actively engage in anti-mosquito behaviour or flee from mosquito-infestedareas. Domestic pigs are confined in small pensand, as they are fed during the day, may sleepduring much of the night. This would lead tonumerous full bloodmeals by mosquitoes,enhancing virus transmission. Finally, feral pigswallow in mud, creating a coat of crusty soil thatmay protect them from mosquito bites (J. Shield,personal communication). Further work is neededto elucidate the enigmatic role that ferals pigs playin the ecology of JE virus in Cape York.

Mosquitoes and Pigs - the critical mix 29

Feral Pigs and Disease Threats

Veterinary status of feral pigs on Cape York

SHIELD, J.

Queensland Department of Primary IndustriesCairns

INTRODUCTION

Although it is the most abundant of the mammalson Cape York and is relatively large and conspicu-ous, the feral pig has not been well studied as asubject of veterinary interest. There have beenonly a few surveys on its health; much of what weknow is based on assumption and speculation.Some of the assumptions and speculation aresoundly based on the knowledge that the animalis the same basic Sus scrofa as our domesticpig, and we do know a lot about the diseaseproblems in our piggeries. Many of the otherperceptions on feral pig health, however, arepoorly founded on emotive and non-scientificjudgements not unlike those that keep half of theworld’s population from eating any pig meat. Thecurrent situation is slightly ridiculous, where most(non indigenous) people on Cape York wouldrather starve than eat even the choicest porkerrunning wild on his property, and yet foul-smelling old boars from the same place find aready market on sophisticated dinner tables inEurope.

Feral pigs carry a number of zoonotic diseasesof high importance to public health; they are alsoof inestimable significance as potential hosts ofexotic animal disease plagues like foot and mouthand swine fever. With these qualificationshowever, infectious disease does not appear tobe a huge factor in the population dynamics ofthe feral pig. On Cape York Peninsula, infectiousdiseases have a minor impact on populationscompared with the effects of seasonal(un)availability of food. The reproductive success

of a Cape York sow is more dependent on herability to find food, and on the predations of birds,dingoes and snakes than it is on her exposure toleptospirosis or parvovirus.

Notwithstanding the above, the following attemptis made to list and discuss some infectiousdisease conditions that do or may infect the feralpigs of the region.

ENDEMIC DISEASESTuberculosis (TB):TB has just recently been eradicated from theQueensland cattle herd, with the use of massivetesting programs and slaughter of large numbersof cattle. Throughout the program (20 years),there was a fear that feral pigs would be infectedthrough close contact with tuberculous cattle orthrough eating the bodies of infected slaughteredanimals. There was never any evidence that thishappened in Queensland although it was afeature of some of the infected areas of the North-ern Territory where there was a lot of TB in thebuffalo and a very close association between pigs,swamps and buffalo.

Brucellosis:The same national (BTEC) disease eradicationscheme that eradicated TB also removed bovinebrucellosis. Pigs however have their own brucel-losis and the bacterium Brucella suis is presentin many feral populations including those on CapeYork. It causes reproductive disease and otherproblems in the pigs. The main impact on ushowever is that Brucella suis can:

30

ABSTRACT

Feral pigs carry several zoonotic diseases of significance to publichealth. They are also of inestimable significance as potential hosts ofexotic animal disease plagues like Foot-and- Mouth and Swine Fever.This papers lists and discusses some infectious disease conditionsthat are prevalent in feral pigs on Cape York.

Veterinary status of pigs on Cape York

• Cause human disease• Infect cattle where it may create ‘false

positive’ tests in disease surveys for bovinebrucellosis

Leptospirosis:Feral pigs commonly contract leptospirosisbecause it is a disease of wet conditions and pigschoose such conditions. Leptospirosis is also veryimportant because it is a dangerous zoonosis:people handling live or dead feral pigs should takeprecautions to prevent becoming infected particu-larly through contact with the pigs’ urine or otherbody fluids.

Melioidosis:This is another bacterial disease found in feralpigs and capable of causing serious illness in man.In pigs it causes usually small abscesses and nosignificant illness; in humans however theabscesses can be severe and sometimes fatal.Humans are more likely, though, to contract thedisease from infected soil in cuts and scratches,than from contact with even an infected pig.

Sparganosis:This is a condition caused by a small tapewormthat cycles through cats, invertebrates, tadpolesand frogs. Pigs become infected through eatingone of the above smaller animals carrying anintermediate stage of the tapeworm; humans canbecome infected similarly by ingesting one of thesmaller hosts or while slaughtering or eating aninfected pig. Sparganosis is listed because it isfrequently discussed and historically quite signifi-cant; it is however of little clinical importance andnow rarely found.

There are a host of other endemic diseases thatmay affect the feral pig; in some cases, serologi-cal tests on blood samples show that exposuredoes occur but without any evidence that this isof more than academic interest.

EXOTIC DISEASESIt must be assumed that every pig disease that isnot already in Australia is an exotic threat andthat it is capable of infecting the feral pigs of CapeYork. Some of these, like foot and mouthdisease, represent agricultural disaster of thehighest order, while others would have such asmall impact that their introduction would not evenbe noticed. For the purpose of brevity, I haveselected, from a long list, only a few diseases for

special mention: Foot and mouth disease, Swine

vesicular disease, Vesicular exanthema, Vesicu-

lar stomatitis, Classical swine fever, African swine

fever, Aujeszky’s disease, Porcine reproductive

and respiratory syndrome, Transmissible gastro-

enteritis. These are the most feared of the exoticviral diseases of pigs. They are highly infectious,spreading from pig to pig through close contact.They occur throughout many parts of the worldand some of them are relatively close to us inparts of South East Asia. All of these diseaseswould be expected to produce high morbidity inpigs. More importantly, their presence in Australiawould cause immediate trade effects, which woulddamage the embattled pig meat industry severely.

Screw worm fly (SWF):SWF is endemic in Papua New Guinea in areasonly a few kilometres from Australia’s northern-most islands. This lethal blowfly lays it’s eggs ina wound on any warm-blooded animal and theresulting fly ‘strike’ can cause severe injury andoften death. If SWF reached Cape York it couldbe expected to affect feral pigs as well as cattle,horses and other domestic, wild and feral animals;humans too can become victims.

Japanese encephalitis:This mosquito-borne disease reached the main-land of Australia in 1998 when it was found in theMitchell River area and near the ‘Tip’. It has notyet been determined if this was a transient incur-sion or if JE is now endemic in Cape York. Whileprimarily a human disease problem, JE does alsoaffect pigs (causing reproductive problems) andhorses (causing encephalitis), but the impact ofthis effect is not likely to be noticed in the CapeYork population. The paper by Dr Ritchie elabo-rates on the interdependence between the JEvirus, pigs and the Culex mosquito.

The arrival of JE has once again focussed publicattention and imagination on a desire for CapeYork to be free of feral pigs. Like any diseasecontrol agency, the Department of PrimaryIndustries would like nothing better.

Cysticercosis and trichinosis:These are serious diseases caused by wormparasites. Cysticercosis is caused by the porktapeworm Taenia solium and it can cause severeillness and death in humans. It is now endemicin Irian Jaya and could spread throughout theisland of New Guinea and threaten the islands ofthe Torres Strait.

31

Feral Pigs and Disease Threats

Trichinosis is caused by tiny Trichinella nema-tode worms. It is widespread throughout the worldand causes disease and even death in humans.

Both of these diseases affect people who eat theimproperly cooked meat of infected pigs: theytherefore are commonest in areas where thesediet habits are accepted in culture and tradition.In the islands of the Torres Strait and in parts ofCape York, cysticercosis and trichinosis could beexpected to survive and cause human suffering.

EXOTIC DISEASE RESPONSEAn Australian outbreak of any of these wouldtrigger a response involving all states and theCommonwealth and the implementation of anexisting framework of actions underAUSVETPLAN, The Australian VeterinaryEmergency Plan. AUSVETPLAN outlines indetail the strategies required for each disease,each industry and each agency.

In the past Australia has successfully eradicatedoutbreaks of a number of these diseases includ-ing Foot and Mouth and Classical swine fever. Inrecent years AUSVETPLAN has been tested insuccessful eradication programs against poultryexotic diseases.

One thing that is repeatedly demonstrated bydisease outbreaks and by simulations, is theextreme difficulty of eradicating a disease once itbecomes established in a wild or feral popula-tion. This is considerably more of a problem ifthe terrain is inaccessible. Cape York, becauseof it’s large pig population and huge areas ofinaccessible country, represents perhaps theworst place to have to eradicate an exoticdisease.

32

33

Feral pig in wallow

Pig damage to lowland swamp vegetationin Edmund Kennedy National Park: the fencedarea has been protected from pigs

Pigs captured in a fence trap inEdmund Kennedy National Park

The ecological damage of pigs tends to behighly concentrated in certain micro-habitatslike the fringes of swamps and wetlands inMalaleuca woodland

Photographs: Jim Mitchell

35

Pigs, when in close proximity to humanpopulations and exposed to mosquitoattack, as in domestic piggeries on TorresStrait Island, increase the risk of transmissionof Japanese Encephalitis to people(Photograph: Jonathan Lee)

Pig caught in a trap operated by the communitybased feral pig trapping program, approximately2000 pigs are killed annually under this scheme(Photograph: Jim Mitchell)

The feral pig (Sus scrofa)(Photograph: Jim Mitchell)

PART 3

POPULATION BIOLOGYAND CONTROL OF

FERAL PIGS

37

The dynamics of feral pig populations inAustralia: implications for management

GILES, J. R.

Conservation and ResearchZoological Parks Board of New South Wales

INTRODUCTIONThere have been several large, long-term studiesof the dynamics of feral pig populations. The mostcomprehensive have been by Giles (1980) andChoquenot (1994) in semi arid rangelands inNSW, Saunders (1988, 1993a) in sub-alpine NSWand Caley (1993) in the Wet Tropics. Thesestudies have yielded broadly similar pictures ofthe biology and ecology of the species withregional and temporal variation in its populationdynamics, and have been drawn upon heavily inthe material presented in this brief review.

DISTRIBUTION AND ABUNDANCEChoquenot et al. (1996) gave an excellentsummary of contemporary information ondistribution and abundance. Feral pigs are widelydistributed in Australia with the major populationsassociated with river systems and floodplains.

Estimates of population densities are summarisedby Choquenot et al. (1996). These range fromless than one pig per km2 in semi-arid rangelands(Giles 1980, Choquenot 1994) to up to 20 km2 inwetlands, swamps and sub-tropical floodplains(Giles 1980, Saunders 1988, Hone 1990, Dexter1990). Local population densities around cropscan reach higher levels.

Giles (1980) reported considerable temporalvariation in population density in both theMacquarie Marshes and on an ephemeralfloodplain in western NSW, with numbers risingsharply after drought, peaking and then decliningas seasonal conditions deteriorated. Pigs bornat around the time of the population peak had lowjuvenile survival rates and grew to be muchsmaller in both adult liveweight and skeletal sizethan those born early in the post-drought eruption.

Essential requirements for permanent populationsinclude water, shelter and suitable food. Asmonogastrics, pigs have a poor capacity to digestcellulose and cannot survive on hayed-off grassesand forbs. They are opportunist omnivores witha strong preference for succulent greenvegetation, a wide variety of animal material, fruitand grain. They also utilise seeds, bulbs, rootsand corms, particularly during dry periods.

Protein requirements of pigs are high comparedto ruminants. At dietary crude protein intake below15%, Giles (1980) found that pre-weaningmortality was high. The energy requirements ofpigs are also relatively high: this is particularlytrue of sows in the last month of pregnancy andduring lactation.

ABSTRACT

This paper reviews studies of the population dynamics of feral pigs, andconsiders the implications of the population biology of pigs for effectivecontrol. Pigs have a capacity for rapid population increase. Unless populationsare reduced by 70% or more, recovery to pre-control levels is likely withintwo years. This rapid recovery means that control efforts must be able toachieve very large reductions over short periods of time if control is to beeffective. Otherwise, control programs are likely to end up as expensive sus-tained yield harvesting operations of little value in reducing unwanted im-pacts of feral pigs.

Dynamics of feral pig populations 39

Table 1. Key reproductive attributes of feral pigs in three habitats in NSW (from Giles 1980, Saunders and

Giles 1995). Numbers of samples are given in parenthesis.

Attribute Kosciusko National Macquarie Marshes Western NSW

Park (ephemeral

floodplain)

Mean litter size at birth 6.25 (66) 6.92 (93) 6.29 (145)

Birth frequency per year 0.84 1.93 1.93

Table 2. Mortality patterns of feral pigs from three habitats in NSW (after Saunders and Giles 1995 andGiles 1980). Statistics tabulated are age class at commencement of the year, probability of surviving to agex (l

x), probability of dying between ages x and x+1 (d

x) and mortality rate q

x, the proportion of animals alive

at age x that die before age x+1 (dx/l

x).

Kosciusko NP Macquarie Marshes Western NSW floodplain

Age class lx

dx

qx

lx

dx

qx

lx

dx

qx

0 1.00 0.85 0.85 1.00 0.89 0.89 1.00 0.94 0.941 0.15 0.06 0.40 0.11 0.03 0.27 0.06 0.02 0.342 0.09 0.02 0.22 0.08 0.02 0.28 0.04 0.01 0.253 0.07 0.02 0.29 0.06 0.03 0.45 0.03 0.01 0.28>4 0.05 0.02 1.00 0.03 0.33 1.00 0.02 0.02 1.00

Periodic protein shortage is probably the factorwhich most commonly limits feral pig populationgrowth in Australia. This is likely to occur in thedry season in the Wet Tropics, during winter athigh elevations and during periods of low rainfallor lack of flooding in the rangelands, particularlywhen populations are at relatively high density.

REPRODUCTIONGiles (1980) found that feral pigs in western NSWfirst bred at 20 -25 kg liveweight if they were underabout 18 months of age, and at 25-30 kg if theywere older. Unlike the Eurasian wild boar, whichhas a distinct breeding season, feral pigs havenot been found to exhibit a non-breeding season,except in the Southern Alps of NSW whereSaunders and Giles (1995) found a seasonalanoestrus in autumn and early winter, as occursin wild boar in northern latitudes.

Table 1 summarises key reproductive data fromthree habitats in NSW presented by Giles (1980)

and Saunders and Giles (1995). It is notable thatmean litter size is much smaller than in domesticpigs.

MORTALITY PATTERNSTable 2 summarises age specific mortality foundby Saunders and Giles (1995) and Giles (1980)in three NSW habitats.

The table shows that, in these habitats:

i) a high proportion of pigs die in the firstyear of life;

ii) few animals survive beyond four years ofage

iii) juvenile mortality on the flood plain washigher than in the other habitats but,otherwise, the mortality patterns arebroadly similar.

Population Biology and Control of Feral Pigs40

RATES OF INCREASEThe instantaneous (exponential) rate of increase(r) is the most common statistic used to describethe rate of population growth. The finite rate ofincrease (er) is the ratio of population size fromone year to the next. A population will grow ordecline in response to births, deaths, immigrationand emigration. The instantaneous rate is zerowhen population size is static, positive when ithas increased and negative when it has declined.

In planning pig control programs, one is centrallyinterested in the rate that the population willincrease after the control program. At this time,resources available per surviving pig are likely tobe high, favouring rapid population growth. Ratesof increase of feral pig populations have beenestimated in several studies, and found to varywith prevailing seasonal conditions and the sizeof the population relative to carrying capacity.

Giles (1980) estimated maximum rates of increaseof pig populations in the Macquarie Marshes andon a semi-arid floodplain in western NSW to liearound 0.6-0.7. Hone and Pedersen (1980)reported a rate of increase of 0.57 over the yearfollowing a 58% population reduction in thepopulation on the same western NSW floodplain.This is equivalent to a finite rate of 1.76 (i.e a176% increase in numbers).

Caley (1993) reported a maximum rate of increasein a Wet Tropics population to be 0.065 per month,

which is equivalent to 0.78 per year (a finite rateof 2.18 per year). Saunders (1993b) reported arate of increase after a helicopter shootingcampaign in the Macquarie Marshes to be 1.34(a finite rate of almost 4.0) but noted thatimmigration into the study area possiblycontributed to this very high rate.

Rates of increase (in the absence of immigrationor emmigration) are highly dependent uponmortality in the first year of life, and mortalityaround the time of weaning varies greatly inresponse to availability of high protein food.

Table 3 shows the relationship of juvenile mortalityrate (q

0.5) to the instantaneous (exponential) (r)

and finite annual rate (er ) of increase in a westernNSW population (Giles 1980). Also shown is thenumerical response two years after aninstantaneous reduction of 70%, as might occurin a control program.

Table 3. The relationships in a western NSW population between juvenile mortality and rate of increase, andbetween rate of increase and predicted population size two years after a 70% reduction in initial populationsize (from Giles 1980).

Mortality rate Instantaneous Finite rate of Predicted population size

in the first year rate of increase (r) annual increase (er) two years after 70%

of life (q0.5)

reduction. (N0=100)

0.5 0.920 2.509 189.0

0.6 0.804 2.235 149.8

0.7 0.663 1.941 113.0

0.8 0.410 1.507 68.11

0.9 0.194 1.214 44.2

0.99 -0.225 0.799 19.1

EFFICACY OF CONTROL PROGRAMSPoisoning, trapping, shooting (from the groundand from helicopters) and dogging are the mostcommon techniques used in Australia tomanipulate feral pig populations. While populationreductions of 90 to 100% were achieved b. McIlroyet al. (1989), reductions of 60-80% have beenmost common (Hone and Pedersen 1980, Hone1983, Hone 1987, Mitchell 1988, Choquenot et al

1993, Saunders (1993b), McIlroy and Gifford1997).

Dynamics of feral pig populations 41

From Table 3, it can be seen that a populationreduction of the order of 70% or below, is likely toresult in recovery to pre-control levels within acouple of years if r is about 0.6 or above. Severalstudies have shown that maximum rates ofincrease of this level and above are likely aftercontrol programs.

Because of the capacity of pig populations toincrease rapidly after a control program, it can bedifficult to avoid control efforts ending up asexpensive sustained yield harvesting operations.Maximum population reduction over a short periodof time is clearly fundamental to effective control.This is often difficult to achieve over large areasand in difficult terrain. A capacity to reduce themaximum rate of increase would be of greatadvantage. This would require one or more ofthe following: reduction in frequency of birth,reduction in litter size and reduction in first yearsurvival rates.

Population Biology and Control of Feral Pigs42

Ecology and management of feral pigsin the Wet Tropics

MITCHELL, J.

Department of Natural Resources and MinesCharters Towers, Queensland

ABSTRACT

Information on the ecological impacts of feral pigs (Sus scrofa) in the world heritage listedtropical rainforests of northern Queensland, Australia, is limited. This study quantifies andqualifies aspects of the ecological impact, spatial and temporal digging activities andhome range and seasonal movement patterns of feral pigs. Digging by pigs variesseasonally and according to microhabitat, with the highest levels of digging occurring inmoist microhabitats at the start of the dry season. Feral pigs in this region have definedsedentary home ranges and their distribution patterns appear to be influenced bymicrohabitat factors, including earthworm populations and water availability. Diggingactivity decreases seedling survival rates in moist microhabitats by 36%. Management

strategies should concentrate on a coordinated, regional, community based approach.

INTRODUCTIONThe World Heritage Area (WHA) listed rainforestof the wet tropics region of northeast Queens-land, Australia, covers over 9,000 km2 and is re-garded as natural heritage of outstanding univer-sal value and one of the most significant regionalecosystems in the world. Feral pigs have beenaccused of posing many diverse threats to theconservation values of the World Heritage Area.To understand the “threat” to World Heritagevalues, information is required on the actualimpact of feral pigs on this environment. Clarifi-cation of the impact of feral pigs in relation toseason, severity, diversity and situation is afundamental component of developing amanagement plan.

This paper summarises research currently beingundertaken to acquire quantitative knowledge onthe ecological impact and ecology of feral pigswithin the rainforests of northern Queensland.The research is presented as four studies:spatial and temporal digging patterns, ecologicalimpact of pig diggings, home range and seasonalmovement patterns, and biological parameters.

STUDY SITEThe study site is situated near Cardwell, northQueensland, Australia (18° 16´ S, 146° 2´ E). Thisarea was selected, as significant feral pig

populations existed in the area (Mitchell andMayer 1997) and a variety of habitat types wereavailable within the region, ranging from highlandrainforests to lowland dry rainforest, open wood-lands, marine swamps and mangroves. Thevariety of habitat types was ideal for assessingthe influence of microhabitat and macrohabitatfactors on feral pig spatial patterns.

The study region was divided into three broadmacrohabitats defined by broad biogeographical“areas”; these include highland rainforests, theecotone between rainforests and croppingsystems, and the coastal lowlands. Within eachof the three areas, a number of key microhabitatstermed “strata” were selected based on the pres-ence of pig activity, previously establishedmicrohabitat preferences (Mitchell and Mayer1997) and the major microhabitat types repre-sented in the area.

RESULTS

Spatial and Temporal Digging patternsPatterns of feral pig digging (rooting) activity, bothspatially (between macro and microhabitats) andtemporally (between seasons) were monitoredover a two year period. Associations betweenthese digging patterns and soil moisture andearthworm population levels were also examined.

Ecology of feral pigs in the Wet Tropics 43

Population Biology and Control of Feral Pigs

Three macrohabitat areas were assessed: high-land rainforests, transitional or ecotone betweenrainforest and lowland open forests, and coastallowland. A range of microhabitat strata wasselected within each macrohabitat area, such asswamps, old logging tracks, dirt roads and creeks.Monitoring at 6 -week intervals, assessed diggingactivity on 195 line intercept transects (50m). Adigging index (proportion of each transectdisturbed) was established for each transect foreach of the 10 recording events. The analysisassessed patterns in the digging index acrossseasons (wet and dry) and across microhabitatstrata, within and between the three macrohabitatareas. An index of earthworm populations wasalso monitored to determine their association withthe digging index.

Spatial and temporal patterns of feral pig diggingactivity were detected within the study area,specific microhabitats were found to have signifi-cantly higher digging activity overall and inspecific seasons. Rainfall (soil moisture) appearedto be the major influence on these digging pat-terns; higher rates of diggings occurred in the earlydry season and predominantly in moistmicrohabitats (swamp and creek microhabitats).The mean digging index over all strata within thethree areas was 6% of the soil surface beingrecently disturbed by feral pigs.

Differences in the mean digging index betweenthe three areas were apparent, with the highlandrainforests recording the highest digging index.This suggests either that there were higher pigpopulation densities in the highland rainforestsor that digging is a more prominent foragingactivity in this area. Differences in diggingactivity between the various microhabitats werealso observed in all of the three areas. In thehighlands there was a significantly higher diggingactivity in the swamp and creek strata. Overallranking of stratum preference was for highestdigging index in swamp (19.3%), creek (9.7%)road (8.0%) track (3.0%) and the lowest prefer-ence for the forest floor (0.1%).

In the transitional area, the creek stratum hadsignificantly higher digging activity (8.5%) thanthe track stratum (0.6%). Digging activity wassignificantly associated with soil moisture in thiscreek stratum and tended to be strongly seasonalwith higher digging activity occurring in the dryseason. Digging activity increased in the creeks

when soil moisture levels dropped sufficiently forearthworm populations to survive and repopulatethe creek bed. Dropping water levels alsoexposed fresh soil surface and flood debrismaterial, which is attractive to foraging pigs.

No significant differences in overall digging activ-ity were detected among the strata in the lowlandarea. Overall ranking of digging preference wasfor most diggings in the swamp (8.0%), creek(6.7%), tracks (4.9%) and the lowest preferencefor the woodland stratum (0.7%). This pattern ofdigging preferences in the moistest strata wassimilar to those observed in the highland and tran-sitional areas.

Patterns of digging activity over time were alsoobserved. In general the highest digging activitytended to occur at the cessation of the wetseason (start of the dry season). A distinct trendof increasing digging activity was evident as thesoil started to dry out; this was very distinct in thetable drains of the road strata. This trend may bedue to the soil becoming more compacted whendry, making digging more difficult, therefore draw-ing pigs to areas where soil is easier to dig suchas the moist soils in the swamps or creeks.Another possibility is that earthworm populationsincrease when soil moisture levels are optimal,thus pigs may be increasing their digging activityto reach this high protein food source (French et

al. 1957). Root mass also increases with optimalsoil moisture levels (especially the fine feederroots) and pigs may be digging more to reach thishigh energy food source.

The importance of earthworms in the diet of pigsis unknown in this region. Dry soil conditions tendto force worm populations to move to deeper soilhorizons, beyond the reach of pigs (Lee 1985).The non-availability of earthworms in some strataduring the dry season may encourage pigs tomove to alternative microhabitats where condi-tions are more favourable. This may be linked tonutritional requirements, especially in relation toprotein. Earthworms have a high protein content(50 to 60%) (French et al. 1957). The effects ofthe seasonal fruiting cycle may also be impor-tant. McIlroy (1993) believed that nutrition maybe the dominant factor influencing pig movementsin this area.

The frequency of diggings within the microhabitatswas also used to compare with the digging index.

44

For the highlands a mean of 77% of transectswere effected by some pig diggings for eachsampling event (transitional area 58%; lowlandarea 85%). Over all three areas, a mean of 59%of transects were affected by some new pigdiggings at each sampling event. Mitchell andMayer (1997) found 63% of their transects hadsome diggings, although this represented totaldiggings and not just recent diggings.

Significant variation in mean frequency of diggingsamong strata was also detected for the highlandand transitional area; no differences were de-tected for the lowland area. In the highland area,all strata had significantly higher frequencies ofdiggings then the ridge stratum. In the transitionalarea the creek stratum had a significantly higherfrequency of diggings than the road stratum.

The magnitude of differences in frequency ofdiggings among strata was influenced by season.For the highland area significant differencesbetween strata were detected in the dry season.For the transitional area significant differences infrequency of diggings between strata alsooccurred in the dry season. In the lowland areaonly one sampling event (dry season) recordedsignificant differences in frequency of diggingsbetween strata. Thus it appears that feral pigsconcentrate in the wetter microhabitats during thedry season and spread out evenly over allmicrohabitats during the wet season.

Ecological Impacts of Feral Pig DiggingsThe aim of this research was to quantify aspectsof the ecology of feral pig diggings usingexclosures. The recovery of selected ecologicalparameters in exclosure plots protected fromferal pig diggings was compared with control plotsnot protected from pig diggings.

Pig-proof exclosures (10m x 10m) were estab-lished in two microhabitat strata (wet areas anddry tracks) in the highland rainforest area only.For each microhabitat three sites were selectedand for each site two replicate exclosures wereestablished giving a total of 12 exclosures.Indices representing four ecological parameterswere measured in each exclosure and in two(unfenced) control plots established adjacent toeach exclosure. The ecological parameters wereseedling germination and establishment, aboveground biomass, below ground biomass andearthworm biomass. Exclosures were monitored

at 4 to 6 weekly intervals over a 2-year period.The exclosures have now been protected fromferal pig impact for 4 years.

Feral pigs were shown to have a substantialimpact on seedlings of rainforest tree species.The mean number of seedlings that were alive ateach sampling event within the exclosures was36% higher then recorded in the control plotswhere feral pig digging activity had been occur-ring (27% dry stratum, 44% wet stratum). Resultsalso indicated that seedling germination andsurvival rates within the exclosures were 20%higher than the controls. No significant differenceswere detected in above ground biomass, belowground biomass and earthworm biomass betweenthe exclosures and the control plots.

Feral Pig Movement PatternsThe aim of this research was to develop a modelof feral pig movements in relation to seasonalinfluences and to document home ranges andhabitat usage in the coastal lowlands. A total of41 feral pigs (19 females and 22 males) werecaptured in traps and fitted with radio collars, 8 inthe highlands, 19 in the transitional zone and 14in the lowlands. Collared pigs were located fromthe ground at least once per month. Any pigswhich could not be located from the ground(generally in the highland rainforests area) werelocated from aircraft.

No evidence of large-scale seasonal movementswas evident for the pigs in the lowlands and thehighlands. The transitional zone did record move-ments from the lowlands to the top of adjacentpeaks (400m), however in all cases the pigsreturned to the lowlands within a couple of days.The 8 pigs in the highlands did move into thewestern rainforest/open forest ecotone (move-ment of 3 km) and spent a considerable amountof time there. Observation revealed that the pigswere using the riparian vegetation along streamsas travel corridors between rainforest andsclerophyll habitats.

For the wet and dry season, individual pigs onthe lowland and transitional areas were radiolocated at 3 hr intervals continuously for 36 hours,and this was repeated each week for 3 weeks. Atotal of 11 feral pigs were collared, with sufficientdata obtained for home range estimations on 7 inthe transitional zone and 3 in the lowland zones.In general, males have a larger mean home range

Ecology of feral pigs in the Wet Tropics 45

Population Biology and Control of Feral Pigs

(8.95 km2) then females (2.35 km2) and both havea larger mean home range in the dry season (9.94km2) then the wet season (3.1 km2).

Biological ParametersA trapping program was instigated in the threeareas to capture feral pigs for biologicalinvestigations. A total of 317 pigs were capturedover a two-year period. Ecological data onmorphometrics, reproduction, diet and growthrates were collected. In the lowland area amark- tag - release program was used for popu-lation estimation. Preliminary data suggest a pigpopulation of 2 per km2 exists in the lowlandareas.

Backdating age estimation of known age pigs(less then 36 months) to determine birth datesrevealed a significant peak in births in January,the start of the wet season. Dietary items seenfrom gross examination include earthworms,fruits, centipedes, grass, roots and plant mate-rial. Growth rates and morphological informationsuggest feral pigs within this rainforest region havefaster growth rates and are on average 10 to 20kg heavier then feral pigs in the dry tropicalregions.

DISCUSSIONFeral pigs have been identified as a major issuefacing the management of the wet tropics WorldHeritage Area. However the ecological effect offeral pig activity is difficult to quantify. Simplemeasurement of soil disturbance (diggings) isinadequate in understanding what “true” damageor otherwise is being imposed on the processesinvolved in the ecosystem. The main visualimpact of feral pig damage is soil disturbance dueto their searching behaviour for plant roots andsoil invertebrates. Although animal signs do notnecessarily correlate with population density oractivity (Hone 1988), rooted ground has beenused as an index of the impact of feral pigs onvarious environments.

Spatial and temporal patterns of feral pig diggingactivity were found within this study area for thetwo years of this study. In general particularmicrohabitats were found to have significantlyhigher digging activity overall and at specificsampling times of the year. Rainfall (through itseffect on soil moisture) appeared to be the majorinfluence on these digging patterns, with higherrates of digging occurring, for all of the strata, in

the early dry season and predominantly in moistmicrohabitats (swamps and creeks). The impactof pig diggings on ecological processes isdifficult to quantify over a short time frame.However pig diggings appear to influence thesurvival of seedlings, especially in the moistareas such as swamps and creeks. Althoughgermination rates were similar in plots exposedto pig digging and plots protected from diggings,the number of surviving seedlings was 36% higherin protected plots. Although rainforest seedlingshave a naturally high attrition rate, an increase inthe attrition rate by 36% when pig digging impactsare added, may have important ecologicalimplications for plant regeneration.

The absence of seasonal movements in the feralpig population is contrary to the general commu-nity perception. Most landholders within theregion believe that feral pigs migrate down fromthe highlands to the coastal lowlands in the dryseason to forage on the ripening sugar cane andbanana crops, and return to the highlands in thewet season when the sugar cane is harvested.The results of this study do not support the exist-ence of such a “seasonal migration”; rather, thisseems to be a perception created by local move-ments by feral pigs inhabiting the transitional area,the rainforest-crop boundary. Home range stud-ies suggests that feral pigs move greaterdistances and forage further when food andwater becomes scarce in the dry season. Thisincreased movement activity would put them ingreater contact with humans especially during thesugar cane harvest season. During the wetseason, feral pigs are more sedentary, food andwater are abundant, and human activity is mini-mised within the crops. Thus human / pig interac-tion is lower in the wet season. This has lead tothe perception of more pigs present in the dryseason compared to the wet and the communityperception that these pigs had to come fromsomewhere: the highland rainforests.

The formation of the WHA is perceived by manymembers of the public as primarily responsiblefor the economic losses incurred by the ruralindustries adjacent to the WHA and attributed toferal pigs. Pigs are an acknowledged pest ofcane, bananas and other small crops on the northtropical coast. However the results of this studysuggest that feral pig adjacent to and in somecases living on landholders’ properties are mainlyresponsible for this economic damage.

46

Control techniques on the WHA fringe need to befully coordinated; poisoning, trapping, hunting andfencing techniques can all be implemented if theactivities are integrated into a control strategy. Forexample large scale trapping programs can beestablished in environmental sensitive areas, orferal pig concentration areas where landholdergroups are available to interact with the program.Hunting by licensed hunters would be suited toareas with low pig populations, or areas inacces-sible to other control techniques. Fencing can beemployed in high-return cropping or intensiveagriculture situations where the cost is warranted.Fencing may also be an option in small sensitiveconservation areas or where rare or endangeredspecies are localised, although the impact onother species needs to be considered.

The implementation of a feral pig managementstrategy must rely on a clear understanding ofthe severity of ecological impact of pig diggingson WHA values and also the level of populationcontrol required to protect these values.Additional research information is required on“best practice” control techniques to developefficient, cost effective and specific populationcontrol. A sound management plan needs tocoordinate, monitor, evaluate and continuallyevolve with developing strategic directions.

Ecology of feral pigs in the Wet Tropics 47

Population Biology and Control of Feral Pigs

Monitoring impacts and control of feral pigs:A case study in Namadgi National Park, A.C.T.

HONE, J.

Applied Ecology Research GroupUniversity of Canberra, A.C.T.

WHAT ARE THE IMPACTS?Ground rooting decreases plant species richnessin grassland in the short-term (Hone unpublisheddata). Rooting can also cause the abundance ofsome plant species to increase and other plantspecies to decrease (Alexiou 1984). Pigs areaccused of other impacts such as erosion but thishas not been studied. Another impact is onvisitor’s enjoyment of the park. Visitors complainto park rangers when they find large areas ofgrassland rooted over by pigs. This is a visualimpact.

WHAT RELEVANT MANAGEMENT ANDRESEARCH HAS OCCURRED?Feral pig control occurs annually in NamadgiNational Park in autumn using wheat soaked inwarfarin. Details of the pig control work, whichstarted in 1985, are described in Hone (1987),McIlroy et al. (1989) and Hone and Stone (1989).The management aim is stated in the ParkManagement Plan of 1986 as “to protect the parkand adjacent areas from damaging effects of pestplants and animals.” The pig control work is basedon considerable field research in Namadgi, andnearby Kosciuszko National Park, and earlier penexperiments. That research has been on theeffectiveness of the poison, warfarin, in pen (Hone& Kleba 1984, O’Brien & Lukins 1990) and fieldstudies (McIlroy et al. 1989), and by theoretical

modelling (Hone 1992); evaluation of baitconsumption (McIlroy et al. 1993, Saunders et al.1993) and other methods such as hunting withdogs (McIlroy & Saillard 1989), and trapping(Saunders et al . 1993); pig demography(Saunders 1993); pig movements (McIlroy et al.1989, Pech & McIlroy 1990, Saunders & Kay1996); and non-target aspects of warfarin poison-ing (McIlroy et al. 1989, McIlroy et al. 1993).

WHAT IS MONITORED?Two parameters are monitored; extant groundrooting and pig abundance. The frequency ofoccurrence of rooting is positively correlated withthe extent of ground rooting in each month,except December, of the year (P<0.05) (Hone1988a) and the frequency of dung counts is posi-tively correlated with observed pig density (R2 =0.66, P<0.01) (Hone 1995). The abundance ofdung and the frequency of occurrence of dungare highly positively correlated (P<0.05) (Hone1998a). It is emphasised that the monitoring wasstarted to answer questions about short-termeffects of pig control on rooting and abundance,largely relating to exotic disease contingencyplanning (especially for foot and mouth disease)rather than to conservation of biodiversity. Alsothe sampling intensity required was estimatedbased on pig abundance in 1985, which was manytimes higher than that it is in 1999.

48

ABSTRACT

This paper is an overview of the monitoring of the occurrence of feral pig (Sus scrofa)impacts, such as ground rooting, and pig abundance in Namadgi National Park, ACT.The park has an area of about 1060 km2 comprising mountain forests and woodlandsranging from 800 m to 1900 m in elevation. A series of commonly asked questionsabout the monitoring and pig control are posed and answered.

Monitoring of other impacts, such as crop and pasture damage and lamb predation,will require different methods. However like monitoring pig rooting they also requireuse of elements of sampling theory and the principles of experimental design.Choquenot et al. (1996) provide an overview of many such monitoring methods, andthey distinguish between operational and performance monitoring. The monitoring inNamadgi is performance monitoring - measuring the effect of management.

Monitoring impacts and control of feral pigs 49

HOW IS IT MONITORED?In summary, there are 700 set plots which weresemi-randomly selected across a range of veg-etation types in the eastern half of the park. Theplacement and orientation of plots were randomlyselected within each of seven sites. The frequencyof monitoring was initially monthly (1985-1986),then seasonally (1986-1988), and now annually(to 1999). The variables are monitored on set plotsusing the line intercept method (for extant root-ing) on lines each 10 m long, or area counts (forfresh dung) on plots 10 m by 2 m. Dung arecleared off plots after counting. In most analysesdata are collated across all 700 plots. Details ofmonitoring methods are described in Hone(1988a), Hone and Stone (1989), Hone (1995)and Hone and Martin (1998). Estimates of theshort-term effects of poisoning on pig abundancehave been compared between data using dungcounts (Hone 1987) and data using radio-track-ing (McIlroy et al. 1989) and found to be verysimilar (about 94%). The monitoring could be usedto detect small changes in pig rooting and abun-dance but would require an increased number ofplots.

WHAT IS THE RELATIONSHIP BETWEENIMPACTS AND PIG ABUNDANCE?Choquenot et al. (1996) discuss the central roleof the relationship between impacts and pigabundance in the effect of pig control on impacts.In Namadgi there is a positive curved (concavedown) relationship across years between theabundance of pigs and the frequency ofoccurrence of pig rooting (R2 = 0.49, P<0.001)(Hone unpublished data). The implication of thisis that a large reduction in pig abundance isneeded to get a substantial reduction in pigrooting. A small reduction of pig abundance, whenthat abundance is initially high, would give littleor no change in pig rooting. There is a negativecurved relationship between plant speciesrichness and extent of pig rooting (R2 = 0.53,P<0.001) (Hone unpublished data).

There is also a positive relationship between pigabundance and the monthly change in rooting (R2

= 0.48, P<0.05) (Hone 1995). Hence when pigsnumbers are high, then rooting increases andwhen pigs are few then rooting decreases. Inbetween there is an estimated equilibrium levelof pig abundance (dung on 3% of plots) at whichthe frequency of pig rooting does not changemonth to month. The relationships between

impacts and costs need to be specified moreclosely. This will aid determination of the mostcost-effective or cost benefit level of pig control.

WHAT ARE THE BENEFITS AND COSTS OFMONITORING?The benefits are that managers learn the shortand long term consequences of their actions,namely that pig abundance and rooting havedeclined substantially since 1985. The observedinstantaneous rate of increase (r) of pigabundance is -0.15 (R2 = 0.36, P<0.01) during1985 to 1999 (Hone unpublished data). Hencemonitoring is important to improving managementand is part of the adaptive management used forpig control in Namadgi National Park. Themonitoring has demonstrated a seasonal patternin pig rooting, with peaks in spring and autumn(Hone 1987), which may be related to seasonalaltitudinal movements of the pigs, and theplanning and evaluation of pig control needs totake account of this pattern.

Monitoring has also identified the sites - highelevation (P<0.05) and low slope (P<0.05) - mostlikely to suffer pig rooting (Hone 1988b, 1995),so pig control can be strategically focussed atthose sites. Pig control did not occur in 1989 and1990 and monitoring showed pig abundanceincreased during those years. During 1996 and1997 distribution of bait from helicopters ceased.Pig abundance did not increase but rooting didincrease. The costs of monitoring are my time andequipment. These are about $2,600 per year. Notethat the monitoring is independent of the pigcontrol work. I do the monitoring and EnvironmentACT does the pig control work. An opportunitycost would be incurred if monitoring ceased, andhistorically would probably have stopped pigcontrol.

PREDICTIONS FOR THE FUTUREThere are several predictions about future trendsin rooting and pigs depending on whether pigcontrol continues or ceases. Monitoring canallow testing of these predictions. If pig controlcontinues there are two short-term predictions;one, pig rooting and pig abundance stay low asthe baiting stays very effective, two, pig rootingand pig abundance slowly increase because ofincreasing bait aversion and or, development ofwarfarin resistance. If pig control ceases, forexample because of funding cuts, there are threeshort-term predictions; one, that rooting and

Population Biology and Control of Feral Pigs50

abundance stay low because pigs are limited bypredation by wild dogs and maybe foxes, two, thatrooting and abundance increase slowly becausepredators reduce rate of increase but do not stopit, and three, rooting and pig abundance increaseat the intrinsic rate of increase (r

m). In the long-

term there are three predictions for pigabundance; one, that it increases then reachesan equilibrium point (logistic growth, as assumedin Pech and Hone 1988), two, that it increasesbut then shows damped oscillations to anequilibrium point (as assumed in Hone 1988b) andthree, that it oscillates over time around, but notreaching, an equilibrium point. Pig rooting wouldbroadly follow the same patterns.

SUMMARYWe have learned that pig control can reduce pigrooting and pig abundance. The frequency of root-ing and pigs are positively related, as are themonthly rate of change in pig rooting. Methodsof ground survey have been developed and partlyevaluated. A set of predictions of future short andlong-term trends in pig rooting and abundancehave been developed and could be tested.

We have not learned much about long-termimpacts on specific plant and animal species,whether the conservation of any species is threat-ened by pigs, whether erosion is significantlyincreased by rooting, or about effects not relatedto rooting, such as from grazing and predation,and about non-target effects of baiting. Researchoverseas, especially in Hawaii and Great SmokyMountains National Park (USA), suggest thesecould be important and worth study.

ACKNOWLEDGEMENTSI thank the manager and staff of NamadgiNational Park for support, and the University ofCanberra for financial assistance.

The effectiveness of trapping in reducing pigabundance in the Wet Tropics of northQueensland

VERNES, K.1, JOHNSON, C. N.1, MITCHELL, J.2

1Department of Zoology and Tropical Ecology, James Cook University, Townsville2Tropical Weeds Research Centre, Queensland DNR, Charters Towers

ABSTRACT

The ability of trapping to reduce pig numbers in the wet tropical lowlands of NorthQueensland was tested and the sensitivity of several indices of abundance of pigs (rate ofproduction of diggings and dung, and disturbance of bait stations) to changes in pigabundance were compared. Reduction in pig abundance over three consecutive cullsyielded an overall 82% decrease in the rate at which new diggings were made. In contrastto digging, the number of dung pellets observed on transects did not show a clearrelationship with changing pig abundance. Visitation by pigs to bait stations did not showany obvious relationship with the reduction in pig numbers, probably due to the problem ofcontagion between bait stations. This study suggests that trapping should be useful inprotecting environmentally sensitive areas from the mechanical impact of feral pigs, andmay also provide a model for monitoring the effectiveness of trapping in intensive controloperations in the Wet Tropics.

INTRODUCTIONFeral pigs (Sus scrofa) are believed to causesignificant environmental damage to therainforests of the Wet Tropics World HeritageArea (WTWHA), through a variety of processesincluding habitat destruction and alteration,dispersal of weeds, promotion of soil erosion(Mitchell and Mayer 1997), competition withnative species (Pavlov et al. 1992; Laurance andHarrington 1997), and egg predation on sensitivebird species such as the cassowary, scrub turkeyand scrub fowl (Hopkins and Graham 1995;Crome and Moore 1990). In addition, pigs arereservoirs of many endemic diseases andpotential vectors of several exotic ones (Pavlovet al. 1992; Mitchell and Mayer 1997), and avariety of agricultural crops bordering theWTWHA are adversely affected by feral pigs(Mitchell 1993).

Options for reducing pig populations in the wettropics are limited. Poisoning may beunacceptable because of its potential effects onnon-target species, shooting may have little effectbecause of the inaccessibility of pigs in the difficultterrain and dense vegetation typical of much ofthe Wet Tropics, and hunting with dogs is

inefficient because of low rates of encounter withpigs (Mitchell 1993). Trapping may be a usefultechnique for pig control, especially in the wettropical lowlands. In this area access for trappingis easier than in the highlands and valuableconservation areas border sugar cane andbanana farms that sustain high impact from pigs.

In this study we carried out a trial trapping programin the wet tropical lowlands of North Queensland.We trapped pigs in three intensive sessionsspread over a four month period. Our objectivewas to reduce the pig population in our study areain three steps, producing four levels of abundance.Several indices of abundance of pigs (rate ofproduction of diggings and dung, and disturbanceof bait stations) were measured before and aftereach trap session, so that we could compare theability of these indices to detect reductions of thepig population. This design also enabled us tocalculate the number of pigs on the study areabefore and after trapping, using the index-removal-index method, and thereby to estimatethe reduction in absolute abundance of pigsachieved by trapping.

Effectiveness of trapping in reducing pig abundance 51

Population Biology and Control of Feral Pigs

METHODSStudy AreaThe study was conducted at Edmund KennedyNational Park (18˚11’S, 145˚59’E), a lowlandcoastal region of the Australian Wet Tropicscomprising a complex mosaic of woodland, dryrainforest, swamp and mangrove forest. Theelevation of much of the park is less than 15 mabove sea level. Swamp and mangrove occur ator near sea level, while rainforest and woodlandoccur along a series of ancient sand dunes (sand-ridges) throughout the park.

Two sites approximately 3 km apart were selected,each consisting of swamp and adjacent ridgehabitats. The area around Dallachy Swamp(hereafter ‘Dallachy’) in the centre of the park wasthe control site, where pig abundance was notmanipulated. The area around Duck Swamp(hereafter ‘Duck’) in the south of the park was theexperimental site, where pigs were trapped. Ateach site, pig activity was surveyed over an areaof approximately 80 ha, and at the Duck site pigswere trapped over an area of approximately 625ha centred on the survey area. Pig control hadnot been attempted within this region of the parkfor several years.

Activity TransectsBelt-transects, each measuring 5 x 50 m, wereestablished to monitor pig diggings and dung. Ateach site ten such transects were established onsand-ridge habitat, and ten in swamp habitat.Transects were spaced at least 50 m apart, andaligned roughly parallel to one another in eachhabitat type. Each transect was partitioned into5x5 m segments, which were then each furthersubdivided into quarters. The percentage coverof pig digging (in 5% increments) was estimatedfor each of these quarters, and the coverage ofexisting diggings marked onto a diagrammaticrepresentation of the transect.

In this way, diggings could be compared betweencensus periods at a sufficiently fine scale to obtainprecise estimates of the area of new diggings.The presence of dung in each 5x5 m transectsegments was also recorded. Dung pellets werecounted and removed at each census. Activitytransects were established in June 1998, and allpig activity recorded. The transects weresurveyed before the first trapping session, thenre-surveyed after each trapping session. Eachsurvey consisted of two passages over the

transects 4-6 weeks apart, with the two passagesused to estimate rate of deposition of dung andrate of production of diggings over that timeinterval.

Bait StationsBait stations were deployed and monitored dailyalong roads and tracks in each site during a periodbefore and after each cull. Each station wasbaited with a small bunch of bananas. Thirty suchstations were deployed at Dallachy and 35 atDuck. Bait stations were positioned approximately100 m apart. Stations were monitored and re-baited daily until the rate of bait-take reached alevel which was maintained for severalconsecutive days. On each occasion, this wasachieved within seven days.

Trapping of PigsNine circular ‘silo’ traps (Choquenot et al. 1996),each with a diameter of approximately 3-4-m,were positioned at Duck during August 1998.Traps were located adjacent to roads and tracks,and were spaced several hundred metres apart.Initially, traps were liberally baited with bananasand wired open to allow pigs free access. Whena trap showed signs of pig usage (generally aftera few days), it was set. Traps were set in the lateafternoon and cleared early the following morning.All captured pigs were killed in the traps. Threeculls were undertaken at approximately 4-6 weekintervals.

Estimates of AbundanceThe index-removal-index method of Caughley(1977) was used to obtain estimates of totalnumbers of pigs present on the Duck site beforeand after each trapping session. Individualestimates were calculated for swamp and ridgeareas, as well as for both habitats combined. Thismethod uses the following equations:

N1 = x

1R/(x

1-x

2) (Equation 1)

N2 = x

2R/(x

1-x

2) (Equation 2)

where N1 is the population estimate before

removal of pigs, N2 is the population estimate after

removal of pigs, x1 denotes the rate at which new

pig-sign is detected before pig removal, and x2

denotes the rate at which new pig-sign is detectedafter removal of R pigs.

52

RESULTS

Removal of pigsPigs were removed from the experimental site(Duck) at three time periods throughout the study(Table 1). The third cull yielded pigs only on thefirst day, and over the following two days, no baitswere disturbed, and no fresh pig tracks were seenaround traps or on the roads which connectedtrap sites. This absence of pig-sign was taken toindicate low pig numbers in the area, and sotrapping was stopped.

Table 1. Trapping periods, showing the number and sex of pigs removed.

Start date of Cull Duration (days) Pigs Males Females

05/08/98 9 9 6 301/09/98 6 8 4 418/10/98 3 2 2 0

Total 18 19 12 7

Table 2. Repeated measures ANOVA tables for the effects of treatment (pigs culled or not culled) andhabitat (swamp or ridge) on the observed rate at which new diggings were added to transects during each ofthe four census periods.

(a) Between Subject EffectsSource SS DF MS F P

Treatment 1.7341 1 1.7341 7.4181 0.01Habitat 4.8766 1 4.8766 20.8609 <0.0001Treatment*Habitat 1.1908 1 1.1908 5.0938 0.03Error 8.4156 36 0.2338

(b) Within Subject EffectsSource SS DF MS F P

Census Period 2.6324 3 0.8775 10.45 <0.0001Census* Treatment 2.5606 3 0.8535 10.16 <0.0001Census* Habitat 0.1570 3 0.0523 0.62 0.60Census* Treatment* Habitat 0.2900 3 0.0967 1.15 0.33Error 9.0704 108 0.0840

Effectiveness of trapping in reducing pig abundance 53

Rate of production of diggingsThe percentage of ground dug by feral pigsthroughout the study prior to any culling of pigswas 2.3 ± 0.4%. Swamps were dug on averagetwice as much (3.1 ± 0.3%) as ridges (1.4 ± 0.6%).Repeated measures ANOVA showed thattreatment and habitat type had a significant effect

on digging rate (Table 2a). The rate of diggingwas significantly lower on the sites where cullingoccurred, and lower on swamp transects than onridge transects (Fig. 1). When the repeat factor(census period) was considered, the onlysignificant effects on digging rate were censusperiod and the interaction between census periodand treatment (Table 2b). Digging rate declinedsignificantly throughout the study, and did so onlyon transects where pig culling had occurred (Fig.1).

There was a significant correlation between therate of digging between census periods and thecorresponding number of pigs removed (r=-0.964,df=2, p<0.05) (Fig. 2).

Population Biology and Control of Feral Pigs

Figure 1. Changes in the rate at which new diggings were made at the Dallachy (control) and Duck (experimental) sites before and after each trapping session (indicated by arrows). Pigs were trapped

only at the Duck Swamp and Duck Ridge sites.

Figure 2. Relationship between the cumulative number of pigs removed and the percentage of ground

disturbed per month by new diggings.

54

Rate of dung depositionIn contrast to digging activity, the number of dungpellets observed on transects did not show a clearrelationship with number of pigs removed.Declines in the number of dung pellets at Duckwere observed throughout the study (r=0.954,df=2, p<0.05), indicating this may have been inresponse to culling. However, declines were alsoobserved on ridge transects at Dallachy, while inthe swamp transects at this site, the amount ofdung detected throughout the study increasedconsiderably. Due to small number of transectswhere dung was actually recorded (most transectsrecorded zero), we were unable to examine thesechanges statistically using repeated measuresANOVA. It is clear, nonetheless, that therelationship between dung and numbers of pigsculled is far weaker than the relationship betweennew diggings and numbers of pigs removed.

Disturbance of bait stationsDisturbance by pigs of bait stations did not showany obvious relationship to the reduction in pignumbers brought about through culling. Beforethe first cull, consumption of baits at both controland experimental sites was 100% . Followingeach cull, however, consumption at theexperimental site continued to peak at or near100%.

Absolute abundance of pigsThe index-removal-index calculation (Equations1 and 2) was used to obtain estimates of totalnumbers of pigs before and after each cull, usingthe rate of production of diggings as the index(Table 3). This suggested that approximately 15pigs were using the transects before the first cull,and that 6 remained afterwards. Approximately10 pigs were present before the second trappingsession, and 2 remained after it. Only 2 pigs weretrapped during the third cull. No dung wascollected after this cull, but the digging datasuggested that pigs were in fact still present inthe area in roughly equivalent numbers to theperiod before the cull. Our traps were distributedover an area of approximately 600 ha, suggestingan initial density of 2-3 pigs per square kilometre.

DISCUSSIONMonitoring techniquesThe study tested several indices of feral pigabundance (rate of production of dung, diggings,and visitations to bait stations) against actualchanges in the number of pigs. The rate at which

Table 3. Estimation of the total numbers of pigs usingswamp and ridge transects at the Duck site before andafter each cull, based on the index-removal-indexmethod. An estimate of total numbers of pigs at theDuck site was also calculated. Estimates could not becalculated on diggings before or after the October cull,due to a slight (and non-significant) increase in diggingactivity, resulting in negative estimated abundances.

Survey Pigs Numbers present

Month Removed

Before After

August 9 15 6

September 8 10 2

October 2 - -

new pig diggings were added to transects wasthe most meaningful of these indices. There wasa significant and approximately linear negativerelationship between the rate at which newdiggings were added to transects and thecumulative number of pigs removed from thestudy area. This shows that production of diggingswas indexing pig numbers. Some past studieshave found a positive correlation between pigdigging and observed pig density (eg. Ralph andMaxwell 1984 in Hone 1988b; Katahira et al.

1993). However, Hone (1988a) observed nosignificant correlation between these variables,and suggested that linear relationships reportedin previous studies were unreliable (Hone 1988b).Two factors, however, distinguish the currentstudy from many others. First, the rate at whichnew diggings were created was measured; otherstudies have simply recorded existing diggingsof unknown age in one-off surveys. Second, thisstudy compared rates of production of diggingswith actual numbers of pigs removed throughtrapping, instead of with estimates such asnumbers of pigs seen along a transect, which maythemselves be subject to bias. Disturbance ofbait stations proved to be of limited use duringthis study. This technique has been used in thepast to monitor the abundance of large mammalsincluding foxes (Thompson and Fleming 1994;Fleming 1997) and feral pigs (Choquenot et al.

1990), but we found that all bait stations weredisturbed even when other evidence suggestedthat the number of pigs had been reduced.

Effectiveness of trapping in reducing pig abundance 55

Population Biology and Control of Feral Pigs

This presumably was due to contagion, such thatbait stations were not independently sampling pigactivity and a small number of pigs were able tovisit all stations. The likely cause of contagion inthis study was the short 100 metre distancesbetween bait stations. Also, bait stations wereplaced along roads. Pigs clearly utilize roads, andthis would have made it more likely that a singlepig could detect a series of bait stations duringnormal wandering.The rate of dung deposition hasbeen used in several other studies as an index ofabundance of feral pigs (Hone 1988; Aplet et al.

1991; Bowman and McDonough 1991; Bowmanand Panton 1991; Anderson and Stone 1993;Katahira et al. 1993; Hone and Martin 1998). Inour study, the number of dung pellets depositedper month at experimental (Duck) sites declinedas pigs were removed from the study area. Thisrelationship, however, was statistically significantprimarily because dung was found in the first andsecond census periods, and not in the third andfourth census periods. Furthermore, dungcollection at the Dallachy Ridge site also declinedas the study progressed, whereas dungdeposition at the adjacent Dallachy Swamp siteincreased markedly. Thus, numbers of dungpellets detected at different census periodsshowed no clear relationship with treatment (ie.sites where pigs were culled versus not culled).The underlying problem with our data on dungwas that many transects recorded no dung at all,so that mean values of rates of dung depositionwere highly variable and difficult to analyse. Hone(1988a, and pers. comm.) found that when largenumbers of small plots are surveyed, theproportion of plots containing dung may be auseful index of pig abundance.

Effectiveness of trappingOur results suggested that almost all the pigsusing the trapped area at the beginning of thestudy had been removed by the end of trapping.This demonstrates that trapping can be veryeffective in reducing the local abundance of pigsin the Wet Tropics. The result of this removal wasthat by the end of the study the rate at which soilwas being disturbed by pigs had fallen to onlyabout 18% of the rate measured before trapping.Many of the ecological impacts that pigs arepresumed to have on ecosystems are related totheir disturbance of soil. It follows that trappingmay significantly reduce the environmental impactof pigs. Trapping of feral pigs has previously beenevaluated by Choquenot et al. (1993) on the

central tablelands of New South Wales andSaunders et al. (1990) in Kosciusko National Park,where it achieved reductions of pig abundance of80% and 71% respectively, results similar to the82% reduction in pig activity produced by ourtrapping program. These figures compare wellwith evaluations of shooting from helicopters (80%reduction, Saunders 1983), and poisoning(reductions of between 60% and almost 100%;see Hone and Pedereson 1980 in Choquenot1996, Hone 1983, Bryant and Howe 1984 inChoquenot et al. 1996, McIlroy et al. 1989,Saunders et al. 1990).

The fact that some evidence of pig activityremained at the end of trapping could haveindicated that there was immigration into the areaduring the study, or that animals living on the edgeof the area made occasional incursions but didnot use the area consistently enough to encountertraps during the trapping sessions. This latterexplanation is very likely, considering that trapswere distributed over an area that was aboutequivalent to the feeding range of an individualpig and that there is a high degree of home rangeoverlap among pigs (J. Mitchell, unpublisheddata): many pigs using the area would have alsoranged outside it to varying degrees. However theeffect of this on levels of pig activity in the trappedarea was quite small over the five months of thisstudy.

Many of the most pressing concerns over theimpact of feral pigs in the Wet Tropics relate totheir effects on environmentally sensitive areas,such as wetlands, habitats of threatened species(for example, the northern bettong Bettongia

tropica, Laurance and Harrington 1997) or areasrich in endemic species. This study suggests thattrapping should be useful in protecting such areasfrom the impact of pigs, and may also provide amodel for monitoring the effectiveness of trappingin intensive control operations.

ACKNOWLEDGEMENTSWe thank Maria Morlin for assistance in surveyingtransects, and Bill Dorney and Warren Price forhelp in trapping pigs. Warren Price also providedvaluable technical support and advice. TheQueensland Department of Environment andHeritage gave permission for the work to beconducted in Edmund Kennedy National Park,and provided equipment and personnel whenrequired. The study was funded by the WetTropics Management Authority.

56

PART 4

IDENTIFYINGMANAGEMENT

OPTIONS

57

Community based feral pig trapping in theWet Tropics of Queensland

DORRINGTON, B.

Land Protection Branch, Department of Natural ResourcesSouth Johnstone

ABSTRACT

The community based feral pig trapping program provides a regionallycoordinated system for control of feral pig populations in the WetTropics. The program enjoys a high level of community support andinvolvement by landholders and members of the public, and benefitsfrom in-kind contributions from registered trappers. The program usesabout 500 traps, and around 200 pigs are trapped each year. Thispaper describes the operation of the program, and discusses itssuccesses and limitations.

Community based feral pig trapping

BACKGROUNDFeral pigs have been recognised as a significantthreat to the conservation values of the WetTropics World Heritage Area (WHA), as well ashaving a substantial economic impact on theneighbouring rural industry. Following the WorldHeritage listing of the WHA in December 1988public perception was that Wet TropicsManagement Authority (WTMA) had caused theproblem by “locking up” the WHA and creating abreeding ground for feral pigs, safe from outsidedisturbances.

In implementing the recommendations of anumber of consultancies to WTMA a managementstrategy was developed and implemented. Thisstrategy was to utilise local community knowledgeand support in addition to the other factorsrecommended in feral pig consultancies toWTMA.

This strategy evolved into a management plantermed the ‘COMMUNITY BASED FERAL PIG

TRAPPING PROGRAM’ and has now beenoperational since April 1993. The program hasundergone continual refinement over the last 6years through an adaptive management processwhich has increased effectiveness and efficiencyof the techniques used and has allowed theadoption of new management information as itbecomes available. This approach has alsoenabled the program to respond rapidly to otherissues as they arise, such as non-target capturesand animal welfare issues.

The program is designed to achieve WTMA feralpig management objectives;

1) To reduce the impact of feral pigs uponthe conservation values of the WHA.

2) To reduce the impact of feral pigs uponneighbours of the WHA.

3) Demonstrate WTMA commitment toferal pig management.

4) Foster public involvement andownership of the problem.

5) To encourage the adoption of BestPractice principals of feral pigmanagement.

The implemented management plan encouragesa regional approach, utilising an environmentalyacceptable control technique. As well as assistingthe conservation and protection of the WHA, theprogram has provided a framework for a regionallycoordinated approach to feral pig management,with more stakeholders becoming involved as theprogram evolves.

METHODOLOGYOrganisation is achieved by dividing the programinto a series of Trapping Systems. A trappingsystem is defined as a group of traps within acertain local area operated by a trained Trapper.The trapper is responsible for organising the dayto day management of traps within his area, whichis performed by either landholders, himself orother feral pig harvesters (hunters). Thispromotion of trapping as a control technique,together with fostering community ownership of

59

Identifying Management Options

feral pig management is of far greater importancethan simply catching large numbers of feral pigs,due to the long-term benefits obtained.

The trapper is also responsible for the collectionof all scientific and management data on hisoperations. The trapper reports to a localcommunity group or rural organisation that act asa Management Group. The individual manage-ment groups direct the trapper in locating thetraps, liaise with landholders and act as a linkbetween the trappers and program management.This approach encourages community ownershipof feral pig management and makes good use ofthe invaluable local knowledge of these people.

The Department of Natural Resources and Minesprovides overall management of the program inconsultation with an advisory committee ofstakeholders, all management groups and theWTMA. The program is closely associated withDNR’s feral pig research unit, lead by Mr JimMitchell, who has received two DNRM achieve-ment awards for establishing the program.

During the 12 months to March 1999, 44 trappingsystems were active. A total of 36 systems wereoperated by 32 program trappers, with sometrappers operating more than 1 system. Trapperfees ($200 per month) were paid for 26 systemsand 10 were operated on a voluntary basis. Ofthe paid trappers 6 received matching paymentsfrom their local management groups, mainly theCane Protection and Productivity Boards. Theremaining 8 systems were operated by others,such as QPWS rangers as part of their duties.

Management groups participating in the programinclude;

• The Cane Protection and Productivity Boards• Queensland Parks and Wildlife Service• Community Environmental Groups• Department of Natural Resources• Local Government• Aboriginal Communities• Non-Aligned Groups

The program also has a commercial contract withthe Australian Defence Force to provide feral pigtrapping systems at all 6 of their training areaswithin and adjoining the WHA.

Approximately 500 traps are available, howeveronly about 60% - 80% are in use at any giventime. The main reason for this is that when feralpig activity in an area ceases the trap(s) therewill often be closed, until the pigs invariably return.An additional 155 private traps are reported asbeing directly within the program, with the sameamount estimated to be operating in the area butoutside the program’s direct control. A largepercentage of these traps have been constructedas a result of the program’s extension activities.

In excess of 8000 feral pigs have beendocumented as being captured from thecommencement of the program up to the end of1998, with the annual total remaining stable atabout 2000 per year since expansion to thecurrent level occurred. A pronounced peak incaptures usually occurs in April/May and October/November, the end of the wet and dry seasons.

COMMUNITY INVOLVEMENTA feature of this program is the high level ofcommunity support and involvement bylandholders and other members of the public. Anattempt to quantify the value of this in-kindcontribution has been made recently, with 20trappers asked to complete a detailed writtensurvey of their in-kind contributions. A total of 13completed surveys were returned and the dataanalysed. Trappers were asked about their weeklycontribution in terms of hours spent, kilometresof vehicle usage and out of pocket expenses. Theaverage trapper was found to spend 29.9 hours(range 14 – 54), 288 kilometres (range 74 – 615)and $13.70 in expenses operating his trappingsystem each week. This information has beenextrapolated in Table 1 to obtain an averageannual in-kind contribution per trapper. Whenmultiplied by 32 trappers the amount to beconsidered is $ 1,018,304, a ratio of $14.51 worthof services provided for every $1 spent on trapperfees. The motivation for the trappers to undertakethis work are many and varied, however a majoritywere feral pig hunters pior to induction into theprogram. This link to the hunting fraternity hasproved invaluable in promoting trapping as acontrol technique by demon-stration.The vastmajority of feral pig hunting conducted in the WetTropics is a recreational pursuit and of little valueto control efforts. In many cases it has a negativeimpact on control by disrupting coordinated controlmeasures.

60

Table 1. In-kind Contribution per Trapper

Per Week Per Year Rate Value

Hours 29.9 1 554.8 $15 p/hr $ 23 322

Kilometers 288 14976 52c p/km $ 7 788

Expenses $13.70 $ 712.40 - $ 712

TOTAL $ 31 822

Management groups were surveyed in person andby telephone, with a sample of landholders beingsurveyed using the same method. Due to the rela-tively small sample number of the latter group thedata obtained could be considered approximate,

however the purpose is only to give an idea ofthe level of community involvement in, and own-ership of the program. The figures obtained havebeen extrapolated to give the annual contributionsdetailed in table 2 below.

Table 2. Annual in-kind Contributions

Group Labour Vehicles Other Total

Trappers $ 23,322 $ 7,788 $ 712 $ 31,822

Man. Groups $ 54,600 $ 18,200 $ 14,352 $ 87,152

Landholders $ 780,000 $ 104,000 $ 12,500 $ 896,500

TOTAL $ 857,922 $ 129,988 $ 27,564 $ 1,015,474

INFORMATION FLOW CHARTA further positive benefit of this level of communityinvolvement and the organisational structure ofthe program is the excellent communicationnetwork that has evolved. As can be seen fromthe flow chart on the following page, there is atwo-way flow of information and data betweenland managers, researchers, landholders and thegeneral community. A detailed communicationplan has been developed to fully exploit thisnetwork.

LIMITATIONSInadequate Monitoring SystemsThe monitoring systems in place do not fullyaddress the program objectives.Action Taken - A detailed monitoring program,“Monitoring Systems for Feral Pigs” is proposedby Mr Jim Mitchell. Funding is being soughtunder the National Feral Animal Control Programcomponent of NHT. An alternative system isbeing developed as a fall back position shouldfunding not be obtained.

Community based feral pig trapping

Environmental Impacts not QuanitifiedAlthough the general perception is that feral pigimpacts upon the WHA are negative andnumerous, these remain unqualified.Action Taken - Current research into the ecologyof feral pigs in Tropical Rainforests is nearingcompletion.

Program Scale InsufficientThe resources allocated to the program are in-sufficient to address its objective on the scale re-quired. In areas where a trapping program hasbeen in place for 2-3 years, there is anecdotalevidence from landholders and land managersthat visible feral pig impact has been reduced byat least 50%, however these areas represent onlya small percentage of the WHA.Action Taken - Initiatives of the Feral Pig Advi-sory Committee have identified and are pursuingseveral alternative funding sources. It is proposedthat any expansion of the program facilitated bythis funding will be targeted at areas of highenvironmental significance.

61

Identifying Management Options

RESEARCH PRIORITIES

Bait – Bananas are the main bait material usedthroughout the program, with other fruits,carcasses and offal being used in areas wherebananas are unavailable or unsuccessful. Whenpigs are feeding on a crop or have a plentifulnatural food source available, trapping efficiencywill drop off dramatically until the alternative foodsource is exhausted or removed. Several otherattractants including molasses, aniseed, vanillaessence, yeast and even creosote have beenreported to be used with varying degrees ofsuccess. Controlled trials should be conductedto fully investigate these and other potential baitsas an improved bait could potentially result in aquantum leap in trapping efficiency. Shouldbiological control methods proceed to the field trialor implementation stage this research could bevaluable. The program does not have theresources or expertise to conduct this research.

Trap Design – The majority of traps used by theprogram (400) are silo type traps with most havingside swinging doors and pig-specific trigger bars.About 100 are portable or mobile traps with 60 ofthese built to a standard box trap designdeveloped by the program. This design is acompromise between several factors includingsize, weight, ease of use and effectiveness.Should further expansion of the program occurmore portable traps will be required. Designrefinements are proposed to the standard designto improve effectiveness. A trap designcompetition has even been proposed, whichwould have the added benefit of increasing publicawareness and acceptance of trapping as BestPractice management. The program can conductthis development.

Figure 1. Information Flow Chart showing communication network which has evolved from the CommunityBased Feral Pig Trapping Program

62

Prospects for pig control: A biotechnologyperspective

SEAMARK, R. F.

CRC for the Biological Control of Pest Animals, CSIRO Wildlife and EcologyCanberra. A.C.T.

ABSTRACT

Biotechnology may not provide the solution to feral pig control, but biotechnologicalinnovations may increase the effectiveness of existing management programs. Thispaper explores the ways in which an understanding of the chemical signals whichinfluence pig behaviour - their pheromonal biology - and the way in which they sensetheir environment by smell - their olfactory biology - could be used to improve pigcontrol by trapping.

Biotechnology, as the name implies, is atechnology which builds on biological knowledge– the more complete the knowledge base thebetter the technical offering. Given the presentdeficits in knowledge of the feral pig in Australia,the development of biotech solutions will besomewhat constrained!

However, there is a considerable body ofknowledge of pigs generally – their nutrition,reproduction and diseases – gained by studies ofdomesticated pigs which suggests somebiotechnical possibilities.

THE FERAL PIG PROBLEM -CURRENT SOLUTIONSAs discussed elsewhere in the workshop,Australian feral pig populations prosper inresponse to opportunities provided by theenvironment through their natural high fecundity,adaptability and social flexibility, and the lack ofmany natural constraints including predators anddiseases; all possible targets for potentialbiotechnical control measures. These would,ideally, aim to prevent a pig problem rather thanreact to the problem once it is established.

This contrasts with the focus of most currentcontrol measures which are only implementedafter pig populations have built up to levels wherethey are having an adverse impact. Thus mostcurrent control measures, which include physicalbarriers, chemical poisons, shooting and trappingor a mixture of the above, aim at reducing theimpact of an established problem rather thanprevention.

Such reactionary means are costly, often posesignificant environmental (and human) risk and,unless managed in a coordinated, strategicmanner, have limited long term impact.

The chemical poisons used in broad scale pigcontrol baiting program are of particular concern– non-specific poisons such as 1080 for safetyissues and vitamin K antagonists such asPindone, for ethical (humane) reasons.

THE FERAL PIG PROBLEM -PROSPECTIVE SOLUTIONSControl of many insect pest population is nowachieved through a biotechnology approachbased on new knowledge of olfactory processesand in particular, specific knowledge of the mes-senger chemicals (pheromones) involved.Biotechnologists have used this knowledge toattract pest insects to toxic baits or to disrupt theirnormal socio-biological behaviour. Could we usea similar approach to feral pig management?

Pigs live in a world which they primarily sensethrough olfaction, thus management of the impactof feral pigs through manipulation of the olfactorysenses could be a sound base on which to builda biotechnological approach. Many mammalianpheromones (Singer 1991) are known whichexcite or depress physiological and socio-biologi-cal activities and can be used to disrupt feeding,mating and social interactions. Altered socio-biological parameters are also known to adverselyaffect the health status and impact of disease onmany mammal populations, including pigs.

Prospects for control - A biotechnology perspective 63

Identifying Management Options64

It is of interest that the earliest mammalianpheromone identified, in 1960, was from the pig.This was a factor identified in the scent producedby boars which caused female pigs in estrus toadopt and hold a mating stance allowing the boarto mount. This substance, a simple testosteronemetabolite, androstenone, is commercially avail-able and is already being used to modify the mat-ing behaviour of domesticated pigs (see Boothand Signoret 1992).

If not androstenone, then other pheromones orolfactory substances could be useful in a feral pigmanagement campaign. Recent developmentsin biotechnology allow putative substances to berapidly screened. One approach is to instrumenta pig so that when it is excited by a particularsubstance, it can be sensed by the neuro-physiological recording instruments as a changein the activity of a key brain centre coordinatingthe particular response eg. feeding, mating etc.The active substance can then be isolated andnow easily identified using mass-spectrometry orother analytical techniques. Alternativeapproaches, under development, plan to useartificial noses to identify such active substancesthrough coupling isolated olfactory-cell receptorsto micro electronic sensors (see Bargmann 1997,for links to the extensive literature in this area).

These substances, once identified, can then besynthesised in the amounts required for field useusing available biotechnology. Such develop-ments would offer completely new horizons ofpossibility in the management of the pig problem.Similar approaches are being considered for rab-bit (Mykytowcz 1985), fox (Whitten et al. 1980)and rodent control (Gao and Short 1993). An earlytarget for such studies would be to identify attract-ants to draw pigs to (or steer them from) specificareas eg. bait stations or traps.

Pheromones and other olfactory stimuli can behighly species specific and attractants could besought that would in themselves provide a highdegree of species-specificity in use, for exampleby only attracting pigs to a bait station, thusreducing the risk of non-target bait uptake orinadvertent trapping of native animal species. Ifadditional safeguards are required these couldbe engineered into the bait containing the toxin(or vaccine) as the attractiveness of the bait wouldnow be independent of the usual food character-istics. Manipulating the bait size, texture and

digestibility could all be used to discourageingestion by non target species.

Alternatively, more sophisticated solutions to theissue of bait safety could be developed throughengineering additional features into the bait toensure target specificity. This is now a realprospect offered through the new knowledge ofdetails of the molecular and cellular biology of thepig which distinguish it from other target species.

The olfactory trail is one of many pathways thatcould be explored in the search for newapproaches towards pest management. Asdiscussed by Mike Holland in the following paper,the Pest Animal Control CRC is exploring otherpossibilities raised through recent discoveries inmolecular medicine of how host organismsrespond to diseases. On the basis of this emergingknowledge, the Pest Animal CRC is now engagedin the development of a new generation of pestcontrol agents aimed at controlling the fertility ofpest animals. The aim is to create fertility controlvaccines which can dispensed in baits or be self-distributing through using naturally disseminating,infectious organisms as vectors. The feasibilityof this approach has recently been demonstratedin laboratory studies with mice. Hopefully thisimportant demonstration can now be built on todevelop control agents and managementstrategies for a range of Australia’s introducedanimal pests, including the feral pig,. Hopefullytoo, these agents will be more efficacious, saferand more humane than those presently in usethrough being designed, from the onset, to preventrather than cure pest problems.

Fertility control for feral pigs: options and status

HOLLAND, M. K.

CRC for the Biological Control of Pest Animals, CSIRO Wildlife and EcologyCanberra. A.C.T.

ABSTRACT

Reduction of the fertility of wild pigs would make control by techniques suchas trapping or poisoning more effective, by slowing the rate of populationrecovery after application of control. Population-wide fertility reduction canpotentially be achieved by immunocontraception. In this approach, theimmune systems of females are ‘tricked’ into raising antibodies againstproteins found normally in sperm or oocytes. Vaccination could be achievedby oral delivery of antigens in baits, or by the dissemination of antigens in agenetically engineered organism. These approaches are being developedfor use against wild rabbits and mice, and might also be applied to pigs.

The damage caused by feral pigs to the environ-ment and the threat they represent to the domes-tic pig industry as potential reservoirs of exoticdisease have been reviewed by Choquenot et al.

(1996). Other concerns such as the possibilityferal pigs might host human diseases such asJapanese encephalitis that might impact on in-dustries such as tourism have more recentlyemerged. There are thus strong and clear gainsto be achieved in a range of industries throughcontrol of feral pig populations.

FERAL PIGS - CURRENT CONTROLPROCEDURESCurrent control procedures for the minimising thedeleterious impacts on the environment of fecundspecies such as feral pigs rely either on increas-ing the rate of pig mortality through shooting (withor without the use of dogs), poisoning or trapping,or exclusion methods such as fencing or habitatmodification, or some combination of these. Thereare a number of disadvantages to theseapproaches ranging from target species specificityto concerns regarding humaneness and accept-ability, to practical issues such as cost andmaintenance. Biological control can overcomesome of these objections but in the case of pigsno agent has been identified which is of highlethality, humane, and which would not present aproblem either to our own domestic pig industryor to our export markets.

FERTILITY CONTROL - IS IT AN OPTION?Another option for control of fecund species, whichhas received varying degrees of attention, is todecrease the birth rate. Feral sows produce twoweaned litters every 12 –15 months with an aver-age litter size of 4.9-6.3 piglets (Giles, 1980,Pavlov, 1983). This fecundity has been directlycompared with that of rabbits rather than otherferal ungulates (Choquenot et al, 1996). Recentstudies (Williams & Twigg 1996), have indicatedthat fertility control of rabbits affects populationdynamics and over time causes a decline inrabbit numbers. Could a similar result be achievedfor pigs?

Studies on pig population dynamics in differentenvironments show that rates of increase arehighly variable from year to year and are drivenby factors like rainfall which determine green feedavailability and thus protein intake (Giles 1980,Hone 1987, Saunders 1988, Caley 1993,Choquenot, 1994). Insufficient protein intake bysows compromises lactation with obviousimplications for survival of piglets, and alsoincreases susceptibility to disease. Other factorssuch as predation by dingoes or wild dogs canalso have an impact on rate of increase. Thusfertility control applied judiciously in seasons ofnaturally low reproduction might be expected tohave a significant negative impact on populationdynamics. A similar response would be expected

Fertility control for feral pigs - options and status 65

Identifying Management Options

if conventional control procedures were used tolower pig numbers and then followed with fertilitycontrol. How successful fertility control might bein preventing pig population increases in goodseasons in the absence of other control proce-dures remains for the moment a matter ofconjecture. However, fertility control technologiesin general should not be viewed as replacementsfor all existing control methodologies but ratheras additional methods which provide further powerto an integrated management approach.

FERTILITY CONTROL - HOW MIGHT ITWORK?Fertility control can be achieved in a number ofways. In selecting the optimal approach thereproductive biology of the target species needsto be well understood. Fundamental decisionssuch as whether targeting the male or the femaleor both need to be considered. In highly monoga-mous species, for example, it may not matterwhich sex is targeted but generally the female isthe prime target because she is the breeding unit.Given that decision we need to carefully examinewhat we know about reproduction in the female.Are there species specific aspects that could betargeted? Options such as disruption of game-togenesis, prevention of ovulation, inhibition offertilization, interference with implantation or eveninhibition of lactation need to be considered. Inthe case of pigs we have substantial backgrounddata on these processes thanks to intensiveattempts to optimise reproduction in domesticpigs.

One option would be to use agonists or antago-nists of hormones regulating the process ofovulation such as leutenising hormone releasinghormone (LHRH), or steroid analogues thatinterfere with endogenous steroid function thusaffecting gametogenesis or ovulation. A similarapproach could also be developed to interfere withimplantation (Nie et al. 1997). The limitingfeature of all these approaches is that they utilisechemicals which must be delivered to the targetanimal. This would have to be through some sortof bait. This immediately raises difficulties, theprincipal one of which is to ensure species-specific delivery. This is critical as many of thesehormone analogues are active across a widerange of mammalian species and perhaps evenin some non-mammalian species. Practical con-siderations like palatability, number of dosesrequired and cost of both the baits and also their

distribution means that such an approach tofertility control applied to feral pigs probably haslittle chance of success.

A different approach can be developed whichrelies on the fact that the reproductive system,and in particular the gametes, are shielded fromthe animal’s immune system. Thus componentsof the sperm and oocyte will provoke an immuneresponse in immunised animals and this responsecan cause infertility. This approach can potentiallybe made species-specific by selecting appropri-ate proteins to be used as antigens to provokethe immunocontraceptive response. The limita-tion comes in selecting the delivery system. Twobroad options exist:

• Oral delivery systems in which the antigen iseither packaged in an inert system such asmicroencapsulation or is incorporated in a livebut genetically crippled delivery system suchas the bacterium S. typhimurium or a virussuch as vaccinia;

• A disseminating micro-organism which isgenetically engineered to express the antigenof interest.

Oral delivery of a genetically engineered vacciniavirus has been successfully used to immunizefoxes in Europe and now in North America againstrabies, and so this approach has had somesuccess at the field level. Nevertheless, the samelimitations that apply to conventional baiting alsoapply here and it will need a new approach suchas the pheremone based strategy described bySeamark in the previous paper before baiting willhave broad applicability to feral pig control inAustralia.

The second approach is one in which the CRCfor the Biological Control of Pest Animals hasdeveloped expertise and is applying to control ofwild rabbits (Holland & Jackson, 1994) and mice(Chambers et al, 1997, Jackson et al, 1998).Species specificity would be achieved boththrough selection of a pig specific pathogen andpig specific antigen. This minimises the risk to non-target species. In the case of pigs an appropriatevirus might be swinepox. This virus can begenetically engineered to express foreignantigens, is species specific and is endemic toAustralia. Domestic pigs can be prevented fromcontracting swine pox simply through good

66

husbandry procedures and the disease is almostnever a problem in commercial piggeries. Thedisease itself is not fatal, although infectedanimals show significant loss of condition anddevelop papules on the abdomen and legs whichbecome pustular then scab and crust.Development of a vaccine that would offerprotection against infection is possible. Thus thevirus would represent little threat to our domesticor export industry. Swinepox is normallytransmitted by the pig louse (Haematopinus suis)

although presumably other arthropod vectorsmight also play a role. Thus it is conceivable thatferal pigs, which are social animals, would transmitthe infection. Indeed, it is possible that swinepoxinfects feral pigs at high incidence already. Noserological data on this point are available. If mostanimals are already seropositive, techniques mustbe developed which ensure that theimmunocontraceptive virus can transmitsuccessfully in competition with wild type strainscurrently circulating in the environment. This isthe same situation currently being addressed bythe CRC for Biological Control of Pest Animals inthe case of rabbits and myxoma virus. Much ofwhat is learned there will be applicable to pigs.However, if seropositivity is low in feral pigpopulations the immunocontraceptive virus willestablish more easily. Thus a potential species-specific delivery system could readily bedeveloped. If the search, currently underway withPRDC support, for a specific antigen succeedsthe basic components for development of a viralvectored pig immunocontraceptive vaccine exist.

LIMITATIONS TO FERTILITY CONTROLBomford (1990) in her comprehensive review offertility control listed four major limitations to thesuccess of fertility control:

• The lack of a long acting agent, whichmakes repeated dosing necessary

• High cost of delivery, especially ifbaiting is involved

• Less effect on the population thanwhen an equivalent number of animalsare killed

• Potential effects on non target species

Viral-vectored immunocontraception, which hasbeen developed subsequent to 1990, largelyovercomes these limitations. Immuno-contraception is potentially a technology with longterm efficacy. Jackson and coworkers (1998) have

reported mice remain infertile for at least 12months after one exposure to a recombinant virus.In their system the response can be boosted bysubsequent re-exposure to recombinant virus. Ifa disseminating virus is used the issue of oraldelivery is obviated, although there may still besituations (see below) where oral delivery maybe economic. The question of impact is harder toassess although the Williams and Twigg (1996)data for rabbits suggest there are significant long-term benefits. This would need to be investigatedearly in any project to develop fertility control forpigs. The real limitations to lethal procedures arecost, the need for regular reapplication to maintainefficacy, and the issue of humaneness. Theseproblems are not as applicable to viral vectoredimmunocontraception. Finally in response toBomford’s final point, I have already dealt withhow species-specificity can be maintained inimmunocontraception through judicious choice ofboth antigen and delivery system.

In conclusion, viral vectored immunocontraceptionrepresents for some species an importantpotential addition to the integrated suite oftechniques required to ensure adequate controlof a pest species.

POTENTIAL ADDITIONAL BENEFITSDevelopment of viral delivery systems that arepig specific provides opportunities beyondfertility control. Swinepox could be engineered tocontain antigens that would induce an immuneresponse which would be protective againstendemic diseases like brucellosis, leptospirosis,tuberculosis and Murray Valley encephalitis orexotic diseases such as foot and mouth, Africanswine fever, Aujeszky’s disease or classical swinefever. In such a case the utility and economics oforal delivery change and it becomes possible toconsider a recombinant vaccinia systemanalogous to that used for rabies control describedearlier. Thus fertility control combined withdisease prevention through the use of geneticengineered delivery systems provides a newopportunity for development of a novel approachto feral pig management

Fertility control for feral pigs - options and status 67

Identifying Management Options

Control of feral pigs in the Wet Tropics

STORK, N. E.1, STANLEY, T.2

1CRC for Tropical Rainforest Ecology and Management, Cairns2Tropical Weeds Research Centre, Queensland DNR, Charters Towers

68

The papers in this volume identify problemscaused by the feral pig population in the WetTropics. Feral pigs cause widespread impacts onthe ecology of tropical rainforests and also causedamage to agricultural systems neighbouringrainforest areas. By far the most importantconcern is the potential for feral pigs to transmitdiseases such as foot and mouth, Japaneseencephalitis and rabies should any of these reachAustralia. The impact of such outbreaks could bedevastating for the Australian livestock industry.The ability to control populations of feral pigs istherefore of vital importance to the AustralianQuarantine Inspection Service, to the farmingindustry, and to other industries such as tourismwhich benefit from the Wet Tropics World Herit-age Area and Cape York wilderness area.

There is a strong push from the hunting commu-nity to be allowed access to hunting in NationalParks and the World Heritage Area. There aredefinite concerns that if this were allowed, it wouldcause major impacts on the environment andcould even lead to an increase in pig impacts bydisrupting their movement patterns and socialorganisation.

Internationally, feral pigs are a problem for anumber of countries such as Malaysia, NewZealand and USA (Hawaii). In some researchareas we have greater expertise and haveadvanced further than researchers in some ofthese countries. In other areas we are lessadvanced.

Feral pigs are a problem for a wide range oforganisations and communities. Currentlyresources being used to tackle feral pigs arespread thinly and we need to produce a combinedprogram with wide support in order to produceeffective action and to give greater leverage toexisting activities. Joint international programswould enhance our own effectiveness as well ascreating the potential for new partnerships in otherareas of environmental concern.

The goal of research into feral pigs in the WetTropics should be to develop strategies to reduceor control feral pig populations in the WetTropics, or parts of the Wet Tropics, and reducetheir impact on natural ecosystems and onagricultural systems.

The following crucial issues can be identified:

1. Potential for spread of human and animal

diseases by feral pigs2. Biodiversity and ecosystem impact of feral

pigs3. Land use and other socio-economic value

impacts4. Development of management strategies5. Rehabilitation of areas affected by feral pigs6. Development of a community education

program

STRATEGIES

Potential for spread of diseases by feral pigs• The likelihood of diseases being introduced

to feral pig communities in the Wet Tropicsand Cape York needs to be addressed. Col-laboration with AQIS sentinel pig program isessential here. An education program isneeded to make land owners and legislatorsaware of the scale of the problem.

Biodiversity and ecosystem impact offeral pigs• Some research has been carried out by Jim

Mitchell to show the impact of feral pigs onnative vegetation and invertebrate communi-ties (particularly earthworms) but a morecomplete understanding is needed. Forexample, how do pigs impact on the speciescomposition of forests through time? Howdoes their feeding affect the pattern anddynamics of vegetation mosaics? Whatthreats are there from pigs to particularlyvulnerable species and communities?

69

Land use and other socio-economic valueimpacts• Land owners in some parts of the Wet Trop-

ics and Cape York are concerned about thedamage caused by feral pigs to their cropsand infrastructure. At present few data areavailable to allow an evaluation of the extentand cost of this damage and its impact onlivelihoods. Pigs also affect the aesthetic valueof Wet Tropics areas and this needs to beevaluated.

• Prediction of the social and economic costsof introduction of diseases to feral pig com-munities is a priority as the potential impactcould be huge.

Development of management strategies• A monitoring system for feral pig numbers and

impact needs to be created and implemented.Some progress has been made towards thedevelopment of a monitoring index.

• There is a need for the development of thedelivery of a safe baiting system that will tar-get pigs and not affect native species. Thismeans that a variety of alternative attractantsthat are easy to deliver will need to be tested.

• There should be an evaluation of the existingpig management programs to determinewhether they reduce pig numbers and theirimpact in the short, medium and long-term.For example does pig trapping/shooting causea breakdown in pig social structure and even-tually lead to increase in pig numbers?

• Immuno-contraception is being developed asnew technology to reduce numbers of foxes,rabbits and mice by the CRC for the Biologi-cal Control of Pest Animals. Some initial trialshave been undertaken to identify the poten-tial use of this technology to control feral pigs.

Rehabilitation of areas affected by feral pigs• Once feral pigs have been removed from an

area there is a need to rehabilitate the area toprevent weed development. Some evaluationof the extent of this problem needs to beundertaken.

Development of a community educationprogram• There are many misconceptions about feral

pigs and it is essential that the facts are pre-sented to communities in order to allay theirfears and to gain greater community supportfor the course of action being undertaken.

Alexiou, P.N. (1983) Effect of feral pigs (Sus

scrofa) on subalpine vegetation at Smok-ers Gap, Australian Capital Territory.Proceedings of the Ecological Society of

Australia 12: 135-142.Anderson, S.J. and Stone, C.P. (1993). Snaring

to control feral pigs Sus scrofa in a remoteHawaiian rain forest. Biological Conserva-

tion 63, 195-201.Aplet, G.H., Anderson, S.J. and Stone, C.P.

(1991). Association between feral pigdisturbance and the composition of somealien plant assemblages in Hawaii Volca-noes National Park. Vegetatio 95, 55-62.

Bargmann, C.I. (1997). Olfactory receptors,Vomeronasal Receptors, and the organi-zation of olfactory information. Cell 90,585-587.

Bomford, M. (1990) A role for fertility in wildlife

management? Bulletin 7 Bureau of RuralResources. Australian Government printingOffice, Canberra.

Bowman, D.M.J.S. and McDonough, L. (1991).Feral pig (Sus scrofa) rooting in a monsoonforest-wetland transition, Northern Australia.Wildlife Research 18, 161-765.

Bowman, D.M.J.S. and Panton, W.J. (1991). Signand habitat impact of banteng (Bos

javanicus) and pig (Sus scrofa), CobourgPeninsula, northern Australia. Australian

Journal of Ecology 16, 15-17.Brown, B.N. (1976) Phytophthora cinnamomi as-

sociated with patch death in tropical rain-forest in Queensland. Australian Plant Pa-

thology Society Newsletter 5: 1-4Burke, D.S. and Leake, C.J. (1988). Japanese

encephalitis. Ch. 28 In: Monath, T.P. (Ed.)The arboviruses: Epidemiology and ecol-

ogy. CRC Press, Boca Raton, Fla.Caley P (1993) The ecology and management of

feral pigs in the wet-dry tropics of the North-

ern territory. M. Appl Sci thesis, Universityof Canberra

Caley, P. (1993) Population dynamics of feral pigs(Sus scrofa) in a tropical riverine habitatcomplex. Wildlife Research 20, 625-636.

Caughley, G. (1977). Analysis of Vertebrate

Populations. (Wiley & Sons: London.)

Caughley, G. and Gunn. A. (1996) ConservationBiology in Theory and Practice. BlackwellScience. London.

Chambers L., Singleton, G.R. and Hood G.M.(1997) Immunocontraception as a potential

control method of wild rodent populations.

Belg J. Zool 127 (Supplement 1) 145-56.Choquenot, D. (1994) The dynamics of feral pig

populations in Australia’s semi-arid

rangelands. PhD Thesis, University ofSydney.

Choquenot D., McIlroy J. and Korn T. (1996) Man-

aging vertebrate pests: feral pigs. Bureauof Resource Sciences, Australian Govern-ment Printing Service, Canberra.

Choquenot, D., Kay, B. and Lukins, B. (1990).An evaluation of Warfarin for the control offeral pigs. Journal of Wildlife Management

54, 353-359.Choquenot, D., Kilgour, R. J. and Lukins, B. S.

(1993) - An evaluation of feral pig trapping.Wildlife Research 20, 19-22.

Choquenot, D., McIlroy, J. and Korn, T. (1996)Managing Vertebrate Pests: Feral Pigs.Bureau of Resource Sciences, AustralianGovernment Publishing Service, Canberra.

Chu, R.M. and Joo, H.S. (1992) ‘Japanese B en-cephalitis’. Chap. 21 In: Diseases of Swine

7th Ed. 286- 291. Wolfe Publishing,London.

Crome, F.H.J. and Moore, L.A. (1990).Cassowaries in north-eastern Queensland:report of a survey and a review and assess-ment of their status and conservation andmanagement needs. Australian Wildlife

Research 17, 369-385.Dexter, N. (1990) Population density and man-

agement of feral pigs at Aurukun, NorthQueensland. Bureau of Rural Resources,Report R/11/91, Canberra.

Dorrington, B. and James, P. (1999) Community

Based Feral Pig Trapping Program – An-

nual Report 1998. Report presented to theWet Tropics Management Authority. DNR,South Johnstone.

Fleming, P.J.S. (1997). Uptake of baits by redfoxes (Vulpes vulpes): implications for ra-bies. Wildlife Research 24, 335-346.

French, C.E., Liscinsky, S.A. et al. (1957). Nutri-ent composition of earthworms. Journal of

Wildlife Management 21, 348.

REFERENCES

References 71

Gao Y and Short R.V. (1993) The control of Ro-dent Populations. Oxford Reviews in Re-

productive Biology 15, 265-310Giles J.R. (1980) The ecology of feral pigs in

western New South Wales. PhD Thesis,University of Sydney

Hall, R.A., Burgess, G.W. and Kay, B.H. (1986)Detection of Australian flavivirus antigensusing monoclonal antibodies in ELISA. Ar-

bovirus Research in Australia, Proceedings

Fourth Symposium May 1986: 118 - 122.Hanna, J., Ritchie, S., Phillips, D., Shield, J. et al

(1996) An outbreak of Japaneseencephalities in the Torres Strait, Australia.Med. J. Aust. 165: 256-260.

Hanna, J., Ritchie, S., Phillips, D., Shield, J. et al.

(1999). Japanese encephalitis in northQueensland, Australia, 1998. Med. J. Aust.

170: 533-536.Hassall and Assc. P/L. (1998). Economic Evalu-

ation of the Role of Bounties in Vertebrate

Pest Management. Prepared for the Bureauof Resource Sciences. BRS, Canberra.

Holland M.K. & Jackson R.J. (1994) Virus

vectored immunocontraception for control

of wild rabbits: identification of target anti-

gens and construction of recombinant vi-

ruses. Reproduction, Fertility & Develop-ment 6: 631-42

Hone J. (1987) Theoretical and practical aspects

of feral pig control. PhD Thesis, AustralianNational University

Hone, J. (1983). A short-term evaluation of feralpig eradication at Willandra in western NewSouth Wales. Australian Wildlife Research

10, 269-275.Hone, J. (1988a). Evaluation of methods for

ground survey of feral pigs and their sign.Acta Theriologica 33, 451-465.

Hone, J. (1988b). Feral pig rooting in a moun-tain forest and woodland: distribution, abun-dance and relationships with environmen-tal variables. Australian Journal of Ecology

13, 393-400.Hone, J. and Kleba, R. (1984). The toxicity and

acceptability of warfarin and 1080 poisonto penned feral pigs. Australian Wildlife

Research 11, 103-111.Hone, J. and Martin, W. (1998). A study of dung

decay and plot size for surveying feral pigsusing dung counts. Wildlife Research 25,255-260.

Hone, J. and Stone, C. (1989). A comparison andevaluation of feral pig management in two

national parks. Wildlife Society Bulletin 17,419-425.

Hone, J. (1987) Theoretical and practical aspects

of feral pig control. Ph.D. Thesis, Austral-ian National University, Canberra.

Hone, J. (1990) Notes on seasonal changes inpopoulation densities of feral pigs in threetropical habitats. Australian Wildlife Re-

search 17, 131-134.Hone, J. (1990) How many feral pigs in Australia?

Australian Wildlife Research 17: 571-572Hone, J. (1992). Modelling of poisoning for verte-

brate pest control, with emphasis on poi-soning feral pigs. Ecological Modelling 62,311-327.

Hone, J. (1995). Spatial and temporal aspects ofvertebrate pest damage with emphasis onferal pigs. Journal of Applied Ecology 32,311-319.

Hone, J. (1998) Island biogeography theory andferal pig disturbance of grassland.Abstracts. Australasian Wildlife Manage-ment Society 11th Annual Meeting, GattonCollege, The University of Queensland.

Hone, J. and Pedersen, H. (1980) Changes in aferal pig population after poisoning. InProceedings of the 9th Vertebrate Pest Con-

ference. Ed. P. J. Clark. Davis: California.Hopkins, M.S. and Graham, A.W. (1985) A

preliminary assessment of problems andcontrol measures of feral pigs in the rain-forests of the Field Force Battle School,Jarra Creek via Tully, north Queensland.Internal Report P5/297, CSIRO Division ofLand and Water Resources, Canberra.

Jackson R.J., Maguire D.J., Hinds, L.A. andRamshaw I.A. (1998) Infertility in mice in-duced by a recombinant ectromelia virusexpressing mouse zona pellucida glycopro-tein three. Biology of Reproduction 58: 152-59.

Katahira, L.K., Finnegan, P. and Stone, C.P.(1993). Eradicating feral pigs in montanemesic habitat at Hawaii Volcanoes NationalPark. Wildlife Society Bulletin 21, 269-274.

Kay, B.H., Boreham, P.F.L., and Williams, G.M.(1979). Host preferences and feeding pat-terns of mosquitoes (Diptera: Culicidae) atKowanyama, Cape York Peninsula, north-ern Queensland. Bull. Ent. Res. 69: 441-457.

Kliejunas, J.T. and Ko, W.H. (1976) Dispersal ofPhytophthora cinnamomi on the island ofHawaii. Phytopathology 66: 457-460.

References72

Laurance, W.F. and Harrington, G.N. (1997).Ecological associations of feeding sites offeral pigs in the Queensland Wet Tropics.Wildlife Research 24, 579-590.

Laurance, W.F. and Grant, J.D. (1994) Photo-graphic identification of ground-nest preda-tors in Australian tropical rainforest. Wild-

life Research 21: 241-248.Laurance, W.F., Garesche, J. and Payne, C.W.

(1993) Avian nest predation in modified andnatural habitats in tropical Queensland: anexperimental study. Wildlife Research 20:711-723.

Lee, K.E. (1985). Earthworms: Their Ecology and

Relationship with Soils and Land Use. Syd-ney, Academic Press.

Masters, K. (1979) Feral pigs in the south-westof Western Australia. Final Report to FeralPig Committee. Agriculture ProtectionBoard and Department of Conservation andLand Management, Western Australia.

McGaw, C.C. and Mitchell, J. (1998). Feral Pigs(Sus scrofa) in Queensland. Pest StatusReview Series – Land Protection. DNR,Brisbane.

McIlroy, J.C. & Saillard, R.J. (1989). The effect ofhunting with dogs on the numbers andmovements of feral pigs, Sus scrofa, andthe subsequent success of poisoning exer-cises in Namadgi National Park, A.C.T. Aus-

tralian Wildlife Research 16, 353-363.McIlroy, J.C. (1993). Feral pig management prob-

lems in the Wet Tropics of QueenslandWorld Heritage Area. Cairns, Wet TropicsManagement Agency.

McIlroy, J.C. and Gifford, E.J. (1997) The “Judas”pig technique: a method that could enhancecontrol programs against feral pigs, Sus

scrofa. Wildlife Research 23, 483-491.McIlroy, J.C. Braysher, M. and Saunders, G. R.

(1989) Effectiveness of a warfarin-poison-ing campaign against feral pigs, Sus scrofa,

in Namagdi National Park. Australian Wild-

life Research 16, 195-202McIlroy, J.C., Gifford, E.J. and Forrester, R.I.

(1993). Seasonal patterns in bait consump-tion by feral pigs (Sus scrofa) in the hill coun-try of south-eastern Australia. Wildlife Re-

search 20, 637-651.McIlroy, J.C. (1993). Feral Pig Management in the

Wet Tropics of Queensland World Heritage

Area. Final report to the Wet Tropics Man-agement Authority, Cairns. CSIRO, Can-berra.

McIlroy, J.C. (1990) Feral pig. In: C.M. King, (Ed.),The Handbook of New Zealand Mammals.Oxford University Press, Auckland, NewZealand.

McIlroy, J.C. (1993) Feral pig management prob-

lems in the wet tropics of Queensland World

Heritage Area. Final Report to the Wet Trop-ics Management Agency, Cairns.

Mitchell, J.L. (1993.) Feral Pig Management in

the Wet Tropics. Systematic assessmentof Feral Pig damage and recommended Pigcontrol methods in the Wet Tropics WorldHeritage Area. Department of Lands, Char-ters Towers.

Mitchell, J.L. (1998) The effectiveness of aerialbaiting for control of feral pigs (Sus scrofa)in North Queensland. Wildlife Research 25,297-303.

Mitchell, J.L. and Mayer, R. (1997). Diggings byferal pigs within the Wet Tropics World Her-itage Area of north Queensland. Wildlife

Research 24, 591-601.Mykytowcz R. (1985) Odour signals in the life of

the rabbit. Australian Science Magazine.

Issue 1Nie, G.Y., Butts A.R., Salmonson and Findlay J.K.

(1997) Hormonal and noon-hormonalagents at implantation as targets for con-traception. Reproduction, Fertility & Devel-

opment 9:65-77Noble, K. 1994. Sustainable Best Practice Feral

Pig Management in the Queensland Wet

Tropics. Final report to the Vertebrate PestProgram. DNR, Townsville.

O’Brien, P. H. & Lukins, B. S. (1990). Compara-tive dose-response relationships and ac-ceptability of warfarin, brodifacoum andphosphorus to feral pigs. Australian Wild-

life Research 17, 101-112.Pav Ecol. (1992) Investigation of feral pig

populations and control measures; Cape

Tribulation Section of the wet tropics World

Heritage Area. Consultancy report to WetTropics Management Authority, Cairns.

Pavlov P.M. (1983) Feral Pigs. In: (Ed) R Strahan.The Australian Museum Complete Book of

Australian Mammals. Angus & RobertsonPublishers, Sydney.

Pavlov, P.M., Crome, F.H.J. and Moore, L.A.(1992). Feral pigs, rainforest conservationand exotic disease in north Queensland.Wildlife Research 19, 179-193.

References 73

Pech, R. & McIlroy, J.C. (1990). A model of thevelocity of advance of foot-and-mouthdisease in feral pigs. Journal of Applied

Ecology 27, 635-650.Pech, R.P. & Hone, J. (1988). A model of the dy-

namics and control of an outbreak of footand mouth disease in feral pigs in Australia.

Journal of Applied Ecology 25, 63-77.Pieris, J.S.M., Amerasinghe, F.P., Arunagiri, C.K.,

Perera, L.P. et al. (1993). Japanese en-cephalitis in Sri Lanka: comparison of vec-tor and virus ecology in different agro-cli-matic areas. Trans. Roy. Soc. Trop. Med.

Hyg. 87: 541-548.Pullar, E.M. (1950) The wild (feral) pigs of Aus-

tralia and their role in the spread of exoticdiseases in Australia. Australian Veterinary

Journal 26: 99-109.Richards, S.J., McDonald, K.R. and Alford, R.A.

(1993) Declines in populations of Austral-ia’s endemic tropic frogs. Pacific Conser-

vation Biology 1: 66-77.Ritchie, S., van den Hurk, A., and Shield, J. (1997).

The 1995 Japanese encephalitis outbreak:Why Badu? Arbovirus Res. Aust. 7: 131-136.

Ritchie, S.A., Phillips, D., Broom, A., Mackenzie,J. et al. (1996). Isolation of Japanese en-cephalitis virus from Culex annulirostris inAustralia. Am. J. Trop. Med. Hyg. 56: 80-84.

Roberts, A., Klomp, N. and Birckhead, J. (1996)Monitoring marine and terrestrial hunting inan Aboriginal community in north Queens-land. In: M. Bomford and J. Caughley, eds,

Sustainable Use of Wildlife by Aboriginal

Peoples and Torres Strait Islanders. Bureauof Resource Sciences, AustralianGovernment Printing Service, Canberra.

Saunders G (1988) The ecology and management

of feral pigs in New South Wales. MSc The-sis, Macquarie University

Saunders, G. and Giles, J. (1995) - Ecologicalcomparison of two wild pig populations insemi-arid and sub-alpine Australia. IBEX,

Journal of Mountain Ecology 3: 152-55.

Saunders, G. & Kay, B. (1996). Movements andhome ranges of feral pigs (Sus scrofa) inKosciusko National Park, New South Wales.Wildlife Research 23, 711-719.

Saunders, G. (1988) The ecology and manage-

ment of feral pigs (Sus scrofa) in New South

Wales. M. Sc. Thesis, Macquarie Univer-sity, Sydney.

Saunders, G. (1993). Observations on the effec-tiveness of shooting feral pigs from helicop-ters in western New South Wales. Wildlife

Research 20, 771-776.Saunders, G. (1993). The demography of feral

pigs (Sus scrofa) in Kosciusko NationalPark, New South Wales. Wildlife Research

20, 559-569.Saunders, G., Kay, B. & Nicol, H. (1993). Factors

affecting bait uptake and trapping successfor feral pigs (Sus scrofa) in Kosciusko Na-tional Park. Wildlife Research 20, 653-665.

Saunders, G., Kay, B. and Nicol, H. (1993). Fac-tors affecting bait uptake and trapping

Saunders, G., Kay, B. and Parker, B. (1990).Evaluation of a Warfarin poisoning programfor feral pigs (Sus scrofa). Wildlife Research

17, 525-533.Singer, A.G. (1991). A chemistry of mammalian

pheromones. J. Steroid Biochem. Molec.

Biol. 39, 627-632Statham, M. and Middleton, M. (1987) Feral pigs

on Flinders Island. Papers and Proceedings

of the Royal Society of Tasmania 121: 121-124.

Thompson, J.A. and Fleming, P.J.S. (1994).Evaluation of the efficacy of 1080 poison-ing using visitations to non-toxic baits as anindex of fox abundance. Wildlife Research

21, 27-39.Tisdell, C.A. (1984) Feral pigs threaten native

wildlife in Australia. Tigerpaper 11: 13-17.Whitten, W.K. (1980). Induction of marking be-

haviour in wild red foxes (Vulpes Vulpes

L.) by synthetic urinary constituents. J.

Chem. Ecology 6, 49-55Williams CK & Twigg L (1996) Responses of wild

rabbit populations to imposed sterility. In: RFloyd, R Shepherd nas PJ DeBaro (Eds)Frontiers of Population Ecology. pp547-60.CSIRO Publishing, Melbourne.

Wilson, G. Dexter, N. O’Brien, P. and Bomford,M. (1992) Pest animals in Australia - a sur-

vey of introduced wild animals. KangarooPress and Bureau of Rural Resources,Canberra.

References74